mm ‘lelllllllllllml 1293 008550 This is to certify that the thesis entitled Dependent Patients and Their Physicians: A Sociological Study of the High Use of Primary Care presented by Rita Schlesinger Gallin has been accepted towards fulfillment of the requirements for Ph . D degree in mm flame @344 Major professor Date 2/ 6 1979 0-7639 LIBRARY MiChigifl w Umvzmity OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. © 1979 RITA SCHLES l NGER GALLI N ALLRICHTS RESERVED DEPENDENT PATIENTS AND THEIR PHYSICIANS: A SOCIOLOGICAL STUDY OF THE HIGH USE OF PRIMARY CARE SERVICES By Rita Schlesinger Gallin A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1979 ABSTRACT DEPENDENT PATIENTS AND THEIR PHYSICIANS: A SOCIOLOGICAL STUDY OF THE HIGH USE OF PRIMARY CARE SERVICES By Rita Schlesinger Gallin This study is about female patients who use high levels of primary care services and their physicians' response to them. The purpose of the study was to explore the way in which (l) variations in the background factors of patients and the context of the medical profession influence the nature of the interaction between physician and patient; and (2) variations in the use of primary care services are generated by the dynamics of their interaction in a medical situation. To explore these variations, a retrospective research plan was developed. A sample of young "healthy" females, stratified according to patterns of use during a six-month period. was selected and the data contained in their medical records for that period abstracted. These patients were interviewed in their homes using a structured survey instrument. In addition, a sub-set of these patients was interviewed in-depth. Finally, the physicians who cared for these patients completed a self-administered questionnaire. Pearson product-moment correlations and multiple regression analyses were used to analyze the data collected from the samples. “.1 11".} Ila-I ii! Rita Schlesinger Gallin The data showed that the poor, less educated, separated or divorced were likely to consider themselves financially and socially deprived and to have experienced a series of stressful events that challenged their resources and skills. Furthermore, the data showed that such women were likely to feel that their psychological assets were limited and their social assets only nominally effective in mediating between them and the consequences of life difficulties. It was found that differences in women's life situations were predictive of variations in their medical and non-medical needs as patients. women who considered the quality of their life poor tended to express a need to cope with failure and a need for catharsis, as well as to present many "symptoms of dependency." Perceived financial and social deprivation exerted the most con- sistent influence on these patient needs. Psychological and social assets available, however, also exerted an effect on the women's needs as patients. The data also showed that physicians' decisions influenced patterns of utilization and the total volume of use of medical services. Patients who were assigned a low proportion of supported diagnoses and received a high proportion of psychotropic and sympto- matic therapies had a high volume of self-initiated use. Patients who received many requests for follow-up visits and a high propor- tion of prescriptions for curative and symptomatic therapies had a high volume of physician-initiated use. The total volume of use was highest, however, among patients who (1) expressed a need to cope Rita Schlesinger Gallin with failure and/or a need for catharsis; (2) presented many "symptoms of dependency;" and (3) received a high proportion of psychotropic and symptomatic therapies. It was concluded that the high use of primary care services was a consequence of the conjunction of the needs of the patients and physicians' responses to these needs. Some patients frequently sought care in the hope that the physician would legitimate their feelings of failure and provide the support that was lacking in their personal environment. When the physician responded sympa- thetically to these needs, the patient was encouraged to return for the purpose of additional reassurance and support. When the physician responded unsympathetically to these needs, the patient was moved to return in a continuing attempt to satisfy her frus- trated needs. To Bernard For reasons that continue to grow 11' “l“ 'w m ACKNOWLEDGMENTS A large number of people have been generous in providing not only their time and assistance, but also their encouragement throughout this project. Although I would have it otherwise, only a few can be acknowledged here by name. The project was supported by Grant Number l R03 HS 02854-01 from the National Center for Health Services Research, HRA. I am grateful to the Center for providing me with the opportunity to conduct the study. Without the cooperation and support of Lewis E. Simoni, M.D., Medical Director of the residency training program at the family health care center, this study could not have been done. I am indebted to Dr. Simoni for opening the practice to me. A very special thank you goes to G. William Given, Ph.D. who offered me friendship, guidance and assistance during the six years I was his research assistant. His encouragement, commitment and insights contributed immeasurably to the success of the research effort and to my intellectual growth. I owe him a debt that can never be adequately repaid. A debt also is owed to the members of my graduate comnittee who supervised my training and this research. Elianne Riska, Ph.D., served as my thesis director. She set a standard of scholarship iii SU‘ which I hope one day to emulate. Bo Anderson, Ph.D., my major professor throughout most of my graduate training, and my teachers, Harry Schwarzweller, Ph.D., and Joseph Spielberg, Ph.D., provided me with supurb training and served as whetstones on which I have tried to sharpen my thinking. Peter Manning, Ph.D. was a reader of the thesis. I am grateful to him for taking time from a very busy schedule to offer assistance. Appreciation is expressed to the faculty and staff of the Department of Community Health Science, College of Human Medicine, Michigan State University, who contributed to the study in many direct and indirect ways. Sister Mary H. Kroger, Ph.D., handled a myriad of administrative details with patience and dispatch. Robert Sprafka, Ph.D., provided invaluable assistance with the sample design and other methodological problems, thereby helping me to avoid many errors. Jacquard Guenon, M.S., and Brijesh Srivastava, M.S., responded to my innumerable requests for computer programming with skill and good humor. Gayle Galmish, Barbara McWilliams and Nancy Vanderwest shared the task of typing various drafts of the research instruments and report. I am grateful to these people and to all the other members of the department for their generous support and help. Particular mention also is merited by the following people whose assistance contributed so much to the success of the study. Rhina Griffel and Jane Richardson did the bulk of the interviews with patients. Their dedication and perseverance made the writing iv i 5., ,:. if . out... “I! \ Fr ‘ Ulari of this report possible. Marcia Kettler, Karolyn Lee and Jane Wisner performed the difficult task of "deciphering" physicians' handwriting and abstracting data from medical records. Richard E. Sturm, M.D., coded the data and provided me with valuable insights. William Crano, Ph.D., gave me invaluable help with the analysis of the data. I am grateful to these peOple for their efforts on my behalf. Finally, I am indebted to my family who absorbed much of the stress associated with my graduate training. My sons, Marc and Peter, helped me immeasurably by their forbearance in the face of my not infrequent distraction and ill humor. Without doubt, however, the strongest and steadiest help has come from my husband, Bernard. His support and patience created an environment in which the realization of my goal became possible. H l! 44 TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES LIST OF MAPS LIST OF APPENDICES Chapter I. II. III. INTRODUCTION AND LITERATURE REVIEW Introduction Literature Review . Utilization: A Dependent Variable . Utilization: An Independent Variable THE RESEARCH PROBLEM Theoretical Framework . Parsons' Theory of the Social System Theoretical Rationale . . Basic Concepts . Overview of Concepts . Concepts Relevant to the Women as Individuals. Concepts Relevant to the Women as Patients Concepts Relevant to the Physician Conceptual Model . . . Hypotheses to be Tested Summary. . . DATA AND METHODS . Research Design Assumptions . The Samples . . . The Survey Sample . . . The In-Depth Interview Sample . The Physician Sample . . The Data . . Data Collection Procedures . Medical Records vi Page ix xii xiii xiv Chapter IV. Survey Questionnaires In- Depth Interviews . . Self-Administered Questionnaires . Major Variables . . Characteristics of the Women as Individuals Characteristics of the Women as Patients Characteristics of the Physician . Outcome . . . . . . . . Analysis Plan . Summary . THE SETTING A Brief Social History of the Community Community Hospital and its Family Health Care Center I The Center Population . . The Family Practice Residency Training Program The Operation of the Center . . . . Summary . . . . . . . THE WOMEN AS INDIVIDUALS AND PATIENTS . Part 1: The Women as Individuals Life Circumstances: Social Position Life Circumstances: Quality of Life Life Circumstances. Social Stressors Psychosocial Assets . . Relationship Between Life Circumstances and Psychosocial Assets . Summary: The Women as Individuals Part 2: The Women as Patients . Comparison of the Patients with Others in the Community. . . . . . . . . Medical Needs . . Non-Medical Needs . Relationship Between Medical and Non-Medical Needs : Relationship Between the Womens' Life Situations and Their Characteristics as Patients Tests of Hypotheses . . Patient Orientations to the Physician- -Patient Relationship. . The Patient' 5 Contribution to the Total Use of Medical Services . Summary: The Women as Individuals and Patients vii Page lOO l7l 20l 206 207 208 208 2ll 2l3 223 234 249 252 fralfl I Ifl ”TY h t In. HPPE Chapter VI. THE PHYSICIANS AND THE PHYSICIAN-PATIENT INTERACTION . Part l: The Physicians Overview: The Physicians . . The Physician and the Patient Roles . Physicians' Attitudes Toward the Dependent Patient I Summary . Part 2: The Physician- Patient Interaction Overview: The Physician- Patient Interaction . . Relationship Between Elements of the Physician- Patient Interaction . . . . . . The Physician- Patient Interaction. Tests of Hypotheses . . Outcomes of the Physician- Patient Interaction. Tests of Hypotheses . Patients' Views of the Physician- -Patient Interaction . Summary: The Physicians and the Physician- -Patient Interaction . . . . VII. SUMMARY AND CONCLUSIONS: THE HIGH USE OF PRIMARY CARE SERVICES . . . . . . . . . . Summary of Findings . The Setting. . The Women as Individuals The Women as Patients The Physicians . . The Physician- -Patient Interaction Summary of Test of Hypotheses . The Problem of Generalizability . Implications: Sociology Implications: Medical Care APPENDICES LIST OF REFERENCES viii 328 329 330 331 333 336 339 346 347 352 354 438 t ‘(V I 141' I ‘1‘“ his. 45. 01‘. PH-w LIST OF TABLES Number and Percent of Patients in the Populations Number and Percent of Patients in the Sub-Populations and Average Number of Visits per Patient . Number and Percent of Patients in the Stratified Samples and Average Number of Visits per Patient Number and Percent of Patients Interviewed and Not Interviewed and Average Number of Visits per Patient . Criteria for Poverty Level Income for Non-Farm Families Living in the County Number and Percent of Patients of Family Health Care Center by Age and Sex, l976 . . . Average Number of Physician Visits per Person Annually by Age and Sex: United States, 1974; Family Health Care Center, l976 . . . . . . . . . Percent of Women by Measures of Social Position Zero Order Correlations Between Measures of Social Position . . . . _ Percent of Females 18-40 Years of Age by Selected Measures of Social Position: County, l976; Study Sample, l977 . . . . . . . . . . Zero-Order Correlations Between Measures of Quality of Life . . . . . . . . . . Zero-Order Correlations Between Measures of Social Position and Quality of Life . . . Zero-Order Correlations Between Measures of Social Position and Social Stressors Percent of Women by Characteristics of Kin Networks . ix Page 88 91 93 94 105 141 143 162 163 164 167 168 170 175 Table 5.8 5.14 .20 .21 .22 .23 .24 .25 010101010101 Zero-Order Correlations Between Characteristics of the Kin Network . . . . . . . . . . Percent of Women by Characteristics of Friend Networks . Zero-Order Correlations Between Characteristics of the Friend Network Characteristics of the Friend Network by Measures of Social Position . . . . . . . . Zero- Order Correlations Between Frequency of Contact with Kin and Friends and Quantity of Psychological Support. . . . Content of Functional Support by Provider of Support Zero-Order Correlations Between Measures of Social Positions and Feelings of Failure . Percent of Women Needing Help with Problems by Social Position . . . . . . . . . . . . Zero-Order Correlations Between Need for Help with Problems and Measures of Quality of Life . Percent of Women Needing Help with Personal or Family Problems by Measures of Social Position . Zero-Order Correlations Between Live Circumstances of the Individual and Characteristics of the Patient . Zero-Order Correlations Between Psychosocial Assets of the Individual and Characteristics of the Patient . Step-Wise Regression of Symptoms of Dependency Step-Wise Regression of Need to Cope with Failure Step-Wise Regression of Need for Catharsis . Characteristics of the "Good Doctor“ . Characteristics of the "Good Patient" Step-Wise Regression of the Patient's Contribution of the Total Use of Medical Services . . . Page 176 180 181 183 188 192 196 199 203 205 216 220 226 228 230 236 244 250 (h a (‘I‘ (h Table 6.1 .10 .11 .12 01010101 A.2 D.1 G.1 Physicians' Reasons for Choosing a Residency in Family Medicine . . . . . . . . . . . . . Proportion Matrix for Six Functions of the Physician Judged in Terms of Effective Medical Practice Proportion Matrix for Six Functions of the Patient Judged in Terms of Effective Medical Practice Percent of Residents' Responses to Three Sentence Stems . Zero-Order Correlations Between Patient Characteristics and Physical Findings Zero-Order Correlations Between Diagnoses and Patient Characteristics . . . . . . . . . Zero-Order Correlations Between Patient Characteristics and Management Plans . . . Zero-Order Correlations Between Diagnoses and Management Plans . . . . . . . . . . . . Step-Wise Regressions of Diagnoses Assigned Patients . Step-Wise REgressions of Therapies Prescribed Step-Wise Regression of Requests for Follow-up Visits Step-Wise Regressions of the Physician's Contribution to Patterns of Use of Medical Services . Step-Wise Regression of the Physicians' and Patients' Contribution to the Total Use of Services . Percent of Population in County by Area and Selected Sociodemographic Characteristics, 1976 Percent of Interviewed and Non-Interviewed Samples by Area and Selected Sociodemographic Characteristics . Distribution of Sample Using Three Different Methods to Categorize Illness Level . . Percent of Residents by Selected Sociodemographic Characteristics . . . . . . . . xi Page 261 264 267 272 283 285 287 289 293 296 299 310 314 358 359 416 437 CA.) (11 Figure 2.1 3.1 5.1 F.1 LIST OF FIGURES A Social System Model of the Use of Medical Services . Defining Criteria of a Physician-Initiated Visit Relationship Between the Women's Life Situations and Their Characteristics as Patients Comparison of Different Methods of Defining a Population of Patients . xii Page 76 125 214 431 LIST OF MAPS Map Page 4.1 City, Regular Patients per 1000 Dwelling Units, 1976- 1977. . . . . . . 139 4.2 Out- County, Regular Patients per 1000 Dwelling Units, 1976-1977. . . . . 140 xiii LIST OF APPENDICES Appendix Page A. A COMPARISON OF THE INTERVIEWED AND NON- INTERVIEWED SAMPLES. . . . . . 355 B. PROBLEMS IN THE FIELD . . . . . . . . . . . 361 C. RESEARCH MATERIALS . . . . . . . . . . . . 367 D. PROBLEMS IN DEVELOPING A LEVEL OF ILLNESS LEVEL . . 411 E. DEVELOPMENT OF A DIAGNOSTIC INDEX . . . . . . . 421 F. PROBLEMS IN DEVELOPING A DEFINITION OF A PATIENT POPULATION. . . . . . . . . 428 G. BACKGROUND INFORMATION ON THE RESPONDENTS . . . . 434 xiv CHAPTER I INTRODUCTION AND LITERATURE REVIEW Introduction A growing concern for the inequitable distribution and rising costs of health services in the United States has led to considerable research over the last decades on the utilization of medical care. Much research has been conducted on the influence of patient social, economic and attitudinal characteristics on the utilization and under- utilization of medical care services. Little research, however, has been carried out on identifying the components of high utilization of health services or on the way in which the behavior of health care providers affects patterns of use. Furthermore, the notion that utilization behavior may be consistent with and an adjustment to a general life situation virtually has been ignored in past research. Most researchers have accepted the values, questions and problems defined by health planners and providers as the reference point of their work (Gold, 1977; Roth, 1962). They have focused on the individual and explained utilization behavior on the basis of sociocultural or psychological attributes which predispose people to behave in certain ways. For example, underutilization of services is suggested to be due to lack of knowledge, trust or responsibility by the individual and an impediment to the efficient functioning of the medical system. an This focus on the individual patient or potential patient is but one example of the social scientists' acceptance of the viewpoint of medical personnel as the norm for research. Since physicians usually prefer to accept their own actions as a given (Roth, 1962:47), researchers rarely have considered the behavior of the practitioner and the way in which it affects patterns of utilization. This omission leaves untouched important issues that need to be considered: for example, the expectations practitioners bring to the medical encounter; the definitions they apply to patients and their conditions; and the effect these expectations and definitions have on the manage- ment process and patterns of utilization. The exclusion of the provider of care as a variable in research has led to incomplete analyses of interaction within the health care delivery setting, a setting that includes the provider as well as the patient. Most researchers also have tended to focus on a few specific characteristics of the individual and have examined their relation- ship to the use of medical services. For example, variations in patterns of utilization have been explained in terms of social class. However, the particular elements of class (e.g., income, life style, values or beliefs) that relate to utilization behavior have not been isolated. Furthermore, the contribution that the life situation associated with a social position makes to the utilization of medical services rarely has been explored systematically in past research. This study attempts to examine these neglected issues in research on utilization behavior. It takes into account both the patient and the physician and their contribution to the volume of use of medical services. Furthermore, it places these individuals within the different contexts that condition their orientations and behavior. The study is about female patients who use high levels of medical services for multiple, diffuse symptoms and their physicians' response to them. It is an exploration of the interaction between physicians and patients and the way in which the nature of this interaction influences utilization behavior. It is argued that the doctor-patient relationship cannot be examined as an isolated dyad, but that background factors of the patient and the context of the medical profession influence this relationship. Thus, variability in the life situations of individuals and in the orientations of physicians generate important variations in utilization behavior. Literature Review The following review is not a compendium of the available literature on utilization. Several such summaries already are available (Aday and Eichhorn, 1972; Anderson, 1963; Graham, 1957; Kasl and Cobb, 1966; McKinlay, 1972; Stoeckle, et al., 1963). Instead, it is a discussion of selected works and it is divided into two sections. The first section reviews works that focus on utiliza- tion as a dependent variable and that identify specific factors associated with the use of medical services. This section is organized within a framework of five different models that have been employed in the study of utilization behavior: (1) the sociodemo- graphic model; (2) the social-psychological model; (3) the social structural model; (4) the delivery system model; and (5) the social system model.1 The second section reviews works that focus on utilization as an independent variable and that discuss the outcomes of physician encounters with patients who frequently seek care for multiple, vague and diffuse symptoms. This section is organized within a framework of the two most frequent results of such physician- patient interaction: (1) a psychiatric diagnosis; and/or (2) a stigmatizing label. Utilization: A Dependent Variable The Sociodemographic Model Utilization of medical services is related to a variety of sociodemographic variables. Age, for example, is an important determinant of utilization, primarily because of its association with patterns of morbidity (Aday and Eichhorn, 1972; Anderson, 1963; Sussar and Watson, 1962). The old tend to use more services because of a higher prevalence of chronic disease and the very young more services because of a higher prevalence of acute conditions (Anderson, 1968; Dice and White, 1969; White, 1968). Similarly, data indicate that sex influences the differential use of services. Women tend to use more services than men, primarily because of their need for obstetrical care (Anderson and Anderson, 1967; Feldstein, 1966). However, even when utilization of services associated with pregnancy and childbirth is excluded, the volume of use for women is greater than it is for men (Nathanson, 1975). Marital status also is related to the use of physicians' services since (1) married and widowed persons tend to be older; and (2) married women use more obstetrical services (Anderson and Benham, 1970; Avnet, 1967). The separated and divorced, particularly those raising children, however, show more physical morbidity and use more medical services than married women living with spouses (Abel- Smith and Titmuss, 1965; Berkman, 1969). Available data furthermore suggest that socioeconomic status is associated with the use of physicians' services (Antonovsky, 1967; Cartwright, 1967; Koos, 1954; National Center for Health Statistics, 1975a). The poor are sicker than other segments of the population but make less use of health services than do the non-poor (Richardson, 1969; Ross, 1962). Kadushin (1964, 1967), however, argues that the inverse relationship between socioeconomic status and illness (and, by extension, use of medical services) is spurious. He suggests that lower class people tend to express anxiety in physical terms and, therefore, over-report morbidity in health surveys. Antonovsky (1967) and McBroom (1970), in critical assessments of Kadushin, find no evidence to support this position. Several researchers have shown, nevertheless, that both the lower and higher social classes tend to see physicians at the same rate when illness occurs (Anderson, 1968; Graham, 1957; Richardson, 1970). Moreover, new methods of financing medical care for the poor are believed to have caused the gap in use between classes to narrow (Bice and Eichhorn, 1972; Kriesberg, 1963; Muller, 1965; National Center for Health Statistics, 1975a). Finally, race and education have been shown to be related to the use of health services. Whites and the better-educated utilize more physician services than non-whites and the less-educated, primarily because of the former's preventive care-seeking behavior (Anderson, 1968; Anderson and Anderson, 1967; Feldstein, 1966; Richardson, 1970; Ross, 1972). All of these findings appear to account for differences in utilization behavior of some groups vis-a-vis others. The socio- demographic model, however, has a major shortcoming; sociodemographic findings do not explain why differences in use occur gmggg individuals within a categorical group. For example, Berkman (1969) has suggested that spouseless mothers frequently use physician services because of the stresses involved in performing dual roles. This explanation, however, fails to take into account the large number of spouseless mothers who do not exhibit such behavior. Sociodemographic findings, then, do not tell us about the human situations that contribute to the individual's decision to visit the doctor. Thus, other models must be employed to explain why variations within groups occur. The Social-Psychological Model Two psychological concepts seem to provide promising approaches to the study of high utilization: perception and motiva- tion. Researchers who adopt a perception approach focus on the individual's view of specified phenomena as a basis for understanding behavior. Those who adopt a motivation approach focus on the indi- vidual's needs, assuming that people behave in such a way as to satisfy their needs. The Perception Approach. Twaddle (1969:106) has stated that illness is in large part "a matter of social definition which vary- ingly reflects cultural and individual differences in orientation toward the biological organism." Put simply, illness is a subjective reaction to a physical state (Kirscht, 1971); individuals exhibit a "selective attention" to symptoms and, accordingly, behave dif- ferently. One person may be extremely concerned about a set of symptoms and seek professional help for them. Another may completely disregard the same symptoms and fail to do anything at all about them. Mechanic (1972a) has suggested that such differences in perception and behavior are a function of culturally and socially learned responses. Several researchers have examined the health and medical beliefs of Mexican and Spanish Americans and found them different from the Anglo-Saxon model (Clark, 1959; Saunders, 1954). Similarly, Zborowski (1958) in a study of hospitalized Jewish, Italian, and Anglo-Saxon patients, and Zola (1966) in a study of Italian, Irish, and Anglo-Saxon ambulatory care patients, demon- strated the differential reaction to and concern with similar symptoms by different groups. In addition, Koos (1954) found significant differences in the way in which the lower and upper classes perceived physical symptoms. All of these findings lead to the inference that the presence of the same symptoms in these popu- lations would result in vastly different rates of use of physician services: for example, immediate concern and consultation versus dismissal and abstention from seeing a doctor. Social scientists also have long recognized that the ascrip- tion of sex status provides a complex of culturally defined behaviors that are relevant to action (Bierstedt, 1963; Davis, 1962). Phillips and Segal (1969:58), for example, argue that "the ethic of health is masculine" and that women report more illness than men because it is more culturally acceptable for them to be expressive about their difficulties. Men are socialized to ignore symptoms and are expected to exert more self-control, to be more stoical (Zelditch, 1960). Illness is considered a feminine char- acteristic to be shunned by men (Barker, 1953; Chesler, 1972; Mechanic, 1965). Gurin, et a1. (1960:209) also maintain that the male role is an active one and that "a man's masculine identity is closely linked to his success in coping with his environment." As a result, they posit that men will resist giving in to symptoms and, even if they perceive symptoms, will be less likely than women to report them. Sex differences in perception of symptoms, then, may strongly influence an individual's readiness to seek medical care. Another factor that has been studied as a determinant of responses to symptoms is stress, that is, interpersonal crises that signify or require change in an individual's ongoing life pattern (Rabkin, et al., 1976). McKinlay and Dutton (1974:255) argue that "stress heightens symptom sensitivity, producing an increased awareness of illness symptoms that might otherwise have been ignored." Mechanic (1964) found that mothers under temporary psychological stress tended to report more symptoms for both a; 1&1 SE! in; themselves and their children. Similarly, Berkman (1969) found that the stresses of spouseless motherhood--i.e., inadequate income, unfulfilled financial expectations, and poverty-area residence-- were significantly related to reports of physical illness. In addition, Coe, et a1. (1969) and Hochstim (1968), in studies con— trasting the stress experienced by those living in poverty areas to that experienced by their counterparts living in better neighbor- hoods, concluded that the conditions of the poverty neighborhood were a source of stress and related to perceptions of poor health. Since it has been shown that defining oneself as "ill" usually preceeds seeking care (Antonovsky, 1972; Suchman, 1965), all of the factors discussed above have a relationship to the decision to seek medical care. Stress not only plays a role in increasing symptom sensitivity, it also increases the likelihood that an individual will seek professional medical care (McKinlay and Dutton, 1974; Mechanic, 1963; Stoeckle, et al., 1963). Indeed, a number of studies have demonstrated that psychological distress is associated with higher levels of utilization of medical services (Cooper, 1964; Cummings and Follette, 1968; Shepherd, et al., 1966; Tessler, et al., 1976). Mechanic (1974), for example, reporting on a study of 151 women living in London, indicated that those who defined their health as fair or poor, reported more stress and used more medical services than those not experiencing stress. In addition, Mechanic and Volkart (1961) found that psychological distress played a significant role in the decision to seek medical care and was 10 associated with higher levels of utilization of medical services. They suggest that persons who are distressed or who otherwise face life problems with which they have difficulty in coping, deal with such situations in part by seeking medical care. Perception, then, is interdependent with motivation. A person perceives selectively in accordance with his motives (Secord and Backman, 1964). The Motivation Approach. Researchers who employ the motiva- tion approach emphasize social-psychological needs as triggers for physician utilization. Roghmann and Haggerty (1975) suggest that the use of services is a coping device by persons who have a need to come to terms with the stress of their situation or the discom- fort they are feeling. Similarly, Mechanic (l972a:1l9) has argued that utilization may be seen as "an attempt to make an unstable, challenging situation more manageable for the person who is encountering difficulty." The presentation of a complaint may be "one way of seeking assurance and support through a recognized and socially acceptable relationship when it is difficult for the patient to present the underlying problem in an undisguised form" (l972a:124). Stoeckle, et a1. (1964) report such a situation among patients seeking care at a medical center. In a study in which they assessed psychological distress in medical patients via sociological and clinical judgments they found that (1) distress seemed to influence the decision to see a doctor because the individual's usual capacity to cope with changing environmental stress was 11 temporarily impaired; and (2) patients under stress seemed to have expectations of how the doctor could help them during this period, that is, how he could satisfy their needs. One such need they describe is the need to communicate aggrieved feelings to a sympa- thetic, understanding, and supportive listener. The need for social contact that permits or encourages such free communication has been labeled "the need for catharsis" by Shuval, et a1. (1970). In a study exploring utilization behavior among immigrants in a prepayment medical insurance system that serves the majority of the population of Israel, Shuval and her colleagues found that the need for a person with whom to communicate correlated positively with high utilization rates, regardless of whether the physician satisfied this need or not. Furthermore, they found that "the need to cope with failure," which they defined as the utiliza- tion of illness as a legitimizing mechanism, also correlated positively with high frequency of use of medical services. Indi- viduals faced with increasing helplessness in coping with personal problems were more likely than others to seek solutions within the framework of the medical institution (also Stoeckle, et al., 1964). These findings support Parsons' formulation (1951) that illness represents one way of coping with failure in Western society. Once a person's illness is legitimized by medical sanction (or significant others), the individual occupies a special role in society, i.e., the sick role. During that time he or she is relieved of the usual demands and obligations that accompany his or her occupational and/or familial roles and the sick role takes 12 priority. Such a favored role, then, introduces important variations into the realm of utilization behavior since it may motivate the individual, consciously or unconsciously, to seek and adopt the sick role as an escape mechanism. Behavior, then, emerges out of motivation and, by implica- tion, is influenced by the course of action available to the indi- vidual. Health-related behavior may be determined by motives that are not health related (Rosenstock, 1960). Minor complaints may be used as an excuse for seeking medical care, when in fact the reasons for seeking care are (l) a need to air personal problems; and/or (2) a need to cope with the stress of an untenable situation. Important as these findings are, few researchers employing the social-psychological model have comprehensively examined the bases of these non-medical needs. A promising start has been made, how- ever, by those using the social structural model. The Social Structural Model Two concepts frequently considered within the context of the social structural model seem applicable to the origin of non-medical needs: isolation and feelings of failure. Researchers who deal with isolation tend to focus on social supports as a basis for under- standing behavior. Those who deal with feelings of failure tend to focus on roles to account for differences in behavior. Isolation. Several studies have suggested that isolation is an objective state that arises from an absence of an adequate social support system (French, at al., 1974; Langner and Michael, 13 1960; Segal, et al., 1965). Such systems consist of enduring inter- personal ties to a group of people who (1) can be relied upon to provide emotional sustenance, assistance, and resources in times of need; (2) provide feedback; and (3) share standards and values (Caplan, 1974). They serve to ameliorate or buffer the effects of stress for the individual, mediating between stresses and his or her response to them (Kaplan, et al., 1977: Rabkin and Struening, l976). Presumably, then, the effects of exposure to stressful events will be exacerbated when the supports accessible to the individual within his or her social environment are insufficient or inadequate. Furthermore, the individual not effectively embedded in social networks or support systems might rely upon the physician as a substitute to satisfy basic social needs not gratified through social interaction with significant others. The findings of several studies dealing with differential use of medical services support this reasoning. Hoppe and Heller (1975), working with a sample of Mexican-American women living in a barrio of Texas, found that familism (as measured by frequency of visitation with relatives) and occupational stability were negatively correlated to consulting a physician. The stable poor who were more familistic turned to relatives for advice in times of trouble. The unstable poor, in contrast, "alienated, anxious, and void of family support," responded to "day-to-day living crises" by seeking medical help (1975:312). Miller, et a1. (1967) studying a sample of patients of a Scottish general practitioner, also found that people with few friends were more likely to develop symptoms and to seek 1'22 10. w QEr 14 medical care in the presence of threatening life events than those with many friends. And, in a somewhat similar vein, Fanning (1967) working with British housewives, attributed the higher consultation rates among those living in flats to their relative isolation and loneliness, and the lower consultation rates of those living in houses to the greater level of social integration among neighbors that this arrangement permitted. These findings, then, suggest that kin, friends, and neigh- bors are particularly useful to the individual as a field for expressing and coping with feelings and unmanageable difficulties. People who are socially well~integrated and have favorable relation- ships with others that are marked by emotional and material support, will be less likely to seek outside help for their problems. Those without such viable relationships, will be more likely to seek to cope with the problems that confront them within the framework of the medical system and will be characterized by high frequency of utilization. Feelings of Failure. Feelings of failure represent a sub- jective state that can arise from different situations: role strains and conflicts; inadequate role performance; or powerlessness to control events. Numerous studies have employed these variables to account for differentials in health behavior both between men and women and among women in the United States. Gove and Tudor (1973) provide a good illustration of the general orientation of those who adopt the role strain and conflict Go th in en 15 approach. They argue that the higher prevalence of emotional prob- lems among women in the United States (as measured by first admission rates to psychiatric hospitals and out-patient clinics) is due to five factors: (1) housewives have no alternative sources of gratifi- cation outside the family; (2) housework is unskilled and of low prestige; (3) the housewife role is relatively unstructured and invisible, giving women leisure to brood over their troubles; (4) working wives have less satisfactory jobs and working places them under more strain than men; and (5) role expectations confronting women are generally unclear and diffuse. There seems to be some indirect support for Gove and Tudor's thesis that women with more "leisure" tend to show higher levels of use than women with a large number of role obligations. Cartwright and Jefferys (1958), found that among workers and non-workers alike those with fewer children had a higher volume of use than those with more children. Similarly, Geertsen and Gray (1970) showed that women who had heavy responsibilities at work or with their children were less inclined to adopt the sick role than others. Petroni (1969), in addition, demonstrated that family size was related to use inversely. The fact that women have more absences from work than men (Blumberg and Coffin, 1956; Enterline, 1961) also seems to support Gove and Tudor's argument that working places women under more strain than men. Yet, data indicate that working may serve as a source of gratification for women. Feld (1963), for example, found that employed married women were less likely to report major illnesses 16 and "minor neurotic symptoms" than unemployed married women. Similarly, Rivkin (1972) showed that working women reported less morbidity, less anxiety and fewer disability days than women who did not work. Among non-working women, then, illness may be used as a reaction to "exclusion from the life open to a man" (Parsons and Fox, 1958). Between men and women, however, working may con- tribute to differential rates of use by placing greater strains on women who perform both integrative and functional roles as well. This notion that strains induced by the performance of dual roles may contribute to differentials in health behavior is supported by two studies about "spouseless" women. Berkman (1969) found that physical illness was more prevalent among spouseless mothers than among their married counterparts. Similarly, Renne (1971) showed that divorced women were more likely than married women to report physical disability, neurosis, and depression. Role strain and conflict, then, can lead to feelings of frustration and failure, and the adoption of the sick role may be used as a mechanism to cope with such feelings. Feelings of failure also can arise in situations in which an individual perceives his or her role performance as inadequate. Beresford, et al. (1977) found that among women in charge of "run- ning the household" (i.e., the widowed, separated, and divorced), those who found it difficult to manage the task visited the doctor significantly more often than those who did not. Similarly, Cole and Lejeune (1972) found that black women who felt inadequate in their performance of roles as wives and mothers were more likely to SE if 17 define their health as poor than were women who considered them- selves relatively successful in these roles. Furthermore, feelings of failure can arise in situations in which the individual feels powerless to control events. Phillips' (1965) research on the concept of self-reliance (defined as the ability to cope with life's problems) is an example of this orienta- tion. Working with a sample of married white females in a suburb, he found that those who placed the highest value on self-reliance were least likely to indicate a willingness to adopt the sick role. In a somewhat similar vein, Gurin, et al. (1960) and K005 (1954) found that a frequently cited reason for not seeking help for illness was the respondent's attempt to cope with it himself. Finally, Berle (1958), writing about Puerto Rican immigrants in New York noted (1958:206): . . illness may be an aspect of lack of success and may therefore become a justification for failure. Failure is almost inevitable when there is a discrepancy between an individual's aspirations and the limited opportunities open to him. All of these studies, then, suggest that the sick role can provide a refuge when an individual feels (1) the well role is unsatisfactory; or (2) inadequate in performing that role; or (3) unable to cope with problems of life. To be sick implies an inability to fulfill tasks and an acceptable avoidance of responsi- bility. Sensing failure, the individual may resort to becoming ill to escape from the pressures of life. Illness provides a rationale for failure and the use of medical services a mechanism to legitimize this failure. 18 0f the three models already discussed, the social structural model appears the most promising avenue via which to understand the bases of high utilization of medical services. This model, however, as the other dealt with earlier, only considers the contribution of the individual to utilization behavior. And, as Coe and Wessen (1965:1027) point out: "Our understanding of the medical behavior of patients is incomplete if the influence of the physician is omitted from consideration." The Delivery System Model McKinlay (1972) provides a seminal review of the literature published in the fifties and sixties on works that employ the delivery system model. Most of these earlier works focused on those characteristics of the medical institution that tend to inhibit utilization. More recently, a body of literature has appeared which, though not necessarily concerned with the use of health services, deals with factors that impinge on physicians and induce high utilization. This literature is scattered and nonaccumulative. Yet, it suggests that the decisions that physicians make strongly influence the direction of utilization. As Shortell (1972:2) has noted: . . once the patient has made initial contact with the medical care system, the physician becomes the principal decision maker and allocator of medical resources. He, in effect, decides how much and what types of additional medical services the patient should receive. Most analysts would agree that these decisions are influenced by a host of factors other than the physical signs of illness alone. (.1.— C], r I’M (Ll 11 53 Ge 19 One important factor among these is the basic assumption that under- lies diagnosis and treatment. That is, "it is better to impute disease than to deny it and risk overlooking or missing it" (Freidson, 1970:225; also Scheff, 1972). Backwin's study (1945) of medical decisions regarding the need for tonsillectomy in children and Garland's studies (1959) of diagnostic error in roentgenology all suggest that the pervasive attitude among physicians is that it is better to suspect illness than not. This implicit norm, then, creates a bias, for the physician feels impelled to do something and tends to overtreat and schedule multiple tests and follow-up appointments. As Freidson (1970:257) notes: The aim of the practitioner is not knowledge but action. Successful action is preferred but action with very little chance for success is to be preferred over no action at all. Medical decisions of the probability type, however, may be determined by factors other than this implicit norm. Haney (1975) notes a variety of factors that affect the physician in his decision making, factors that derive from the social background of the patient, the value system of the physician, and the interaction of the two. In addition, Hershey (1972), in a study of 17 physicians practicing in an urban area, found that most of those interviewed thought that liability considerations had a relatively strong influence on their practice; the majority took into account their assessment of a patient's likelihood to sue when making medical decisions. 20 Medical decision making, and resultant use, also may be affected by the way in which the system of care is organized (Simon, 1966). Freidson (1973:475), for example, found that a ". . . pre- paid service contract, which freed the subscriber from having to pay for each service he wished . . . encouraged many physicians to manage patient demands by increasing referrals and reappointments." His finding is supported by data collected at a prepaid group prac- tice affiliated with John Hopkin's which indicate that 47.2 percent of visits there were initiated by providers of care (Gertman, 1974: 364). Moreover, the amount of time available to practicing physicians to treat patients also may affect decisions. Muller (1972:489) suggests that the ". . . doctor's contribution to over- medication must be seen in the context of need to balance time avail- able for practice against time demanded for medical care." Thus, the longer the doctor's day, the more drugs prescribed (Hemminki, 1974:355).2 In summary, then, the implicit norm underlying medical practice, the perception of patients, and the organization of prac- tice all may influence physicians to order unnecessary tests, superfluous appointments, and unneeded medications. Such actions may bring about use of medical services by encouraging individuals to enter the sick role (Warren and Wolter, 1954). As Kuch, et a1. (1977:647) have noted, when a ". . . physician 'organifies' a psychosomatic complaint . . . and labels it a disease for which treatment is prescribed, then the physician reinforces the patient's feelings and beliefs that his or her body is sick and supports the 21 patient in the sick role." In addition, such actions also may cause use of medical services by exposing patients to the hazards involved with some diagnostic procedures or therapeutic regimens, thereby causing iatrogenic disease (Cartwright, 1974; Illich, l976; Mapes, 1977; Mulroy, 1973; Waldron, 1977; Warren and Walter, 1954). These hazards and their consequences have been examined in depth in a recent book by Illich (1976) in which he discusses the "crises" in our western system of medicine, i.e., the growth of iatrogenesis. Illich argues that the roots of iatrogenesis are found in modern technology which determines the underlying assump- tions of physicians and patients as well as the organization of medicine. Moreover, he posits that the consequences of this tech- nology not only are clinical, but social (1976:39-124) and cultural (1976:127-208) as well. For Illich, medicine is a monopoly that fosters a "patient population" for the "medicalization of health," turning self reliance, mutual care, and personal responsibility into "misdemeanors or felonies."3 Thus, although analysts may not agree on the bases of physicians' decision, all do concur that their actions affect the use of health services. This consequence is not always intended or recognized by physicians. But, in fulfilling their manifest function (i.e., the prevention and cure of disease), doctors do induce utilization. Moreover, as Shuval, et a1. (1970) have demonstrated, physicians also fulfill latent functions and these too bring about the use of services. Chi: 22 These latent functions stem "from the professional role of the physician and from the nature of the sick role" (1970:4) and open up the possibility for the individual to turn to the physician to satisfy "non-medical" needs. On the one hand, the physician's role is ideally suited to the individual's needs for catharsis since it involves communication in an uncritical, supportive atmosphere. On the other hand, it is suited to the individual's need to cope with failure since the legitimization of illness enables the individual to use the sick role as an escape mechanism. In their study, Shuval and her colleagues demonstrated that the latent functions of the physician motivated individuals to use medical facilities and increased utilization rates. When the individual's needs were not satisfied, high utilization occurred as the patient's desire to obtain such satisfaction was intensified. When the individual's needs were satisfied, high utilization occurred as the patient was motivated to “enjoy" the latent function (1970). The notion that the latent functions of the physician encourage individuals to use his services or that non-medical needs activate these functions suggests that consumers and providers con- jointly have an influence on utilization behavior, that they consti- tute a social system. The Social System Model This approach first was proposed by Henderson (1935) in a challenge to the medical profession to study the physician and 23 patient as a social system. His challenge was accepted, not by physicians, but by sociologists and other social scientists who have examined the interaction between the physician and patient. Probably the most widely recognized of these investigations is that of Parsons who developed a theoretical model of the social roles of the two parties and the norms governing them in his book Ihg_ Social System (1951:429-479). According to Parsons, the social system is disrupted (or in a state of disequilibrium) when individuals cannot fulfill normal social roles. Society, therefore, uses illness as a mechanism to isolate such deviants and channel their behavior into another role (the sick role) that has less dangerous consequences for it. Moreover, society designates the physician as its agent to serve the system's need to keep deviance within bounds and to maintain viability. The physician's role is conceptualized by Parsons as a social role that is universalistic, functionally specific, and affectively neutral. Universalism implies that the physician views patients as equal members of a universe of health and that categori- cal criteria (e.g., sex, race, or religion) are irrelevant to the process of care. Specificity of function implies that the physician limits his attentions and activities to areas that are strictly medical and within the purview of his technical competence. Affective neutrality implies that the physician adopts an impersonal attitude toward patients and avoids emotional involvement with them or their plight. 24 The sick role is defined by Parsons by reference to physician expectations of patients: that is, the ill person has the obliga- tion to seek technically competent help and to try to get well as quickly as possible.4 Moreover, although the patient initiates the relationship, the values and norms of the physician take precedence in and pattern the interaction. As Wilson (1963:279-280), in a discussion of Parson's formulation, notes: . . the practitioner has a nearly exclusive monopoly of psychological and social leverage . . [H]e is active doing something for and to the patient, while the patient is passive, asking that something be done for him. Parson's active-passive construct is similar to a model of the physician-patient interaction formulated by Szasz and Hollender (1975) to describe the relationship between a physician and a patient with an acute infection.5 In this model, which they call "guidance- cooperation," both participants are active in the sense that they contribute to the relationship and what ensues from it. However: The main difference between the two participants pertains to power and its actual or potential use. The more power- ful of the two (. . . physician) will speak of guidance or leadership and will expect cooperation of the other member of the pair (. . . patient) (1975:434). Empirical data are available to support interactional con- structs based on a theory of power (Ford, et al., 1967, Ort, et al., 1964). Ort and his colleagues (1964:31), for example, found that medical students, their faculty, and practicing physicians described the physician-patient relationship as ". . . interaction that takes place within a specifically medical setting and is characterized by the doctor's active involvement and by cooperation he 9r Nh Ca 25 on the part of the patient." On the basis of their data, they derived a model of the physician-patient relationship in which the doctor is the locus of control. The physician and patient, then, theoretically and empirically, are seen as constituting a dyadic system in which interaction is patterned by the congruence of the norms for the patient with the expectations of the professional. Bloom (1965) argues, however, that much more of reality could be traced intelligibly if the physician-patient relationship was defined, not solely in terms of the physician's orientation, but in terms of the patient's orientation as well. Because the physician's orientation results from participation in a medical culture and the patient's orientation from participation in a socio-environmental culture, ". . . the patient's culture leads him to expect what the doctor's culture does not suggest" (Freidson, 1962:215). As a result, the two are always in potential conflict with each other (1962:209; also Bloom, 1965; Bloor and Horobin, 1975; Sussar and Watson, 1962).6 The patient, holding distinct beliefs about the nature of his problems, feels obliged to try to judge and control what is happening to him. The physician, unaware of the patient's back- ground, feels convinced that he is "right" and most able to decide what is "best" for the patient. At best, the clash results in the castigation of the patient for his illogicality and the assignment of pejorative labels (Duff and Hollingshead, 1968; Freidson, 1975; Lorber, 1975). At worst, it results in missed or ignored cues and J- Ext, behe the 26 inappropriate or misdirected responses (Zborowski, 1975; Zola, 1975). In either case, utilization may be increased (1) as the patient embarks on a career of "doctor-shopping" to find the help he or she needs; or (2) as the patient returns for care in response to iatrogenic disease. Utilization: An Independent Variable The "high use" patient has long been recognized as a unique medical phenomenon. Indeed, reports on this type patient appear annually in the medical literature. Most of these reports are the products of studies about patients who frequently present for care with vague and multiple complaints (Fabrega, et al., 1969; Kuch, et al., 1977; Matarazzo, et al., 1961; Stewart, et al., 1975) that are located in diffuse body systems (Farmer and Harvey, 1975; Lipsitt, 1968, 1970; von Mering, et al., 1976). Such patients usually undergo extensive and expensive diagnostic procedures with few objective findings (Kaywin, 1973; Kuch, et al., 1977; Malcom, et al., 1977). Moreover, although they receive more therapies, these patients usually are found to be refractory to treatment (Kaywin, 1973; Lipsitt, 1968, 1970; Schuller, 1977; von Mering and Early, 1969). As a result, their problems are considered difficult to define and diagnoses arrived at usually are not reproduceable (Farmer and Harvey, 1975; Kuch, et al., 1977; Lipsitt, 1978; Schuller, 1977). Unfortunately, the reports on "high use" patients rarely extend beyond descriptions of their clinical and, less frequently, behavioral traits. Little, if any, information is provided about the management strategies adopted to care for them or the outcome 01‘ die DIG Cla 0111 27 of their repeated visits to the physician. It is not difficult to infer what this outcome is, however. The titles given to the reports leave little doubt that it is a psychiatric diagnosis and/or a stigmatizing label. The Psychiatric Diagnosis A patient is at high risk to be assigned a psychiatric diagnosis if he or she frequently seeks care (Banks, et al., 1975; Bellin, et al., 1976; Locke and Gardner, 1969) for a large number (Matarrazzo, et al., 1961; Stewart, et al., 1975) of vague and diffuse symptoms (Farmer and Harvey, 1975; McCranie, et al., 1978; Stimson, 1976) that are presented in a helpless and complaining manner (Fabrega, et al., 1969; McCranie, et al., 1978; Mechanic, l976; Zola, 1963). Most frequently, because the volume of use of these patients is considered disproportionate to their actual physical findings, their complaints are assumed to be of emotional origin (DeVaul, et al., 1977; Farmer and Harvey, 1975; Hesbacher, et al., 1975; Kellner, 1966). The assignment of a psychiatric diagnosis " . . . often serves as a disguise for the fact that no proper diagnosis can be affixed" (Balint, l970:7). The psychiatric classification is used as a residual category by the somatically- oriented physician after he has ruled out organic bases for symptoms (Bellin, et al., 1976; Kessel, 1965; Mechanic, 1972b). Estimates of this type of patient range from a low of ten percent (Locke and Gardner, 1969) to a high of eighty percent (Kaufmann and Bernstein, 1957; Stoeckle, et al., 1964). Most 1‘6: 1'1“ 1'11 a 28 estimates, however, are in the neighborhood of thirty percent (von Mering and Early, 1966). The proportion of such patients is reported to be larger among those of the lower classes (Dohrenwend and Dohrenwend, 1967; Petras and Curtis, 1968). Furthermore, females have been found to predominate among such patients (Bellin, et al., 1976; Hesbacher, et al., 1975; Kellner, 1976; Locke and Gardner, 1969; Matarrazzo, et al., 1961). The preponderence of women diagnosed as psychiatrically ill may be due to the kinds of symptoms they report, the ways in which they express distress, or to sex-role stereotypes physicians may hold about women and their characteristics (Mechanic, 1976). A number of researchers (Campbell, 1973; Corea, 1978; Ehrenreich and English, 1973) have asserted that male physicians have a general tendency to stereotype and devalue women relative to men. Frequently, they diagnose women's presenting physical symptoms as psychogenic rather than organic in origin, thereby implying that the symptoms are not reflective of "bona-fide" or serious illness.7 Furthermore, the preponderence of high users of services among all patients diagnosed as having psychological illness may be due to the physician labeling of patients itself. Shepherd, et a1. (1966:20) raise this possibility: If two of the factors influencing doctors in labeling patients as 'neurotic' are the frequency of attendance and the multiplicity of complaints, it would hardly be surprising to find that psychiatric patients were fre- quent attenders with many bodily complaints. Whatever the reason for the assignment of a psychiatric diagnosis, if the physician does little else than label and dismiss the patient, ti: 51‘ the of "Pr the dlf t1, Y0” 29 then his or her subsequent behavior is likely to remain substantially similar to the original pattern. The Stigmatizing Label Patients at high risk to be assigned a stigmatizing label share many of the same clinical characteristics as patients diagnosed as psychiatrically ill. They frequently seek medical care for multiple and diffuse symptoms and are considered to have less or no "bona-fide" illness (Fabrega, et al., 1969; Fox, 1957; von Mering, 1972). Often, in fact, once all plausible medical conditions are ruled out, a psychiatric diagnosis is assigned to these patients as well as a label such as "problem," "crock," "turkey," or "fat file" (Bellin, et al., 1976; Kaufman and Bernstein, 1957; Lipsitt, 1968, 1970; Schuller, 1977). They do not fit the medical model which is inappropriately applied to them (Malcom, et al., 1977) and are patients "from whom the diagnostic sheen has been worn" (Martin, 1957:192). The label of "problem” however, does not denote primarily the presence of a diagnostic problem. Rather, it is a declaration of the difficulties associated with the care of these patients. "Problem" patients ". . . make extensive and intensive demands on the physician's time and role, pose administrative and managerial difficulties to the medicalprofession, and not infrequently frustrate those who attempt to help them" (Fabrega, et al., 1969:334; also von Mering and Early, 1966). 30 In large part this frustration lies in the fact that the conduct and values of "problem" patients are not congruent with normative medical expectations (Ima, et al., 1970; Lipsitt, 1970; von Mering and Early, 1969). They are patients with "deviant" attitudes in the areas of trust, cooperation, complaints, and demands (Glogow, 1973; Kaywin, 1973; Lorber, 1975; Schuller, 1977) who do not respond to the efforts of the professional in the way that he would like. Labeling a patient, however, does little to ameliorate the physician's frustration. Moreover, it does little to mitigate the patient's "demands" on the physician. As researchers in the sociology of deviance have suggested (Becker, 1963; Erickson, 1964), the process of labeling an "offender" and making him con- scious of himself as a deviant may evoke the very behavior that was thought to be "inappropriate." Summar The review of the literature suggested the following points which seem to be especially important with respect to this research. 1. Differences in use of medical services cannot always be explained by differences in clinical morbidity. Individuals exhibit a "selective attention" to symptoms and the person whose capacity to cope with environmental stress is impaired is likely to be more sensitive to symptoms and to seek medical care for minor complaints. 2. Such individuals are likely to seek solutions to their problems within the framework of the medical institution as a result of a need for the kind of social contact that permits or encourages 31 free communication (i.e., a need for catharsis); and/or a need to legitimize feelings of lack of success (i.e., a need to cope with failure). 3. The need for catharsis is likely to arise in situations in which the individual is not effectively embedded in a social net- work or the members of the network are unable to serve as a satisfactory field for expression and coping with feelings and unmanageable difficulties. 4. The need to cope with failure is likely to arise in situations in which the individual suffers role strain or conflict; feels inadequate in the performance of roles; and/or feels powerless to control events. 5. Physicians may strongly influence the direction of utilization since once initial contact is made with the medical care system, the doctor decides how much and what types of additional services the patient should receive. 6. These physician decisions are affected not only by the physical signs of illness but also by the implicit norm underlying medical practice, the perception of patients, the physician's own values, and the interaction of the two. 7. The interaction of physician and patient takes place within a specifically medical setting. However, each interactant brings to it an orientation resulting from a different social sphere which serves to define his or her expectations of the others' behavior. When these expectations are not congruent, high use of medical services is likely to occur. 1’] ti: th Ch l“E 32 The studies and analyses reviewed above and summarized here, demonstrate that many factors have been identified as contributing to frequent utilization. In addition, this review suggests that a conceptual framework which focuses on (1) the sources of the expectations that influence the medical behavior of patients and physicians; and (2) the conjoint influence of physicians and patients on use of medical services is likely to provide insights of value in understanding this phenomenon. Plan of Report This report is divided into two main parts. Part I presents the research problems and methods of investigation; Part II, the findings and conclusions. In Part 1, Chapter 2 outlines the con- ceptual framework adopted in this study and defines the main con- cepts used. Chapter 3 discusses the data sources and research design. Chapter 4 describes the setting in which the study was conducted. In Part II, Chapter 5 presents findings relative to the individual and to the patient and discusses the way in which indi- vidual characteristics are related to patient attributes. Chapter 6 discusses the physician and his orientation and presents data on the interaction of physician and patient. Emphasis is placed in this chapter on the management process and the outcome of it. In Chapter 7, the findings are summarized and implications for further research are discussed. TL) ‘3] 4.0 d_ 361 19E 96r Laz (lav 197. an! CHAPTER I: NOTES 1The models were selected because they appear to represent promising approaches to the study of high utilization. Not all the models are mutually exclusive and the assignment of a work to one model rather than to another may appear arbitrary. In the writer's opinion, assignments best reflect the model under which works are included. (See McKinlay, 1972 for a discussion of additional models that have been employed in the study of utilization behavior.) 2The use of drugs, particularly prescription drugs, increased substantially in the United States during recent decades. According to Silverman and Lee (1974:18): "In 1950, an average of 2.4 prescriptions per person were filled in community pharmacies, in 1960 an average of 3.5, and in 1972 an average of 5.5" 3See Navarro (1976) for a critique, developed within a Marxist framework, of Illich's argument. 4To the best of my knowledge, Parson's formulation of the patient (or sick) role has been systematically investigated only twice: by Gordon (1966) and by Twaddle (1968). 5Szasz and Hollender also posit an active-passive model, but consider it applicable only to the relationship between a physician and a patient in coma or under anesthesia. In addition, they posit a mutual participation model that they consider applicable to the relationship between a physician and a patient with a chronic disease. In this latter relationship, the physician helps the patient to help himself. According to Szasz and Hollender, however, this model is "essentially foreign to medicine" (1975:435). 6This conflict has been documented in studies of patients' satisfaction with and acceptance of medical personnel (Freidson, 1961; Koos, 1954. 1955; Reader, et al., 1957) as well as in more general studies of physician-patient interaction (Clark, 1959; Lazare, et al., 1975; Paul and Miller, 1955; Saunders, 1954). 7Two sources of such sex-role stereotypes among physicians have been suggested: (1) the socialization which they receive during their medical school training (Howell, 1974; Safilious-Rothschild, 1974); and (2) the images of women communicated in medical drug advertisements (Prather and Fidell, 1975; Seidenberg, 1971). 33 CHAPTER II THE RESEARCH PROBLEM As suggested in Chapter I, this study is concerned with (l) the contribution of the patient and the physician to the fre- quent use of medical services; and (2) the circumstances that affect their orientations and behavior. These are research problems which have been studied separately, but, they seldom have been examined together. The central questions posed by this study, therefore, can be defined as: Is the high use of medical services related to non- medical needs arising from the individual's environment? Is the response of the physician to these needs associated with attitudes arising from membership in a medical profession? Is frequent utilization of medical services related to patients' needs as well as to physicians' patterns of response to them? To further specify the research problem, this chapter under- takes three related tasks. First, the theoretical framework within which the problem is analyzed is described and the rationale for its adoption set forth. Then, the basic concepts used in the study and the range of questions to be explored are discussed. Finally, a model of the use of medical services, including the hypotheses to be tested, are presented. 34 va Pa SL1 35 Theoretical Framework In order to examine the influence of physicians and patients on the use of medical services, a basic theoretical framework is needed. Specifically, a framework is needed that will relate seem- ingly disparate realms of social action to each other: social relationships; the medical profession; role behavior; and cultural values concerning appr0priate health behavior in the United States. Parsons' theory of the social system would seem useful in providing such a framework. Parsons' Theory of the Social System Parsons offers a definition of the social system at several points in his work: . . a social system consists in a plurality of individual actors interacting with each other in a situation which has at least a physical or environmental aspect, actors who are motivated in terms of a tendency to the 'optimization of gratification' and whose relations to their situations, including each other, is mediated in terms of a system of culturally structured and shared symbols (Parsons, 1951: 5-6 . A social system is a system of the actions of individuals, the principal units of which are roles and constellations of roles. It is a system of differentiated . . . roles (Parsons and Shils, 1959:197). A social system . . . is a system of interaction of a plurality of actors in which action is oriented by rules which are complexes of complementary expectations con- cerning roles and sanctions (1959:195). These quotes suggest key elements in Parsons' theoretical framework-- e.g., action, orientation, motivation, rules, roles--elements on which the following discussion focuses. 36 The System of Social Action Action is the starting point for the social system and is defined as behavior that is oriented to the attainment of ends in situations by means of the normatively regulated expenditures of energy (Parsons and Shils, 1959:53). It involves a choice among alternatives and the use of criteria derived from past experience to make that choice. The choice is oriented to the attainment of goals and is made from among alternatives that may be particular to a given situation or specific to all situations. In choosing and pursuing alternatives, the actor employs two distinctive systems of orientation: a motivational orientation and a value orientation. There are three "modes" or forms of motivational orientation: the cognitive, the cathetic, and the evaluative. The "cognitive mode" consists of behaviors by which an actor perceives an "object" and defines its characteristics and consequences for the gratification or deprivation of his needs. The "cathetic mode" consists of behaviors by which an actor invests an "object" with affective significance in accordance with its conse- quences for his needs. The "evaluative mode" consists of those behaviors by which an actor " . . . allocates his energy among the various actions with respect to various cathected objects in an attempt to optimize gratification" (Parsons and Shils, 1951:59). The motivational orientation, then, essentially provides a framework for analyzing "problems of interest” to the actor (Parsons, 1951:14). In the course of this analysis, the actor employs what are variously called norms, standards, and criteria that define towa the ‘ l“ c 0 CI 37 appropriate methods of choosing among alternatives, i.e., the value orientation. There are three modes of value orientation, each corresponding to a mode of motivational orientation: The cognitive mode of value-orientation involves the various commitments to standards by which the validity of cognitive judgments is established . . . . The appreciative mode . . . involves the various commitments to standards By which the appropriateness or consistency of the cathexis of an object or class of objects is assessed . . . . The moral mode of value-orientation involves the various commitments to standards by which certain consequences of particular actions and types of action may be assessed with respect to their effects upon systems of action (Parsons and Shils, 1959:60, emphases in original). The value orientations, then, provide the standards of what consti- tutes "satisfactory solutions" to the actor's "problems" (Parsons, 1951:14). When a choice is made in a situation involving other actors, the actor-subject also employs schemes for classifying social objects on the basis of their significance to him. He must decide how to act toward them in terms of what they are (their quality) or in terms of what they do (their performance). And, he must decide how to act toward them in terms of their "scope of significance" to him. If the objects have: . . . a broad and undefined significance for the actor- subject . . . [then] he feels obliged to grant them any demand they make of him, so long as the granting of the demand does not force him to fail in other obligations higher on a priority scale of values. In this case we may say the object has for the actor-subject a broad scope of significance. Its significance is diffuse . . . . [If the objects have a] narrow and clearly defined significance for the actor-subject . . . [then he] does not feel obliged to grant them anything that is not clearly called for in the definition of the relationship which obtains between them. In this case we say the scope of signifi- cance of the object . . . is s ecific (Parsons and Shils, 1959:57-58, emphases in origina . UP ar th th of is 5C1 to' EXP: 38 Parsons contends that an actor must select from these types of motivational and value orientations and evaluative schemes when he makes a choice in a situation: Specifically . . . the actor must make five specific dichotomous choices before any situation will have a determinate meaning. The five dichotomies which formulate these choice alternatives are called the pattern variables because any specific orientation (and consequently any action) is characterized by a pattern of the five choices (Parsons and Shils, 1959:76, emphasis in original). These five pairs of choices are: affectivity versus affective neutrality; self-orientation versus collective-orientation; universalism versus particularism; ascription versus achievement; and specificity versus diffuseness. The "system of action" described above is organized into three sub-systems: the social system; the personality system; and the cultural system. The three systems are not reducible in terms of one another. However, they interpenetrate one another and each is indispensible to the others in the concrete manifestation of action. The Social System The basic unit in a social system is the role, the dynamic aspect of status. Status is a collection of rights and duties and to put these into effect is to perform a role (Parsons, 1951:25). Expectations are the primary components of roles and have two aspects: On the one hand . . . expectations . . . concern and in part set standards for the behavior of the actor, ego, who is taken as the point of reference; these are his 'role- expectations.‘ On the other hand, from his point of view there is a set of expectations relative to the contingently 39 possible reactions of others (alters)--these will be called 'sanctions,’ which in turn may be divided into positive and negative according to whether they are felt by ego to be gratification-provoking or depriving (Parsons, 1951:38, emphasis in original). A crucial component of expectations, then, is the probable reaction of others to an actor's possible action, a reaction which is antifipated in advance and affects choices. Roles within a social system must be complementary if goals, both private and collective, are to be pursued effectively. "Although conflict can exist within a social system and, in fact always does, there are limits beyond which it cannot go and still permit a social system to exist" (Parsons and Shils, 1959:198). An important characteristic of a social system, then, is that it must possess a minimum degree of integration based on a system of generalized and institutionalized patterns of value-orientation. No social system is ever completely integrated, however. There always are some discrepancies between role expectations and performances of roles, i.e., tendencies to deviance. On the one hand, new members coming into the system have not "learned" their given role orientations and thus tend to act in ways that are unsatisfactory (Parsons, 1951:205). On the other hand, established system members may: . . . depart from conformity with the normative standards which have come to be set up as the common culture. A tendency to deviancein this sense is a process of motivated action, on the part of an actor who has unquestionably had a full op ortunity to learn the requisite orientations (1951:2061. nu... t? to C07 DUI The "lot COU ba: 318' than T 111 T 40 Both forms of "deviance" are a threat to the "equilibrium" of the system. And, since the "first law of social process" is to maintain the interaction process (1951:205), the system contains these prob- lems through two mechanisms: mechanisms of socialization and mechanisms of social control. Socialization refers to the process by which an actor learns the requisite orientations for satisfactory functioning in a role (1951:205). It is a life-long process that requires actors to adjust to changing roles and shifts in role expectations. Social control refers to the process by which the social system motivates actors to abandon tendencies to deviance and resume conformity. The most common mechanism employed to check disruptive tendencies is "defense;" sufficient rewards are accorded for conformity and punishment for deviance to tip the balance in favor of conformity. The mechanisms of socialization and social control function to motivate actors to conformity with a given set of expectations, to counteract deviance and other strains in the system and to bring it back to a stable state, thereby maintaining equilibrium. If the system's mechanisms of control do not work adequately, the system changes its state or disintegrates. The Cultural and Personality Systems and Their Relationship to the Social System The cultural system constitutes the standards for conducting action and gives the action system its primary “sense of direction." It is not a system of interaction as a social system is, but, rather, is a system of values, norms, and symbols that guide the choices of SO: Org ex; of and a m. lnsi Of a 911m one Sltu “in. 41 individual actors and limit the types of interaction. The person- ality system constitutes a constellation of "need-dispositions" that “. . . control a system of concrete orientations and actions aimed at securing for the personality certain relations with objects" (Parsons and Shils, 1959:159). It is not a system of action organized about the relation of actors to each other as a social system is, but, rather is a system organized about the "living organism," the individual. The relationship between personality and social systems lies in the fact that the need-dispositions of the former and the role expectations of the latter are both derived from the same patterns of value orientation. The relationship between the cultural system and these two systems lies in the fact that, on the one hand, the values, norms, and symbols of the former are internalized and become a main part of the personality and, on the other hand, they are institutionalized and become part of the social system. In a system of action, then, the personality system provides the motivational energy and orientation, the cultural system provides the value orientation and the social system provides the structure of the situation. Action, Parsons contends, is understandable only when all three systems are considered. Theoretical Rationale The social system is a concept that was developed in order to analyze macrosocial phenomena. It is a concept that refers to a complex of interdependencies between parts as well as to an 42 interdependency between such a complex and its surrounding environ- ment. This environment is not just physical, however. It includes the cultural and personality sub-systems as well. The social system, generated by the process of interaction, is the primary link between culture and the individual. The concept is adapted in this study in order to analyze microsocial phenomena. Specifically, it is used to analyze the relationship between physician and patient in order to understand (1) the whole complex of interdependencies of which the relationship forms part; and (2) the consequences of the relation- ship. Three elements of Parsons' theory guide this report's dis- cussion: the cultural system; the personality system; and the social system. The cultural system, represented by the traditional belief system in America, is considered since dominant values theoretically guide the choices of individual actors. The person- ality system, represented by the individual located within a social or a professional context, is considered since motivations and orientations are organized and develop at this level. And, the social system, represented by the doctor-patient relationship, is considered since interaction takes place at this level. The cultural tradition of American society is that "medical practice is organized about the application of scientific knowledge to the problems of illness and health, to the control of 'disease'" (Parsons, 1951:432). In order to effect this idea, the physician role is culturally defined as oriented toward universalism, functional specificity, and affective neutrality. That is, the 43 physician: (1) views patients as equal members of a universe of health without regard to categorical criteria that are irrelevant to the process of care; (2) limits his attentions and activities to areas that are strictly medical and within the purview of his technical competence; and (3) adopts an impersonal attitude toward patients and avoids emotional involvement with them or their plight. Similarly, the patient role is culturally defined as functionally specific and affectively neutral in orientation. That is, the patient role is confined to the sphere of health and "is focused on an objective problem not on . . . orientations to an emotionally disturbing problem, although this may be instrumentally and other- wise involved" (1951:438). The interaction of physician and patient, then, is culturally formulated in terms of role complementarity, of parallel expectations. This prototype of the physician-patient interaction, however, may be modified by varying orientations that derive from the different contexts in which the physician and patient are located. The physician's definition of his role may be shaped by the type of medical school he attended and/or the structure of the medical setting within which he provides care. He may believe his functions are limited solely to biophysiological problems. Or, he may believe his authority extends to psychosocial pathology. The patient's definition of her role may be affected by the resources she has available to cope with her general life situation. She may focus on her medical problem only. Or, she may consider personal problems apposite to the situation. The expectations brought by the Dr be or 44 interactants to the medical encounter may complement or may contra- dict each other. They are a product of a host of factors experienced at the individual level. Furthermore, the interaction of the physician and the patient may be expected to be influenced by the extent to which the expecta- tions are or are not congruent. If their expectations do not parallel each other, the physician may demand conformity to the ideal patient role by adopting certain management strategies to con- tain deviance and ”reequilibrate" the system. If their expectations do parallel each other, he may disallow the ideal patient construct by adopting different management strategies that are congruent with the orientations guiding the interaction. In either case, as the literature reviewed in Chapter I indicated, differential use of medical services might be expected to occur. The theoretical scheme proposed by Parsons appears to pro- vide a good framework within which to analyze the outcomes of differ- ent types of physician-patient interaction. First, at the level of the social system, the theory deals with action that is oriented toward gratification of the attainment of goals within a situation. The opposites of gratification and deprivation and, therefore, of the two basic tendencies of action-~seeking and avoidance--are inherent in this conception. The formulation thus seems to hold promise as a guide to the analysis of the outcomes of physician- patient interaction that might or might not be mutually goal- oriented. 45 Second, at the level of the personality system, the theory deals with the organization of goals and means to the accomplishment of goals. Furthermore, although understanding of the influence of experiential factors on the organization of goals is not necessary at the societal level (Parsons, l951:9), the formulation does not preclude an exploration of these factors at the sub-population level. It thus seems to hold promise as a guide to the analysis of the way in which contextual factors, relevant to the interaction, affect variations in orientations and outcomes. Third, at the level of the cultural system, the theory deals with appropriate behavior of physician and patient in a situation "organized about the application of scientific knowledge" to the control of somatic problems (Parson, 1951:432). The formulation thus seems to hold promise as a guide to the analysis of interaction in which deviance from this traditional American belief is likely to occur, the "mechanisms of social control" (1951:432) employed to contain it, and the subsequent effect of these mechanisms on outcomes. Finally, the theory deals with the interaction of the three sub-systems. The formulation thus seems to hold promise as a guide to the analysis of seemingly disparate, yet intertwined phenomena. The interdependency of these sub-systems is reflected in the following definition: the doctor-patient relationship is conceived of as a system of social roles, derived from culture, and influenced by the actors' personalities. 46 Basic Concepts The foci of this study are patients who frequently seek medical care for multiple diffuse symptoms and their physicians' response to them. The review of the literature in Chapter I sug- gested that an individual who perceives her life style as inadequate and whose social and economic environment is unstable may use high levels of medical services. Furthermore, the review suggested that this pattern of utilization may be an adjustment to the person's particular life situation. When alternative sources of gratification are blocked or less available elsewhere, the individual may seek to satisfy non-medical needs through the medical institution. Thus, she may rely on the physician to provide the support that is lacking in her personal environment. This study refers to such an individual as a dependent patient. Specifically, she is defined as an indi- vidual who frequently uses medical services to satisfy needs over and above the need for specific diagnosis or treatment of illness. The study provides an investigation into the factors that contribute to the dependent patient's high use of services, as well as an exploration of the influence of her behavior and that of the physician on differential patterns of utilization. Specifically, the study examines the influence of (l) the life situation of the individual on non-medical needs that may affect the use of medical services; and (2) the professional experience of the physician on orientations that may evoke responses to these needs. The basic concepts employed to explore these issues and the variations in pa ce; the to dis ori set desc and deft 47 health and professional behavior that might be found are presented in this section. Overview of Concepts Three major sets of concepts are employed in this study: (1) a set relevant to the individual; (2) a set relevant to the patient; and (3) a set relevant to the physician. The set of con- cepts relevant to the individual is used to provide a description of the life situation of study subjects. The set of concepts relevant to the patient is used to provide a description of the need- dispositions that arise from this situation and that influence orientations toward the doctor-patient relationship. Finally, the set of concepts relevant to the physician is used to provide a description of the way in which the doctor performs his functions and the way in which his professional experiences influence his definition of the physician-patient relationship. Concepts Relevant to the Individual and to thelPatient It has long been recognized that difficult life circum- stances or stress situations lead to varied responses and that these variations are due to individuals' differing definitions of and capacities to cope with or adapt to these stimuli (Caplan, 1974; Cobb, l976; Dohrenwend and Dohrenwend, 1974; Mechanic, 1972a). Three separate, but interdependent, phenomenon are inherent in this statement: (1) the life circumstances of the individual; (2) the crises she may experience; and (3) the resources she has available 0t Qt be (181* vhi env set is l lndi (NU: lndi her 48 to cope with or adapt to the demands of life difficulties. All three phenomenon are considered in this study. Descriptions of life circumstances may be subjective or objective. The individual's perception of her life, i.e., its quality, reflects a subjective description of conditions that might be assumed to influence her life experiences. The individual's marital status, income or level of education reflect an objective description of characteristics that might be expected to affect her life chances. Both types of descriptions are investigated in this study. In addition, the presence of stressor events that might impose hardships on the individual and affect her sense of well being also are investigated. All three phenomena represent potential demands with which the individual may be called upon to cope or to which she must adapt. Cobb (1976:311) defines caping as "manipulation of the environment in the service of self" and adaptation as "change in the self in an attempt to improve person-environment fit." This study is concerned with the psychosocial assets that contribute to an individual's ability to cope with deleterious life circumstances (Nuckolls, et al., 1972). Specifically, it is concerned with the individual's social support systems (i.e., social assets) and with her feelings concerning herself (i.e., psychological assets). A large body of literature exists on the way in which social supports mediate the individual's response to difficult life situa- tions by "buffering" or "cushioning" her from their consequences (Cobb, 1976; Crogg, 1972; Henry and Stevens, 1977; Kaplan, et al., 49 1977; Rahe and Arthur, 1978). Caplan (l974:6) suggests three ways in which such support systems help the individual to manage long- term burdens and privations or acute stress situations. First, the members of such systems help the individual mobilize her psychological resources and master her emotional burdens. Second, they share tasks with her. And third, they provide her "with extra supplies of money, materials, tools, skills and cognitive guidance to improve . . . [her] handling of . . . [her] situation" (1974:6). The nature and strength of the social supports available to the individual, then, may (1) help her come to terms with the pressures of her particular situation and/or (2) provide tangible assistance in managing demanding or frustrating aspects of her environment. Three categories of people that constitute potential sources of such support are considered in this study: (1) members of the kin network; (2) members of the friend network; and (3) neighbors. Links with all three categories of significant others add up to a potential system of psychological and functional support that can facilitate the individual's ability to cope with life difficulties or stressor events. In addition to these social assets the individual's psycho- logical assets also can contribute to her ability to cope with the demands of environmental pressures. Henry and Stevens (1977) and Nuckolls, et al. (1972) emphasize the importance of the way in which feelings about oneself are related to the definition of a situation and subsequent behavior. "The assurance of personal worth in being able to do socially valued work or to completely maintain a home 50 . . . [or to solve problems independently] ranks as an important asset" that is a "defense" between distressful life changes or deprivations and an individual's ability to cope with them (Henry and Stevens, 1977:221. In this study, the individual's feelings of self-worth are defined by the presence or absence of feelings of failure. Such negative or positive feelings about herself can con- tribute to the way in which the individual manages difficult life circumstances with which she may be confronted. It is argued in the study that in the absence of viable psychosocial assets certain needs arise, i.e., the need for catharsis and the need to cope with failure. The frequent use of medical services for multiple diffuse symptoms is an adjustment to this privation, a mechanism by which the individual attempts to gratify, needs not satisfied within her personal environment. The dependent patient's emphasis on the socioemotional aspects of the doctor- patient relationship is consistent with her deleterious life circumstances as an individual. Concepts Relevant to the Physician Merton (1957a:51) distinguishes between the manifest and latent functions of an institution, defining the former as conse- quences that are intended and recognized by participants in a system and the latter as "consequences which are neither intended nor recognized by participants." Both functions are considered in this study. The manifest function of the physician is defined as the prevention and cure of disease, the latent functions as the 51 satisfaction of the need for catharsis and the satisfaction of the need to cope with failure. It is argued in the study that the physician's performance of these functions does not always conform with the primary cultural tradition, that his personal orientation determines, at least in part, his approach to medicine, his definition of disease, and even his determination of treatment. These orientations may be derived from the background in which he was raised, may emerge during his years in medical school or may be influenced by his experiences in practice. Regardless of their source, they will have an effect on his response to the dependent patient and, consequently, the out- comes of physician-patient interaction. Concepts Relevant to the Women as Individuals It has been argued that non-medical needs arise from the individual's social and economic environment, that needs are a product of experiential factors and perceptions of them. The differ- ential use of medical services, then, may be a function of broad aspects of the patient's life situation. This study focuses on six such aspects: (1) quality of life; (2) social position; (3) social stressors; (4) social networks of kin and friends (i.e., social assets); (5) relations with neighbors (i.e., social assets); and (6) feelings of failure (i.e., psychological assets). 52 Quality of Life Quality of life refers to the individual's perception of the objective circumstances of her life. The term quality inevitably is a comparative measure and, hence, is used to define the indi- vidual's sense of well-being about her situation (i.e., the “good- ness" of it). It is assumed that pe0ple living in different circumstances will express different patterns of well-being and that these will reflect the peculiar quality of the situation in which they live. In the study, the individual's sense of well-being is described in terms of her life as a whole as well as in terms of specific domains of life into which her experiences are segmented. The phrase general state of life is used to refer to the overall life situation of the individual. The term family life is used to refer to the way in which the family functions as a social unit, i.e., the interpersonal relations among and between its members. And, the term family money situation is used to refer to the family's financial position. The family is considered to consist of two or more pe0p1e who share living facilities and a common coffer and who perceive themselves to have a meaningful interpersonal relationship. Although this definition is not necessarily congruent with tradi- tional sociological definitions applied to the prototype of the family,1 it was adopted in order to account for pseudo-families such as those consisting of friends, as well as non-intact families such as those in which the marriage has been disrupted. 53 The family is the unit of "living" and it might be expected that the degree to which it is socially or financially disadvantaged will affect the individual's health as well as her perception of it; the health of the family as a functioning unit and the health of the individual are assumed to be interdependent (Haggerty, 1965; Mechanic, 1962). Part of the research problem, then, focuses on the following questions: Does the individual's perception of the general state of her life relate to the presence or absence of non- medical needs and/or her use of medical services? Does the individual's perception of her family's life and money situation relate to the presence or absence of non-medical needs and/or her use of medical services? Social Position One of the most glaring inequalities in the United States is the differential distribution of income. At the extreme lower end are the poor, those with little command of economic resources and little power to shape their own destiny. Such people usually are black and divorced or separated females heading households. The median income of families with children headed by women 25 to 44 years of age was $4,000 a year in 1969, compared to a median of $11,600 for all families with two parents (Stein, 1970). This dis- crepancy was greater for non-whites than for whites; fifty-eight percent of all non-white female headed families were poor as com- pared to thirty-two percent of their white counterparts (Ferris, 1971). 54 Furthermore, although labor force participation for women with preschool children is increasing for all females, those whose marriages have been disrupted are more likely to be working; in 1971, 20 percent of married mothers, 41 percent of separated mothers, and 62 percent of divorced mothers worked (Waldman and Gover, 1972). The jobs held by these women tended to be low-paying and insecure since they frequently enter the job market with less educational credentials essential for most well-paying occupations and are given less job training (Bluestone, et al., 1971). The distinctions of income, race, marital status, family structure, educational and social role, then, produce crucial differ- ences in the life situations of individuals. Low-status people have less money, power, knowledge and other resources with which to cope with life problems. In addition, working places greater strains on spouseless mothers who must perform both integrative and functional roles. Individuals near the bottom of the social hierarchy, there- fore, might be expected to attempt to manage the stress of their lives by seeking solutions within the medical institution. Part of the research problems, then, focuses on the following questions: Do sociodemographic characteristics relate to the presence or absence of non-medical needs and/or the use of medical services? Social Stressors Social stressors refer to any set of events that involve new demands, changes in life routines, and breaks in established life patterns (Mechanic, 1974; Rabkin and Struening, 1976). Such events 55 may be socially or personally desirable such as marriage or the birth of a planned child, or they may be events that are experienced as threatening or demoralizing such as divorce or an unwanted pregnancy. In this study the focus is on the latter type events since they imply the greatest challenge to the individual's coping resources and skills. No change in circumstance is the same for any given indi- vidual, however. Its impact ranges according to the hardships that may accompany it. A disrupted marriage may force the woman to work and to be both mother and father to her children. But the disruption also may produce a more relaxed family life. Clearly, stressor events are modified by the meanings that people attach to them, as well as by their ability to handle these events effectively. Furthermore, the impact of such events may be presumed to be additive; more events are likely to have a greater effect on the individual (Rabkin and Struening, 1976:1014). An individual might be able to adapt to one event that required or signified change, but be unable to cope with the demands of a series of events that challenged her resources and skills. Finally, not all stressor events precede the occurrence of illness and not all people who are exposed to them become ill. The response to bacterium as well as to symptoms that have a very high prevalence in the population are highly variable. It might be expected, therefore, that the individual who is least able to adjust to stressor events will be most likely to focus on symptoms as a 56 way of managing them. Part of the research problem, then, focuses on the following questions: Does the presence of stressor events relate to the presence or absence of non-medical needs and/or the use of medical services? Social Assets: Social Networks The phrase a "network of social relations" frequently has been used to describe a set of interrelationships or "links" between persons. This use of 'network,‘ however, is purely metaphorical and is very different from the notion of a social network as a specific set of linkages among a defined set of persons, with the additional property that the characteristics of these linkages as a whole may be used to interpret the social behavior of the persons involved (Mitchell, l969:l). The concept of social networks, then, is different from that of social relationships because in network analysis the nature or characteristics of the interconnections between people are used to interpret social actions. The analytical, rather than just metaphorical, use of the concept social networks seems to have been adopted in the 19505 by anthropologists dissatisfied with the ability of structural- functional analysis to explain social change and phenomena in com- plex societies. The idea was introduced by Barnes (1954) in a study of social action in a Norwegian community and developed by Bott (1957) in a study of the allocation of conjugal roles within London families. Unfortunately, Bott's study had the initial effect of associating the notion of social networks almost exclusively with 57 conjugal roles (Aldous and Straus, 1966; Nelson, 1966; Turner, 1967; Udry and Hall, 1965). More recently, the concept has been used by British social anthropologists to study interaction in a variety of situations.2 It has not been so readily adapted by sociologists in the United States, however. Researchers have conducted many studies of kinship (Aldous, 1967; Cohen and Hodges, 1963; Gans, 1962, 1967; Litwak, 1965; Rainwater, 1964; Sussman and Burchinal, 1962), a few of friendship (Adams, 1967; Cohen, 1961; Lazersfeld and Merton, 1954), and a few studies in which relations with neighbors and friends as well as those with kin were examiend (Axelrod, 1956; Bell and Boat, 1957; Litwak and Szelenyi, 1969). But, few explicit studies about social networks have been carried out.3 Furthermore, although nearly everyone who has written on utilization behavior "seems to have paid lip service" to the influence of kinship and friendship networks on the use of medical services, there have been "remarkably few attempts to study . . . [their] nature" (McKinlay, 1972:129).4 In part, the dearth of research employing the concept may be due to the fact that general theoretical insights about social networks gained in other disciplines have not diffused to any great extent into areas of sociological research. In part, it may be due to the fact that the influence of kin and friends on the way in which individuals "define and act (or fail to act) upon symptoms or life crises" is considered "truistic" (1972). The relationship of significant others to the use of medical services cannot be assumed a priori, however. This study, therefore, 58 examines the "nature" of kin and friend networks in order to see if existent variations relate to differences in utilization behavior. The personal network is the focus of the study with the individual as the reference point in the network. Only the kin and friend regions of the network are examined. Neighbors, of course, provide a pool of potential friends and may overlap with the kinship and friendship categories. Nevertheless, since it cannot be taken for granted that relationships with them imply the services and support connoted by the primary relationships of kinship and friend- ship, neighbors are viewed as a unique conceptual unit.5 Several characteristics of networks are viewed as germane to the understanding of utilization behavior: density; content; and frequency of contact. Density is a structural characteristic of networks and contributes to the shape of the network, the pattern of the links in the network in respect to one another (Mitchell, 1969:12). Content and frequency of contact are interactional characteristics and define the nature of the links themselves (1969:12).6 Density. The density or "connectedness" (Bott, 1957) of a network refers to the extent to which the people known by an individual know and meet with one another independently of the individual (l957:59).7 A close-knit (i.e., dense) network is one in which there are many relationships among people; a loose-knit (i.e., sparse) network, one in which there are few such relationships (1957:59). The consequence of the density of the 59 network, so defined, is the extent to which the norms of the network members are or are not reinforced through intercommunication: When many of the people a person knows interact with one another, that is when the person's network is close-knit, the members of his network tend to reach concensus on norms and they exert consistent informal pressure on one another to conform to the norms, to keep in touch with one another, and if need be, to help one another . . . . [Conversely], when most of the people a person knows do not interact with one another, that is when his network is loose-knit, more variation on norms is likely to develop in the network, and social control and mutual assistance will be more fragmented and less consistent (Bott, 1957:60). Mitchell argues, however, that: The degree to which a person's behavior is influenced by his relationships with others often turns on the extent to which he can use these relationships to contact people who are important to him or alternatively, the extent to which people who are important to him can con- tact him through these relationships. This is the general idea of reachability in a segment of a network . . . [and] should be differentiated . . . from density . . . which refers rather to the extent to which everyone is a set of ego's contacts knows everyone else (1969:15). For Mitchell, then, the degree to which norms are reinforced or pressure brought to bear on some individual, is likely to depend on the number of people who must be contacted to reach the person con- cerned. "A dense network may imply that this enforcement is more likely to take place than a sparse one, but this cannot be taken for granted" (1969:18-19). Given Mitchell's reservation about the correctness of the definition of density developed by Bott and used in this study, the following clarification of the meaning attached to the concept is provided. The network functions as a channel for the transmission of information. The degree to which the information communicated 60 includes judgments and opinions that are likely to serve to rein- fbrce norms and bring pressure to bear on some specified person remains an underlying assumption in this study rather than a datum to be examined empirically. Thus, it is assumed that a close-knit network is more likely to function as an effective channel of information than is a loose-knit one. Content. The content of the network refers to its purpose, "the meanings which the persons in the network attribute to their relationship" (Mitchell, 1969:20). This content may be, as suggested above, the flow of information. The primary purpose of networks specified in this study, however, is the provision of social support, psychological as well as functional. The term psychological support is used to refer to any behavior that makes an individual feel better or improves her Spirits. Serving as an emotional outlet so that difficulties can be acknowledged and feelings expressed, listening and helping to identify and objectify a problem, or sustaining and validating another's self-image are a few, among several possibilities, such support may take. The term functional support is used to refer to any tangible assistance that is provided for the individual such as financial assistance, or help with child care, with problems when there is illness in the family, with transportation or with home repairs. Frequency of Contact. Frequency of contact is used to refer to the number of times individuals contact each other or have 61 interaction. Since frequency of contact may be affected by the degree to which significant others are available for interaction, location of network members also is considered in this study. The neighborhood--the area within walking distance of the individual's front door--is the territorial unit likely to provide a pool of non- household persons actually available for interaction.8 Thus, this area is contrasted with that outside it. Furthermore, since access to travel facilities, as well as the use of the telephone, may minimize the importance of geographic distance for frequency of interaction, type of contact also is considered in this study: that is, telephone vis-a-vis personal, face-to-face contact. The density and content of the network and the frequency of contact between kin and friends might be expected to be related to the individual's need for catharsis and/or need to cope with failure. The psychological and functional support of kin and friends may serve as a mechanism to ameliorate the strain of problem situations and/or to mitigate the impact of problems on an individual's sense of well-being. Furthermore, the density of the network and the frequency of contact between network members may affect the degree to which such support is forthcoming. Part of the research problem, then, focuses on the following questions: Does the nature of the individual's social networks relate to the presence or absence of non-medical needs and/or the use of medical services? De th- 62 Social Assets: Relations with Neighbors The term neighborhood refers to a bounded geographic unit that usually is homogenous in terms of socioeconomic criteria; in the study this territorial unit is defined as the area within walking distance of the individual's front door. The term neighbor- hood also may refer to a social unit in which communicative and/or functional ties exist between and among people; spatial proximity provides individuals with the opportunity to meet together, to form friendships, and to establish relationships of mutual obligation. The term neighborliness is used to refer to the extent to which such intimate and cooperative relationships exist between people living within a defined geographical area. Neighborliness is a consequence not only of contact, but of common interest and continuity as well. The homogeneity and popula- tion stability of the neighborhood, therefore, might be expected to be related to the degree of interaction between neighbors. In addition, the psychological support and functional aid available from neighbors might be expected to affect the impact of problems on an individual.9 Furthermore, the degree to which the neighborhood is socially and physically disorganized might be expected to be a source of stress and related to the individual's health or her perception of it. Part of the research problem, then, focuses on the following questions: Does the degree of neighborliness relate to the presence or absence of non-medical needs and/or the use of medical services? 63 Do the social and physical conditions of the neighbor- hood relate to the presence or absence of non-medical needs and/or the use of medical services? Psychological Assets: Feelings of Failure Feelings of failure refer to the individual's belief that she has fallen short in some domain of life, that she is unsuccess- ful. In a society such as America, in which achievement and self- reliance are values that receive strong emphasis, it might be expected that inadequate role performance or needing help with . problems might give rise to such feelings. The inability to perform socially prescribed roles--as measured by one's own expectations or against the performance of others--or the inability to solve prob- lems through one's own efforts may be sensed as indications of a lack of skill or competence to function independently. The individual who fails to fulfill a socially defined role expectation or fails to handle problems by herself, however, is not blamed for these deficiencies if she is in poor health. Thus, it might be expected that individuals who feel they fall short of the dominant culture's view of success might be likely to perceive their health as poor to justify their self-defined failure. Part of the research problem, then, focuses on the following question: Do feelings of failure relate to the presence or absence of non-medical needs and/or the use of medical services? 64 Concepts Relevant to the Women as Patients It has been suggested that individuals may rely upon the physician as a substitute to satisfy basic social needs not gratified through alternative sources. Furthermore, it has been suggested that in United States society the physician has the ultimate authority to legitimate illness and to sanction the adoption of the sick role. The use of medical services, then, may be a function of differential needs and predispositions. This study focuses on two such needs: the need for catharsis; and the need to cope with failure. Need for Catharsis The need for catharsis refers to the need for the kind of social contact that permits or encourages free communication. Theoretically, the physician's role seems ideally suited to this need since it involves communication in an uncritical, supportive atmosphere; during the course of diagnosing or managing patients' problems, the physician may provide the opportunity to talk about subjects that may or may not be related to the medical problem for which care is sought. Individuals who feel a need to express and cope with feelings and unmanageable difficulties, therefore, might be expected to seek medical care in an attempt to alleviate their concerns and problems. Part of the research problem, then, focuses on the following question: Does the individual's need for catharsis relate to the use of medical services? 65 Need to Cope with Failure The need to cope with failure refers to the use of illness as a legitimizing mechanism. In western society, once an individual's illness is legitimized by medical sanction, the person occupies a special role, i.e., the sick role. According to cultural expecta- tions, the incumbent of the sick role is exempted from responsibility for his or her condition and relieved of the usual demands and obligations that accompany his or her occupational and/or familial roles. Individuals who feel a need to escape from frustrating or untenable problems of life, therefore, might be expected to define their health as poor and to seek medical care in an attempt to obtain sanction for the avoidance of duties and responsibilities. Part of the research problem, then, focuses on the following question: Does the individual's need to cope with failure relate to the use of medical services? Concepts Relevant to the Physician It has been suggested that attitudes play a role in the orientation and responses of one actor to another (Parsons and Shils, 1959:159).10 Furthermore, it has been suggested that the differ- ential use of medical services may be a function of differences in physician's orientations and responses to patients. This study focuses on the physician's orientation toward dependent patients and on his response to them as he performs his manifest and latent func- tions. 66 Orientation to the Dependent Patient The phrase orientation to the dependent patient refers to the position the physician adopts toward the patient who frequently seeks care in an attempt to satisfy non-medical needs. In dealing with patients, the physician is expected to maintain a balance between identification and dissociation and to deal with problems of health and disease by means of professional knowledge and skill: that is, he is expected to be affectively neutral and functionally specific. These are ideal norms, based on the expectation that the patient seeks care for an essentially medical problem. Medical problems may be varyingly defined, however. Some physicians may define them in their narrowest, somatic sense. Others may define them broadly, considering socioemotional problems highly relevant to health. To the extent that the patient deviates from or conforms to the physician's expectations, it might be expected that his attitude toward her will change and that this change will be reflected in his orientation: that is, his affectivity will increase. On the one hand, the physician who believes his authority is limited to organic medical problems may feel he is being taken advantage of and dissociate himself completely from the patient and her plight. On the other hand, the physician who believes socioemotional prob- lems are within the purview of his technical competence may increase his identification and involvement with the patient and her problems. In either case, it might be expected that the physician's orienta- tion will influence the way in which he performs his manifest and 67 latent functions. Part of the research problem, then, focuses on the following question: Does the physician's orientation to the dependent patient relate to his response to her? Manifest Functions of the Physician The phrase manifest functions of the physician refers to management activities related to the prevention and cure of disease, i.e., laboratory procedures and x-rays ordered, requests for return visits, and therapies prescribed. As suggested above, the physician is an individual who (1) is susceptible to emotional responses which as a professional he is supposed to inhibit; and (2) is likely to manage patients' problems differently on the basis of his orientation toward them. The physician who is unfavorably oriented toward the dependent patient might be expected to adopt one of two possible management strategies. He may order extensive tests and follow-up visits in an attempt to find an organic problem, thereby "making" the patient conform to his expectations. Or he may prescribe a psychotropic drug and dismiss the patient, thereby implying that her symptoms are not reflective of "bona-fide" or serious illness. Con- versely, the physician who is favorably oriented toward the dependent patient might be expected to adopt a more psychologically- oriented management strategy, listening to the patient and exploring her difficulties with her. Regardless of which management strategy is adopted, however, high use of services might be expected to occur. A continuing search for organic illness obviously will generate use as the patient must 68 return to the physician for diagnostic testing. A dismissal with a prescription for a tranquilizer might be expected to generate use as the patient returns in an attempt to satisfy her frustrated needs. Similarly, an "invitation" to communicate personal problems might be expected to generate use as the patient returns for additional support and reassurance. Part of the research problem, then, focuses on the following question: Do the management strategies adopted by physicians relate to the use of medical services? Latent Functions of the Physician This study focuses on the relationship of the physician's role to two latent functions of the medical institution: the satisfaction of the need for catharsis; and the satisfaction of the need to cope with failure. Satisfaction of the Need for Catharsis. The phrase satis- faction of the need for catharsis is used to refer to the extent to which the physician feels he allows the patient who wants, to talk and unburden herself of problems. In the course of diagnosing and treating illness, the physician is required to devote time to listening to the patient's description of her problems. This aspect of his role, then, opens the way for the patient who feels a need to express difficulties to talk about them, i.e., to attempt to satisfy her need for catharsis. The extent to which her attempt is success- ful, however, might be expected to vary with the physician's defini- tion of his role. 69 Some physicians may feel that listening to personal problems is a legitimate professional obligation, that satisfaction of the need for catharsis is part of their function. Others may feel that this is a function of other professionals such as a social worker or psychiatrist, that fulfilling this need is not within the purview of their role. Physicians define their role in different ways and it might be expected that they will respond differently to patients who wish to talk about problems that, from the doctor's viewpoint, may or may not be related to their health. The physician who defines his role broadly is likely to allow patients freedom to talk. The physician who defines his role more narrowly is likely to impose limits on the subjects to be discussed during interaction. The extent to which the physician feels he satisfies the need for catharsis, therefore, might be expected to be related to the management strategies he adopts and to the use of services. Part of the research problem, then, focuses on the follow- ing questions: Does the physician's perception of his satisfaction of the need for catharsis relate to the management strategies he adopts and the use of medical services? Satisfaction of the Need to Cope with Failure. The phrase satisfaction of the need to cope with failure is used to refer to the extent to which the physician feels he legitimates illness which he believes is being used to justify failure. Illness cannot be self-defined in Western society but must be certified by a physician. The doctor, therefore, plays a major role in helping a patient to 7O cope with failure when he defines a set of symptoms as a disease and prescribes a therapeutic regimen. Patients often present with vague and ambiguous symptoms which the physician, rightly or wrongly, may assume are real or imaginary. Some physicians may feel that a patient of this type is a malingerer, demanding legitimation for her complaints out of some personal need. Such physicians are likely to dismiss the patient with a clean bill of health, thereby negating her right to adopt the sick role. Other physicians may feel that indeed the patient is not diseased but be sympathetic to her need to be sick. Such physicians are likely to respond medically to the patient, thereby sanctioning the adoption of the sick role and permitting her to cope with failure. The extent to which the physician feels he satisfies the need to cope with failure, therefore, might be expected to be related to the management strategies he adopts and to the use of services. Part of the research problem, then, focuses on the following questions: Does the physician's perception of his satisfaction of the need to cope with failure relate to the management strategies he adopts and the use of medical services? Professional Experiences of Physician The cultural definition of the physician's role has been a dominant theme throughout the above discussion. Yet, there have been indications that physicians do depart from this prototypic role and that such departures may be a function of a number of factors. This study focuses on three such factors: the type of medical school attended by the physician; his career stage; and the structure of the setting in which physician and patient interact. 71 Type of Medical School Attended. It has been argued that there are two distinct segments in American medicine--the scientifically-oriented and the practice-oriented (Miller, l970)-- and that different medical schools train students for careers in one or the other of these segments. The term academically-oriented medical school has been used to refer to the institution that (l) emphasizes the basic sciences and the acquisition of knowledge; and (2) prepares students for a career in scientific medicine (Perlstadt, 1972a, 1972b). The term practice-oriented medical school has been used to refer to the institution that (l) stresses the behavioral as well as the basic sciences and the application of knowledge as a service; and (2) prepares students for careers as practicing physicians (Perlstadt, 1972a, 1972b). Kendall and Selvin's (1957) study of Cornell medical students and Becker and his colleagues' (1961) study of Kansas medical students both demonstrate the impact that different institutional curricula and emphases have on the career choices of young graduate physicians. It also has been argued that the medical school is the major socializing influence on physicians, that the culture learned there has primary significance for what he or she does or does not do in later practice (Becker, et al., 1961; Bloom, 1965). Indeed, Miller (1970:216) reports that administrators at a teaching hospital com- plained that graduates of academically-oriented medical schools "ordered more laboratory work than necessary" because of their "academic way of thinking." Graduates of academically-oriented schools, then, might be expected to be more technically-oriented, 72 directing their efforts primarily to physical pathology at the expense of interpersonal aspects of patient care. Conversely, graduates of practice-oriented schools might be expected to be more human-oriented, channeling their efforts to physical as well as to social-psychological pathology and taking the time necessary to con- sider the social context within which illness develops. Part of the research problem, then, focuses on the following question: Does the type of medical school a physician attended relate to his orientation toward and response to dependent patients? Career Stage, Hall (1948) has conceived of the medical career as a set of more or less successful adjustments to the net- work of institutions, formal organizations, and informal relation- ships within which medicine is practiced, and distinguished four stages of a medical career: (1) generating of an ambition; (2) gaining admittance to the various medical institutions; (3) acquiring, retaining, improving, and, perhaps, transferring a clientele; and (4) developing a set of informal relationships with colleagues. The physicians who participated in this study either were completing stage two of their career, or entering stage three: that is, they were residents in training or recent graduates of the training program just beginning a practice. The term career stage, therefore, is used to refer to the year of residency of the physician. Much of the research on the impact of career stage on pro- fessional attitudes has been conducted with medical students and has emphasized the extent to which "humanitarism" decreases as the 73 student progresses in his or her training (Becker, et al., 1961; Eron, 1955; Fox, 1957; Reisman and Platou, 1960). Millin (l975:4), in fact, has suggested that "medical schools admit incoming students who are interested in people and transform them into doctors interested only in diseases." Gray, et a1. (1966), argue, however, that affiliating with patients during practice modifies the decrease in "humanitarianism" that develops in medical school. In a study of physicians with comparable "humanitarianism" scores at the time of their graduation, they found significant differences in scores after several years of practice. Those who had entered "high- interaction" specialities indicated greater "humanitarian" attitudes than those who entered "low-interaction" specialties. Gray and his colleagues attributed this increase in humanitarianism to more physician-patient interaction. The stage or level of training of a physician, then, might be expected to affect his orientation toward dependent patients and response to them. A first year resident might be less favorable toward them because they frustrate his efforts to achieve success during the period his technical skills are given their first real test (Martin, 1957). A more advanced physician might be more favorable toward them because extended contact with patients may have heightened his awareness of the need to treat the "whole" person rather than just her symptoms. Part of the research problem, then, focuses on the following question: Does the career stage of the physician relate to his orientation toward and response to dependent patients? 74 Structure of the Medical Setting, It has been argued that the structure of the medical setting affects the behavior of physicians toward patients. Freidson (1960) suggests that in bureaucratic settings physicians have less freedom to deviate from ideal norms because they are more visible to colleagues than doctors in solo practice. Shuval, et al. (1970) agree with Freidson but suggest that the bureaucratic setting also may exert certain pressures on physicians to modify their professional roles. The inordinate amount of paper work required and the large numbers of instructions to be followed--including the requirement to see a given number of patients--often causes physicians to deviate from their "ideal nonaffectively oriented role" in response to their dissatisfaction. This study was conducted with residents in training and it might be expected that the structure of the program will affect their orientations and responses to patients and, perhaps, modify the effects of other influential factors. First, faculty, intent on achieving defined goals, may stringently enforce norms. Second, these medical educators may continually scrutinize and monitor the trainees' medical practice. Residents, therefore, may be pressured to accept certain norms and less free to define their own professional role. Part of the research problem, then, focuses on the following question: Does the structure of the medical setting influence the orientation and response of physicians toward dependent patients? 75 Conceptual Model The concepts presented above provide a starting point for the development of a social system model of the use of medical services. They suggest that variability in the life situations and needs of individuals and in the orientations and responses of physicians introduce important variations into patterns of utilization. Furthermore, they suggest that the social dynamics of the doctor-patient relationship influence the use of medical services. Figure 2.1 shows the social system model of the use of medical services developed. In this figure, the doctor and patient can be seen acting in their respective roles (A and B) in a medical situation (C). The motivational-orientation of these interactants is pictured as deriving from the individual contexts in which they are located (0 and E). For the patient, this orientation is seen to be influenced by her life situation (D): for the physician, the medical profession (E). Both patient and physician (A and B) and the separate contexts in which they are located (D and E) are based in the dominant culture (F). Together, these five (A-tB-tD-+E-+F) form the environment in which the social system (C) exists, a system generated by the process of interaction between physician and patient in a medical situation. It is argued that the use of medical services (G) is a product of the dynamics of this system. 76 .mmu*>cmm peeved: mo mm: mg» yo quoz Seaman pmpoom <1-.F.N «gamed .u. muu.>¢um ab “a. «EOUOE>m odommo .au_oo_o:u>na oe:._cb ;u_3 once on $00: . «gonna .0.uOm n.ocozuoo so. one: . accenoLBo _o_uom "do ouconao Lo coconocs a... do >u..n:a co_u_non .o_00m 77 hypotheses to be Tested The conceptual model presented above provides a mechanism by which the specific hypotheses to be tested are derived. A crucial question posed by this model is the extent to which varia- tions relative to one sub-system will influence the others. This is an empirical question which only can be answered by the testing of the following hypotheses.n 1. Individuals who occupy a low social position will have a need to cope with failure and present multiple diffuse symptoms. 2. Individuals exposed to more social stressors will have a need to cope with failure and present multiple diffuse symptoms. 3. Individuals who perceive the quality of their life as poor will have a need to cope with failure and present multiple diffuse symptoms. 4. Individuals who sense feelings of failure will have a need to cope with failure and present multiple diffuse symptoms. 5. Individuals with limited or non-supportive social networks will have a need for catharsis and present multiple diffuse symptoms. 6. Patients with greater ndn-medical needs will use more medical services in the presence of multiple diffuse symptoms. 7. Physicians who attended an academically-oriented medical school will have a less favorable attitude toward dependent patients. 8. Physicians in an early career stage will have a less favorable attitude toward dependent patients. 9. Physicians will order more diagnostic procedures in the presence of patients with multiple diffuse symptoms. 78 GIVEN THE ABOVE: 10. Physicians' decisions will result in different patterns of use of medical services. 11. The conjunction of the needs of the patient and the response of the physician to these needs will result in high use of medical services. Summary This study is centrally concerned with variations in health and professional behavior and their effect on the use of medical services. The patient and the physician are seen as located in separate contexts from which their orientations toward the medical situation derive. Because the physician's orientation results from participation in a medical profession and the patient's orientation from participation in a socioeconomic millieu, their expectations may be in conflict with each other. Attention is focused on the nature of the expectations and goals these participants bring to the medical encounter, the sources of these phenomena and their subsequent effect on outcomes. Parsons' formulation of the social system provides a base for the conceptual framework of the study. The social system, along with the cultural and personality systems, are seen as con- stituting three units in a system of action. The three systems interpenetrate and are interdependent with one another. Physician- patient interaction within a medical situation (i.e., the social system) is viewed as influenced by need-dispositions developed within a specific context (i.e., the personality system) and 79 theoretically guided by societal standards of appropriate behavior (i.e., the cultural system). Attention turns in Chapter 3 to the body of methods devised to explore the research problem in a concrete setting. CHAPTER II: NOTES 1Traditionally, definitions are offered that refer to the procreational unit, including father, mother, and offspring. Commonly these emphasize interacting personalities and roles, structural arrangements of positions, or institutions representing the performance of cultural tasks (Hill and Aldous, 1967; Nye and Berarado, 1966). 2See the papers included in Mitchell (1969) for a sampling of the situations to which the concept has been applied. 3See the second edition of Bott's 1957 book (1971:248-343) for a review of some of the research on social networks that has been done since her work first was published. 4See Salloway's study (1973) of the use of the urban medical system by Gypsies and McKinlay's research (1973) with lower working class pregnant women as examples of works that have attempted this task. These works, however, do not examine the impact of social networks on non-medical needs that may motivate use. Rather, they analyze the relationship between utilization of medical services and social networks in terms of Freidson's (1961) formulation of the "lay-referral" system. 5Primary consideration in the study was given to diffuse relationships that implied services and support of a general nature. Links with categories of people who might be expected to provide individuals with some specialized type of service or support were not investigated. Furthermore, no distinction was made between active or latent links in the network, that is, between relation- ships that were being used or those that could be mobilized for a specific purpose and then become inoperative when that purpose has been met. 6See Mitchell (1969:1-50) for a discussion of the character- istics of networks likely to be relevant to descriptions of social behavior. 7Bott (1971) in a new final chapter included in the second edition of her 1957 book, adopts the word "density" as a substitute for the term "connectedness." 80 81 8The neighborhood also is likely to be related to the density of social networks. That is, the density of a network is likely to be close-knit where an area is homogeneous (socially and economically similar) and where there is continuity (population stability or low population turnover) (Bott, 1957; Gans, 1962; Young and Willmott, 1957). 9Whether this support is supplementary to that received from kin and friends or compensatory remains an empirical question. 10Parsons (1951:109) equates the word attitude with the "need-dispositions" that form the core of the personality system. nHypotheses one through five each contain two dependent variables. However, in testing, the relationship between the independent variable and each dependent variable was treated as a separate hypothesis. CHAPTER III DATA AND METHODS In order to explore the research problem presented in Chapter II, this study uses a retrospective research design employing data from four sources: (1) survey interviews; (2) in-depth inter- views; (3) self-administered questionnaires; and (4) medical records. These data were provided by patients, and physicians in training at a family health care center in a mid-western, industrial city. This chapter presents a discussion of the research design and the assump- tions that underlie it; the samples; the data; the measurement of variables; and the analysis plan. Research Design It is argued in this study that the doctor and patient are not an isolated dyad, and that social and professional factors pro- vide a constant field in which they interact. Furthermore, it is argued that high use of medical services, in the absence of clinically defined morbidity, is a consequence of the non-medical needs of the patient and the response of the physician to these needs. In order to test the validity of these premises, a retro- spective research plan was developed. First, a sample of patients, stratified according to patterns of use during a six-month period, was selected and the data 82 83 contained in their medical records for that period abstracted. Second, these patients were interviewed personally in their homes using a structured survey instrument. Third, a sub-set of these patients was interviewed in-depth using a standardized, but unstructured, format. Fourth, all residents at the center were asked to complete a self-administered questionnaire. The study thus was designed to examine the extent to which observed patterns of utilization of medical services could be accounted for by (l) aspects in the patient's social environment; (2) factors related to the patient qua patient; and/or (3) management strategies adopted by physicians to care for them. Two assumptions underlie this design. Assumptions 1. Patterns of use of medical services are relatively consistent over time; patients have a tendency to repeat past utilization patterns in the future. 2. The objective circumstances of patients' lives remain relatively consistent over time; their life situations are not likely to change within a short time. The findings of Densen, et a1. (1959) lend support to the first assumption that patterns of utilization remain fairly constant over time. In a study of a random sample of families enrolled in a prepaid health care plan in New York, they found that (1) members followed for a three-year period were likely to remain at the same utilization level from year to year; and (2) the chances that individuals would have the same level of use from year to year were greater than would be expected if one year's experience was 84 independent of the previous year's. Furthermore, Densen and his colleagues demonstrated that consistency of use was true for all patients, low utilizers, medium utilizers, and high utilizers. The findings of several researchers lend support to the second assumption that the life situations of individuals tend to remain fairly immutable over time (Berle, 1958; Brandwein, et al., 1974; Fine, 1974). Individuals who lack the advantages of educa- tional and vocational skills negotiable on the job market have poor objective chances to improve their economic situation. Limited economic resources to meet financial demands engender strains which place a burden on individuals and the family as a functioning unit. Victims of a poor distributive order, then, tend to be confronted by life problems from which they have difficulty escaping and which they cannot solve in a short time. With these assumptions in mind, attention turns to the selection of the samples for the study. The Samples The interview data for this study were provided by three categories of respondents: (l) a sample of patients who responded to a survey interview; (2) a sub-set of this sample which was selected for the purpose of in-depth interviews; and (3) a group of physicians who completed a self-administered questionnaire. The purpose of this section is to describe the way in which these respondents were selected. 85 The Survey Sample A sample of 120 women (60 dependent and 60 non-dependent patients) was planned for this study. The selection of this sample was carried out in three stages. First, two populations of women, 18 to 40 years of age, who regularly sought medical services at the family health care center were created. Second, criteria were established for selecting two sub-populations. Third, two samples, stratified according to patterns of use, were drawn from the sub- populations. These three procedures are described in the following section and the rationale for the age and sex categories chosen is discussed. Rationale for Selection of Females 18 to 40 Years of’Age The sex and age categories included in the study were chosen for both theoretical and pragmatic reasons. First, the members of this age group tend to visit a physician less than the old who consume more services because of a higher prevalence of chronic diseases, and the very young who use more services because of a high prevalence of acute conditions (Aday and Eichhorn, 1972). Thus, there was an interest in examining the possible roots of unexpectedly high utilization rates among some individuals in this presumably "healthy" age category. Second, excluding utilization of services associated with pregnancy and child birth, women should tend to visit a physician less than men since they have lower morbidity rates for most causes of death (Anderson, et al., 1972). Again, there was an interest in examining the possible roots of an unexpected phenomenon. 86 Third, women eighteen years of age are likely to be leading lives independent of their parents and those forty years of age, more likely to be premenopausal. Thus, these lower and upper age limits would eliminate the possible intervention of parental influence and physiological and/or psychological problems associated with menopause on visits. Finally, the majority of the patient population at the center is female. Thus, if men were included in the sample, too few cases would have been available for analysis.1 The Pppulations The populations used in this study were defined by means of computerized data from the health information system in use at the center since January, 1975.2 The first population was created prior to the survey phase of the study and the second, two weeks after interviewing had begun. Attempts to locate patients were only partially successful because (1) the addresses and telephone numbers on record at the center were highly inaccurate; and (2) the indi- viduals selected were relatively mobile. Therefore, in order to ensure the required size sample, a second population was created from which potential respondents could be selected. The first population was composed of patients who showed evidence of contact with the center before and after the defined observation period; it is a population ofregular, rather than tran- scient, patients. Since data contained in the health information system were entered only through the middle of 1977, the most current period, bounded by a prior- and a post-six month period, 87 was selected as the period of observation: July l, 1976 to December 31, 1976. The following criteria were developed and used to select the population: (1) female only; (2) age eighteen or older and age forty or less on July l, 1976; (3) one visit of record before July 1, 1976;3 (4) one personal visit (PV) between July 1, 1976 and December 31, 1976; and (5) one visit of record after December 31, 1976. The second population was composed of patients who showed evidence of contact before the defined observation period only; data contained in the health information system were not available to demonstrate a visit of record after the period. The six months that followed the first--January l, 1977 to June 30, l977--were selected as the second period of observation. The following modified criteria were used to define this population: (1) female only; (2) age eighteen or older and age forty or less on January 1, 1977; (3) one visit of record prior to January 1, 1977; and (4) one PV between January 1, 1977 and June 30, 1977. The populations, created by using these two sets of criteria, are described in Table 3.1. Population A (7/1/76 - 12/31/76) is smaller than Population B (1/1/77 — 6/30/77) and recorded frequencies are somewhat different. The differences are due to the less rigorous selection criteria used to define the later population as well as to the exclusion of 141 patients from Population B; 115 of those excluded were acutely ill patients who had visited the center during both periods of observation but had been selected for inclusion in '1) 88 TABLE 3.l.--Number and Percent of Patients in the Populations. Population Type of Patient A: 7/1/76 - 12/31/76 8: 1/1/77 - 6/30/77 Number Percent Number Percent Accutely 111a 209 60.6 280 50.5 Chronically 111b 29 8.4 27 4.9 Pre- or Post-Natal 73 21.2 132 23.8 Health Maintenance 34 9.8 115 20.8 Onlyc TOTALS 345 100.0 554 100.0 aAn acutely ill patient was defined as one who had visited the center one or more times during the observation period primarily for the care of short-term problems. This category, of course, also includes dependent patients whose problems may be more long-term. bA chronically ill patient was defined as one for whom any single diagnosis was entered as a label at 60 percent or more of all her visits. cA health maintenance patient was defined as one who had visited the center only once or twice during the period of observation for routine or gynecological check-ups. 89 the first sample. Twenty-six were chronically ill patients who also had visited the center during both periods of observation.4 The Sub-Populations This study focuses on the high use of medical services for diffuse symptomatology that cannot be explained by clinical morbidity. Therefore, all health maintenance, pre- or post-natal, and chronically ill patients were excluded from the populations. Such patients either had sought preventive care in the absence of somatic complaints or had sought care primarily for long-term conditions that might have accounted for high use. For purposes of the study, a health main- tenance patient was defined as one who had visited the center only once or twice during the period of observation for routine or gynecological check-ups. A long-term or chronically ill patient was defined as one for whom any single diagnosis was entered as a label at 60 percent or more of all her visits. This latter criterion separated out those with injuries, hypertension, diabetes, multiple sclerosis, malignancies, epilepsy, hyper- and hypothyroidism, systemic lupus erythematosis, and pylonephritis. In addition, it separated out patients with diagnostic listings such as anxiety, depression, migraine or tension headaches, back pain, and other ill-defined conditions. Since the literature had suggested, however, that emotional or social problems often were a contributing factor to some frequent utilization--and these diagnoses may be suggestive of such problems-~all patients with the latter diagnostic listings in the populations were retained.5 90 In short, the exclusion of healthy, pre- or post-natal and chronically ill patients left two sub-populations of patients who had visited the center one or more times during two six-month periods for a combination of routine physical exams, acute illnesses, and/or psychosocial problems.6 Table 3.2 contains data on the total number and the rate of medical care utilization of these patients. Although the sizes and the total use of the sub-populations are different, the average number of visits is approximately the same: 2.8 visits for sub- population A and 2.3 visits for sub-population B. These rates are comparable to those reported for national samples, if one accepts the assumption that a majority of patients will have a similar pattern of use during the consecutive second six-month period. For example, excluding utilization of services associated with pregnancy and childbirth, averages of 5.6 visits a year (National Center for Health Statistics (1975b:25) and 4.8 visits a year (Nathanson, 1975:58) have been reported for women in the United States. The Stratified Samples In order to differentiate between average and frequent utilizers in the two sub-populations, a criterion of frequent use was established. Three or more visits during a six-month period was determined to indicate high use. But, these high users could not all be assumed to be dependent patients because some sought care due to recurrent episodes of acute illnesses. A group of average utilizers was included in the stratified samples, therefore, as a 91 m.~ oee omm m.~ mam mom mo are: to m m.F mom ea, m.F we, NF, muwmc> ~-P u=a_uaa Lea mpemv> co naeaeuaa peaeeaa Lea me_me> Lo macawpaa m»_me> co amass: ea mapme> co amass: ca Lwnsaz .m>< page» Lamazz Loganz .m>< pouch cmassz 3 AN\OM\e - kk\_\P "m ek\PM\~P - ak\P\k u< an: L5 as _o> cowpmpznoatnsm .ucawuaa Lea mpemw> $0 LwnEzz OOMLO>< Uta mcomumpznomlaam ms“ cw muzmmuwm $9 HCUULwQ DEM LQDE=ZII.N.M m4m wee: co m m.p um um o.~ em mm muwmw> Nip pea_aaa Lea wepac> Lo au=a_3aa Seaeuaa Lea mpeme> co apeaeeaa m3_me> Lo Leassz co muem_> ea Laae=z Lo conssz .m>< Peach cwnszz Longsz .m>< peach cmasaz mm: mo manpo> NN\GM\m 1 N~\F\— "m om\Fm\~— u mN\P\m u< apasam eaecepaapm .eeawpaa aaa weenm> $0 LmnEzz mmm&m>< 0:6 mwFQEmm UmwkamLum ms“ cm. mucwwumm $0 “chLmn— Dem LUQE:ZII.m.m m4m<._. 94 N.m aa_ me e.m mmm ea m3 ago: to m a._ ma mp e.F Ne om aueae> ~-, mezmheea omzmd>muezH-zoz m.m com mm m.m oeN me m3 N-_ mezuHemmezfi Seaaaaa tea mpem_> ea meeaeaaa Seaweaa Laa mu_me> ea meeacaaa muwme> ca Laaeaz Lo mu_ae> to range: to gossaz .m>< Pouch Longsz cwnssz .m>< quoh emnsaz m o 05: AA\OM\e - NN\P\_ "m ek\_m\~_ - ek\_\k ”< a: L Pa> aPQEam eaecmeaapm .ucmwpma can mupmw> eo consaz omacm>< can umzmw>cmu=~ uoz ace umzmv>cmucn mucowuaa mo acoucma new conszztt.e.m m4m<~ 95 interviewed and non-interviewed samples on the basis of selected sociodemographic characteristics.) Table 3.4 also illustrates the consequences of surveying primarily lower-class respondents (1) who are highly mobile; and (2) for whom addresses and telephone numbers on record are inaccurate. The completion rate of the survey was only 48 percent, for both samples as well as for the total sample.7 Of the 52 percent of patients not interviewed, less than 10 percent refused to partici- pate in the study; the remainder could not be located. The problem in locating potential respondents in part was due to incorrect telephone numbers and/or addresses listed at the center. In part, it also was due to lack of a supplementary information source. For example, the telephone company was unable to supply a new or correct telephone number in 59 percent of the cases. In 19 percent of the cases, the new telephone number was not given because it was unlisted. In 14 percent of the cases, someone was reached at the telephone number or address on record. But, this person was unable to give the caller the required information. Common responses were: "She picked up and pulled out. I don't know where"; or "I don't know where she is now. I get her mail and usually tear it up.“8 The information listed at the center for respondents who finally were interviewed was also in most cases inaccurate. The address, telephone number, or both were incorrect for 46 percent of these patients (13 percent, 13 percent, and 20 percent respectively). Only the perseverance of the interviewers made even the low comple- tion rate possible. (Appendix 8 includes a discussion of some of the 96 problems involved in working with a sample of primarily lower class women.) The In-Depth Interview Sample In order to provide insights into the data collected through the survey interview, twenty-five respondents were selected for in-depth interviews. These women were chosen on the basis of an expressed interest in the study (as documented during the survey interview) and willingness to share their experiences. Problems also were encountered in completing interviews within these respond- ents. They were difficult to reach and almost one-third of the time not at home when an interviewer appeared for a scheduled appointment. The Physician Sample Thirty-three physicians were in training at the center in 1977-1978 when this study was conducted; seven first year residents; thirteen second year residents; and thirteen third year residents. All were asked by the medical director of the training program to participate in the study by completing a self-administered question- naire. Twenty-five (75.8 percent) agreed: four (57.1 percent first year residents; twelve (92.3 percent) second year residents; and nine (69.2 percent) third year residents. In addition, eight physicians, graduates of the training program who had cared for patients during 1976-1977, were contacted by letter and asked to complete a questionnaire. Six (75 percent) completed and returned a questionnaire. 97 Data are available, however, for those physicians who chose not to participate in the study, but who cared for patients during 1976-1977. A pilot study had been conducted in 1976-1977 and eleven of these physicians had completed a questionnaire then: four first year residents; three second year residents; three third year residents; and one graduate of the program. The data, voluntarily provided in the pilot study, are used in this study. In total, then, data were collected over a two-year period from forty-two physicians: eight (19.0 percent) first year residents; fifteen (35.7 percent) second year residents; twelve (28.6 percent) third year residents; and seven (16.7 percent) graduates of the program. Five of these physicians, however, had not cared for patients in 1976-1977: four of the first year residents in 1977-1978 and one second year resident who was new to the program the year the study was conducted. Thirty-seven of the physicians who completed questionnaires, therefore, were at the following career stages when they cared for patients in 1976-1977: fifteen (40.5 percent) in the first year; twelve (32.4 percent) in the second year; and ten (27.0 percent) in the third year of training. The Data The data for this study were derived from four sources: (1) patient medical records; (2) personal interviews with patients in their homes using a standardized, structured interview; (3) in- depth interviews with patients in their homes using a standardized, 98 but unstructured format; and (4) self-administered questionnaires completed by physicians who cared for patients at the center. This section describes the data, the way in which they were collected, and their advantages and limitations. Data Collection Procedures As indicated above, potential respondents were identified employing computerized data from the health information system in use at the center. A letter was sent to these individuals by the medical director of the center. This letter (1) described the study; (2) requested participation in it; and (3) indicated that an interviewer would be calling to make an appointment to meet and speak with the woman. (Appendix C contains copies of the research materials used in the study.) When a potential respondent was contacted, the study again was described to her and she was asked when it would be convenient to spend the thirty to forty-five minutes necessary to respond to the survey questionnaire. An agreement to meet with the interviewer was considered tacit consent to participate in the study. More formal consent was obtained, however, at the time of the interview itself; respondents were asked to sign a form indicating their willingness to be interviewed and to have their past and future medical records reviewed. Medical records were abstracted at the center. All inter- views were conducted in the patients' homes by female interviewers who ranged in age from twenty-three to forty-four years of age. 99 Answers were recorded on the survey instrument at the time of the interview. In addition, notes were taken during the in-depth inter- views. The quotes of patients used in this report, therefore, are verbatim statements. Medical Records Medical records provide a chronological account of patient visits, their purposes, diagnoses, management, and outcomes; they reflect the patient's description of her illness as well as the physician's assessment of it. Thus, they offer a unique opportunity (1) to relate the presence of disease and the seriousness of illness to utilization of medical services; and (2) to link the treatment of patient problems to outcomes and use of services. The abstract schedule used in the study was designed to organize data relevant to visits made during the observation period. A separate form was completed for each visit and information such as the following was abstracted: symptoms; physical findings; diagnoses; laboratory procedures and/or x-rays planned and performed; results of diagnostic procedures; and dispositions; i.e., requests for follow-up visits. In addition, any information that was included in the medical record about the patient's background and/or difficulties she might be encountering was abstracted. Most records, however, contained limited, if any, social information about the patient. This omission represents a significant limitation of the medical record as a source 100 of data. Thus, a survey questionnaire was developed to collect information about the individual and her situation. Survey Questionnaires Survey questionnaires provide an opportunity to collect data that can help link attitudinal and environmental factors to use of services and medical outcomes. The relatively structured nature of such instruments also provides an opportunity to collect comparable data from all respondents for use in quantitative analyses. The questionnaire developed for this study consisted of a standardized interview guide and recording form that included many close-ended and some open-ended items. The instrument was designed to stimulate a conversation-like interaction between the interviewer and the respondent. This technique not only permitted probing, but also elicited relatively natural responses. These responses repre- sented the individual's description of her social and personal con- dition and her perception of it. The structured quality of this questionnaire, however, had its limitations. By its very nature, the instrument created a situation that provided few opportunities to obtain the serendipitous insights possible in a more free-flowing interview situation. Thus, in order to provide insights into the structured data, supplementary information was collected in a series of in-depth interviews. In-Depth Interviews In-depth interviews provide an opportunity to collect data that can illuminate and add richness to reported quantitative data. 101 In addition, they provide an opportunity to collect data that can help suggest alternative analytic strategies. The interview format used in this study was standardized, but flexible. During the hour in which the interviewer and respondent interacted, the following tapics were covered: the role of the doctor; the role of the patient; the action taken in the presence of certain symptoms; and satisfactions and dissatisfactions with the center. The breadth of these t0pics led the interviews into far-ranging areas such as: the individual's situation and personal problems; her coping strategies; and her view of her life. These data, however, are qualitative and represent the views of only a few, albeit relatively representative, sub-set of the sample. Self-Administered Questionnaires Interviews with physicians offer a unique opportunity to develop data that can be used to relate the backgrounds and attitudes of doctors to their behavior in a clinical setting. Since such data usually are not available to researchers, their use in this study represents a distinct advantage. The questionnaire developed con- tained both close- and open-ended questions and was designed to elicit data that could be quantified as well as data that could be used to suggest potential insights into responses. It covered t0pics such as: the role of the physician; the role of the patient; attitudes about dependent patients; reasons for selecting family practice as a specialty; and satisfactions and dissatisfactions with medical practice. 102 hejor Variables Four major categories of variables are considered in this study: (1) characteristics of the individual; (2) characteristics of the patient; (3) characteristics of the physician; and (4) out- comes. These variables are tied to the concepts discussed in Chapter 11; they are concepts translated into measures. The purpose of this section is to explain how this process was carried out. Characteristics of the Women as Individuals The characteristics of the individual reflect her life situation and her perception of it. Six major aspects of her situa- tion were considered: (1) quality of life; (2) social position [i.e., (a) race; (b) education; (c) income; (d) marital status; (e) social role; and (f) family structure]; (3) social stressors; 4) social networks; (5) relations with neighbors and characteristics of the neighborhood; and (6) feelings of failure. Quality of Life Four attributes of respondents' perception of the objective circumstances of their lives were measured: (1) general state of life; (2) comparative family life; (3) comparative money situation; and (4) difficulty in living on income. General State of Life. The respondent's perception of her overall life situation was operationalized as a question which asked if, in general, she felt her life was smooth or troubled. If a respondent described her life as smooth with little trouble, or 103 smooth with occasional trouble, she was classified as having a smooth state of life. If a respondent described her life as fre- quently upset by trouble or most of the time upset, she was classified as having a troubled state of life. Comparative Family Life. The respondent's perception of the functioning of her family as a social unit was operationalized as two questions that asked how well she felt her family life compared with that of her relatives and with that of her friends. Answers were recorded on a five-point scale with a score of one assigned to a response of "a lot better" and a score of five to an answer of "a lot worse.” A low score, therefore, indicates that the respondent feels her family life is good in comparison with that of others she knows. Comparative Money Situation. The respondent's perception of her family's financial condition was operationalized as two questions that asked how she felt her family's money situation com- pared with that of her relatives and with that of her friends. Answers were recorded on a five-point scale with a score of one assigned to a response of "a lot better" and a score of five to an answer of "a lot worse." A low score, therefore, indicates that the respondent feels her family's financial situation is good in compari- son with that of others she knows. Difficulpy in Living on Income. The respondent's perception of her family's ability to manage on its income was operationalized 104 as a question that asked, in general, how hard or difficult it was for her and her family to live on their present income. Answers were recorded on a five-point scale with a score of one assigned to a response of "very difficult" and a score of five to an answer of "not difficult at all." A low score, therefore, indicates that the respondent feels her family lacks economic stability. Social Position Six traits that might be expected to affect the individual's life chances were measured: (1) race; (2) education; (3) income; (4) marital status; (5) social role; and (6) family structure. Race. Race was operationalized by interviewer observation. Respondents were categorized as caucasian or black. Education. Education was operationalized as a question that asked respondents the last year they had completed in school. For purposes of analysis, responses were classified into the following six categories: (1) none to eight years; (2) some high school; (3) high school graduate; (4) some college; (5) college graduate; and (6) graduate school. Ihephe, Income was adjusted for family size and was operationalized as two questions that asked respondents (1) to name all the people who lived in their house full-time; and (2) to classify their total family income for the past twelve months before any taxes or other deductions. For purposes of analysis, respondents were classified as having: (1) an inadequate income; (2) a marginal 105 income; or (3) an adequate income. The three categories were developed on the basis of criteria for poverty level income for non- farm families living in the county (see Table 3.5). TABLE 3.5.—-Criteria for Poverty Level Income for Non-Farm Families Living in the County.3 b Family Size Income 1 $3,140 4,160 5,180 6,200 7,220 8,240 0101wa aSource: Bureau of Community Services, Department of Labor, the State, May 1978. bFor families of more than six members, $1,020 is added for each additional person. An inadequate income was defined as a total family income that was below poverty level. An adequate income was defined as a total family income that was above poverty level. A marginal income was defined as an income that could not be classified as either below or above poverty level; income was operationalized as a categorical variable in this study and, in several cases, the poverty level criterion fell between the upper and lower levels presented in 106 the study's variable. For example, a respondent living in a two person household with an income between $3,000 to $4,999 was not classifiable as living below or above poverty level on the basis of the criterion, i.e., $4,160. Rather than assume at which end of the continuum her income was, the category "marginal income" was created for purposes of classification. Marital Status. Marital status was operationalized as two questions that asked respondents (1) their relationship to the head of the household; and (2) if they were single or had not reported a husband as the head of the household, had they ever been married. For purposes of analysis, respondents were classified into three categories: (1) single; (2) married; and (3) disrupted marriage. A gjhgle_woman was defined as a female who never had been married and lived alone or with other relatives or friends. A married woman was defined as a female who lived with her husband or with a common-law husband whom she identified as the head of the household. A woman with a disrupted marriage was defined as a widowed or divorced female or a woman who was separated from her husband. Social Role. The social role of a respondent was operation- alized as a question that asked what she usually did most of the time. For purposes of analysis, respondents were classified into three categories: (1) housewife; (2) worker; and (3) dual roles. A housewife was defined as a woman who reported that she kept house only. A worker was defined as a woman who only reported that she 107 was employed for pay at a job. A woman with dual roles was defined as one who reported that she engaged in several activities, e.g., kept house and worked. Family Structure. Family structure was operationalized as a 9 question that asked respondents who was the head of the household. For purposes of analysis, respondents were dichotomized into the categories male-headed household and female-headed household. A male-headed household was defined as one in which a father, a husband, an uncle or a male friend was reported as the head. A female-headed household was defined as one in which the respondent, a mother, or an aunt was reported as the head. Social Stressors The presence or absence of social stressors in respondents' lives was operationalized as a set of eleven questions that asked if any of several stress-related events had happened to them during the last year. The events selected were intended to represent family, personal, occupational, and financial situations that require or signify change in ongoing adjustment. Answers were recorded on a dichotomous scale: a positive response was assigned a score of one and a negative response, zero. The scores then were added together and a single index assigned to each respondent which ranged from zero to seven. 108 Social Networks Five characteristics of kin and friendshjp_networks were measured: (1) density; (2) content; (3) frequency of contact; 10 (4) type of contact; and (5) location of kin and friends. Each characteristic was measured separately for kin and for friends. Density. The density of networks was Operationalized as two questions that asked respondents (1) if they knew whether any of their friends and relatives got together when they were not with them; and (2) if so, how often.11 A close-knit network was defined as one in which people got together regularly, a loose-knit network as one in which people got together only occasionally, not at all, or never. If a respondent was unable to identify the fabric of her networks, they were classified as loose-knit since it was assumed that in such a case her kin and friends probably did not get together regularly. Content. The content of networks was operationalized as a series of questions that asked respondents: (1) whether they received psychological support suchas friendliness, attention, encouragement, and understanding;12 (2) whether they received func- tional support such as financial assistance, help with child care and personal problems, and aid when there was an illness in the family; and (3) if such support had been received, who had provided it. If a respondent mentioned kin and friends, she was classified as receiving support from these categories of people. 109 Frequency of Contact. The frequency of contact with kin and friends was operationalized as a set of four questions that asked respondents (1) how often they had talked on the phone with kin and friends during the past month; and (2) how often they had met and talked with their kin and with their friends in the last month. Answers were recorded on a five-point scale that included the follow- ing categories: never; once; two or three times; one or two times a week; and nearly every day. Type of Contact. The type of contact respondents had with kin and friends was operationalized by deriving variables from responses to questions about frequency of contact. If a respondent said that she had Spoken with kin and friends on the telephone but had not met with them, she was classified as having had phghe_ contact only. If a respondent said that she had met and talked with kin and friends but had not spoken with them on the telephone, she was classified as having had personal contact only. If a respondent had Spoken with kin and friends on the telephone and also met and talked with them, she was classified as having had both phone and personal contact. Location of Kin and Friends. The location of significant others was operationalized as a set of questions that asked respondents if their kin and friends lived mostly within their neighborhood or not. The neighborhood was defined for respondents as the area within walking distance of their home. If a respondent said that most of her kin and friends lived in the neighborhood, she 110 was classified as having neighborhood-based networks. If a respondent said that some, but not all, of her kin and friends lived in the neighborhood, she was classified as having location-mixed networks. If a respondent said that none of her kin and friends lived in the neighborhood, she was classified as having hphf neighborhood-based networks. Relations with Neighbors and Character- istics of the Neighborhood The neighborhood was defined for respondents as the area within walking distance of their home. Neighborliness as well as the nature of the neighborhood in which respondents lived were measured. Three aspects of this nature were considered: (1) stability; (2) homogeneity; and (3) frequency of contact. Neighborliness. Neighborliness was operationalized as a set of questions that asked respondents how easy or hard it was to get the following four tasks done in their neighborhood: (1) having a neighbor watch the children for a few hours if she was called away suddenly; (2) getting a ride from a neighbor if she couldn't use other transportation; (3) getting information or help from a neighbor if she couldn't do something herself; and (4) getting together with neighbors for something like a barbecue or party. Answers were recorded on a five-point scale and an index of neighborliness created in the following manner.13 Responses such as "always very easy" or "fairly easy" were considered to indicate that the respondent felt free to call upon 111 neighbors for assistance when she needed it and were given a score of three. A response such as "depends on the situation" was con- sidered indicative of an unsurity about the value of the relation- ship and was assigned a score of two. Responses such as "fairly hard" or "always very hard" were considered to indicate that the respondent felt the relationship was almost completely dysfunctional and were given a score of one. The scores for the questions were added and a single index assigned to each respondent, on the basis of the following criteria:14 (a) high neighborliness - a score of eight or nine; (b) equivocal neighborliness - a score of six or seven; (c) low neighborliness - a score of three, four, or five. Stability of Neighborhood. The stability of the neighborhood was operationalized as a question that asked respondents if, during the past year, there had been many people moving in and out of the neighborhood, a few or hardly any. A giep1e_neighborhood was defined as one in which there had been hardly any movement of people, an unstable neighborhood as one in which many or a few people had moved in or out. Homogeneity of Neighborhood. The homogeneity of the neigh- borhood was operationalized as a question that asked respondents if they thought people in the neighborhood had many things in common, some things in common, or only a few things in common. A homogeneous neighborhood was defined as one in which neighbors had many things 112 in common, a heterogeneous neighborhood as one in which neighbors had some or only a few things in common. Frequency of Contact. The frequency of contact between neighbors was operationalized as a question that asked respondents if, during the year, neighbors got together many times, a few times, or hardly ever. The nature of the neighborhood was defined as characterized by high frequency of contact if neighbors got together many times, and by low frequency of contact if they got together only a few times or hardly ever. Feelings of Failure Two aspects of feelings of failure were measured: (1) comparative role performance; and (2) need for help with problems. Comparative Role Performance. Comparative role performance was measured in terms of overall performance of tasks of daily living as well as in terms of performance of three major social roles: (1) housewife; (2) worker; and (3) student. Respondents who performed dual roles were asked to respond to all questions relative to them. 0verall_performance of tasks of daily living was operation- alized as a question that asked a respondent how well she felt she managed in doing all that had to be done in the day. Answers were recorded on a five—point scale with a score of one assigned to a response of "a lot better" and a score of five to an answer of "a lot worse." A low score, therefore, indicates that the respondent 113 feels that she is successful in her overall performance. Performance of major social role was operationalized as a set of three questions that asked how good the respondent felt she was at: (1) taking care of a home; (2) doing the kind of work she did; and/or (3) getting along at school. Answers were scored on the same five-point scale as that used to measure overall performance of tasks of daily living. Need for Help with Problems. The need for help with problems was operationalized as a follow-up question to one that asked respondents whether they received help from outside the household with certain day-to-day concerns. For example, if a respondent indicated that she did not receive help with an activity, she was asked whether she could use help with it. A positive response was assigned a score of one and a negative response, zero. The scores then were added together and an index assigned to each respondent which ranged from zero to six.15 Characteristics of the Women as Patients The characteristics of the patient reflect her medical and non-medical needs. Both needs were operationalized as perceptual variables in the study. Non-medical needs, by definition, is a perceptual concept. Medical needs, however, may be a subjective or an objective concept. In this study the concept was measured sub- jectively by accepting the symptoms presented at physician visits as a measure of the individual's perception of her medical needs.16 114 Medical Needs The symptoms presented by patients during the six-month observation period were classified in three ways: (1) by number; (2) by type of complaint; and (3) by number of different parts of ‘7 Additive as well as the body in which symptoms were located. average scores were calculated. That is, some variables were adjusted for variations in the number of visits among patients in order to account for the fact that those who had sought care more frequently had had more opportunities to report symptoms. In all cases, a summary index was assigned to each patient. Number of Symptoms. The number of symptoms presented by the patient during the six-month observation period was operationalized in four ways: (1) number of symptoms; (2) average number of symptoms; (3) number of different symptoms; and (4) average number of different symptoms. 1. Number of Symptoms - total number of symptoms presented during the six-month period. 2. Average Number of Symptoms - total number of symptoms presented during the six months divided by the number of visits. 3. Number of Different Symptoms - total number of distinct symptoms presented during the six months. 4. Average Number of Different Symptoms - total number of distinct symptoms presented during the six months divided by the number of visits. Iype of Complaint. The type of complaints presented by the patient during the six-month period was operationalized using a typology developed by Zola (1966, 1973). In this typology, two types 115 of complaints are distinguished: one type represents a rather specific organic dysfunctioning and reflects the way in which the patient restricts her difficulties; the second type represents a more global malfunctioning and reflects the way in which the patient emphasizes the more diffuse nature of her complaints: Two broad groups of symptoms are incorporated in this typology. The first may be designated as 'specific dysfunction' and includes the separate categorTes of malfunctions (difficulty in hearing, seeing, etc.); discharge (drain, drip, bleeding, sneeze, etc.); inappropriate (hear rings, hums, sees spots, etc.); and involuntary (tics, blinks, throbs, etc.). The second group is called 'diffuse anomalies' and includes the distinct categories of hurts (aches, pains, burns, sore, tender, etc.); form (swelling, breakout, puffing, etc.); countenance (color, weight, general appearance, etc.); energy level (tired, weak, etc.); and bodily state or condition (dizzy, sick, dry, nauseous, hot, nervous, etc.) (personal communication, I. K. Zola, December 9, 1977). All symptoms were coded, classified according to Zola's typology, and operationalized in four ways: (1) number of symptoms of specific dysfunction; (2) number of symptoms of diffuse anomalies; (3) pro- portion of symptoms of specific dysfunction; and (4) proportion of symptoms of diffuse anomalies. 1. Number of Symptoms of Specific Dysfunction - total number of symptoms of specific dysfunction presented during the six-month observation period. 2. Proportion of Symptoms of Specific Dysfunction - total number of symptoms of specific dysfunction presented during the six-months divided by the total number of symptoms. 3. Number of Symptoms of Diffuse Anomalies - total number of symptoms of diffuse anomalies presented during the six-month period. 4. Proportion of Symptoms of Diffuse Anomalies - total number of symptoms of diffuse anomalies presented during the six months divided by the total number of symptoms. 116 Number of Different Parts of the Body in Which Symptoms Located. The number of different parts of the body in which symptoms were located during the six months was operationalized in two ways: (1) number of different parts of the body; and (2) average number of different parts of the body. 1. Number of Different Parts of the Body - total number of distinct parts of the body in which symptoms located during the six-month observation period. 2. Average Number of Different Parts of the Body - total number of distinct parts of the body in which symptoms located during the Six months divided by the number of visits. Non-Medical Needs Two non-medical needs of the patient were measured: (1) the need for catharsis; and (2) the need to cope with failure. Need for Catharsis. The need for catharsis was operation- alized as three questions that asked respondents: (1) if they knew enough people with whom they felt free to talk about personal prob- lems; (2) if they sometimes felt that people around them were too busy to talk; and (3) if they sometimes felt that people were not 19 A positive response to paying attention to them when they talked. the first question, and negative responses to the second and third were considered indicative of the lack of a need for catharsis. That is, these responses were considered to indicate that the indi- vidual had sympathetic, understanding, and supportive listeners among the members of her social networks and did not need an alterna- tive source to satisfy her need to express feelings. The following 117 set of responses to the three questions above, therefore, operation- alized the extent of the need for catharsis. 1. No Need for Catharsis - all three responses were congruent with the standard. 2. Some Need for Catharsis - one response was not congruent with the standard. 3. Greater Need for Catharsis - two or three responses were not congruent with the standard. Need to Cope with Failure. The need to cope with failure was operationalized as a question that asked respondents how they would describe their general health. Responses of excellent or good were considered indicative of the lack of a need to use illness to justify failure, i.e., no need to cope with failure. Responses of fair or poor were considered indicative of a need to use illness to cope with failure.20 Characteristics of the Physician The characteristics of the physician reflect his attitude toward the dependent patient and his response to her. Three aspects of his attitude were considered: (1) his orientation toward the dependent patient; (2) his perception of his satisfaction of the need for catharsis; and (3) his perception of his satisfaction of the need to legitimate failure. Four major aspects of physician response were considered: (1) assessment or diagnosis of patient problems; (2) diagnosis procedures ordered; (3) therapies prescribed; and (4) disposition of the patient. In addition, three characteristics that might be expected to relate to the physician's orientation were 118 considered: (1) his career stage; (2) the type of medical school he had attended; and (3) the structure of the medical setting in which he functioned. Physician's Attitude Toward the Dependent Patient As indicated above, the physician's attitude toward the dependent patient was measured in three ways: his orientation toward her, and his perception of his satisfaction of two of her non-medical needs. These measures do not reflect the physician's actual response to the patient; they measure his perception of his behavior toward her. Physician's Orientation Toward the Dependent Patient. The physician's orientation toward the dependent patient was operation- alized as three sentence stems residents were asked to complete: (1) "when a patient persistently presents with vague or inconsistent symptoms . . ."; (2) "when physical laboratory examinations do not account for a patient's complaints . . ."; and (3) "to me, dealing with the social and emotional problems of patients is . . . ."21 Responses were coded according to the following three categories: 1. Favorable - feeling tones such as "challenging" or "interesting" or approach tendencies such as "delve into social and/or psychological factors;" 2. Neutral - no definite, or ambivalent, feeling tones or no definite approach-avoidance tendencies expressed. This category included ambivalent responses such as, "may be rewarding but are time-consuming"; a re- evaluation is necessary"; or "differential diagnosis should be reviewed." 119 3. Unfavorable - feeling tones such as “frustrating," "disinterest," or "less satisfying," or avoidance tendencies such as “approach the situation from a medical viewpoint." For purposes of analysis, favorable responses were assigned a score of one, neutral or ambivalent responses a score of two, and unfavorable responses a score of three. Thus, the lower a physician's score, the more favorable his orientation toward the dependent patient. Physician's Perception of His Satisfaction of the Need for Catharsis. The physician's perception of his satisfaction of the need for catharsis was operationalized as a question that asked residents how they responded, under ordinary working conditions, when patients talked about problems which were only indirectly related to their hea1th.22 If residents indicated that they encouraged such patients to talk, they were classified as satisfying the need for catharsis. If they indicated that they tried to limit such discussion in general to subjects directly connected to the medical problem, they were classified as not satisfying the need for catharsis. Ehysician's Perception of His Satisfaction of the Need to Legitimate Failure. The physician's perception of his satisfaction of the need to legitimate failure was operationalized as a question that asked residents how they generally behaved toward patients who 23 became ill to justify failure. If residents indicated that they responded to such patients with sympathy and understanding, they 120 were classified as satisfying the need to legitimate failure. If they indicated that such patients annoyed them, they were classified as not satisfying the need to legitimate failure. Physicians' Response to the Dependent Pfitient The response of physicians to dependent patients was measured four ways: (1) assessment of problems or diagnoses; (2) diagnostic procedures ordered; (3) therapies prescribed; and (4) disposition of patients. These measures reflect physicians' actual behavior, the management strategies they adopt to care for patients.24 Assessment of Problems. Physicians' assessments of patient problems were operationalized by having three judges code diagnoses on the basis of criteria that differentiated (1) those that were psychological or symptomatic from (2) those that were organic.25 (See Appendix E for a detailed description of the coding criteria and procedures and a discussion of the development of the diagnostic index described below.) The two diagnostic categories reflect physicians' appraisals of patients' problems and, with one exception, the extent to which these diagnostic appraisals were supported or were not supported by clinical evidence. That is, unsupported diagnoses refer to labels that indicated the presence of symptoma- tology that could not be clinically-defined and validated (e.g., headache, pain, malaise) and/or organic diagnoses that were not supported by positive physical findings or diagnostic procedures. Supported diegnoses refer to labels that indicated the presence of 121 symptomatology that was supported by clinical evidence (e.g., urinary tract infection, strep throat). Eeycholpgical diagnoses, however, refer to labels that indicated the presence of socioemotional prob- lems, whether or not these diagnoses were supported by recorded findings (e.g., anxiety, depression). After the diagnostic data for the six-month period were coded, proportional scores for supported diagnoses and for psychological/unsupported diagnoses were calculated per patient using the following formula: total number of diagnoses within category divided by total number of diagnoses in the two categories. A diagnostic index for each of the two categories then was assigned to patients. When the findings of the study are discussed, there- fore, the following two variables are used to refer to physicians' assessments of patients' problems: 1. Proportion of Psychological/Unsupported Diagnoses - number of psychological and/or unsupported diagnoses divided by the total number of diagnoses recorded for the six-month period;25 2. Proportion of Supported Diagnoses - number of supported diagnoses divided by the total number of diagnoses recorded for the six months. Diagnostic Procedures Ordered. The diagnostic procedures ordered by physicians were measured in two ways: (1) the number of laboratory tests and/or x-rays planned per patient; and (2) the number of laboratory tests and/or x-rays actually performed per patient. 122 1. Number of Laboratory Tests and/or X-rays Planned - total number of diagnostic procedures recorded as part of the physician's management plan during the six-month observation period. 2. Number of Laboratory Tests and/or X-Rays Performed - total number of diagnostic procedures with results recorded during the six-month observation period. Therapies Prescribed. Three types of therapies prescribed by physicians were considered: (1) psychotropic therapies; (2) symptomatic therapies; and (3) curative therapies. The term psychotropic therapies is used to refer to drugs that have an effect on the mind or are capable of modifying mental activity such as tranquillizers or antidepressants. The term symptomatic therapies is used to refer to drugs that have a palliative effect on symptoms such as analgesiacs, antihistomines, or antacids. The term curative therapies is used to refer to drugs that have a healing or remedial effect on disease such as antibiotics or antibacterials. These therapies were measured per patient in six ways: (1) number of psychotropic therapies; (2) proportion of psychotropic therapies; (3) number of symptomatic therapies; (4) proportion of symptomatic therapies; (5) number of curative therapies; and (6) proportion of curative therapies. 1. Number of Psychotropic Therapies - total number of psychotropic drugs prescribed during the six-month observation period; 2. Proportion of Psychotropic Therapies - number of psychotropic drugs prescribed divided by the total number of drugs prescribed during the six months; 3. Number of Symptomatic Therapies - total number of symptomatic drugs prescribed during the six-month observation period; 123 4. Proportion of Symptomatic Therapies - number of symptomatic drugs prescribed divided by the total number of drugs prescribed during the six months; 5. Number of Curative Therapies - total number of curative drugs prescribed during the six-month observation period; 6. Proportion of Curative Therapies - number of curative drugs prescribed divided by the total number of drugs prescribed during the six months. Disposition of Patients. Physicians' disposition of patients was measured by the total number of return visits requested per patient during the six-month observation period. Characteristics Expected to be Related to Physician'sTAttitude Three characteristics that might be expected to relate to the physician's attitude toward the dependent patient were considered: (1) career stage; (2) type of medical school attended; and (3) structure of the medical setting in which physician and patient interact. Career Stage. The physician's career stage was operation- alized as a question that asked in which year of residency the doctor was. Answers were recorded on a five—point scale with a score of one assigned to a first-year resident and a score of five to a graduate of the program who was involved in practice. Iype of Medical School Attended. The type of medical school attended was operationalized by having two physicians--a dean and an associate dean of a medical school--classify the twenty different 124 medical schools from which the residents had graduated into one of two categories: (1) academically-oriented; and (2) practice-oriented. Agreement between the physicians was 100 percent. Structure of the Medical Setting. Data on the setting in which physician and patient interact were collected through a personal interview with the director of the residency training program and observations at the health care center. A description of this set- ting, and a discussion of the way in which its structure affects the behavior of physicians in training is presented in Chapter IV. Outcome Outcome is the dependent variable in the study. It refers to the number of face-to-face contacts between patients and providers that occurred during the six-month observation period, i.e., the actual use of medical services. Three aspects of this use were considered: (1) the total number of visits to the family health care center; (2) the number of visits that were doctor-initiated; and (3) the number of visits that were patient-initiated. Total Number of Visits The total number of visits to the family health center simply is a count of the number of times the patient sought care from a physician at the center during the six-month observation period. Number of visits ranged from one to ten. 125 Number of Physician-Initiated Visits A physician-initiated visit is defined as one in which the patient visited the center at the physician's request. Figure 3.1 includes the criteria that were used to determine if a visit was physician-initiated. Number of physician-initiated visits simply is a count of the number of times the patient sought care in response to a physician's request for a follow-up visit. The number of physician-initiated visits ranged from none to six. Number of Days in Which Patient Requested to Ctherion Return to Center 7 days If the patient visits the center within 6 to 10 days of the specified time, i.e., one day earlier or 3 days later, the visit is defined as physician-initiated. 14 days If the patient visits the center within 12 to 22 days of the specified time, i.e., 2 days earlier or 8 days later, the visit is defined as physician-initiated. 21 days If the patient visits the center within 18 to 29 days of the specified time, i.e., 3 days earlier or 8 days later, the visit is defined as physician-initiated. 1 month If the patient visits the center within 25 to 38 days of the specified time, i.e., 5 days earlier or 8 days later, the visit is defined as physician-initiated. Figure 3.l.--Defining Criteria of a Physician-Initiated Visit. 126 Number of Patient-Initiated Visits A patient-initiated visit is defined as one in which the patient visited the center of her own volition. Number of patient- initiated visits simply is a count of the number of times the patient presented for care without being requested to do so by the physician. The number of patient-initiated visits ranged from none to seven. Analysis Plan Two statistical techniques were used to analyze the data: Pearson product-moment correlations and multiple regression analysis. Pearson correlations indicate the way in which two variables vary 27 They provide a single number which summarizes the together. strength and the direction of the linear relationship between two variables. (Since this is done, however, at the sacrifice of the detail which is provided in cross-tabulations, data sometimes also are reported in contingency table form or in lists of frequencies.) The technique provides an easy means by which the strength of relationship between one pair of variables and a different pair can be compared. Thus, the correlations were used to select from among all the variables, those that were significant and might be expected to be good predictors of the differential use of medical services. These selected variables, then, were entered into the regression equations. Multiple regression is a technique that evaluates and measures the overall dependence of a criterion variable on a set of independent 29 or predictor variables. The technique served a major research 127 objective of the study by indicating how much of the total variation in the dependent variable was explained by all of the independent variables acting together. The stepwise method was used in all analyses. This method admitted variables into the regression equa- tion according to their contribution to explain of variance in the dependent variable. Thus, the first variable admitted was the one which explained the most variance when all other independent vari- ables were controlled, the second admitted was the one which explained most of the remaining variance and so on until the pro- cedure was halted because the remaining independent variables no longer made a significant contribution to the explanation process. Summary A study was designed to examine the extent to which past utilization behavior could be accounted for by individual, patient, and physician characteristics. A sample of young “healthy" females, stratified according to patterns of use of medical services, was selected and the data for a six-month period abstracted from their medical records. Structured questionnaires were used in survey interviews with these women and standardized, but flexible, question formats used in in-depth interviews with a sub-set of them. In addi- tion, self-administered questionnaires were completed by the physicians who had provided medical care for the women. The data collected permitted an examination of the relationship between four sets of variables: (1) individual; (2) patient; (3) physician; and (4) use of medical services. Analyses focused on patients' needs and 128 their sources, physicians' responses to these needs, and patterns of utilization behavior. CHAPTER III: NOTES 1The ratio of females to males in the 18 to 40 year old age category was approximately four to one. 2Data for this information system are collected through a one- page, multiple copy document that is an integrated billing and patient encounter form. The form that contains modified elements of the patient's chart (e.g., diagnosis, laboratory procedures and/or x-rays, treatments, provider of care) is attached to the patient chart, com- pleted by the physician at the end of each patient visit, and the data then edited and submitted to computer processing. 3A visit of record is defined as any entry in the data system that indicates a contact with the center. Such an entry may Show that an actual visit was made or that an appointment was can- celled or broken. 4If these patients were to be included in Population 8, the percentages within its categories become more like those in Popula- tion A: that is, acutely ill, 56.8 percent; chronically ill, 7.6 percent; and pre- or post-natal, 19.0 percent. The percentage of health maintenance patients, however, remains higher (16.6 percent) in Population 8, a reflection either of annual use for purposes of preventive care or transient use. 5Non-somatic complaints, of course, also can influence use patterns among pre- and post-natal and chronically ill patients. Nevertheless, since the possibility existed that these conditions themselves might exacerbate the predisposition to or create emotional problems, such patients were excluded from the sampling pool. 6Visits made for the purposes of preventive care were excluded in all analyses. 7The completion rates for the different categories of patients were not significantly different. In sub-sample A, 43.4 percent of the low users and 50.0 percent of the frequent users were interviewed. The respective figures for sub-sample B were 51.3 percent and 46.4 percent. 129 130 8The remaining 8 percent of the cases were accounted for by patients who had moved out of the county or state and patients who had no phone and for whom an accurate address could not be located. 9No respondent asked to have this term defined. Although the head of the household usually is considered the person who earns the most money, in this study respondents defined the term themselves. Thus, if a woman identified her husband as the head of the household and she was employed and he was not, her designation was accepted. 10The characteristics of the network were measured only from the point of view of the respondent. Furthermore, information about these characteristics was derived from survey interviews not by direct observation. Mitchell (1969) considers the latter a more reliable and accurate method to collect such data. 11The measures of the density of social networks were developed by Professor John B. McKinlay, Department of Sociology, Boston University. Professor McKinlay sent me a copy of an interview schedule being prepared for a study of family planning among young married couples in the Boston area. The questions on social networks included in the instrument were lar ely drawn from an earlier exploratory study he had conducted IMcKinlay, 1973). I am grateful to Professor McKinlay for sharing his materials with me. 12The measures of psychological support used in this study were developed by Caplan, et a1. (1976). 13Answers to only three out of the four questions in the set were used to create an index of neighborliness: i.e., having a neighbor watch the children for a few hours if you were called away suddenly; getting a ride from a neighbor if you couldn't use other transportation; and, getting information or help from a neighbor if you couldn't do something yourself. 14If a respondent did not have children, her index was based on responses to only two questions. 15The "need for help with problems" variable has an inherent weakness: that is, it is based on the assumption that those who receive assistance with specified activities are satisfied with the amount of help they receive and do not feel a need for any additional assistance. This, of course, may not be a valid assumption. It is suggested, therefore, that researchers who might wish to use this set of questions ask respondents who receive functional help whether more support would have been helpful. 131 16An attempt was made to measure medical needs objectively by developing a "level of illness" index based on clinical evidence recorded in the medical chart, e.g., physical findings and results of diagnostic procedures. The attempt was not successful. See Appendix D for a description of the techniques used to develop this index and a discussion of the reasons for the failure. 17Symptoms were coded employing the classification system developed for use in the National Ambulatory Medical Care Survey (National Center for Health Statistics, 1974). 181 am indebted to Professor Zola for taking the time when he was on sabbatical leave to provide me with the nine categories of malfunctioning by which he classified presenting complaints. 19The three questions used to operationalize the need for catharsis were developed by Shuval, et a1. (1970). In a sense, these questions are a measure of the individual's perception of the "intensity" of her relationship with significant others. Intensity refers to the strength of the ties that bind individuals together, the willingness with which they are prepared to carry out the obliga- tions associated with these ties (Mitchell, 1969:27-28). Shuval and her colleague's questions, then, imply that if significant others are perceived as not honoring the obligation implied in their link with the individual, she will feel a need for catharsis. 20None of the patients were chronically-ill. Furthermore, even those who sought care primarily for acute illnesses were, from the viewpoint of the physician, not very sick. Thus, a response of fair or poor health was not congruent with the patient's clinically- defined health state. 21The sentence stems were developed by Ford, et a1. (1962). In addition, the coding techniques used were a modification of those suggested by these authors. That is, the coding of responses was based on the degree to which they expressed a positive or negative reaction to the subject matter. Both feeling tones and approach- avoidance tendencies were considered in assigning a code to a response. The coding was done by three people: (1) a graduate physician about to begin his residency training program; (2) a professor of sociology involved in the training of the residents; and (3) a graduate student in psychology. The agreement among these coders was 89 percent. Differences in coding were resolved through discussion. 132 22The question used to define the satisfaction of the need for catharsis was developed by Shuval, et a1. (1970). In their study in Israel, these researchers used three questions to operationalize this variable (1970:103-104). In a pilot study at the center, how- ever, it was found that only the one question used here succeeded in differentiating among residents; 95 and 75 percent of residents indicated that they did not feel they satisfied the need for catharsis in response to the first and third questions respectively developed by Shuval and her colleagues. 23The question used to define the satisfaction of the need to legitimate failure was developed by Shuval, et a1. (1970). In their study in Israel, these researchers used three questions to operation- alize this variable (1970:119-120). In a pilot study at the center, however, it was found that only the one question used here succeeded in differentiating among residents; 100 percent of the residents indicated that they did not feel they satisfied the need to legitimate failure in response to the second and third questions developed by Shuval and her colleagues. The American residents were classified as not satisfying the need because they felt they would attempt to deal with any under- lying psychosocial components of the problem rather than just treating the patient medically. Shuval, et al., however, felt that only those doctors who managed such patients with medical therapies exclusively, satisfied the need to legitimate failure. 24Physicians' responses to patients were measured not on the basis of strategies adopted by individual doctors across patients, but rather on the basis of the strategies adopted by one or more doctors for individual patients. That is, all measures were opera- tionalized as summary indices that were assigned to each patient in the sample, not to each physician. The discussion, then, focuses on the management strategies adopted by unidentified physicians to treat categories of patients. The small number of physicians in the sample in combination with the relatively small number of patients and relatively short observation period, made it impossible to analyze the responses of individual physicians to patients in a statistically, meaningful way. A qualitative analysis was made, however, of the responses of several physicians with different orientations toward the dependent patient in order to see if these were related to different management styles. 251 am grateful to Professors c. w. Given and 8. Given for developing the classificatory criteria used to code diagnoses. 26Diagnoses that indicated the purpose of the visit was health maintenance were not included in the total number of diagnoses. 133 27The technique assumes bivariate normal distribution, linearity and a set of interval-level variables. Several authors state, however, that the Pearson correlations can be used with "imperfect data" that do not meet these assumptions (Asher, 1976:26; Harris, 1975:7). 28The theory of multiple regression analysis assumes multi- variate normal distribution and equal variance-covariance within groups. In practice, however, the technique is very robust, that is, insensitive to any but gross departures from the assumptions and they need not be strongly adhered to (Harris, 1975:7). CHAPTER IV THE SETTING The social system approach adopted in this study views behavior of individuals in a medical situation in terms of phenomena beyond that immediate context. One, the behavior of patients is interpreted in terms of their life situations. Two, the behavior of physicians is interpreted in terms of their location within a medical profession. And three, the behavior of both is interpreted in terms of dominant cultural values. A key component of reality, however, is omitted in these three forms of interpretation: the setting within which the behavior occurs. This chapter, therefore, examines the site in which the study was conducted in order to provide a general social context for the material that follows. First, a brief history of the community is presented. Then, the medical facility in which the sample of patients seeks care is described. Finally, the structure of the family practice residency training program and the operation of the health center in which residents care for patients are examined. A Brief Social History of the Community_ The study was conducted in a large, mid-western industrial city that is the home of one of the major American manufacturing corporations. The city's economic life centers around the activities 134 135 of this corporation which primarily involve the manufacture of automobiles and related products. As a result, the community has been very sensitive to national and international economic cycles. The city's history, as that of all American cities, is characterized by a transition from the exploitation of natural resources to the development of industries. In the early nineteenth century, the area became a post for fur traders but as settlers started to move in, small enterprises were established to meet local demands for goods and services. By the 18505, the area had become an important lumbering center which attracted laborers as well as other small entrepreneurs producing wagons and carts. When logging ceased to be a profitable enterprise in the late 18005, many workers were able to find employment in the city's bur- geoning transportation trade. The destruction of forests made agri- culture an important sector in the state and farmers needed vehicles. The vehicle industry flourished as demands for its products grew and a reservoir of skilled craftsmen began to settle in the city as its reputation as an important employment center increased. With the advent of the motor car, however, the industry's viability became questionable and local entrepreneurs began to direct their energies toward the new and profitable automobile business. The first factory to manufacture autos was established in the city in 1903 and soon affiliated industries began to settle in the city as demands for direct access to supplies and control over their produc- tion grew. By 1909 a major corporation, organized locally, had assumed complete control and substantial ownership in twenty 136 automobile or accessory companies in the city in order to ensure stable profits. The subsequent growth and decline of the city's population has been a reflection of its almost complete dependence on this corporation and its enterprises. During the early years of this century, numbers of workers, many foreign born, were recruited into the industry and the city's population increased. In the l930$, however, the city's population decreased as a decline in auto pro- duction led to the unemployment of many workers. With the outbreak of World War 11, local factories ceased manufacturing civilian goods and became one of the principal sources of supply of war materials. As young men were drafted, women and unskilled workers moved in to replace them in the city's factories. The population of the city increased throughout the l940$ as a stream of migrants, many from the South, settled there to work. After the second World War the city's factories returned to automobile manufacturing and large numbers of workers were attracted there for employment.1 In the 19505, however, the inner city began to experience a decrease in population as people moved out into the burgeoning suburbs in the previously rural out-country. This popula- tion decline has continued in the last twenty years as unemployment-- resulting from national and international trends as well as the increasing mechanization of automobile plants--has impelled numbers of displaced workers to seek employment elsewhere.2 By the mid-19705 the slow, but steady, loss of inner-city residents had shifted the balance of population in the county and 137 the populations of the city and out-county were l74,218 and 275,386, respectively.3 Many of those left behind were the disadvantaged. More than twice as many people in the city lived below poverty level than in the out-county (16.9 and 6.8 percent, respectively). In addition, living below poverty level was part of a constellation of characteristics, i.e., poor, female heading a household, child living with one adult, and biack.4 In the next section the relationship of this population to that served by the center is considered and the ambulatory care center of Community Hospital is discussed. Community Hospital and its Family Health Care Center Community Hospital is a 423-bed, acute care facility in which all medical specialties are represented.5 Its major residency is based in the Family Practice Training Program which was organized in l97l to "improve the quality of care" at the hospital and, to train family physicians who, hopefully, would enter practice in the area or in the state. Although the program provides experiences in the management of hospitalized patients, its major objective is the training of family practice physicians who will provide primary care in an ambulatory setting. The Family Health Care Center occupies the entire third floor of the administration building which is located adjacent to the hospital. Its forty-six rooms are divided among practice "modules," that contain a consultation room with two to four adjacent examining and/or treatment rooms, and specialized rooms for performing ambulatory surgery, electrocardiography, sigmoidoscopy, cryotherapy, 138 physiotherapy and laboratory procedures. In addition, the center has a large central waiting room, contiguous to the room in which the reception desk is located and the medical records are filed, and four internal waiting areas. The Center Population A large proportion of the patients at the center live in the general location of Community Hospital, an area predominantly inhabited by people of the caucasian race, living on middle incomes in privately-owned homes (see Maps 4.1 and 4.2). The center tends to attract fewer patients from areas in which other hospitals and large concentrations of physicians are located.6 Admissions to the center are open and patients are not 7 During 1976, a population of 1,042 screened for teaching purposes. patients had regular contact with the center.8 Seven out of ten of these patients were women and more than half were twenty-five years of age and over (see Table 4.1). The higher proportion of adults in the population may be explained by the fact that many of these people had been patients of the director of the training program; when the center opened they followed him there. Records on race and income of patients are not maintained at the center. It was estimated, however, that about 60 percent of the patients were white. Furthermore, in 1976, 18 percent of the patient population lived in areas marked by unemployment and poverty and approximately 20 percent were covered by Medicaid. In the county as a whole, during this same period, 77 percent of the residents were 139 IEOULRI PRTIENTS (PER 1000 DHELLINO UNITS). 19784977 I 20.00 - 63.99 Ea 10.00 - 19.99 fig 7.00 - 9.99 E] 4.00 - 6.99 E] 0.00 - 3.99 Map 4.1.--City, Regular Patients per 1000 Dwelling Units, l976-1977.a aMaps of the city and out—county are produced separately for reasons of clarity. Census tract boundaries are used in these maps. The cross (+) represents the location of Community Hospital. 140 REWLEI PNIIENTS (PER 1000 DUELLIND UNIT”. 1878-1977 I 20.00 - 63.99 $10.00 - 19.99 mm 7.00 - 9.99 [3 4.00 - 6.99 E] 0.00 - 3.99 Map 4.2.--0ut-County, Regular Patients per 1000 Dwelling Units. l976-1977. 141 .ucmucmn n.n .cmopo new menu» me new mucougmq «.mp .mcmma qolmc mucmucma m.m~ .mcmm» eeumm mucmucma m.~P .mcmmz emu», mucoucmn m.mm .mcmma up cove: "mm: aucsoo pupa» mgu com cowuan_cumwu can maps o.oo_ «cop m.ao NNN 5.0m cum mmm< FP< qu El .3 Mm. mm mm. as 25 28> a. m.~_ NmF m.ap mu. «.m_ we mtam> em-me m.- mmm e.¢~ ma_ ¢.m_ No menu» we-m~ o.P~ m.~ e.m~ Pap a.» mm mama» ¢~-~F o.m~ Pam m.- .0. 9.4m o_. mama» a, tune: unmucmm smasaz acmucma gmassz acmugma gmnsaz mmmxmm :uom oposmm m—mz mm< xmm .oha, .xwm use wm< an cmgcmu menu supmmz xPPEmm mo mucmvuoa mo ucmucma ucm smasazuu.P.e u4mrmume ac: mmou u N .ueoaasm mm>wmume u p "ueoaaam we ucmpcoo mama xeo>m argue: n m .xmm: m mos?» 93M go one u w .mmswu «menu go o3» u m .muco u N .em>mc u — "panacea Fecomemg was mcogm eo aucmacmen "comma coogeongmwmcuco: u m .uwstlcopumuop u N .cwmmnlcoogeongmvmc n P "corumoou newcxummoop n N .upcxlmmopo n p " “emcee "mcvoou moo. v a .3. 8. v a i mp. «enm.- eNN.- mp. No.- peoaasm pmcowuucau $3.. 2.8.- 3.. 9:. 28%... P332263 "acmucou «euo. «wNN.- No.1 uuoucou pucomema mo xocmsamem 4N_.- No.1 pumucou mecca we zucmacmeu no. cowpeuOJ ueoagzm uumucou pueucou covuwqu zapmcmo —muvmopozuzma chomema «song we mo zucmzomem xucmscmem ovumvemuueemso $203va :3. we; .._.0 murumwkmuummeu cmgumm mCOwHMFwLLOu LmULOIOLwNII.m.m udmgr 177 that women with kin in the neighborhood were just as likely to be members of loose-knit networks as they were to be members of close- knit ones (47.6 and 52.4 percent, respectively). Living within walking distance, however, did encourage personal contact among relatives. Women with kin in the neighborhood met with them more often than those without relatives living in close proximity (r = -.27, p < .003). Furthermore, women with kin in the neighbor- hood had more frequent telephone contact with them than those who did not (r = -.17, p < .05). Two explanations for the correlation between the regularity of telephone and personal contact appear plausible. Propinquity may facilitate the establishment of closer ties between kin and thus bring about frequency of contact. Frequency of telephone contact may facilitate the transmission of invitations to visit or requests for assistance. Indeed, women who had more regular telephone com- munications with their relatives were more likely to receive functional support from them than those without such frequent contact (r = -.22, p < .01). Members of close-knit networks were more likely to receive psychological support from kin than were members of loose-knit net- works (r = .17, p < .05). There was no statistically significant relationship, however, between the receipt of psychological support from kin and the receipt of functional support from them. Forty-two percent of those who received psychological support did not receive functional support and 44 percent of those who received functional support did not receive psychological support. Furthermore, members 178 of close-knit networks tended to be slightly less likely to receive functional support from relatives than members of loose-knit net- works; that is, 50 percent of the women who belonged to close-knit networks received functional support from kin in comparison with 58 percent of those who belonged to loose-knit networks. This finding is difficult to explain. The data suggest, however, that the lack of relationship between density of network and receipt of functional support may be a reflection of the varying resources available to the women due to their different social positions. Those with more education were likely to be members of close-knit kin networks (r = -.l9, p < .04) and there was a slight tendency for the more educated to receive less functional support from kin. (Fifty percent of those who had graduated from high school or attended college received tangible assistance from rela- tives in comparison to 59 percent of those who had less than a high school education.) Furthermore, the more educated were likely to be living on adequate incomes and the latter also tended to be members of close-knit networks and not to receive functional support from kin. (Fifty-one percent of those living above poverty level received tangible assistance from relatives in comparison to 59 percent of those whose financial situation was less secure.) The lack of relationship between density of network and receipt of functional support, then, may be a consequence of differ- ing resources. Women who were members of close-knit networks were more advantaged, educationally and financially, and probably more able to locate and pay for the services of non-relatives if they 179 needed tangible assistance. Thus, they may have tended to report less functional aid from their kin. Social Assets: Friend Networks Table 5.9 presents data on the distribution of women by the type of friend networks in which they were enmeshed. The networks of most tended to be loose-knit and a majority of their friends were not located in the neighborhood. Approximately half of the > women were in daily phone contact with friends and two-thirds saw them at least weekly. Only three women had had no contact with friends during the month in which they were interviewed. Four- fifths received psychological support from friends, but only half as many received functional support from them. Location of friends had no statistically significant rela- tionship to frequency of contact with them, but frequency of tele- phone contact was correlated with personal contact (see Table 5.10). Women who talked with their friends regularly also met with them frequently (r = .57, p < .001): "A phone call usually means a visit." Regularity of personal contact, however, did not ensure the individual of functional support. Nevertheless, those who met with their friends nearly every day tended to receive more tangible assistance from them than those without such frequent contact (54.8 and 38.5 percent, respectively). Furthermore, women who had frequent telephone and personal contacts with friends were likely to receive both psychological and functional support from them (r = .20, p < .02). 180 TABLE 5.9.--Percent of Women by Characteristics of Friend Networks. Characteristic Number Percent Density Close-knit 41 33.1 Loose-knit 83 66.9 Location Neighborhood 25 20.2 Location-mixed 23 18.5 Non-neighborhood 76 61.3 Type of Contact Phone only 5 4.0 Personal only 4 3.2 Both phone and personal 112 90.3 No contact 3 2.4 Frequency of Phone Contact Nearly every day 61 49.2 1-2 times per week 33 26.6 2-3 times per week 16 12.9 Once 7 5.6 Never 7 5.6 Frequency of Personal Contact Nearly every day 30 24.2 1-2 times per week 55 44.4 2-3 times per week 25 21.0 Once 5 4.0 Never 8 6.5 Content of Relationship Psychological support 98 79. Functional support 52 41. 400 Location of Friends ProvidingpSupport Psychological Support: Neighborhood 25 25.5 Location—mixed 20 20.4 Non-neighborhood 53 54.1 Functional Support: Neighborhood 24 46.2 Location-mixed 6 11.5 Non-neighborhood 22 42.3 181 mo. v a .11 Noo. v a .1 ««ON. mo.- mo.- ao. vo.- pecaaam pmcovuucau «eo~.- «mN.- No.- No.- Seoaaam Pauemo_o;u»ma uucmpcou pumucou enm. mo.- «epN.- Pecomema mo absences; -.- kenp.- pumpcou mzoga we xocmscmem ewN. cowumuod ueognsm uuepcoo pueucou :owuaqu auwmcmo Pmuwmopoguxmm pmcomema econ; mo mo xucmaomeu Aucmacmem ovumwemuueecsu .xkgumz 0:073.- 0—3. .+0 mowumewuumLmsu cmgumm mcowumpmshou L00LOIOL¢NII.O_..m u._m<._. 182 Members of close-knit networks were likely to have neighborhood-based friends (r = .28, p < .002) and to be in regular telephone contact (r = -.l7, p < .05) and personal contact (r = -.21, p < .02) with them. Paradoxically, however, these women were no more likely to receive psychological and functional support from their friends than were members of loose-knit networks. Eighty percent of those who were members of close-knit networks and 78 percent of those who were members of loose-knit networks received psychological support from friends and 43 and 40 percent, respectively, received functional support from them. This finding is difficult to explain. No statistically significant relationships were found between social position and density of network or the receipt of psychological and functional support. Women with differing educational, financial and marital statuses were as likely to be members of close-knit networks as they were to be members of loose-knit ones. However, those who were single tended to receive more psychological (95.0 percent) and functional support (51.8 percent) from friends and to have more frequent contact with them than other women (see Table 5.11). Two explanations for the relationships between an unmarried status and receipt of support from and frequency of contact with friends appear plausible. First, the single may, in the process of asserting their independence from their family of procreation, establish more solidary relationships with friends. Thus, they may have frequent contacts and intense relationships with them that facilitate the provision of support. Second, the single may 183 p.5m m.o~ n.P~ ¢.mm m.Nm o.Nu o.ow hpxmuz ummmp p< "pumucou pmcomema No Nucmacmeu o.pm m.mo ¢.au o.mm e.pn ¢.wm o.mm apxmm: pmmwp u< “puepcoo mcocmupmk we mucosameu ugoaaam m.o¢ N.m¢ o.e¢ ¢.¢m o.mm m.mm w.Pm Pecowuucse mm>wmuom «Leanna o.Pm o.mN o.mm F.mN N.mm N.mo o.mm Pmquo_o;uxma mm>Pmuma uncom>mcwm m>vueoaaam mumzcmu< —m=rmenz Lm>o ace paceum ompnaemwo voweeez mpmcmm vcm :OFumaumew sup: emergent mumsamumcfi Poocum gap: cash mmwg xeozumz azure; oeoucu cowumuzuu maumum paupem: mg» ea muwumwemuumemgu coepemoa paeuom .cowurmom pmwuom mo mmgsmmmz an xeozumz ucmwgm mgu $0 muwumwemuumemnonn.P—.m m4mwwume mos?» mo eonssc ..m.Pv ueoansm pmuwmoposuamm we Nuwpcumc mxmu xem>m apemm: . m .xmm: m mmewu oz» so «so u e .mmsva omegp co o3» u m .muco .mmswu egos N .Fpm an no: n — "Azom>emusr uuumucou we aucmacmem umcwuou mo. v a .11 moo. v a .1 «emN. «eNN. m_. N_. mcvucoumemucz cmzogm eNm. 9mm. mp. co. vomoeaoucm «emN. «mN. mo. No. xppammemu vacuumwg eeON. «eoN. «NN. NF. vopnaoeh can: mmmcppucmweu uueucoo pumucou uueucou pumucou chomemm econ; Pacomema ozone mucmweu =_x ueoaaam Pmowmopoguzma No “coucoo pumacoo eo oaah use xucmacmeu .ueoaqam qupmopogoamm eo apwucmzo use mvcmvem use :wx xumz Homucoo we Nucmzameu cmmzuwm meowumpmeeou coveoloemN--.N—.m N4moea emspo on» we so“ xgw>m mo use mcwze mewmame mmao; m.m_ m m.nm F_ m.oN m use copucueonmcmep saw: QFm: mew—noea e.m_ m o._m mN m.om mp appsmw eo chomemg gar: ape: m.m¢ mom m.Np mp m.Nm eN menu vagu gov: QFm: m.~m PF m.m P m.mm NF memapms amcos sue; a_m: xppsmw cw m.m m m.mN eN m._o Nm mmmcppp can; mempnoeq new: ape: pcmuema smasaz ucwuema consaz ucmuema Lungs: emzuo mucmwcm ueoqqzm we ucmucou ugonasm mo emvw>oea .yeoaazm 0o emuw>oea ma ueogasm pmcowuucsm mo p:wu:ou--.mp.m u4mPumemaaou "Amcv>?P mo mmmem snow :wlmsmpnoea new: cmumm: upmsv Lao» o» Aumummc upm; ocv cemN sage mmmcme "msmpnoem cow: upm: ummz "mcwuoo po. v a 3. Bo. v a .1 mo.- mo. ~—.u oo.- FF. mcp>wm aprmo No mxmmp mp. «mm.- eo.- mm. po.- acmczum Np. «NN.- «NN.- mo.- m_. emxeoz mo.- No. «eNN. Po. «em.- mewzmmzo: ”mucmseomema mpom m>wumemaeou .SN.- No.- ««MN.- ,mm.- ,mN. "mem_noea sue: apmx ummz mezpuaeum zuw>_pm< cowumuzum meoucm magnum u_o;mm=oz Lone: Pmuwemz mezpwmm mo mmcppmmm copuwmom _mvuom .me=__mu mo mmcw—mmm use meowuwmoa pmwmom mo mmeammmz cmmzumm mcowumpmeeou emueouoLmNuu.ep.m mmm

wmume on: cmso: mnhm mmmppom emeemmm< new p.¢_ m_ N.e4 me P.mm em mmaaeaem Fooeom sue: cowumzcmgw m._m N m.em NP e.oe m_ Pooeum 5mm: can“ mam; coppmuaum e._F a ¢.m¢ mm ~.mm Fm amazeme< ~m=0memz o.m~ __ e.gm e, m.mm Re can moaacmeaem msoucH e.mm m_ _.NN m, m.mm m. emmaztmeo mmaeteaz ~.m 4 ..mm RN N.em mp em_eeaz _._P m m.mm a N.mm m_ mpmcem namepm Pemeeaz ucmuemm emnE:z ucmuemm emnssz ucmuema emnsaz npmz mummz upm: ummz poz mmoo mmsm_noem msmpnoem Pecomemm saw: upmz m>wmumm uoz mmoo Pecomemm saw: upm: mm>wmumm msmpnoem chomemm new: upm: Low nmmz coem_moa _aeuom .eoepemoa meuom mo mmesmmmz an mempnoea xpwsmu Lo Pecomemm guy: upm: mcwummz cmsoz mo u:muema--.np.m umm umcwmpaxm op acmuema mco ummmp um mcpuznrepcou mmpnmwem> .1 om_.m _oo. om.m m_. mm. P_. ewe om umeaaeou me__ pream m .NP.N Poo. 4m.N __. mm. op.- msouce :0 m=e>ep Napaueemeo N mvcmvem NNF.~ coo. mo.m No. NN. mm. 502m meoaaam Paco.mu==m F me a a we a «4m m_naeea> emceemucm mmpamvem> mpaempaz ea emeeo .zucmucmamo mo msouaszm mo cowmmmgmmm mmvznamum--.ON.m N4m

vmcoopoxm om ucmuemo mco ummmp um mcwuonweucou mmpnmweo>k Npp.o ooo. oN.o om. oo. m_.- ooooouoom o mnemoem op oFF.m ooo. No.o NN. No. N_. ooeoosou mew. N__som o op_.o ooo. es.“ Fm. oo. o_. mem_ooeo our: opmo toe oooz o oe_P Npeoo mo momma oop.m ooo. oo.o op. mo. Fm. mo moooeeoeeoo oseooeooaoo m FNP.N ooo. oo.op o_. om. MN.- o_x zoom oeoooom Fooo_ooooa N msooco NNP.F ooo. Np.op NF. om. oN.- oo ooe>,_ »o_ooomm.o P we a N Nm N «am mpnmwem> «mcpemucm mmpnoweo> opo.opoz mo emoeo .meapomu goo: moou on ummz mo :oommmemma mmvzuomum--.PN.m m4m

vmcwmpaxm o» ucmucmo mco mmmmp um mcopanpeucou mmpomwem>a _N_.N ooo. oN.o oo. oo. op. owe eo ooeooeou mew, apogee N N~_.F moo. Fm.o No. No. MN. weep mo oooom Foeoomo P to o a mm o 99o mpoaoeo> «oooeooom mmpnmwem> m_ooo~oz mo emoeo .mmemzuwu L0$ 0002 $0 commmwsmmm 0m wIIamumll.NN.m MAmgP 231 Discussion: Tests of Hypotheses About the Relationship Between the Women's Life Situations and Their Characteristics as Patients A woman's sense of well-being about her life circumstances exerted an effect on her medical and non-medical needs. Perceived financial and familial instability were consistently good explanatory variables. The psychological and social assets available also exerted an effect on the woman's needs as a patient. For example, the lack of functional support from friends was a good predictor of "symptoms of dependency." By contrast, the receipt of functional support in combination with a poor assessment of overall performance of tasks were good predictors of the need to cope with failure. Thus, the data supported some, but not all, of the hypotheses. H1. Individuals who occupy a low social position will have p_peed to cope with failure and present multiple diffuse_§ymptoms. The data provide little support for this hypothesis. Social position contributed nothing to the explained variance in "symptoms of dependency." Furthermore, only education contributed to the explained variance in the need to cope with failure. But, this con- tribution added only two percent to that explained by the five variables previously entered into the regression equation. It appears, then, that the contribution of social position, when used in conjunction with other individual variables, is minimal. These findings most probably derive from the close associa- tion between social position and perceived life quality and psycho- logical and social assets. That is, those who occupied the lowest 232 social position had the least sense of well-being about their circum- stances and limited psychological assets. Furthermore, the poor, less educated and separated or divorced tended to receive more functional support from kin than their more advantaged counterparts. Thus, social position most likely was unable to contribute appreciably to the variance already explained by contextual variables with which it was closely associated. H3. Individuals who perceive the quality of their life as poor will have a need to cope with failure and present multiple diffuse symptoms. The data provide strong support for this hypothe- sis. Perceived financial instability was a most powerful explanatory variable, followed in strength by feelings of poor family function- ing. Both variables accounted for more than half of the variance explained in the need to cope with failure and almost half of that explained in "symptoms of dependency." The efficacy of these variables most probably derives from the interdependency that exists between the health of the family as a functioning unit and the health of the individual. Since the family is the unit of "living," the degree to which it is socially or financially disadvantaged implicitly affects the individual's health as well as her perception of it. Thus, the strong contribu- tion of perceived financial and familial instability to variance in medical and non-medical needs is not surprising. 233 H4. Individuals who sense feelings of failure will have a need to cope with failure and present multiple diffuse symptoms. The data provide only partial support for this hypothesis. Feelings of failure were not a significant predictor of "symptoms of depen- dency" and thus were not included in that regression analysis. They were, however, a significant predictor of the need to cope with failure and were entered into the regression equation. Together, a poor assessment of overall performance of tasks and a need for help with problems accounted for one-quarter of the variance explained in the need to cope with failure. The emergence of feelings of failure as an explanatory vari- able in the analysis is not surprising. Past research (Beresford, et al., 1977; Cole and Lejeune, 1972) has shown them to be good pre- dictors of need. Furthermore, in a society such as America, where the values of achievement and self-reliance receive strong emphasis, it is not surprising that individuals who feel they do not meet society's definition of success, would perceive their health as poor in order to rationalize their self-defined failure. In summary, variations in the personal environment exerted considerable effects on the medical and non-medical needs of the patient. Not all elements of individual life situations contributed to explained variance across needs. Nevertheless, the data provided support for the hypothesis that patient needs are a product of experiential factors and perceptions of them. In the next section, attention turns to the expectations the women bring to the medical 234 encounter and to the way in which these are influenced by varying life situations. Patient Orientations to the Physician- Patient Relationship Descriptions of the physician-patient relationship offered in Chapter I emphasized its asymmetry, the explicitness of the norms governing the behavior of the participants and the complementarity of their expectations. This section presents the findings on the women's views on this relationship.12 The Physician's Functions The physician gives care to sick people by diagnosing and treating their problems. The women considered these institution- alized role obligations of the physician and described their expectations in predictive as well as normative terms. That is, the physician wili_diagnose problems and the physician §p921g_ ‘3 On occasion, however, these expectations were diagnose problems. violated, giving rise to concern and/or vexation. Women went to the physician for relief of their symptoms. But, they also were concerned with having their condition diagnosed and labeled.14 For some, the physician's inability to diagnose the cause of their problem was a source of fear. The patients had tentatively defined their problem as a serious disease, and although they hoped this was not the case, most strongly feared that it was. The relationship between uncertainty and fear was articulated well by one woman who said: "Uncertainty seems to be the most damaging thing because people then imagine the worst." 235 For others, the physician's inability to name the etiology of their symptoms was a source of frustration and irritation. The patients tended to feel that the doctor was interested, not in the cause of their problem, but in their symptoms. The following excerpts illustrate the emotions aroused when the doctor was unable to fulfill expectations associated with diagnosis: I want to know the cause. There's a cause for every effect . . . . I'm so uncomfortable and that makes me irritable. It bothers me. I'm so frustrated. I get no answer to the cause, just treatment. Doctors should not push pills. They shouldn't just write a prescription and say, 'This will take care of it.' They should get to the real cause. It seems as if there's got to be something other than pills to make you feel good. He doesn't really know so he says, 'These pills will help your symptoms.‘ I really want to know so I'll know there's something to work with! The women, then, had specific expectations of the physician role. When the physician did not fulfill these role expectations, they were inclined to lose faith in him. But the women also had less formalized expectations of the physician. When he performed these tasks, he earned the appelation, "good doctor." The "Good Doctor." The women were asked the question: "What makes a good doctor?" Table 5.23 shows that a majority mentioned attributes which characterize the physician's ability to show a personal concern for the patient. Only one-fourth of the character- istics relate to the physician's technical knowledge or skill or to the breadth of his authority. These data support those of Friedson (1961), Ben-Sira (1976), King (1962), and K005 (1955), which indicate that the patient is more interested in the affective than the 236 TABLE 5.23.--Characteristics of the "Good Doctor." Characteristic Eggggggegf Takes his time and is friendly 24 Communicates 19 Respects the patient 18 Devoted to people, not personal enrichmenta 12 Inspires trust and confidence 8 Concerned with preventive as well as curative medicine 8 Skilled and knowledgeable 8 Admits his fallibility and is comfortable with referring 4 aThis characteristic is similar to that which Parsons (1951) had in mind when he used the phrase collectivity orienta- tion; the physician sets aside his immediate interests and needs, as well as other considerations, that might conflict with his acting in the patient's best interest. Thus, the doctor does not capitalize on the patient's situation to his own financial advantage. 237 instrumental component of the physician-patient interaction. More- over, these data contrast the belief of the director of medical education at the family health center who noted: Patients are really not interested in a nice kind of doctor. They would prefer to feel within them that he's very knowledgeable and intelligent.15 According to the women, a "good doctor" takes his time with patients and is friendly. He conveys his willingness to take time by his physical behavior and his friendliness by his verbal behavior. For example, the women characterized these ideal behaviors as: He doesn't move fast. He sits down. If a doctor stands up I don't say much because he looks like he's in a hurry. He looks me right in the face when I talk to him. He doesn't make me feel like I was taking his time. He doesn't just come in and say, 'What's the problem?‘ He takes time to put you at your ease. He asks, 'How are you feeling? How's everything else?‘ Then, after friend- ship he'll ask, 'Why are you here? What's the problem?‘ Thus, the physician's willingness to extend the encounter, coupled with a conversational manner, is labeled as a personal physician- patient relationship. In addition, the women also felt that a "good doctor" respects his patients. He conveys this respect by "listening" and responding to what they say and by empathizing with them and their problems. An empathic physician is one who is sensitive to the needs and values of others: for example, he ". . . tries to put himself in my position like how he would feel." Furthermore, he is sensi— tive to what the patient is saying and does not dismiss her 238 complaints as products of her imagination.16 The following excerpts explicate the meanings women attached to these attributes: He couldn't see my pain but I was hurting . . . . No matter what I said, he wasn't listening . . . . A good doctor listens fully to what you have to say and attempts to do something about it. Sometimes you hurt but they can't believe you hurt because they can't find anything wrong. One time the doctor was doggarned irritated. He was mad at me because he didn't know why I was hurting. He made me feel stupid for being there. These women's problems may have been induced by the mind, rather than by a pathogen. Nevertheless, they were ailments that caused real pain and were undeserving of the contempt with which they were met. The physician's failure to "hear" what the patient is saying and/or to explore alternative ways to help her, then, impede the therapeutic relationship. The women also thought that a "good doctor“ communicates with the patient by giving understandable explanations: I like them to explain so I can understand. He can give me a big long word but then talk to me so I can understand. I have a right to know. It's my body. If you don't know what's happening you're scared. You need to cooperate for effective treatments. The treatment might do you good, but you might not think it was doing good. A person is scared when she goes to the doctor and needs to get over it. As other studies have shown, when a patient is kept in ignorance, she is excluded from being a partner in the therapeutic process and a dearth of knowledge makes a less "compliant" patient (David, 1968: Rosenstock, 1975; Sapolsky, 1965).]7 239 Finally, the women felt that a "good doctor" inspires trust and confidence in the patient: "He has the responsibility to see that the patient has trust in him." He "earns" this trust by indi- cating that he is interested in the person as a whole, not just her body. When the physician establishes a relationship of trust and understanding with the patient, she is more likely to cooperate with him in treatment without fear or skepticism. In summary, the women saw a distinction between the physician role and the "good doctor." That is, diagnosis and treatment were behaviors required of the physician. By contrast, personalized care was a behavior desired of the physician. Thus, the "good doctor" was defined as one who (1) was affective; (2) showed a relaxed manner: (3) provided health education; and (4) inspired trust in the patient. The Breadth of the Physician's Authority Almost all of the women (88 percent) believed that the physician should be responsible for socioemotional as well as for 18 Among these women, 60 percent felt that the physical problems. physician should accept this responsibility himself, rather than referring patients to another professional. Most women, however, said that they did not spontaneously initiate a discussion with the physician of non-medical problems. Instead, they waited for the physician to create a climate in which they felt free to divulge their concerns. The following excerpt loLoos so maun noux plnous uoioisfiud aun noun panailaq uamoM aun ‘smalqoud uiaun noaun puo asoufioip Alaninoanna on uapuo u; ‘nuous uI '4941939 40 1315 59139 911M 0P 01 sou noun no nsow 'sfiuiun nnoqo lean puo uuiun noK mou uo dsoufi o aAou puo Kuonstu [limo] unofi noun plnous uonoop aul 'nuainod aun no unloau louauafi aun u; nuod 6Lq o [old nuamuoninua puo alxns 6u;n;1 'Jau uo 6uifiuidwi sassauns aun puonsuapun on Japuo u; nuamuouinua s,nuainod aun no 6uipuonsuapun p006 o anou J0130p aun noun nuonuodw; S; n1 :panoool auaM Kaun uoiuM u; nuamuouinua aun no afipalnoun o anou uoioisfiud aun noun nuonuodm; Kuan son n; noun nla; uamom aun ‘uosoau slu1 Jog z°amin aun [lo sauoopoau anou Kaun uo nunu suoowons Jigun paiuuom an,£aun ;; ‘spuaiun Km unIM paoinou an,1 'alqouaulnn noK saxow puo Kfiuaua unox sdos Kano" 'XPIS 39 ll!“ [PDQ 391 ‘X315 51 P01” 991 JI :ssaull; agnowos u; glasn; snsaniuom [inuanbaun ssaunsip lootfiol -ouo£sd noun paziufiooau uawom aun ‘spuon uauno u1 °smalqoud Kliwon JO [ontuow puo suuaouoo [enououin "6°a ‘suotnonnts aJll uiaun unIM paugmnua Klqooiunxau; auaM swalqoad unloau noun paziufiooau uamoM aun no [lo ‘nai °suoinonoadxa s,uoio;sfiud aun on puodsauuoo plflOM n; noun [on o uons u; alou nuetnod JLaun Buiuinap auaM uawoM aun ‘snul 'dlau panqu Kaun uoiuM Jon swalqoud [ouognoma 61 -oLoos pou Kuom noun snatnuanu; no snas unoq u; nuapina soM n; ‘swalqoud [ooisxud uniM one: no; panuasaud uawom aun u6noun1v '6uixlon nuons Allooinowonno 1 as DUOJM s.nouM aw suso aH '6uoun s,6u;unamos asuas on swaas uonoop Kw :suuaouoo uans Buissaudxa Jon panuoufi uoissimuad puo panoauo soM anowtlo SlWl UGLUM u; JBUUPW aun saqiuosap 092 241 beings located in a social and economic environment. All believed that such information was important in the diagnosis of both somatic and socioemotional problems. Variations in Definitions of the BhySiEian's‘Functions Twelve percent of the women interviewed did not believe that the doctor should be responsible for the care of socioemotional problems. These women were not dependent patients, as most of those who held the opposing view were. They were patients who had sought care during the two six-month periods for acute somatic problems only. In order to explore the representativeness of these women's different beliefs, quantitative data obtained from the total sample was examined. The survey interview concluded with a question asking the women: "If you had a half-hour during which you could talk to your doctor about anything you wanted, what would you talk about?" The women's answers had been coded into three categories: (1) personal problems; (2) other; and (3) don't know. Those who mentioned personal problems (48 percent) mentioned problems such as marital or sexual difficulties, specific problems with children, and other interpersonal problems. Those who mentioned other problems (38 percent) mentioned subjects such as special health problems and general health information about menopause, pregnancy or child rearing. It was hypothesized that women whose life situations were different would wish to discuss dissimilar subjects with the 242 physicians and that their desires would reflect varying orientations to the breadth of the physician's authority. The data provided support for this hypothesis. For example, women who had experienced more stressful events during the year in which they were interviewed were likely to want to communicate personal problems to the doctor (r = -.26, p < .003). Furthermore, the desire to talk about personal problems with the physician was particularly strong for those who reported sexual problems (r = .21, p < .02), marriage trouble (r .30, p < .01) and trouble with children (r = .24, p < .02). Women who felt that the general state of their life was troubled or who reported a need for help with personal or family problems also were likely to want to talk with the doctor about these difficulties (r = -.27, p < .002 and r = .39, p < .001, respectively). Similarly, those who had a need for catharsis and a need to cope with failure expressed a desire to talk about personal problems with the physician (r = .27, p < .003 and r = -.18, p < .04, respectively). In addition, although "symptoms of dependency" were not significantly related to type of subject mentioned, 61 percent of those who presented an above the average number of "symptoms of dependency" wanted to talk with the physician about personal problems. In short, those patients with greater non-medical needs and those women whose life situations were difficult were likely to want to discuss personal problems with the physician. On the basis of these surrogate data, then, it was concluded that dependent patients are likely to consider personal problems apposite to the 243 physician-patient interaction. By contrast, non-dependent patients are likely to believe the physician's authority is limited to bio- physiological problems. The Patient's Functions A patient is any individual who is sick, needs help, and wants to get well.21 Individuals know they are sick and require pro- fessional help when: (1) they are in severe pain: (2) their com- plaints persist; (3) their symptoms interfere with normal function- ing; and (4) they are unable to cope with their problems themselves, i.e., home remedies are not efficacious in relieving symptoms.22 In discussing the meaning of illness and their response to it, the women frequently spoke of the need to triage their com- plaints, to accept responsibility for the care of minor problems. Self-care for "little aches and pains" was necessary because the physician ". . . is very busy and is there to take care of sick people." Sick people almost always were defined as individuals with organic illness. Yet, as indicated above, there was ample evidence in the interviews to suggest that the women felt that socioemotional illness also was within the breadth of the physician's authority. Furthermore, there was evidence to suggest that the women held beliefs about the role obligations of the patient and if fulfilled, she was a "good patient." The "Good Patient." Table 5.24 presents the women's responses to the question: "What makes a good patient?" The data 244 TABLE 5.24.--Characteristics of the "Good Patient." Characteristic Percent of Responses Follows the doctor's advice 30 Communicates honestly 25 Pays her bills 17 Accommodates the physician's busy schedule 16 Acknowledges the physician's superior knowledge 12 show that the women defined the "good patient" in much the same way that social scientists and practicing physicians do, that is, as a cooperator with a guiding partner (Ford, et al., 1967: Lorber, 1975: Ort, et al., 1964; Szasz and Hollender, 1975). According to the women, a "good patient" is an individual who facilitates the doctor's work by communicating her symptoms to him frankly and honestly (25 percent). The following example illustrates the way in which the women felt honesty facilitates the therapeutic relationship: Tell him your problem. Don't keep him guessing if you have an idea what's wrong. If you have a problem with your boyfriend or kids it could make you sick to your stomach. He could be giving you pills for nausea but if he knew you were nervous he could look another way at it. In addition, the women thought that a "good patient" cooperates with the physician by accepting and following his advice (30 percent). The physician is more able to cure if the patient complies with his instructions, as the following excerpt illustrates. 245 A doctor is there to help people. If the patient comes in with a problem they should work on it together. Some pe0ple sit back and expect a cure without any effort. The patient should accept her responsibilities and follow the doctor's orders. Thus, not only does the good patient cooperate with the physician, but she also acknowledges that the doctor is in control (see also Ford, et al., 1967; Ort, et al., 1964; Parsons, 1951; Szasz and Hollender, 1975). She trusts and relies on his judgment because he has special skill and knowledge. Responses such as the following indicate the degree to which the "good patient“ accepts a subordinate role in the therapeutic relationship: They're the docs -- they know. She listens to the doctor and goes with whatever he says . . . . He knows. Accept what's wrong with you and trust that your doctor will take care of you. The women also thought that a "good patient” is punctual, avoids cancelling or breaking appointments and acknowledges that the physician's time is restricted (16 percent).23 Finally, the women felt that a "good patient" acknowledges the contractual nature of the physician-patient relationship and pays her bills (17 percent). In short, the elements of a "good patient" seemed to be the follow- ing in terms of importance: (1) follows the physician's advice; (2) communicates honestly: (3) fulfills her financial obligations; and (4) acknowledges the physician's superiority. 246 Summary: The Physician-Patient Relationship The physician-patient relationship has been defined in the sociological literature as a dyadic system in which interaction is patterned by the congruence of the norms for both actors. Further- more, the interaction is defined as functionally specific, i.e., oriented toward bi0physiological phenomena, and affectively neutral. As observed above, the women's definitions of the roles of the physician and the patient partially supported this definition. There was ample evidence in the interviews to suggest that the women considered the relationship asymmetrical but desired it to take a symmetrical character. Their definition of the patient role indicated that its adoption involved a loss of self-determination. In addition, their characterization of the physician's actual role also suggested that the physician exerted social control in the interaction. For example, when they spoke about physicians using "big words" that made medicine incomprehensible to them, the women in an oblique way had identified the strategy used by physicians to assert such control. By shrouding their work in mystery, physicians conveyed to patients the message that they possessed knowledge unreachable by laymen, thereby not only maintaining but reaffirming 24 Similarly, the women's descriptions of the unequal relationship. the physicians' approach to problems that they were unable to diagnose implied a loss of control by participants who considered themselves more powerful. That is, according to physicians, their resources are superior to patients, a circumstance that should 247 enable them to discern the cause of a problem and institute effective treatment. Thus, when a physician could not exercise control, he became angry with the patient, making her "feel stupid . . . like he had better things to do with his time." There also was evidence in the interviews to suggest that the women had internalized the view that medicine is oriented toward "the application of scientific knowledge to the problems of illness and health, to the control of 'disease'" (Parsons, 1951: 432). They explicitly stated that the diagnosis and treatment of physical problems was an institutionalized role obligation of the physician and became angry when these expectations were not fulfilled. Furthermore, although most felt that socioemotional problems consti- tuted a legitimate condition for which to seek care, the fact that so few spontaneously presented their concerns implies that they viewed non-somatic problems as illegitimate. Similarly, the data show that the women had internalized the norm that defines the physician's role as affectively neutral. Personalized care, including expressions of human warmth and interest, was a desired behavior, not a required behavior. Thus, the physician who responded to the patient in an expressive manner earned the appelation "good doctor." Data from the interviews also document that certain expecta- tions guided the women's definitions of their role as patients. Their descriptions of both their specific and general role obliga- tions suggested that they had internalized the traditional belief that the "ideal" patient is: 248 . . someone who is able to assess symptomatology with sufficient expertise to know which conditions he should present, and when he should present them to the GP, but at the same time one who, having assessed . . . his con- dition, will defer to the doctor's assessment on presentation (Bloor and Horobin, 1975:276, emphases in original; see also Parsons, 1951:441 and 1964:275). Furthermore, the data suggest that the women felt they could be negatively sanctioned if they did not adhere to the physicians' norms for behavior. For example, a "good patient" pays her bills: "If you want him [the physician] to do anything next time, you have to pay your bills." Moreover, she fits the medical model: "Some doctors if they don't know what's going on they don't want to deal with it." Finally, there were indications, both in the in—depth and survey interviews, that orientations to the physician-patient relationship were modified by different life situations. Dependent patients were likely to define the physician's functions broadly, extending his authority to socioemotional phenomena. Non-dependent patients were likely to define his functions narrowly, limiting his authority to somatic problems only. The findings, then, support the view that varying need-dispositions derive from the different con- texts in which the patient qua individual is located. In the final section, attention turns to the way in which these differing need- dispositions are related to the total use of medical services. 249 The Patient's Contribution to the Total Use of Medical Services It was hypothesized in Chapter II that patients with greater non-medical needs would use more medical services in the presence of multiple diffuse symptoms (see hypothesis six). In order to test this hypothesis, a step-wise regression of the total use of medical services during the two six-month periods of observation was carried out. Three variables were entered into the equation: (1) "symptoms of dependency"; (2) need to cope with failure; and (3) need for catharsis. The results of this analysis are presented below. Test of Hypothesis Inspection of the data indicates that the medical and non- medical needs of the patient exerted an effect on the total use of services (see Table 5.25). Those who had a high volume of use during the two six-month periods presented many "symptoms of dependency" (B = .24) and had a need to c0pe with failure (8 = .31) and a need for catharsis (B = .14). Together these variables accounted for 16 percent of the variance in the use of services. More than half of the variance was explained by the first variable entered into the equation. That is, the need to cope with failure was the most powerful explanatory variable (R2 = .09). Many "symptoms of dependency" explained an additional six percent of the variance. But, the need for catharsis contributed only one more percent to explained variance. The inability of need for catharsis to account for more explained variance is difficult to unravel. The relationship between 250 .ugmmmz mums Lo acmNUPmmmou :owmmmemme umNPueoucmum n m .11 .umueogme moo mucmwem> umcwmpoxm op ucmuemn mco «momp um acousaFeucou mmpamoem> k. oNF.o ooo. oo.N op. oo. o_. m_meooooo eom oooz N _N_.N ooo. om.op mp. om. oN. Nooooooooo mo meooosam N NN_.P Poo. mo._F oo. oN. Pm. oeopooe ope; oooo oo oooz P mo o o No N woo opoooeoo eoooeooom mmw o_o_opoz ewoewmww .mmo_>emm Fooeoez mo omo _oooe moo oo oooooooeoooo m.oooooea one mo ooemmmeooo omoz-ooom--.oN.m moooe 251 need and use is virtually a truism in the literature and Shuval and her colleagues (1970) did demonstrate a relationship between need for catharsis and high rates of utilization.25 Perhaps, as suggested earlier, the need to communicate feelings of distress is greatest for those who also are burdened with the need to cope with feelings of failure. Thus, because a high need for catharsis was closely associated with the need to cope with failure, it was unable to con- tribute more to explained variance. The efficacy of the need to cope with failure and "symptoms of dependency" as predictors of use of services is not surprising. Self-defined poor health by implication suggests a concomitant self- defined need for medical care. Furthermore, if, as Cole and Lejeune (1972:355) have suggested, ". . . people who define their health as poor are more likely to search for symptoms to justify their taking the sick role," the frequent use of medical services for large numbers of symptoms that are difficult to deny appears a plausible eventuality. Although the amount of variance explained in total use of services was not great, the findings represent an important attempt to test a set of axioms that pervade the literature. They provide empirical support for the view that health-related behavior may be strongly influenced by motives that are not health related. Further- more, they provide support for the view argued in this report that utilization behavior may be an adjustment to a general life situa- tion. When opportunities for the satisfaction of basic social needs are blocked or not fully available in the personal environment, the 252 individual with limited resources is likely to attempt to satisfy these needs through the medical institution. Summary: The Women as Individuals and’Patients In the foregoing pages, the life situations and need- dispositions of the study sample have been examined. Approximately one-third of the sample was poor, relatively uneducated, separated or divorced. These women tended to have little sense of well-being about their circumstances and to have experienced a series of stressful events. Furthermore, these women tended to feel that their psychological assets were limited and their social assets only nominally effective in mediating between them and the consequences of life difficulties. It was shown that there was a marked association between these life situations and patient characteristics. Women who per- ceived the quality of their life as poor tended to have a need to cope with failure and a need for catharsis, as well as a tendency to present many "symptoms of dependency." Multiple regression analysis indicated that perceived financial and social deprivation exerted the most consistent influence on medical and non-medical needs. Psychological and social assets available, however, also exerted an effect on the women's needs as patients. Furthermore, it was shown that differing life situations were related to orientations toward the physician-patient relation- ship. An analysis of the qualitative interview data, and a more limited analysis of the quantitative data, indicated that dependent 253 patients were more likely to consider personal problems apposite to the physician-patient interaction than non-dependent patients. Finally, it was shown that the high use of medical services was influenced by the medical and non-medical needs of the patient. Multiple regression analysis indicated that the need to cope with failure and many "symptoms of dependency" were good predictors of the total use of medical services. The contribution of the need for catharsis to explained variance, however, was minimal. In the next chapter, attention turns to the way in which the response of the physician to patient needs influences patterns and volume of use of medical services. CHAPTER V: NOTES 1The following discussion is based on data collected from 124 women in the sample. Although 127 patients were interviewed, the data for three women were omitted from all analyses after information from their medical records was processed. One of these women had cancer in situ and the other two had sought care at the center for one or two routine checkups only. Thus, they did not meet the criteria for inclusion in the sample. Race was not included as a variable in any of the analyses because of the small number of blacks in the sample, i.e., 14 women. Fewer of the blacks were separated or divorced (28.6 percent vs. 40 percent for whites) but more were living on inadequate incomes (35.7 percent vs. 23.9 percent for whites). The distribution of blacks across education levels, however, was more even than it was for whites. Fifty percent of the blacks had at least a high school education and 50 percent a college education. The percentages for whites were 74.5 percent and 25.5 percent, respectively. 2Twenty-one women (16.9 percent) responded to the question, "Did anything else happen that I didn't mention?” Nine out of every ten reported the occurrence of an event with negative impact. Only two (9.5 percent) mentioned an event that connoted a positive impact on their lives, i.e., the start of a love affair. 3Since almost nine out of ten of the women had both telephone and personal contact with kin and friends, type of contact was not included as a variable in any of the analyses (see Table 5.7 and 5.9). 4Although the relationships were not statistically significant, members of close-knit friend networks also tended to report receiving most types of support more frequently: friendliness when troubled (r = -.05); listened carefully (r = -.O4); and encouraged (r = -.O6). The only item which did not follow this pattern was "showed under- standing" (r = .10). 5A Specific question was not asked about these latter two activities, however. Reports of such help were elicited by a question that asked if the respondent received help from outside the household with anything else besides the activities mentioned. Perhaps, then, if the question had been asked directly, the distributions would have been different. 254 255 6No relationship was found between the women's assessment of their role performance and their social assets. The discussion, therefore, focuses on the relationship between feelings of failure that arise from a need for help with problems and the social support systems of the women. 7None of the relationships between measures of quality of life or social stressors and supportive behaviors of kin and friends were statistically significant. 8A direct comparison of the volume of use of the two groups was difficult because of the way in which the country data were recorded. Six to fourteen visits and fifteen or more visits were treated as categorical variables in the county study and it was impossible to analyze how many women had six, fourteen or "n“ number of visits during the year. Therefore, in order to calculate the averages reported: (1) all persons with fifteen or more visits were excluded; and (2) the median number of visits for those reporting six to fourteen visits was used in calculating the average per community resident. 9Visits ranged from one (14.5 percent) to ten (0.8 percent), with 3.4 visits the average per patient. Seventy-six patients (61.3 percent) visited the center between one to three times and 48 (38.7 percent), four or more times. The figures for the sample patients interviewed are comparable to those reported in Chapter III, Table 3.4. The slight differences noted are due to the distinct sources from which the data were obtained. The material used in this chapter was derived from patient medical records: that in Chapter III, from data contained in the center's health information system. 10Since the correlation between the various classifications of symptoms was so high, a reliability coefficient was computed. The results of this computation yielded an alpha of .89. Given this high coefficient, an index of symptoms of dependency was calculated and assigned to each patient. This index was used in all analyses. nVariables were omitted from discussion if they did not meet one of two criteria. The first criterion omitted variables that were never entered into the regression equation because they did not meet three statistical criteria specified by the computer program (Nie, et al., 1975:345-346). The second criterion omitted variables that had been entered in to the equation, but had added less than one percent to the total variance explained. 256 12The following discussion is based on the responses of 25 women to questions asked in the in-depth interviews. Approximately three-quarters (72 percent) of these women were dependent patients. In part, the discussion relies on their responses to queries about the responsibilities of the doctor and the patient and the character- istics of a “good doctor" and "good patient." In part, it also relies on responses to other questions posed during the unstructured interview. 13T'wo out of every three of the doctor's responsibilities cited by the women were "diagnoses." In the survey interviews, less than ten percent of the women did not mention diagnosis when asked, "What did you want the doctor to do the last time you visited him or her?" Those who did not mention diagnosis answered that they wanted advice and/or reassurance. 14Reader, et a1. (1957) found that the major concern of most of the study patients seeking care at the New York Hospital-Cornell Medical Center was learning about the seriousness of their condition and a name for it. Less than one-third were not concerned with a diagnosis. In a similar vein, Ley and Spelmafi_(1965) noted a tendency in patients interviewed after a medical visit to recall the diagnosis better than the advice. 15The extent to which the director's belief receives emphasis in the training of residents is unknown. Given his beliefs, however, it is not surprising that a number of women felt a lack of personal care. 16The lack of male physicians' empathy for and sensitivity to female patients is discussed in Campbell, 1973: Cooperstock, 1971: Corea, 1978: Ehrenreich and English, 1973. 17Reader, et a1. (1957) found that the patients they studied also wanted explanations from the doctor but did little, if anything, to obtain information. They hypothesized that patients do not expect doctors to attempt to fulfill this need. The study data suggest, however, that patients frequently do not seek information because the doctor does not give them the opportunity (see Tagliacozzo and Mauksch, 1972, for a discussion of this point). 181n the survey questionnaire, 56 percent of the women said that they felt the doctor was a suitable person to talk with about personal problems. (Thirty-one percent responded that he was not suitable and 13 percent were uncertain or qualified a positive answer with remarks such as ". . . if you feel comfortable with him.") Sixty percent of the survey sample, moreover, indicated that they had 257 talked with their doctor about personal problems. Gurin, et a1. (1960) report that 30 percent of patients with acknowledged "mental problems" first sought the aid of a general practitioner. Such explicit use, documented by surveys, suggests definite expectations of the doctor to deal with the psychological and social problems of patients. 19Perhaps because they did not feel free to discuss situa- tional problems with the physician, only four (1.0 percent) out of the 422 visits made during the two six-month periods were solely for socioemotional problems. However, although only four visits were made solely for socioemotional problems, data in the medical records indicate that non-medical problems were discussed at 53 (12.6 percent) of the visits. The most frequently discussed problems recorded were marital problems (23.1 percent) followed by economic problems (18 percent) and parent-child conflict (15.4 percent). 20Although all of the women agreed that worry can make you sick, several indicated that they had been surprised when their physical problem had been interpreted for them in this way by their doctors. All also agreed that they would attempt to cope with mild depression by situational adjustments before seeking care from a physician. Responses such as the following were common: "Take the day off": "Do something I like with people I enjoy being with"; or "Occupy my mind and not dwell on it." 2‘The women's definition of the patient role closely parallels that formulated by Parsons (1951). Using physician expectations of patients as a reference, he states that the ill person has the obliga- tion to seek technically competent help and to try to get well as quickly as possible. 22See Baumann (1961) and Apple (1960) for material relative to the definition of "being sick." 23Many women talked about the long wait they must endure to see the doctor at scheduled appointments. Only two, however, suggested that their time also was valuable and that an effort should be made to contact them and apprise them of a delay. 24See Moore and Tumin (1949) for a discussion of this point: "Ignorance on the part of a customer of specialized services (for example, medical or legal advice) helps to preserve the privileged position of a specialized dispenser of these services" (1949:788-789). 258 25The findings of Shuval, et a1. (1970), however, are reported in frequency distributions only. It is difficult to assess. therefore, if the observed differences were tested for statistical significance. CHAPTER VI THE PHYSICIANS AND THE PHYSICIAN-PATIENT INTERACTION The focus of this study is the nature of patient and physician behavior and the effect of this behavior on the use of medical services. The material presented in the previous chapter indicated that differ- ences in the women's life situations influenced their need-dispositions as patients and their overall use of medical services. This chapter examines the influence of the medical profession on physicians' atti- tudes and the way in which the residents' definitions of disease and determination of treatment influence the patients' utilization of medical care. The chapter is divided into two parts. Part one presents the findings relevant to the physicians and explores the way in which role definitions and orientations toward the dependent patient are affected by the context of medical work. Part two examines the data concerning the physician-patient interaction as well as analyzes the way in which medical decisions affect the use of medical services. Part 1: The Physicians According to Parsons' (1951) formulation, the role of the physician is oriented toward the application of scientific knowledge to the control of somatic illness. The physicians in this study, 259 260 however, were training for careers in a medical specialty--family medicine--which has been defined by its proponents as one where the "doctor spends most of his time thinking about the patient and the impact of various forces [i.e., social and psychological] on his health or illness over a period of time" (Alpert and Charney, l973:3). Thus, departures by the residents from the expected norms of tradi- tional physician behavior would in fact be in line with the definition of their specialty. It was argued in Chapter II that departures from prototypic role definitions are a product of a host of factors experienced at the level of medical practice. Furthermore, it also was argued that differences in attitudes toward the dependent patient derive from such contextual variables. The findings apposite to this argument are presented below. Overview: The Physicians A majority (56.8 percent) of the 42 residents were graduates of academically-oriented medical schools. Thirty-six were males and six females.1 Their decision to choose a residency in family medicine, in large part, had been prompted by the variety in patient types and medical problems the specialty provided (see Table 6.1). This attrac- tion had drawn one-quarter of the residents into the field by default. They, as for example one of them said, ". . . liked all aspects of medicine to some degree, but no one specialty enough to spend a life- time in it." The residents' emphasis on personal, rather than humanitarian, reasons for choosing family medicine was reflected, in general, in 261 TABLE 6.1.--Physicians' Reasons for Choosing a Residency in Family Medicine.a Reason Percent of Responses Variety 44.9 By default 20.4 Personal affiliation with patients 16.3 Impact on pe0ple's health as a primary 14.3 care physician Need for physicians who treat the whole person 4.1 aThese reasons were offered in response to the question: "Why did you decide on a residency in family practice?" their answers to the question: "What do you enjoy most about being a doctor?" Approximately one-third (36.4 percent) of the responses emphasized the satisfactions derived from the technical aspects of the role, i.e., diagnosis, treatment, and completion related to therapeutic results, and one-fifth (20 percent) indicated the satis- factions derived from the professional position, i.e., independence, prestige, power, and financial rewards. Only one-third of the responses (34.5 percent) stressed the satisfactions derived from the help-giving or supportive elements of the role and one—tenth (9.1 percent) the satisfactions obtained from an "on-going" relationship with patients. Most of the residents' dissatisfactions with medicine derived from the work environment. Two-thirds (63.4 percent) of the responses 262 to the question “What do you like least about being a doctor" referred to night calls, hours of work, paper work, and lack of privacy or time for themselves and/or their families. A further one- fourth of the responses referred to dissatisfactions created by problem patients, i.e., "crocks" (21.9 percent) and potential liti- gants (4.9 percent). Finally, one-tenth (9.8 percent) referred to dissatisfactions that derived from the uncertainties involved in medicine and their attendant responsibilities. As one resident wrote: "The thing I like the least about medicine is the constant pressure to perform, produce, and do the right thing in a field where most aspects are fraught with uncertainties." In summary, the residents had chosen a specialty in which emphasis was placed on the total management of the patient's physical and socioemotional needs. Yet, only a few indicated that they had any interest or responsibility in this domain. We must, however, consider one limitation of this finding. The residents had been asked to express their views on medicine as a career rather than on the role obligations of a family practitioner. In the next section, attention turns to their views on the ideal behaviors expected of the physician and the patient. The Physician and the Patient Roles The physician interview schedule contained two sets of pair- comparisons (see Appendix C). One set contained six items relevant to the physician's functions. The other set contained six items apposite to the patient's responsibilities. Each set of six items 263 was presented to the residents in all possible pairs and they were asked to judge which one of the pair helped them to practice medicine effectively. The results of the analysis of their responses are presented below.2 The Physician's Functions Table 6.2 presents the data on the residents' judgments of the relative importance of the six physician functions which they were asked to compare. Inspection of these data indicates that the residents considered health education and the establishment of rapport their two most important functions and the greatest con- tributors to their effectiveness as physicians (scale value [sv] = .7735 and sv = .7297, respectively). Furthermore, the data show that the residents attached more significance to these two items than to items that referred to the technical aspects of their functions. For example, a greater distance separates the values given the first two items on the scale from those given the fourth item (insists on a thorough work-up, sv = .1726) and the sixth item (requires the patient to return for follow-up, sv = .0000). Thus, the residents judged affectivity, rather than instrumentality, more helpful in effective practice. The data also show that the residents attached more signifi- cance to the first two items on the scale-~health education and the establishment of rapport--than to items that referred to the breadth of their authority. For example, a great distance separates the values given the first two items from those given the third item 264 oooo. meo. man. mmmm. mama. mmnm. m4oo .N oN. o_. oN. No. No. oo. mooooooo co momeoeoooo ooo moooooxm .P m m e m N F :omuucam .muwaumeo Pmuwumz m>vummeem mo msemh cw vmmuow newcomxco ms» mo mcommucam xwm so» xweum: cowueoooeaiu.N.o mom<~ 265 (devotes time to social and emotional factors, sv = .3639) and the fifth item (focuses on the treatment of physical problems, sv = 0331). Furthermore, the values assigned to the third and fifth items suggest that the residents defined their role broadly. That is, they considered socioemotional problems within their purview and indicated that directing attention to such problems was far more important for effective practice than focusing on physical problems. As one resident noted elsewhere in the interview, dealing with social and emotional problems ". . . is a cure for more than 70 percent of my patients' complaints." In short, the elements of the physician's role the residents considered most helpful in terms of effective practice were, in order of importance: (1) health education; (2) establishment of rapport: (3) treatment of socioemotional problems: (4) diagnoses: (5) treatment of somatic problems; and (6) continuous care. On the basis of these rankings, then, the following conclusions seem warranted: First of all, the residents identified themselves as practitioners, not research scientists. Second, as practitioners, they considered their primary obligations to be the solution of patient problems, not technical problems. Finally, they thought that the success with which they solved these problems principally was a function of the mode of their response to patients, not its content. 266 The Patient's Responsibilities Table 6.3 presents data on the residents' judgments of the relative importance of the six patient functions they were asked to compare. The data indicate that the residents considered the first four items in the scale relatively inseparable and attached almost equal importance to them as contributors to effective practice. A very short distance separates the four items from each other: (1) gives an honest and accurate history (sv = 1.2871): (2) is self- 1.2165): (3) does what reliant and willing to help him/herself (sv the doctor advises (sv = 1.2002): and (4) entrusts him/herself to the care of the doctor (sv = 1.1463). Together, however, these four items suggest an inherent contradiction in the residents' expectations (see also Bloor and Horobin, 1975). On the one hand, they expect the patient to be an active partner in the therapeutic process: he or she is supposed to evaluate and present symptoms in a way that allows the doctor to arrive at a diagnosis and institute proper treatment. On the other hand, the residents expect the patient to be a passive cooperator in the process: he or she is supposed to defer to the doctor's opinion, follow his advice, and accept responsibility for adherence to the prescribed regimen. The data also show that a great distance separates the first four items on the scale from the fifth, i.e., responds to treatment (sv = .7346). The relative unimportance that the residents attached to the patient's response to treatment is not surprising, however. Logically, the patient should respond to the doctor's intervention 267 oooo. oomN. moop.p NOON.F moFN.~ prN._ moum eoouou moo poo: mmoo .m opmmemo\soo mp. mm. on. oo. oo. Nm. opmo co meoppoz ace ocmopmeumpmm mH .N op. mm. Nm. mm. mo. oo. xeoomoo momesuum moo ammooo cm mm>ow .p o m o m N p coouucam m>ooomomm mo meomo co omoooo oomoomm moo mo .moooooem omooomz mcooomoou xom so» xoeumz coooeoooeoii.m.o uom

oo.: mocmmmea apoemomomemo ucmopmo o cmoz .p ocmuemo emosoz Homoemo emaE:z ocmuemo emnE:z m—nmeo>me:= pmguamz mpnmeo>mm so pompm>vns< mmmeoommm memom .mEmom mucmoomm mmeo» oo mmmocommm .moomuommm mo o:mummou-.¢.m 33P 273 The relationship between career stage and response to the patient who presents vague symptoms cannot be explained as a product of the training received in medical school. Although 50 percent of the more advanced physicians and 55 percent of the first year resi- dents had attended academically-oriented schools there was no rela- tion between this education and a negative view of the patient. Instead, physicians in advanced stages of their careers might have given unfavorable responses because of frustrations experienced with the realities of family practice. For example, the data showed that the more unfavorable a physician's response to patients who present vague complaints, the more likely he was to derive his greatest satisfactions in medicine from the technical aspects of his role (r = .37, p < .05). The data also showed that more advanced residents tended to derive their greatest satisfaction in medicine from this aspect of their role. That is, 40 percent of the more advanced physicians, in contrast to 17 percent of the first year residents, indicated that the major satisfaction they found in medicine was in diagnosis and treatment. If one accepts the assumption that one cohort of residents will share similar attitudes with other cohorts of residents, then it might be deduced that experiences in practice temper the "humanitarianism" the residents bring with them to their residency. Thus, residents become more technically-oriented as they advance in their training and less favorably oriented toward patients who impede them from satisfying their scientific interests (see also Adler and Shuval, 1978). 274 In summary, the data suggest that career stage may have an influence on a physician's orientation toward the dependent patient. They do not support, however, the relationship between career stage and attitude hypothesized in Chapter II and found to exist by Gray and his colleagues (1966). More extended contact with patients in a "high-interaction" specialty did not appear to heighten a physician's awareness of the need to treat the "whole" person rather than just her symptoms. Rather, extended contact with patients, particularly those with vague and inconsistent symptoms, appeared to intensify a physician's: (l) frustration because of the absence of the professed sources of work satisfaction; and (2) disinterest in dealing with any socioemotional problems that might underlie patients' problems. Perception of the Performance of Latent chtions in Medical Work As observed earlier, the latent functions of the physician's role examined in this study are (1) the satisfaction of the patient's need for catharsis; and (2) the satisfaction of the patient's need to c0pe with failure. The satisfaction of the need for catharsis was measured by the extent to which the physician felt he allows the patient to talk about herself and her problems. The satisfaction of the patient's need to cope with failure was defined as the extent to which the physician views his work as part of legitimation of illness which is being used to justify the patient's failure. The data indicated that slightly more than half of the residents (54.8 percent) felt that they satisfied the patient's need for catharsis. By contrast, almost three-quarters (71.4 percent) 275 felt that they did not satisfy the patient's need to cope with failure. In other words, residents were more likely to extend sympathy to patients who talked about problems related only indirectly to their health than they were to patients who used illness explicitly to justify failure. No statistically significant relationship was found between perceived satisfaction of the patient's need for catharsis and per- ceived satisfaction of the patient's need to cope with failure. This lack of relationship most probably was due to the small number of physicians, i.e., 12, who felt they satisfied the patient's need to cope with failure. In addition, no statistically significant relationship was found between the physicians' orientations toward the dependent patient and perceived satisfaction of the patient's non-medical needs. This lack of relationship might be a function of the questions used to operationalize these concepts. Apparently each set tapped a different cognitive dimension. The residents' career stage was not found to be related to perceived satisfaction of the patient's need for catharsis. More advanced physicians were as likely to feel that they satisfied this need (50 percent) as residents in the early stages of their career (55 percent). But, perceived satisfaction of the need for catharsis was found to be related to the type of medical school the physician had attended. That is, physicians who were graduates of practice- oriented schools were likely to feel that they allowed patients freedom to talk about problems only indirectly related to their health (r = .35, p < .02). This finding supports the view of Becker 276 and his colleagues (1961) and Bloom (1963) that the medical school has a socializing influence on physicians. The emphasis given to the social and behavioral sciences in practice-oriented medical schools apparently does affect the expectations their graduates bring to the physician-patient interaction. The data also showed that the residents considered the use of illness to legitimate failure to be widespread among patients on welfare (83.3 percent), women (65.1 percent) and non-whites (56.1 percent). The dependent patients in this study, then, both poor and female, were at high risk to be considered clients who became ill to escape from life difficulties. No statistically significant relationship was found, however, between perceived satisfaction of the need to cope with failure and contextual variables. For example, more advanced physicians were as likely as those in the early stages of their career to feel they satisfied the patient's need to cope with failure (40 and 35 percent. respectively). Furthermore, graduates of academically-oriented schools were as likely as graduates of practice-oriented schools to feel that they responded sympathetically to patients who used illness to justify failure (26 and 31 percent, respectively). These findings were not surprising because the residents' responses were so nega- tively skewed. In short, then, neither training in medical school nor career stage appeared to have bearing on a physician's perceived response to patients he felt became ill to justify some sort of failure in their lives. 277 Summary The residents had chosen a specialty in which the physician is expected to provide continuous care to patients whom he knows and understands as social beings located in a socioeconomic environment. The data showed that the residents, as a group, held views congruent with this expectation. That is, they described themselves as prac- titioners who had a primary obligation to solve the somatic and socioemotional problems of their patients. By contrast, the data showed that certain individuals among the residents held views that were in conflict with these expecta- tions. That is, some residents were more interested in solving technical, than patient, problems, particularly dependent patients' problems. Only a weak relationship was found, however, between the attitudes of physicians and contextual variables. Residents in a more advanced stage of their career tended to be unfavorably oriented toward the dependent patient. It was argued that their negative attitudes were a product of a decreasing “humanitarianism" that develops as physicians attempt to cope with the realities of family practice. Graduates of academically- oriented medical schools were less likely to feel they satisfied the patient's need for catharsis than graduates of practice-oriented schools. It was suggested that the difficult curricula of medical schools apparently do influence the expectations their graduates bring to the physician-patient interaction. The findings documented the view that physicians' orientations toward patients differ. They provided only partial support, however, 278 for the relationships hypothesized to exist between these different orientations and the context of the medical profession. Part 2: The Physician-Patient Interaction It was observed in Chapter V that during the two six-month periods of observation the patients made 422 visits to the family health care center. Furthermore, it was shown that patients with greater non-medical needs and many "symptoms of dependency" made more visits to the center than did other women in the sample. The major argument of this study, however, is that the high use of medical services is a consequence not only of patient needs but of physicians' responses to these needs as well. This part presents the findings on the way in which physicians' definitions of illness and determination of treatment influenced the women's use of medical services. The findings are organized into four sections. In the first section, the patient-physician interaction is described. Next follows an examination of the relationship between the different elements of the interaction. Then, the outcomes of physicians' decisions about patients are analyzed. Finally, in the last section, the women's views on the physician-patient interaction are discussed. Overview: The Physician-Patient Interaction Of the 422 visits the patients made to the center during the two six-month periods, 284 (67.3 percent) were initiated by the patient and 138 (32.7 percent) initiated by the physician. The 279 number of patient-initiated visits ranged from zero (3.2 percent) to seven (0.8 percent), with an average of 2.3 visits per patient. The number of physician-initiated visits ranged from zero (39.5 percent) to six (1.6 percent), with an average of 1.1 visits per patient. In short, 49 patients (39.5 percent) initiated all of their visits and four (3.2 percent), none of their visits. The physician-patient interaction during these visits followed a pattern which approximated entries in the medical record. First, the patient presented her symptoms. Next, the physician performed an examination and recorded physical findings. Then, he arrived at and assigned a diagnosis. Finally, he decided on a management plan. The following materials are organized in a way that parallels this pattern of interaction. Physical Findingg It was observed in Chapter V that the patients presented their symptoms in different ways. Some tended to restrict their difficulties while others emphasized the more diffuse nature of their complaints. Furthermore, approximately two-fifths (41.9 per- cent) of the patients presented an above the average number of "symptoms of dependency." Regardless of the way in which patients presented symptoms, however, few were considered to be very ill (see Appendix D). The average number of physical findings recorded per visit was 1.2 per patient, although most of these findings were objective (77.2 percent). Furthermore, less than one-quarter (23.4 percent) of the results of the diagnostic procedures performed were 280 positive. In short, the findings indicated that the patients, in general, sought care for ailments that were judged by their physicians to be relatively minor. Diagnoses The faculty at the center require residents to record their assessments of the patient's problems at the time of each visit, be they presumptive or definitive. Approximately three-quarters (76.6 percent) of the diagnoses recorded were considered by the residents to be definitive. Nevertheless, only 55.5 percent of all diagnoses were supported by physical findings and/or positive diagnostic pro- cedures (see Appendix E). Slightly less than one-third of the diagnoses (29.1 percent) were unsupported and ten percent were psychological.4 For purposes of analyses, diagnoses were dichotomized into the categories (1) unsupported/psychological diagnoses; and (2) supported diagnoses. Approximately two-fifths (41.3 percent) of the diagnoses fell into the first category and three-fifths (58.7 per- cent) into the second. The diagnoses of 12 patients (9.7 percent) were all unsupported/psychological and the diagnoses of 46 patients (37.1 percent) all supported. In short, almost two-thirds of the patients had been assigned one or more unsupported/psychological diagnosis during the two six-month periods. Physician's Plan The physician's plan constitutes the management strategies adopted for the care of the patients. As described in Chapter III, 281 three aspects of this plan were considered: (1) diagnostic pro- cedures planned and performed: (2) therapies prescribed: and (3) disposition of patients. Diagnostic Procedures Planned and Performed. Of the 828 diagnostic procedures planned by the residents, 405 (48.9 percent) were performed.5 The number of procedures planned per patient ranged from zero (14.5 percent) to 25 (0.8 percent), with an average of 6.7 procedures per patient. The number of procedures actually performed per patient ranged from zero (16.1 percent) to 25 (0.8 percent), with 3.3 the average per patient. Thus, tests were planned and performed for more than four-fifths of the patients (85.5 and 83.9 percent, respectively). Therapies Prescribed. As described in Chapter III, three types oftherapies prescribed by physicians were considered: (1) psychotropic therapies: (2) symptomatic therapies; and (3) cura- tive therapies. Four hundred and eight therapies were prescribed during the two six-month periods, an average of 3.3 per patient. Half (50.7 percent) of these therapies were symptomatic, two-fifths (40 percent) curative and one-tenth (9.3 percent) psychotropic. Approximately two-thirds of the patients received some symptomatic (63.7 percent) and some curative (64.5 percent) therapy during the six months. In addition, 15.6 percent of the patients received psychotropic therapies.6 282 Only 14 patients (11.3 percent) received no therapies. Two- thirds of these patients (64.3 percent) had sought care at the center only once during the two six-month periods. One-fifth (21.4 percent) had visited the center twice and two patients (14.2 percent) had sought care three and four times, respectively. In short, almost all the patients left the doctor's office with a prescription in hand. Disposition of the Patient. The residents requested patients to return to the center for follow-up visits 225 times during the two six-month periods, an average of 1.8 requests per patient. One- quarter (25 percent) of the patients were never asked to return to the center. The compliance rates with physician requests for follow- up visits was very high. Almost two-thirds (61.3 percent) of the patients returned to the center at the physician's request. In short, residents asked patients to return to the center slightly over half (53.3 percent) of the time they sought care and most patients complied with this request. Relationship Between Elements of the Physician-Patient Interaction Table 6.5 presents data on the relationship between patient characteristics and the results of the residents' physical examina- tions. Inspection of these data indicates that there was no statistically significant relationship between a patient's perceived medical needs and the severity of the condition for which she sought care. For example, patients with many "symptoms of dependency" were as likely to have a dearth of physical findings as those with few 283 mo.v 9.. mo. mo. oo.- mmescmuoeo uwumocamou mupzmme m>wpomoo mo coopeoooeo exp. no.1 Np. mucoucom m>ooumno=m we coooooooeo No.1 oo.- oo.- wagons?» m>ouumnno we cooueoooeo mo. o_.- oo. mmcoucow —muom»oo mo emosac mmmmm>< men—om; goo: momemoomu oucmvomomo moou oo cmmz Loo ummz mo msouosxm monoucom quomxoo muoomoemuumemou ocmoomo .mmooocom ~ouomxoo woo muoomoemouoemou oomoomo cmmzomm mcooompmeeou emceouoemNii.m.o Nomooomee we eeooeeeeeo emoN. eeoN.- mooooooo m>ooomooo mo ooooooooom N—.- N_. mmeoeeom m>ouumwesm we eeoueeeeee «mom. eeom.- mmeoeeoe Pmuomxoe we emosee mmmem>< "wmeoeeou Fmeomxoo mmmeemmoo emoeeeeom mmmeemmoo Pocono—common eeooeeeeeo \emoeeeesmes eeooeeeeee meoomoemoeeemou uemoooe mmmeeamoo .muoomoemoememou aemoooo ecm mmmeemmoa cmmzumm mcowuwpwhgou L02OI20NII.©.W w4mwome=e we eewoeeeeeo oo. oo.- oo.- mmwememoo ewomeeooexm we eeoueeeeeo oo_. me. «No. mmoomemoo ewoeeoeoexme we eewoeeeeeo oo.- oo. op. emEeewemo mmeeemeeeo ewomeemmwe we mmeEa: mmoem>< No.- me. FF. emcempe mmeeemuege ewumeemmwe we emosee mmmem>< meapweu sow: momeeooeu aeemeememo moeu eo emmz eew emmz we mseoosxm mmwmmoeeum uemsmmeeoz memmz pom—emeieez memmz pmeoemz .meopo acmemmmeoz eoe mewomoemuememou uemwooo :mmzomm meewuepmeeeo emeeoieemNii.N.o m4mwome=u we eewueeeeeo «oN.- «ON. mmweoemoo ewomseoosam we eewoeeoeeo eemN.- «omN. mmwememoo eweeeoeoexme we eewoeeoeeo emELewemo oo. oo. mmeoemeeeo uwomeemowe we emosoe mmmem>< emecmpo _o.- Po. mmeoemeeeo ewomeemowe we emesoe mmmem>< mmmeemewo emoeeoeom eewoeeoeeo mmmeemmwo Fmewmepeouamo \emoeeeoomoa oewoeeeeeo mempo oememooemz mmmeemmwo .mcmpo uemEmmeeoz eem mmmeeoowo :mmzomm meewoopmeeeu emego-egm~--.m.e mom emewepexm ea acmeeme mee ommmp om mewoeowepeeu mmpemwee> .1 on.m ooo. we.eN oo. mm. m_.- meepwow oowz moee ea emmz m _N—.N ooo. pm.mm um. oo. oo.- xuemeememe we maeerXm N mmeweeww NN_.— ooo. mm.om Nm. em. Nm. pmewmxoe we omega: mmmem>< F mmmecmewo emoeemeam eewoeeeeee oNF.m ooo. we.mN oo. me. my. meopwew sow: moeu ea emmz m _N~.N ooo. om.mm um. Fe. mp. oeomeomome we mseoesxm N mmeweeww NNP.P ooo. mm.em Nm. om. Nm.- Pmuwmxoo we Lease: mmoem>< w mmmecmwwo wmewmewezexmo \emoeemmmmcz cowoeeeeeo mewemuem Q .1 oo o No mooowooz moo moomoomo mmoomoom> . we emeeo .moemwomo emomwmm< mmmeemmwo we meewmmmemmm mmwzuemomii.m.e Nom emewm—oxm e» oemegme #5. mac ummmp om mewoaaweoeee mmpamwem> a. NN_._ ooo. vo.NN mp. mm. mm.- mmmeemmwe pmuwmepeaeame \emueeeeomoo we :ewoeeeeee p mmwmmemaw m>womeeu we eewoeeaeeo oN..N No. om.m oo. oN. mo. moommomomm mo meooooam N NNo.o No. NN.m oo. oN. oo.- mmmooomoo emueeeoem we eewoeeeeee p mmwmmemow ewumeeuesxw we eewoeemeeo —N~.N woo. eo.w o_. Nm. op. mgewwmw aowz meee eo emmz N NNw.p —oo. mN.~p mo. oN. aN. mmmeemmwe Fmewmepeaezmo \emueeqoamee we eewueeoeeo P mmwmmemaw ewaeeoeaexmo we eewueeeeeo mcwemoem Q a. we o No mooomooz mom moomoom> mmoomommo . we emcee .emaweummeo mmwememaw we meewmmmemmm mmwziemumiu.o—.a mam emewmwoxm ea acmeemo mee ammmw am oewaoaweaeeu mmpamwem>w on.m Noo. oN.o Nw. em. ow.- meowwmw aawz moeu ea emmz m FNF.N woo. oN.w ——. mm. PN.- xuemeemome we mseaosxm N mmmecmmwe Pmewoeweaeomo NNP.P moo. oN.o No. oN. oN.- \emaeeooomeo we cowaeeoeeo w «mewemaem mo m a No maowoooo woo moomoom> mmaoooom> we emeeo .mawmw> oo-:ewweu sew mammoemm we eewmmmeomm mmwzuomamii.F—.o mao emewmwoxm ea acmeeme mee ammmw am oewaoaweaeee mmwamwem>e mmwomemaa on.m ooo. em.m mN. mo. ma. m>wameoe we eewaeeoeeo m mmwomemaa PNF.N ooo. ow.¢p oN. we. oN. ewameeaosxm we cewaeeoego N mawmw> ooizewpew NNP.F ooo. mo.NN mp. No. no. Lew mammoemL we eewaeeoeeo P moomo> mmomaoooo-omaoom»oa mmwomemaa oN_.m No. mm.m mo. oN. FF. owamooeaoao.am we :ewaeeoego m mmoee PNF.N wo. om.o No. oN. ow. ewoeeaeaexmo we cowaeeoego N mmmeemmwe NN_.P wo. ow.m mo. wN. ¢—.- emaaeooom we cowaeeeeeo w mawmw> mmamwawoaiaemwamo oeoemacm Q .1. we a Na maoomwoz «em mwamwem> mmwamwem> . we emeeo .mmew>emm wmewem: we mm: we meemaamo ea oewaoaweaoeo m.omwewmzao moa we moewmmmeomm mmwziomamii.Nw.o mao emewmwoxm ea aemuemo mee ammmw am mewaoaweaeeu mmaamwem> 8. 314 mawmw> oeizepwew oaa.o ooo. oN.m mN. we. ow. Lew mammzeme we eewaeeoeeo o owa.m ooo. nN.o PN. oo. ma. mwmemaame Lew emmz m mmwommmoa owa.e ooo. oo.o ow. mo. Na. uwamseaoEAm we cowaeeoeeo o mmwomemaa on.m ooo. eo.w Na. we. Np. ewoeeaeaexmo we sewageoeeo m _NF.N woo. mm.oa ma. om. aN. aecmeemome we mseaosom N NNw.p aoo. mm.ww oo. oN. oN. meowwmw aaw: moee ea emmz a oocwemaom wo a o No mamowooz moo moomoom> mmoomoom> we emeeo .mmew>emm we mm: pmaew maa ea cowazaweaceu .macmwama com .mcmwuwmxza msa we cowmmmemmm mmwziamamii.mp.o mamea< ee ow mm< meememo o.Nw o.m o.mN ooze apoem moo o.mm _.wo w.¢w awoem mee :maa moo: "my mm< Lowe: emeewwao aawz eweammoez we meoaeoeam w._w _.o N.N_ memwwmz co m.mm o.mo o.Nm memwamz :e aez "memwwmz co meaeammoez o.NN m.m o.oo emaae\xemwm o.NN N.oo ¢.om mawaz mmmmm mamaew xaeoeo aaeoeuiaoo xawu mewamwemaumemao ewaomeoeemeewuem mme< .owow .mmwamwemaememao ewoomeoesmeeweem emaemwmm eom mme< xa zaeoeo cw cowamwooeo we acmeemoui.P.< mao mwaa we mmme maa ea .mame wmoaem maa mom meommme zxoeo maa :mmzama .meee Le momma eza eeeaeeaaowme ewmaa cw em>ww ema xmoa amaa emaeeome maemeeeomme emzmw>emaew moa we acmeeme eoewixaxwmm ¢.N¢ N.¢o m.ae o.om o.m¢ m._m momma e3a :maa mmma o.wm w.mm w.mm mw.mo e.om w.mo meeE Le momma ezw "aremmmoe: we emm: aa mecmewmmm we aamewm o.oe N.e¢ m.mm o.wm o.em N.o¢ ELez zepmm p.om m.mm N.wo o.m¢ «.me m.om ELez m>eam so ea wmoau "emxewmem m>ea< Le ow mm< meememo o.mm N.o¢ w.mN e.em w.o¢ ¢.NN seez 3ewmm e.¢o w.mm m.ww o.mo o.om o.ww eeez m>eam ee ea wmoem "wwwee<,e3w saw: mewwwa mo mo< emeeo emeewwao aaw: meaeammoe: m.om m.om o.No m.om N.wm m.o¢ seez :e—mm N.oo m.mo o.wm w.oo m.ao w.om EL.ez m>eam ee ea —m=em omemwpmz ee meweommoe: m.m o.m m.o o.o o.oa m.w emxwz zwwmwemm o.NN N._N m.NN o.MN o.om o.oa xemam x—aemewEeemeo N.mw o.mw N.wN o.mo N.om a.ow mawaz xwaemcwEeemeo "memo mamaew xaeoeo aaeoeeiaeo zawo mwmaew maeoeo manoeuiaeo zawo emzmw>emaeaicez emzmw>emaoa mewamwemaememao mme< xa mwosmm emaemwmm aa aemew we meow .mewamwemaememoo ewaemeoeeme -eaeem emaemwmm eem mmo< xa mmwosmm emzmw>emaeaieez eem emzmw>emaea we acmeemo--.Ni< uaomp gap; on» on oocmwmmo mm: ozm.wucwoa x_m coco woos co on Peace moo: mucoum mmocpp_ m.Fm=uP>wo:P co co woos co ucoocog _m CH0 .asocm mmmcppm —o>op gap; as» on nocmpmmo mm: ocm .mucvoa xwm can» ocoe co op peace ocoz museum m._m=or>mocp co co ocos co ucoocoa me man .azocm ~o>mp mmocppw now; on» on nocmwmmm mm: osm .cmos venom on» o>onm so on Foacm mm: mcoom came m.pm=or>wocr cm emu a.mm as m.m~ mm a.mm mo mmmccpe co Pa>a_ zoo ~.oo dc m._c mm K.¢e mm mmu=_pc co Pa>mp ems: "muvmw> FP< e.ve om ¢.¢~ FF a.mm mm mmoc—pw mo pm>o— 3o; o.mm mm a.mn am ¢.¢e om mmocppw we pm>op saw: "muwmp> N on F ~.~m mm m.om em p.mm me mmocppv we po>op zoo m.~m me N.mo em m.¢¢ mm mmocppp co po>op any: umppm_> meet co m acoucoa cooszz acoucoa consaz ucmocom conssz muwmw> we coasaz uHHH can a“ ma Fm>m4 mmmcppH opam . ~w>wi_ mmch: meLommumu OH m—uonuwz ucwgmmfiwo $92.: Germ: mPQEmw $0 cowvanwLHmPn—II.P.D m4m<._. 417 clinician, involved in teaching and research; a graduate physician about to begin his residency training; and a second year medical student just entering the clinical years of his training. These people read the medical record abstracts of ten "stable" patients in the high level of illness category and ten "stable" patients in the low level category. The purpose of this reading was to determine if the patients assigned to one (or the other) category shared cer- tain clinical characteristics that could be established as a set of criteria and applied to the data contained in the abstracts of the 43 ”unstable" patients. All three reviewers concluded that most of the patients in the high level category seemed little different from those in the low level category. In addition, although the purpose of the classificatory scheme had been to differentiate between patients on the basis of severity of illness, all three felt that most of the patients assigned to the high level category were not very ill. As one reviewer noted: "They had findings and got a prescription. But, the whole thing was of little medical consequence." The reviewers, then, were unable to establish any criteria on the basis of their readings. Moreover, they concluded that there were several inherent problems in attempting to develop a classificatory scheme based on the data available. First, there was a dearth of physical findings recorded in the medical records. On the one hand, they felt these data might represent the small number of findings detected by the physician or only the number he or she considered significant enough to enter. 418 On the other hand, they felt these data recorded over time and in the presence of vague complaints, might reflect the onset of a . 6 severe organic problem. Differences in modes of recording data and/or the insidious course of some illnesses, then represented obstacles to the development of a level of illness index. Second, there was a great deal of variation in the number of tests ordered for patients. 0n the one hand, the reviewers felt that entries of large numbers of tests might represent a thorough- ness of workup (or an overutilization of tests). 0n the other hand, they felt that entries of small numbers of tests might reflect the physician's ability as a diagnostician (or an underutilization of tests). These differences, regardless of their bases, meant that the probability of having tests was not the same for all patients and thus represented an impediment to the development of an index. Finally, there were large numbers of entries of "all find- ings normal" in the medical records. 0n the one hand, the reviewers felt that such an entry might represent the actual health status of a patient who had complied with her physician's request for a follow-up visit. 0n the other hand, they felt that such an entry might reflect, not the health status of the patient, but the philosophy of the physician toward psychosocial pathology. As one reviewer noted: Many physicians keep their assessment of the patient's psychosocial debility in their minds, not in the medical chart. As a consequence, they reschedule visits, enter a benign organic diagnosis and "normal" in the chart, and use the visit as a counseling session. 419 Entries of "all findings normal," then distorted the scores of some patients, obfuscated the scores of others, and represented a hinderance to the development of an index. In conclusion, an attempt was made to develop an index that would permit illness level to be related to other sociobiological factors. Component variables were selected that were felt to be well-defined and reliable. The data, however, were strongly influenced by physicians' clinical abilities as well as judgments and the effort was thwarted. As Balinsky and Berger (1975:286) have noted: "Even in the class of data referred to as clinical evidence . . . problems of diagnostic accuracy and prognosis often skew measures." NOTES: APPENDIX D 1Diagnoses were examined to see if they could be used as a measure of duration (i.e., an illness episode). These, however, reflected both a knowledge of the course of disease as well as an inference based upon presenting complaints (that is, symptoms presented over time in a similar way). Thus, the idea of including duration as a component of the scheme was abandoned because the diagnostic data available were based on clinical knowledge as well as the doctor-patient interaction. 2In calculations based upon the additive method, one was treated as a zero. "No findings" was assigned the weight of one, however, to permit calculations based upon the multiplicative method. The use of these methods is discussed below. 3The scores described here were calculated using the addi- tive method. Scores also were calculated using the multiplicative method but the range was so large (1.0 to 1296.0), that a few high scores inflated the grand mean for the group. As a consequence, Rule I (see below), when applied to the data, produced a skewed distribution. 4The grand mean (total number of points divided by total number of visits) for all women was 7.48. The mean scores for individuals ranged from 0.0 (5.7 percent) to 19.3 (0.8 percent). SSix points was established as the cutting point in this rule because (I) this number represented one solid objective find- ing; and (2) it was the most frequently occurring score per visit. 6In the instance of one patient, this was the case. The ill-defined abdominal pains for which she sought care over a period of months were diagnosed as ulcerative colitis after emergency surgery. Only one such case occurred, of course, but it does repre- sent an inherent problem when dealing with a paucity of data. Moreover, it should represent a caveat to physicians; dependent patients--as this patient had been labelled by her physician--also get sick. 420 APPENDIX E DEVELOPMENT OF A DIAGNOSTIC INDEX 421 APPENDIX E DEVELOPMENT OF A DIAGNOSTIC INDEX In order to differentiate between patients on the basis of the diagnoses assigned to them, a "diagnostic index" was developed. The index is based on the judgments of three people: two nurse clinicians involved in teaching and research and a graduate physician about to begin his residency training. These people read the medical record abstracts of all patients and coded the diagnoses on the basis of a set of criteria (see below). Proportions then were calculated using the coded data in order to categorize physicians' assessments of patients' problems. This section describes the methods used to develop the diagnostic index. Components of the Index The diagnostic index is based on four sets of data: (I) the diagnoses entered by the physician; (2) the symptoms presented by the patient; (3) the physical findings recorded; and (4) the results of diagnostic procedures. "AM The data for each patient visit had been recorded in a separate medical abstract schedule. The reviewers read the abstracts and, using the criteria listed below, assigned each visit an appropriate code. 422 423 Coding: Codes and Criteria l. Psychological. The diagnoses and findings that the physician recorded and the symptoms that the patient reported are strictly psychological. Diagnoses of anxiety and depression are most common. Findings or symptoms such as moody, labile, crying, or emotional outbursts are recorded. A diagnosis of anxiety or depression automatically are coded psychological, even in the absence of findings to support this diagnosis. 2. Symptomatic. The diagnosis recorded by the physician or the symptoms reported by the patient are not validated by either one physical finding or one positive diagnostic procedure. Symptoms such as headaches, nausea, vomiting, fatigue, tired all over, or weakness are most common. If the physician has entered a definitive diagnosis such as "flu" or "anemia," it is categorized as symptomatic if the diagnosis is not validated. 3. Organic. The diagnosis entered by the physician of the symptoms reported by the patient are validated by physical findings and/or positive diagnostic procedures. For example, if a patient reports a temperature and "achy" feelings and these are supported by findings such as a red throat and/or a "wheezing chest," the diagnosis is categorized as organic. If a diagnosis has been validated at one visit, it need not be validated at follow-up visits, i.e., it is categorized as organic. 4. Psychological and Symptomatic. Two or more diagnoses are entered. In combination, they meet the criteria described in one and two above. 424 5. Psychological and Organic. Two or more diagnoses are entered. In combination, they meet the criteria described in one and three above. 6. Symptomatic and Organic. Two or more diagnoses are entered. In combination, they meet the criteria described in two and three above. 7. Psychological, Symptomatic and Orggnic. Three diagnoses are entered. In combination, they meet the criteria described in one, two and three above. 8. Health Maintenance. The purpose of the visit is pre- ventive care, e.g., physical exam, PAP test, or request for birth control pills. 9. Symptomatic and Health Maintenance. Two or more diagnoses are entered. In combination, they meet the criteria described in two and eight above. lO. Organic and Health Maintenance. Two or more diagnoses are entered. In combination, they meet the criteria described in three and eight above. Five hundred twenty-two diagnoses were coded and there was 92 percent agreement among the three coders. Agreement became total once discrepancies had been resolved through discussion. Of the 522 diagnoses coded, 10 percent were categorized as psychological; 29.1 percent as symptomatic (unsupported); 55.5 percent as organic (supported);and 5.4 percent as health maintenance. 425 Scoring In order to categorize patient diagnoses on the basis of physicians'assessments the following tasks were performed: 1. All diagnoses within the health maintenance category were excluded in the calculations; 2. The number of psychological and symptomatic (unsupported) diagnoses were added to form one category; 3. A proportional score for organic (supported) diagnoses and a proportional score for psychological[symptomatic (unsupported) diagnoses was calculated using the following formula: total number of diagnoses within category divided by total number of diagnoses in the two categories; and 4. A single summary index for each of the two categories was assigned to the patient. Discussion Essentially, the index dichotomizes diagnoses into categories that differentiate (l) those that are psychological or symptomatic (unsupported) from (2) those that are organic (supported). These are categories that reflect physicians' assessments of patients' problems and the extent to which their diagnostic assessments were supported or not supported by clinical evidence. Several caveats are necessary, however. First, the sub-category symptomatic (unsupported) may be interpreted several ways. For example, the sub-category subsumes labels which reflect symptomatology that cannot be definitively 426 diagnosed (e.g., pain, malaise). It also subsumes labels which may reflect physicians' attitudes toward socioemotional problems. That is, the physician may have entered an organic label, rather than a psychological one, and counseled the patient. Such an organic label, however, was categorized as symptomatic because it was not supported by clinical evidence. Finally, the sub-category subsumes labels which may reflect physicians' diagnostic expertise. That is, a physician may have entered an organic label which, because it could not be supported by clinical evidence, was categorized as sympto- matic. The sub-category, then, does not elucidate the cognitive process by which a physician made a diagnostic decision. Rather it indicates that patients did not have clinically defined morbidity. Second, the category organic (supported) indicates only that patients were assigned a diagnosis that could be supported by clinical evidence. Certain diagnoses, however, have a greater tendency to be supported by diagnostic procedures than others.1 For example, a vaginal culture automatically will result in the growth of flora. Vaginitis, therefore, may have been judged an organic diagnosis more often than a urinary tract infection because the former is associated with more positive test results. Further- more, the category organic does not reflect the severity of patients' conditions. Severity ranges along a continuum, and diagnosis may be supported by one, two or "n" number of findings for different patients. In short, the category organic reflects only the presence of definitively defined clinical morbidity. NOTES: APPENDIX E 1I am indebted to R. E. Sturm, M.D., for suggesting this point to me. 427 APPENDIX F PROBLEMS IN DEVELOPING A DEFINITION OF A PATIENT POPULATION 428 APPENDIX F PROBLEMS IN DEVELOPING A DEFINITION or A PATIENT POPULATION] A statistical population consists of the totality of the observations about which relevant inquiries are to be made (Nalpole, 1974:122). Populations in this sense are arbitrary and must be carefully defined: Clearly, there is nothing natural about a statistical population because it is defined to satisfy specific research objectives . . . . Statistical methodology comes into play only after the relevant populations have been defined to suit the research objectives at hand (Thomas, 1976:35). The objective for this research was to define a population with a common characteristic. This was a complex undertaking. Did the population consist of all people who had contact with the center during a specified period of observation? Or did it consist of people who were known to have had contact over a prolonged period of time? What were the parameters of the group? The answer was neither simple nor obvious since the characteristic chosen to define the patient population may have considerable consequences for the data that are developed. This section elucidates these consequences by describing the way in which different definitions create varia- tions in statistics. 429 430 In order to develop a definition of the patient population of the health center, a characteristic that a group of patients might have in common had to be discovered. An obvious one was evidence of contact during a common period of time. Since the period of observation for the sample of this study was the twelve months between July l, 1976 and June 30, 1977, the first criterion established stated that an individual must have shown evidence of contact with the center during this period in order to be considered a member of its patient population. This criterion created a shared quality among individuals. But, it also aggregated together people who had contact with the center over a prolonged period of time as well as those whose contacts had been more ephemeral. Thus, in order to differentiate between regular and transient patients, one additional criterion was established that stated that an individual must have had contact with the center during a six- month prior- and/or post-observation period in order to be con- sidered a member of its patient population. Together, the criteria created a set of mutually exclusive definitions that described all combinations possible. Figure F.l contains the five definitions that reflect the most plausible descriptions of a patient population.2 Method A, in which the only characteristic of interest is the defined period of observation, produces the largest population. This size, however, begins to decrease when additional criteria are included in the definition. In Method B, the decrease is small. But since no selection criteria are applied during the period of 431 .meceebme co cerbmpzeoa a m=F=_cae co meoguez ucmceccce co comecaesoe--.fi.e maeeee .ueeeuemeeee eexece e co unmeem me: ecee Feemeee sews: em uvmw> peeuue ee opeo_eew awe mouse mesh .ceoeee one cup: wees come we; ueeaeeo e peep moueupeew cope: Eeumam euee me» e? mouse Ace me eeewcee mp eceeec we uwmw> < =.eceoem we uwmw>= Eco» eeeeesee use gee eepuev>oceee ee mm mo>e .eevcee one» mewcee ceueeo on» gum: mueeueee we geese: ace ea eee ce ueepeeu e: ee; e>ee e—aeu aeopuee e m_ peso .eepcee es_u e memcee ewe—gee use: emcouwce eewuoepem ee peso ebee_eew ea eeme mp gaze: use: ezbe m.e weep mo> egos so ago mo> wees so one mo> egos ce meo m m.m ommm xe< mo> ages so eeo mo> egos co use a —.e mmwp ao> egos so one xe< ao> wees ce eeo u ~.P Neom xe< xc< mo> wees so use m o.m mmoe >=< Nee) egos co «:0 asc< < beepuee Lee mueepuee umee eewue>cemeo Levee mecm_> co co co eowcea cocoa: ceesez emece>< geesez 1|. cocb_cecee 432 observation, a small mean results. In Methods C through E, the decrease is appreciable but the more rigorous criteria applied produce a larger mean. The demand for evidence of continued contact, then, decreases the size of a patient population. However, the statistic it produces describing the volume of use of this population may reflect reality more accurately. Developing a definition of a patient population is a complex undertaking, and one which the researcher should approach with caution. Including transient patients in the group will increase its size. But, it will also tend to lower the average number of visits reported per patient. NOTES: APPENDIX F 1I am indebted to Robert J. Sprafka for the material described in this Appendix. He made me aware of the problems involved in developing a definition of a patient population and created the methodology and data discussed herein. 2The data included in this figure were derived from the health information system in use at the center. 433 APPENDIX G BACKGROUND INFORMATION ON THE RESIDENTS 434 APPENDIX G BACKGROUND INFORMATION ON THE RESIDENTS The residents ranged in age from 26 to 41 years, with 29.8 years the average age. Almost three-quarters (71.4 percent) were married (see Table 6.1). Slightly over one-third (38.4 percent) came from families in which their fathers had occupied high social positions; five (11.9 percent) were the sons of physicians. An approximately equal percentage (35.9) grew up in homes in which their fathers occupied relatively low social positions. Almost three-quarters (69.1 percent) of the residents, however, felt that they had had difficulty in financing their medical education. Most (95 percent) of the residents had majored in the physical or natural sciences in college. About one-third (31.6 per- cent) felt that they would have chosen another major if they had the opportunity. Of these residents, three-quarters (75 percent) would have preferred to major in the social or behavioral sciences. Only one statistically significant relationship was found between the background characteristics of the residents and their attitudes. That is, physicians who expressed greater difficulty in financing their medical education were likely to feel that they extended sympathy to patients whom they felt used illness to legiti- mate failure (r = .32, p < .03). In short, the environment in which 435 436 residents were raised seemed to have little bearing on their sub- sequent attitudes as physicians. 437 TABLE G.1.--Percent of Residents by Selected Sociodemographic Characteristics. . . Percent of Characteristic Residents Marital Status: Single 26.2 Married 71,4 Marriage Disrupted 2.4 Social Class:a I 20.5 II 17.9 III 25.6 IV 33.3 V 2.6 Type of Place Where Raised: Rural area 16.7 Non-suburban small town 21.4 Suburban area 38.1 Urban area 23.8 aSocial class is based on Hollingshead's (1967) "Two Factor Index of Social Position." Class I represents the highest index of class position on the scale and Class V the lowest position. The index is based on the educational level and occupation of the residents' fathers at the time the residents were 16 years of age. LIST OF REFERENCES 438 REFERENCES Abel-Smith, B. and R. M. Titmuss 1955 The Cost of the National Health Service. Cambridge: Cambridge University Press. Adams, 8. N. 1967 “Interaction Theory and the Social Network." Sociometry 30:64-78. 1968 Kinship in an Urban Setting. Chicago: Markham Publishing Company. Aday, L. and R. Eichhorn 1972 The Utilization of Health Services: Indices and Corre- lates. A Research Bibliography. *DHEN Publication No. (HSM) 73-3003. TRockvilTe, Maryland: National Center for Health Services Research and Development. Adler, I. and J. T. Shuval 1978 "Cross Pressures During Socialization for Medicine." American Sociological Review 43:693-704. Aldous, J. 1967 "Intergenerational Visiting Patterns: Variation in Boundary Maintenance as an Explanation.“ Family Process 6:235-248. Aldous, J. and M. A. Straus 1966 “Social Networks and Conjugal Roles: A Test of Bott's Hypothesis.“ Social Forces 44:576-580. Alpert, J. L. and E. Charney 1973 The Education of Physicians for Primary Care. DHEN Publication No.7(HRA) 74-3113. Washington: U.S. Department of Health, Education, and Welfare. American Medical Association 1971 Reference Data on the Profile of Medical Practice. Chicago: ’Center for HeaTth Services Research and Development. 439 Anderson, 0. 1963 Anderson, R. 1968 Anderson, R. 1967 Anderson, R. 1970 Anderson, R. 1972 Antonovsky, 1967 1972 Apple, D. 1960 Asher, H. B. 1976 Avnet, H. H. 1967 Axelrod, M. 1956 440 N. "The Utilization of Health Services." Pp. 349-367 in H. E. Freeman, 5. Levine, and L. G. Reeder (eds.), Handbook of Medical Sociology. New York: Prentice-Hall. A Behavioral Model of Families' Use of Health Services. Research Series No. 25, Center for Health Administration Studies. Chicago: University of Chicago Press. and O. H. Anderson A Decade of Health Services. Chicago: University of Chicago Press. and L. Benham "Factors Affecting the Relationship Between Family Income and Medical Care Consumption." Pp. 98-107 in H. E. Klarman (ed.), Empirical Studies in Health Economics. Baltimore: The Johns Hopkins Press. , R. McL. Greeley, J. Kravits, and O. H. Anderson Health Service Use: National Trends and Variations, 1953-71. DHEH Publication No. (HSM) 73-3004. Washington: Health Services and Mental Health Administration. A. "Social Class and Illness: A Reconsideration." Sociological Inquiry 37:311-322. "A Model to Explain Visits to the Doctor: With Special Reference to the Case of Israel.“ Journal of Health and Social Behavior 13:446-454. "How Laymen Define Illness." Journal of Health and Human Behavior 1:219-225. Causal Modeling. Beverly Hills: Sage Publications. Physician Service Patterns and Illness Rates. New York: Group Health Insurance, Inc. "Urban Structure and Social Participation." American Sociological Review 21:13-18. 441 Backwin, H. 1945 "Pseudocia Pediatrica." New England Journal of Medicine 232:681-697. Balinsky, H. and R. Berger 1975 "A Review of the Research on General Health Status Indexes." Medical Care 13:283-293. Balint, M. 1957 The Doctorngis Patient and the Illness. London: Pitman Medical Publications. 1970 "Repeat Prescription Patients: Are They an Identifiable Group?" Psychiatryyin Medicine l:3-l4. Banks, M. H., S. A. A. Beresford, D. C. Morell, J. J. Waller and C. J. Hatkins 1975 “Factors Influencing Demand for Primary Medical Care in Women Aged 20-44 Years: A Preliminary Report." International Journal of Epidemiolpgy 4:189-195. Barker, R. 1953 Adjustment to Physical Handicap and Illness. New York: Sbcial Science Research Council. Barnes, J. A. 1954 "Class and Committees in a Norwegian Island Parish." Human Relations 7:39-58. Bauman, B. O. 1961 "Diversities in Conceptions of Health and Physical Fitness." Journal of Health and Human Behavior 2:39-46. Becker, H. 1963 Outsiders: Studies in the Sociolpgy of Deviance. New Yfirk: The Free Press. Becker, H. S. and A. L. Strauss 1956 "Careers, Personality, and Adult Socialization." American Journal of Sociology 62:253-263. Becker, H. S., B. Geer, E. C. Hughes, and A. Strauss 1961 Boys in White: Student Culture in a Medical School. Chicago: University of ChicagoPress. Bell, U. and M. D. Boat 1957 "Urban Neighborhoods and Informal Social Relations." American Journal of Sociology 62:391-398. 442 Bellin, S. S., B. Z. Locke and M. New 1976 ”The Neighborhood Health Center as a Mental Health Diagnostic Service." Public Health Reports 91:446-451. Ben Sira, Z. 1976 "The Function of the Professional's Affective Behavior in Client Satisfaction: A Revised Approach to Social Interaction Theory.“ Journal of Health and Social Behavior 7:3-21. Beresford, S. A. A., J. J. Haller, M. H. Banks and C. J. Hare 1977 “Why Do Women Consult Doctors? Social Factors and the Use of the General Practitioner." British Journal of Preventive and Social Medicine 31:220-226. Berkman, P. L. 1969 "Spouseless Motherhood, Psychological Stress, and Physical Morbidity." Journal of Health and Social Behavior 10:323-334. Berle, B. B. 1958 Eigh;y_Puerto Rican Families in New York City. New Ybrk: Columbia University Press. Bice, T. H. and K. L. White 1969 “Factors Related to the Use of Health Services: An International Comparative Study." Medical Care 8:124- 133. Bice, T. H. and R. L. Eichhorn 1972 "Socioeconomic Status and the Use of Physicians' Services: A Reconsideration." Medical Care 12:261-271. Bierstedt, R. 1963 The Social Order. New York: McGraw Hill. Bloom, 5. H. 1963 The Doctor and His Patient: A Sociological Analysis. NewTYork: Russell Sage Feundation. Bloor, M. J. and G. H. Horobin 1975 "Conflict and Conflict Resolution in Doctor/Patient Interactions." Pp. 271-284 in C. Cox and A. Mead (eds.), A Sociology of Medical Practice. London: Collier- Macmillon. Bluestone, B., H. Murphy and M. H. Stevenson 1971 Low Wages and the Workinngoor. Detroit: Institute of Eabor and Industrial Relations, University of Michigan- Hayne State University. 443 Blumberg, M. S. and J. C. Coffin 1956 "A Syllabus on Work Absence." Archives Industrial Health 13:55-70. Bott, D. 1957 Family and Social Network. London: Tavistock PubTications Ltd. Brandwein, R. A., C. A. Brown and E. M. Fox 1974 "Women and Children Last: The Social Situation of Divorced Mothers and Their Families." Journal of Marriage and the Family 36:498-514. Campbell, M. A. 1973 Why Would a Woman Go Into Medicine? Medical Education in the U.S.: A Guide for Women. TNewTYork: *Feminist Press. Caplan, G. 1974 Suppprt Systems and Community Mental Health. New York: Behavioral Publications. Caplan, R. D., E. A. R. Robinson, J. P. R. French, Jr., J. R. Caldwell and M. Shinn l976 Adhering to Medical Regimens: Pilot Experiments in Patient Education and Social Support. Ann Arbor: Institute fOr Social Research, University of Michigan. Cartwright, A. 1967 Patients and Their Doctors. London: Routledge and Kegan Paul. 1974 "Prescribing and the Relationship Between Patients and Doctors." Pp. 63-74 in R. Cooperstock (ed.), Social Aspects of the Medical Use of Psychotropic Drugs. Toronto: Alcoholism and Drug Addition Research Foundation of Ontario. Cartwright A. and M. Jeffreys 1958 "Married Women Who Work: Their Own and Their Children's Health." Social Science and Medicine 12:159-171. Chesler, D. 1972 Women and Madness. New York: Avon. Clark, M. 1959 Health in the Mexican-American Culture. Berkeley: University of California Press. 444 Cobb, S. 1976 "Social Support as a Moderator of Life Stress." Psychosomatic Medicine 38:300-314. Coe, R. M. and A. F. Wessen 1965 “Social-Psychological Factors Influencing the Use of Community Health Resources." American Journal of Public Health 55:1024-1031. Coe, R. M., J. M. Goering and M. Cummins 1969 Health Status of Low Income Families in an Urban Area. Final Report for the Bi-State Regional Medical Program. St. Louis: Medical Care Research Center. Cohen, A. K. and H. M. Hodges 1963 "Characteristics of the Lower-Blue-Collar-Class." Social Problems 10:303-334. Cohen, Y. A. 1961 "Patterns of Friendship." Pp. 351-386 in Y. A. Cohen (ed.), Social Structure and Personality: A Casebook. New York: Holt, Rinehart and Winston. Cole, 5. and R. Lejeune 1972 "Illness and the Legitimation of Failure." American Sociological Review 37:347-356. Cooper, 8. 1964 "The Epidemiological Approach to Psychosomatic Medicine." Journal of Psychosomatic Research 8:9-15. Cooperstock, R. 1971 “Sex Differences in the Use of Mood Altering Drugs: An Explanatory Model." Journal of Health and Social Behavior 12:238-244. Corea, G. 1978 The Hidden Malpractice: How American Medicine Mistreats Women. New York: Jove Publications,7Inc. Coult, A. and R. Habenstein 1962 "The Study of Extended Kinship in Urban Society." Sociological Quarterly 3:141-145. Cummings, N. A. and W. T. Follette 1968 “Psychiatric Services and Medical Care Utilization in a Prepaid Health Plan Setting: Part 11." Medical Care 6:31-4 . Davis, J. A. 1962 gtipends and Spouses. Chicago: University of Chicago ress. 445 Davis, M. S. 1978 "Variations in Patients' Compliance with Doctors' Advice: An Empirical Analysis of Patterns of Communication." American Journal of Public Health 58:292-306. Densen, P. M., S. Shapiro and M. Einhorn 1959 "Concerning High and Low Utilizers of Service in a Medical Care Plan, and the Persistence of Utilization Levels Over a Three Year Period." Milbank Memorial Fund Quarterly 37:217-249. DeVaul, R. A., S. Zisook and H. J. Stuart 1977 "Patients with Psychogenic Pain." Journal of Family Practice 4:53-55. Dohrenwend, B. S. and B. P. Dohrenwend 1967 "Field Studies of Social Factors in Relation to Three Types of Psychological Disorder." Journal of Abnormal Psychology 72:369-378. 1974 "Overview and Prospects for Research on Stressful Life Events." Pp. 313-331 in B. S. Dohrenwend and B. P. Dohrenwend (eds.), Stressful Life Events. New York: John Wiley. Donabedian, A. 1966 " valuating the Quality of Medical Care." Milbank Memorial Fund Quarterly 44:171-179. Duff, R. S. and A. B. Hollingshead 1968 Sickness and Society. New York: Harper and Row. Ehrenreich, B. and 0. English 1973 Complaints and Disorders: The Sexual Politics of Sickness. New York: Feminist Press. Enterline, P. E. 1961 “Work Loss Due to Illness in Selected Occupations and Industries." Journal of Occupational Medicine 3:405- 411. Erikson, K. T. 1964 "Notes on the Sociology of Deviance." Pp. 9-21 in H. Becker (ed.), The Other Side. New York: The Free Press. Eron, L. D. 1955 "Effect of Medical Education on Medical Students' Attitudes." Journal of Medical Education 30:559-566. 446 Fabrega, H., R. J. Moore and J. R. Strawn 1969 "Low Income Medical Problem Patients: Some Medical and Behavioral Features." Journal of Health and Social Behavior 10:334-343. Fanning, D. M. 1967 "Families in Flats." British Medical Journal 18:382- 386. Farmer, R. O. T. and P. G. Harvey 1975 "Minor Psychiatric Distrubances in Young Adults." Social Science and Medicine 9:467-474. Feld, S. 1963 "Feelings of Adjustment." Pp. 331-352 in I. Nye and L. Hoffman (eds.), The Employed Mother in America. Chicago: Rand McNaTTy. Feldstein, P. J. 1966 "Research on the Demand for Health Services." Milbank Memorial Fund Quarterly 44:128-162. Ferris, A. 1971 Indicators of Trends in the Status of American Women. New York: Russell Sage Foundation.. Fine, S. 1974 "Troubled Families: Parameters for Diagnosis and Strategies for Change." Comprehensive Psychiatry 15:73-77. Ford, A. B., R. E. Liske and R. S. Ort 1962 "Reactions of Physicians and Medical Students to Chronic Illness." Journal of Chornic Diseases 15:785-794. Ford, A. B., R. E. Liske, R. S. Ort and J. C. Denton 1967 The Doctor's Perspective: Physicians View Their Patients and Practice. Cleveland: The Press of Case Western Reserve University. Fox, R. C. 1957 "Training for Uncertainty." Pp. 207-241 in R. K. Merton, G. G. Reader and P. L. Kendall (eds.), [pp gtudent-Physician. Cambridge: Harvard University ress. Frank, A., S. Eisenthal and A. Lazare 1978 "Are There Social Class Differences in Patients' Treatment Conception?“ Archives General Psychiatry 35:61-69. Freeman, H. 1963 Freidson, E. 1960 1961 1962 1970 1973 447 E. and O. 6. Simmons The Mental Patient Comes Home. New York: John Wiley and Sons. "Client Control and Medical Practice." American Journal of Sociology 65:374-382. Patients' Views of Medical Practice: A Studyiof Sub- scribers to a Prepaid Medital Plan in the Bronx. New York: Russell Sage Foundation. "Dilemmas in the Doctor-Patient Relationship." Pp. 207- 224 in A. E. Rose (ed.), Human Behavior and Social Processes. Boston: Houghton Mifflin Campany. Profession of Medicine. New York: Dodd, Mead and Company. "Prepaid Group Practice and the New 'Demanding Patient.'" Milbank Memorial Fund Quarterly 51:473-489. French, J. R. P., W. Rodgers and S. Cobb 1974 Gans, H. J. 1962 1967 Garland, L. 1959 Geertson, H. 1970 Gertman, M. 1974 Glogow, E. 1973 "Adjustment as Person Environment Fit." Pp. 78-101 in G. Coelho (ed.), Coping. New York: Basic Books. The Urban Villagers. New York: The Free Press. The Levittowners. London: The Penguin Press. H .tudies of the Accuracy of Diagnostic Procedures." American Journal of Roentgen 82:25-38. R. and R. M. Gray "Familistic Orientation and Inclination Toward Adopting the Sick Role." Journal of Marripge and the Family 32:638-646. "Physicians as Guiders of Health Services Use." Pp. 362-382 in S. J. Mushkin (ed.), Consumer Incentives for Health Care. New York: Prodist. "T98 'Bad Patient' Gets Better Quicker." Social Policy 4: 2-76. 448 Goering, J. M. and R. M. Coe 1970 "Cultural Versus Situational Explanations of the Medical Behavior of the Poor.“ Social Science Quarterly 51: 309-319. Gold, M. "A Crisis of Identity: The Case of Medical Sociology." 1977 Journal of Health and Social Behavior 18:160-168. Gordon, G. 1966 Role Theoryyend Illness. New Haven: College and University Press. 6055, M. E. W. 1969 "Influence and Authority Among Physicians in an Out- patient Clinic." Pp. 275-291 in A. Etzioni (ed.), A Sociological Reader of Complex Organizations. New YOrk: Holt, Rinehart and Winston, Inc. Gove, W. R. and J. Tudor 1973 "Adult Sex Roles and Mental Illness." American Journal of Sociology 78:812-821. Graham, S. 1957 "Socio-Economic Status, Illness and the Use of Medical Services." Milbank Memorial Fund Quartegly 35:58-66. Gray, R. M., W. R. Newman and A. M. Reinhardt 1966 "The Effect of Medical Specialization on Physicians' Attitudes.“ Journal of Health and Human Behavior 7:128-132. Guilford, J. P. 1954 Psychometric Methods. New York: McGraw-Hill Book Company. Gurin, G., J. Veroff and S. Field 1960 Americans View Their Mental Health. New York: Basic Books. Haggerty, R. J. 1965 "Family Diagnoses: Research Methods and Their Reliability for Studies of the Medical-Social Unit, The Family.” American Journal of Public Health 55: 1521-1533. Hall, 0. 1948 "The Stages of a Medical Career." American Journal of Sociology 53:327-336. 449 Hamburg, D. A. and J. E. Adams 1967 "A Perspective on Coping Behavior.“ Archives General Psychiatry,17:277-284. Haney, J. Allen 1975 "Psychosocial Factors Involved in Medical Decision Making." Pp. 420-432 in T. Millin (ed.), Medical Behavioral Science. Philadelphia: W. B. Saunders Company. Harris, R. J. 1975 A Primer of Multivariate Analysis. New York: ACademic Press, Inc. Hemminki, E. 1974 "The Effect of a Doctor's Personal Characteristics and Working Circumstances on the Prescribing of Psycho- tropic Drugs." Medical Care 12:351-357. Henderson, L. J. 1935 "Physician and Patient as a Social System." New Epgland Journal of Medicine 212:819-823. Henry, J. P. and P. M. Stevens 1977 Stress, Health and the Social Environment: A Socidlogical Approach to Medicine. New York: Springer- VerTOQ. Hershey, N. 1972 "The Defensive Practice of Medicine-Myth or Reality." Milbank Memorial Fund Quarterly 50:69-98. Hesbacher, P. T., K. Rickels and D. Goldberg 1975 "Social Factors and Neurotic Symptoms in Family Practice." American Journal of Public Health 65:148-155. Hill, R. and J. Aldous 1967 International Bibliographyof Research in Marriage and Family. Minneapolis: University of Minnesota Press. Hochstim, J. R. 1968 "Poverty Area Under the Microscope." American Journal of Public Health 58:1815-1827. Hollingshead, A. B. 1967 "Two-Factor Index of Social Position." Pp. 384-388 in C. Bonjean (ed.), Sociological Measurements: An Inventory of Scales and’IndiEes. San Francisco: Chandler. 450 Hoppe, S. K. and P. L. Heller 1975 "Alienation, Familism and the Utilization of Health Services by Mexican-Americans." Journal of Health and Social Behavior 16:304-314. Howell, M. C. 1974 "What Medical Schools Teach About Women." New England Journal of Medicine 291:304-307. Hunter, A. 1975 "The Loss of Community: An Empirical Test Through Replication." American Sociological Review 40:537-552. Ima, K., D. M. Tagliacozzo and J. C. Lashof 1970 "Physician Orientation and Behavior: A Study of Out- patient Clinic Physicians." Medical Care 8:189-199. Illich, I. 1976 Medical Nemesis: The Expropriation of Health. New York: Pantheon Books. Kadushin, C. 1964 "Social Class and the Experience of 111 Health: The Need for Further Research." Sociological Inguiry 34:67-80. 1967 "Social Class and 111 Health: A Reply to Antonovsky." Sociological Inqgiryy37:323-332. Kaplan, B. H., J. C. Cassel and S. Gore 1977 "Social Support and Health." Medical Care 15:47-58. Kasl, S. V. and S. Cobb 1966 "Health Behavior, Illness Behavior and Sick Role Behavior." Archives Envirnomental Health 12:246-267. Kaufman, M. R. and S. Bernstein 1957 "A Psychiatric Evaluation of the Problem Patient: A Study of a Thousand Cases from a Consultation Service." Journal of American Medical Association 163:108-111. Kaywin, P. R. 1973 "Non-Thanksgiving Turkey." New England Journal of Medicine 289:1257-1261. Kellner, R. 1966 "Neurotic Symptoms in Women: Attendances in a General Practice." British Journal of Psychiatry 112:75-77. Kendall, P. 1957 Kessel, N. 1965 King, 5. H. 1962 Kirscht, J. 1971 Koos, E. 1954 1955 451 L. and H. Selvin "Tendencies Toward Specialization in Medical Training." Pp. 153-174 in R. K. Merton, G. G. Reader and P. L. Kendall (eds.), The Student-Physician. Cambridge: Harvard University PFess. "The Neurotic in General Practice." The Practitioner 194:636-641. Perceptions of Illness and Medical Practice. New York: Russell Sage Foundation. P. "Social and Psychological Problems of Surveys on Health and Illness." Social Science and Medicine 5:519-526. The Health of Regionville: What the People Thou ht and Did About it. New YOrk: CdlUmbia UniVersity ress. “Metropolis - What People Think of Their Medical Services.“ American Journal of Public Health 45:1551- 1563. Kriesberg, L. 1963 Kuch, J. H., 1977 Langner, T. 1960 Lazare, A., 1975 Lazerfed, P. 1954 "The Relationship Between Socio-Economic Rank and Behavior." Social Problems 10:334-352. S. S. Schuman and H. 8. Curry "The Problem Patient and the Problem Doctor or Do Quacks Make Crocks?" Journal of Family Practice 5:647-653. and S. Michael Life Stress and Mental Health. New York: The Free Press. 5. Eisenthal and L. Wasserman "The Customer Approach to Patienthood." Archives of General Psychiatry 32:553-558. F. and R. K. Merton "Friendship as Social Process: A Substantive and Methodological Analysis." P . 18-66 in M. Berger, T. Abel and C. H. Page (eds. , Freedom and Control in Modern Society. New York: van—Nostrand. Ley, P. and 1965 Lipsett, D. 1968 1970 Litwak, E. 1965 452 M. S. Spelman ”Communication in an Outpatient Setting." British Journal of Clinical Psychology 4:114-117. R. "The 'Rotating' Patient: A Challenge to Psychiatrists." Journal of Geriatric Psychiatry 2:51-61. ”Medical and Psychological Characteristics of 'Crocks.'" Psychiatry in Medicine 1:15-25. "Extended Kin Relations in an Industrial, Democratic Society." Pp. 292-323 in E. Shanas and G. F. Streib (eds.), Social Structure and the Family: Generational Relations. Englewood Cliffs: Prentice-Hall. Litwak, E. and I. Szelenyi 1969 Locke, B. Z. 1969 Lorber, J. 1975 Malcom, R., 1977 Mapes, R. 1977 Martin, W. 1957 "Primary Group Structures and Their Functions: Kin, Neighbors, and Friends." American Sociologlcal Review 34:465-481. and E. A. Gardner "Psychiatric Disorders Among the Patients of General Practitioners and Internists." Public Health Reports 84:167-173. "Good Patients and Problem Patients: Conformity and Deviance in a General Hospital." Journal of Health and Social Behavior 16:213-225. H. K. Foster and C. Smith "The Problem Patient as Perceived by Family Physicians." Journal of Family Practice 5:361-364. "Aspects of British General Practitioners' Prescribing." Medical Care 15:371-381. “Preferences for Types of Patients." Pp. 189-205 in R. K. Merton, G. G. Reader and P. L. Kendall (eds.), The Student-Physician. Cambridge: Harvard University Press. Matarazzo, R. G., J. D. Matarazzo and G. Saslow 1961 "The Relationship Between Medical and Psychiatric Symptoms." Journal of Abnormal and Social Psychology 62:55-61. McBroom, W. 1970 McCranie, E. 1978 McKinlay, J. 1972 1973 McKinlay, J. 1974 Mechanic, 0. 1962 1963 1965 1972a 1972b 1974 1976 453 H. "Illness, Illness Behavior and Socio-Economic Status." Journal of Health and Social Behavior 11:319-326. W., A. J. Korowitz and R. M. Martin "Alleged Sex-Role Stereotyping in the Assessment of Women‘s Physical Complaints: A Study of General Practitioners." Social Science and Medicine 12:111-116. 8. "Some Approaches and Problems in the Study of the Use of Services--An Overview.” Journal of Health and Social Behavior 13:115-152. "Social Networks, Lay Consultation and Help-seeking Behavior." Social Forces 51:275-292. B. and D. B. Dutton "Social-Psychological Factors Affecting Health Service Utilization." Pp. 251-303 in S. J. Mushkin (ed.), Consumer Incentives for Health Care. New York: Prodist. , “The Concept of Illness Behavior." Journal of Chronic Diseases 15:189-194. "Some Implications of Illness Behavior for Medical Sampling." New England Journal of Medicine 269:244-247. "Perception of Parental Responses to Illness." Journal of Health and Human Behavior 6:253-257. "Response Factors in Illness: The Study of Illness Behavior." Pp. 118-130 in E. G. Jaco (ed.), Patients, Physicianl and Illness. Second edition. New or : The Free Press. "Social Psychological Factors Affecting the Presentation of Bodily Complaints." New England Journal of Medicine 286:1132-1139. "Discussion of Research Programs on Relations Between Stressful Life Events and Episodes of Physical Illness." Pp 87-97 in B. S. Dohrenwend and B. P. Dohrenwend (eds.), Stressful Life Events: Their Nature and Effects. New , York: John Wiley and Sons. "Sex, Illness, Illness Behavior and the Use of Health Services." Journal of Human Stress 2:38-44. Mechanic, D. 1961 454 and E. H. Volkhart ”Stress, Illness Behavior, and the Sick Role." American Sociolpgical Review 26:51-58. Merton, R. K. 1957a 1957b Miller, P., 1976 Social Theory and Social Structure. New York: The Free Press. "Some Preliminaries to a Sociology of Medical Education." Pp. 3-79 in R. K. Merton, G. G. Reader and P. L. Kendall (eds.), The Student-Physician. Cambridge: Harvard UniversityPress. McD., J. G. Ingham and S. Davidson "Life Events, Symptoms and Social Support." Journal of Psychosomatic Research 20:515-522. Miller, S. J. 1970 Millon, T. 1975 Mitchell, J. 1969 Moore, W. E. 1949 Muller, C. 1965 1972 Mulroy, R. 1973 Mumford, E. 1970 Prescription for Leadership: Trainiggfor the Medical Elite. Chicago: Aldine Publishing ompany. "Introduction." Pp. 1-6 in T. Millon (ed.), Medical Behavior Science. Philadelphia: W. B. Saunders Company. C. "The Concept and Use of Social Networks." Pp. 1-50 in J. C. Mitchell (ed.), Social Networks in Urban Situations. Manchester: The University Press. and M. M. Tumin "Some Social Functions of Ignorance." American Socio- logical Review 14:785-791. "Income and the Receipt of Medical Care." American Journal of Public Health 55:510-521. "The Overmedicated Society: Forces in the Marketplace for Medical Care." Science 176:488-492. “Iatrogenic Disease in General Practice: Its Incidence and Effects." British Medical Journal 2:407-410. Interns: From Students to Physicians. Cambridge: HarvardUniversityPress. 455 Nathanson, C. A. 1975 “Illness and the Feminine Role: A Theoretical Review." Social Science and Medicine 9:57-62. National Center for Health Statistics 1974 1975a 1975b Navarro, V. 1976 Nelson, J. 1966 Nie, N. H., 1975 Nuckolls, K. 1972 The National Ambulatory Medical Care Survey: Symptom Classification. Series 2, No. 63. Rockville, Maryland: USS. Department of Health, Education, and Welfare. Health United States 1975. DHEW Publication No. (HRA) 76:1232. Rockville, Maryland: U.S. Department of Health, Education, and Welfare. Current Estimates from the Health Interview Survey: United States--1974. ’Series 10, No. 100. Rookville, ‘Maryland: U.S. Department of Health, Education, and Welfare. "The Industrialization of Fetishism: A Critique of Ivan Illich.“ Pp. 103-131 in V. Navarro, Medicine Under Capitalism. New York: Prodist. "Clique Contacts and Family Orientations." American Sociological Review 31:663-672. C. H. Hull, J. G. Jenkins, K. Steinbrenner and O. H. Bent Statistical Package for the Social Sciences. Second Edition. ’New York: MEGraw-Hill. 8., J. Cassel and B. H. Kaplan "Psychosocial Assets, Life Crises and the Prognosis of Pregnancy." American Journal of_§pidemiology 95:431-441. Nye, I. F. and F. M. Berarado (eds.) 1966 Ort, R. 5., 1964 Parsons, T. 1951 1964 Emerging Conceptual Frameworks in Family Analysis. New Ybrk: Macmillan Company. A. 8. Ford and R. D. Liske "The Doctor-Patient Relationship as Described by Physicians and Medical Students." Journal of Health and Human Behavior 5:25-34. The Socia1_System. Glencoe: The Free Press. Social Structure and Personality. New York: The Free ‘Press. Parsons, T. 1955 Parsons, T. 1958 Parsons, T. 1959 Paul, B. D. 1955 456 and R. Bales Famil Socialization and Interaction Process. Glencoe: The Free Press. and R. Fox “Illness, Therapy and the Modern Urban American Family." Pp. 234-245 in E. G. Jaco (ed.), Patients, Physicians and Illness. New York: The Free Press. and E. A. Shi1s "Values, Motives and Systems of Action.” Pp. 97-275 in T. Parsons and E. A. Shils (eds.), Toward a General Theory of Action. Cambridge: HarvardUniversitylPress. and W. B. Miller (eds.) Health, Culture and Community. New York: Russell Sage FOundation. Perlstadt, H. 1972a 1972b "Goal Implementation and Outcome in Medical Schools." American Sociological Review 37:73-82. "Internship Placements and Faculty Influence." Journal of Medical Education 47:862-868. Petras, J. W. and J. E. Curtis 1968 Petroni, F. 1969 Phillips, 0. 1965 Phillips, 0. 1969 Prather, J. 1975 "The Current Literature on Social Class and Mental Illness in America: Critique and Bibliography." Behavioral Science 13:382-398. A. ”Social Class, Family Size and the Sick Role." Journal of Marriage and the Family 31:728-735. L. "Self-reliance and the Inclination to Adopt the Sick Role." Social Forces 43:555-563. L., and B. E. Segal "Secual Status and Psychiatric Symptoms." American Sociological Review 34:58-72. and Fidell, L. S. "Sex Differences in the Content and Style of Medical Advertisements." Social Science and Medicine 9:23-29. Rabkin, J. G. and E. L. Struening 1976 Rahe, R. H. 1978 "Life Events, Stress, and Illness.“ Science 194:1013-1020. and F. J. Arthur “Life Changes and Illness Studies: Past History and Future Directions.” Journal of Human Stress 4:3-15. 457 Rainwater, L. 1964 "Marital Sexuality in Four Cultures of Poverty." Journal of Marriage and the Family 26:457-466. Reader, G. G. 1957 "The Cornell Comprehensive Care and Teaching Program." Pp. 81-101 in R. K. Merton, G. G. Reader and P. L. Kendall (eds), The Student-Physician. Cambridge: Harvard University Press. Reader, 6. 6., L. Pratt and M. C. Mudd 1957 "What Patients Expect from Their Doctors." The Modern Hospital 89:88-94. Reisman, L. and R. Platou 1960 "The Motivation and Socialization of Medical Students." Journal of Health and Human Behavior 1:174-182. Renne, K. S. 1971 "Health and Marital Experience in an Urban Population." Journal of Marriage and the Family 33:338-350. Richardson, W. C. 1969 "Poverty, Illness and Use of Health Services in the United States." Hospitals 43:34-40. 1970 "Measuring the Urban Poor's Use of Physicians' Services in Response to Illness Episodes." Medical Care 8:132-142. Rivkin, M. 1972 Contextual Effects of Families on Female Responses to Illness. Unpublished Ph.D. dissertation, Johns Hophins University. Roghmann, K. J. and R. J. Haggerty 1972 "Family Stress and the Use of Health Services." International Journal opridemiology 1:279-286. Rosenstock, I. M. 1960 "What Research in Motivation Suggests for Public Health.“ American Journal of Public Health 50:295-302. 1975 “Patients' Compliance with Health Regimens." Journal of American Medical Association 234:402-403. Ross, J. A. 1962 "Social Class and Medical Care." Journal of Health and Human Behavior 4:35-40. 458 Roth, J. A. 1962 "'Management Bias' in Social Science Study of Medical Treatment." Human Organization 21:45-49. Safilious-Rothschild, C. 1974 Women and Social Policy. Englewood Cliffs: Prentice- Hall. Salloway, J. C. 1973 "The Gypsies and the Urban Medical System." Urban Anthropology 2:113-126. Sapolsky, A. 1965 "Relationship Between Doctor-Patient Compatibility, Mutual Perception, and Outcome of Treatment." Journal of Abnormal Psychology 70:70-76. Saunders, L. 1954 Cultural Differences and Medical Care. New York: Russell SageFoundation. Scheff, T. J. 1972 "Decision Rules and Types of Error, and Their Conse- quences in Medical Diagnosis." Pp. 309-323 in E. Freidson, and J. Lorber (eds.), Medical Men and Their Work. Chicago: Aldine. Schuller, A. B. 1977 "About the Problem Patient." Journal of Family Practice 4:653-654. Secord, P. F. and D. W. Backman 1964 Social Psychology; London: McGraw Hill. Segal, B., E. Weiss and R. Sokol 1965 "Emotional Adjustments, Social Organization and Psychiatric Treatment." American Sociological Review 30:584-591. Seidenberg, D. 1971 "Drug Advertising and Perceptions of Medical Illness." Mental Hygiene 55:21-27. Shepherd, M., 8. Cooper, A. C. Brown, and C. W. Kalton 1966 Psychiatric Illness in General Practice. London: oxford University Press. 459 Shortell, S. M. 1972 A Model of Physician Referral Behavior: A Test of Exchange Theory in Medical Practice. lResearchlSeries No. 3l. Chicago: University of Chicago Center for Health Administration Studies. Shuval, J. T. [in collaboration with A. Antonovsky and A. M. Davies] 1970 Social Functions of Medical Practice. San Francisco: Jossey-Bass, Inc. Silverman, J. and P. R. Lee 1974 Pills,,Profits and Politics. Berkeley: University of California Press. Simon, A. J. 1966 "Social Structure of Clinics and Patient Improvement." Pp. 501-508 in W. R. Scott and E. H. Volkhort (eds.), Medical Care. New York: John Wiley and Sons, Inc. Stein, R. L. 1970 “The Economic Status of Families Headed by Women." Monthlnyabor Review 93:19-25. Stewart, M. A., I. R. McWhinney and C. W. Buck 1975 "How Illness Presents: A Study of Patient Behavior." Journal of Family Practice 2:411-414. Stimson, G. V. 1976 "General Practitioners, 'Trouble' and Types of Patients." Sociology Review Monograph 22:43-51. Stoeckle, J. D., I. K. Zola, and G. E. Davidson 1963 ”On Going to See the Doctor, The Contributions of the Patient to the Decision to Seek Medical Aid. ” Journal of Chronic Diseases 16:975-989. 1964 "The Quantity and Significance of Psychological Distress in Medical Patients." Journal of Chronic Diseases 17:959-970. Suchman, E. A. 1964 "Sociomedical Variations Among Ethnic Groups.“ American Journal of Sociology 70:319-331. 1965 "Stages of Illness and Medical Care." Journal of Health and Human Behavior 6:114-128. Sussar, M. W. and W. Watson 1962 Sociology in Medicine. New York: Oxford University Press. Sussman, M. 1953 Sussman, M. 1962 Szasz, T. S. 1975 Tagliacozzo, 1972 Tessler, R., 1976 460 B. "The Help Pattern in the Middle Class Family." American Sociological Review 18:22-28. 8. and L. Burchinal “Kin Family Network: Unheralded Structure in Current Conceptualizations of Family Functioning." Marriage and Family Livipg 24:231-240. and M. H. Hollender "A Contribution to the Philosophy of Medicine: The Basic Models of the Doctor-Patient Relationship.“ Pp. 432-440 in T. Millon (ed.), Medical Behavioral Science. Philadelphia: W. B. Saunders Company. D. L. and H. O. Mauksch "The Patient's View of the Patient's Role." Pp. 162-175 in E. G. Jaco (ed.), Patients, Physicians and Illness. New York: The Free Press. D. Mechanic and M. Dimond “The Effect of Psychological Distress on Physician Utilization: A Prospective Study." Journal of Health and Social Behavior 17:353-364. Thomas, D. H. 1976 Turner, C. 1967 Twaddle, A. 1968 1969 Udry, J. R. 1965 von Mering, 1972 Figurinngnthropology: First Principles of Probability and Statistics. New York: Holt, Rinehart and Winstow. "Conjugal Roles and Social Networks: A Re-Examination of an Hypothesis." Human Relations 20:121-130. C. Influence and Illness: Definitions and Definers of Illness Behavior Among Older Males ianrovidence,_Rhode Island. Unpublished Ph.D} dissertation, Brown University. "Decisions and Sick Role Variations: An Exploration." Journal of Health and Social Behavior 10:105-115. and M. Hall "Marital Role Segregation and Social Networks in Middle- Class Middle-Aged Couples." Journal of Marripge and the Family 27:392-395. 0. "The Diffuse Health Aberration Syndrome: A Bio- Behavioral Study of the Perennial Out-Patient." Psychosomatics 13:293-303. 461 von Mering, O. and L. W. Early 1966 "The Diagnosis of Problem Patients." Human Organization 25:20-23. 1969 "The Ambulatory Problem Patient: A Unique Teaching Resource." American Journal of Psychiatry 126:146-150. von Mering, O., G. Shannon, W. Deal and P. Fischer 1976 "A Long Day's Journey to Health Care." Human Organization 35:381-389. Waldman, W. E. and K. R. Gover 1972 "Marital and Family Characteristics of the Labor Force." Monthly Labor Review 95:4-8. Waldron, I. 1977 "Increased Prescribing of Valium, Librium, and Other Drugs - An Example of the Influence of Economic and Social Factors on the Practice of Medicine." Inter- national Journal of Health Services 7:37-62. Walpole, R. E. 1974 Introduction to Statistics. New York: Macmillan Publishing Co., Inc. Warren, J. V. and J. Wolter 1954 "Symptoms and Diseases Induced by the Physician. " General Practitioner 9:77-84. White, E. L. 1968 "A Graphic Presentation on Age and Income Differentials in Selected Aspects of Morbidity, Disability, and Utilization of Health Services." Inguiry 5:18-30. Wilson, R. N. 1963 "Patient-Practitioner Relationships.“ Pp. 273-295 in H. E. Freeman, S. Levine and L. E. Reeder (eds.), Handbook of Medical Sociology. Englewood Cliffs: Prentice-Hall, Inc. Young, M. and P. Willmont 1957 Family and Kinshipyin East London. London: Routledge and Kegan. Zborowski, M. 1952 "Cultural Components in Responses to Pain." Journal of Social Issues 8:16-30. Zelditch, M. 1955 Zola, I. K. 1963 Zola, I. 1966 1973 K. 462 "Role Differentiation in the Nuclear Family: A Comparative Study." Pp. 307-351 in T. Parsons and E. Bales (eds.), Family, Socialization, and Interaction Process. Glencoe: The Free Press. "Problems of Communication, Diagnosis and Patient Care: The Interplay of Patient, Physician and Clinical Organization." Journal of Medical Education 38:829-838. “Culture and Symptoms - An Analysis of Patient's Presenting Complaints." American Sociolpglcal Review 13:615-630. "Pathways to the Doctor - From Person to Patient." Social Science and Medicine 7:677-689.