gin-v" . a"... _ W .
5 ‘

M . LIBRARY

THESIS

Michigan State
University ~~

 

This is to certify that the

thesis entitled

JI'UL E-I 3"? CF .I'I I-IIII‘I TEL"
I ‘CDICT LG CUTCCI-IL C‘l‘ IULliDiocI LINEN—“(Y

r3137'3."""":3""" I." "T3fif’°I ," 7
LL\'~J.L511A_JA'J.L L. CR K’s.-\K./l _vl\ F.LII\J

Ipresented %y

r‘ *
x.

’ 'fi
Lu.)

[)1

J

GER LD DEIJNIS JUIIR

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Counseling,

 

 

Personnel Services and Educational

Psychology

Q/wflS/X [111/

Major pfétzessor

\ «a «\
Date lay/F0 emes Bngc—lkes, In“).

0-7639

OVERDUE FINES ARE 25¢ PER DAY
PER ITEM

   
   

Return to book drop to remove
this checkout from your record.

 

 

 

Copyright by C)
GERALD DENNIS JUHR

1980

DEVELOPMENT OF AN MMPI SUBSCALE
PREDICTING OUTCOME OF MULTIDISCIPLINARY TREATMENT
FOR CHRONIC LOW BACK PAIN

BY

Gerald Dennis Juhr

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services,
and Educational Psychology

1980

ABSTRACT
DEVELOPMENT OF AN MMPI SUBSCALE

PREDICTING OUTCOME OF MULTIDISCIPLINARY TREATMENT
FOR CHRONIC LOW BACK PAIN

BY

Gerald Dennis Juhr

The primary intent of this study was to explore
the feasibility of developing a subscale of items from the
Minnesota Multiphasic Personality Inventory for use in
predicting outcome of treatment for chronic low back pain.
Previous use of the MMPI in personality assessment of
chronic low back pain patients repeatedly has confirmed
the presence of psychoneurotic involvement among a major-
ity of such patients and strongly indicated that higher
levels of pre-treatment involvement are positively cor-
related with unsuccessful treatment outcome. Here the
attempt was made to develop an MMPI subscale enabling more
effective screening, diagnosis, and matching of these
patients with treatment strategies and resources.

Subjects for the study were 185 former in-patients
of a multidisciplinary back pain clinic, whose treatment
had consisted of neurosurgical (facet injection and facet
rhizotomy), psychological (EMG biofeedback and other

muscular relaxation procedures), and physical therapy.

Gerald Dennis Juhr

Subjects were assigned treatment outcome status based on
their responses to a mailed questionnaire, the Pain Survey,
which had been constructed and piloted expressly for this
research.

Of these subjects, two-thirds were randomly .
selected to comprise a scale-development subsample, and
the remaining one-third became the cross-validational sub-
sample. The former subjects' MMPI records were analyzed
by chi-square to determine which items best discriminated
between successful and unsuccessful subjects. ‘The 24 best
items constituted a tentative Back Treatment Success Scale,
whose ability to discriminate among members of the cross-
validational subsample was then tested by comparing mean
scale scores of successful and unsuccessful subjects with
a one-way analysis of variance. This test proved statis-
tically non-significant.

This result was at least partially attributable to
the size of the scale-development subsample. Therefore,
to provide as comprehensive a basis as possible for future
research, the two subsamples were pooled and the chi-square
test of item-discrimination was applied to the total sample
population. This yielded a total of 45 items discriminat-
ing at the .05 and .10 levels; these were examined for
standard MMPI-scale membership, categorized on the basis

of content, and statistically factor analyzed.

Gerald Dennis Juhr

The majority of the items belonged to just four of
the standard MMPI scales, the Depression, Psychopathic
Deviate, Schizophrenia, and F scales, though all scales
were represented by at least two items. Item-categoriza-
tion based on content permitted identification of the
following factors related to unsuccessful outcome:

(1) denial of social non-conformity, (2) self—deprecatory
attitudes tending to guilt and paranoia, (3) health com-
plaints and disease phobia, (4) impaired faculties of con-
centration, coordination, and awareness, (5) depressed
affect and behavior, (6) repressed hostility and authority
problems, (7) non-affirmation of fundamentalist religious
beliefs, (8) attraction by members of same sex, and

(9) excessive use of alcohol. Statistical factor-
analysis of the 45 items led to identification of eight
response-profiles that appeared similar to personality
profiles frequently associated with chronic low back pain
patients. Among these were hypochondriasis, reactive
depression, somatization, manipulation for secondary gain,
family conflict, and relative freedom from psychopathology
(low rate of item-endorsement in unsuccessful direction).

These findings were not cross-validated on another
sample. The coincidence of the content- and factor-
analyses with previous research findings, however, suggests
that the 45 items here selected and examined may provide
a valid basis for the eventual development of such a sub-

scale as was here intended.

To my forthright and loving parents,
Hans and Anita,
who have dedicated so much
to my growth and education
and
To my dear wife,
Janneke,

beneficent helpmeet

ii

ACKNOWLEDGEMENTS

Tibetan therapeutic lore maintains that there are
three pre-requisites to any successful treatment -- the
belief of the patient, the belief of the healer, and the
right relationship between patient and healer. This dictum
was at least in part substantiated by the research reported
Vin the following pages. At this point, however, I wish to
acknowledge the "right relationships" -- the favorable
disposition and actions of many persons other than myself --
which proved vital to the successful treatment of this
dissertation. I am truly grateful to them.

The conception and design of this research resulted
from conversations with John Jerome, Ph.D., psychologist
of the Ingham Low Back and Pain Clinic. He and James
Bullock, M.D., the clinic's orthOpedic surgeon, encouraged
me to undertake this project and did much to make it
possible. They solicited their patients' co-operation,
and allowed me access to their records. I was made to feel
welcome in their offices and given timely assistance by
their clerical staff.

Many chronic pain patients whom I have never met

responded to the survey. Some shared their continuing

iii.

pain; others shared their relief. Their willingness to
communicate was indispensable, and is sincerely appreciated.

The guidance of Dr. James Engelkes greatly improved
the quality of this undertaking and assured its successful
completion. Likewise, Drs. Gregory Miller and Bob Winborn
gave generously of their time to serve on my guidance
committee and review my research.

My special thanks to Lee Burkhardt for her competent,
efficient preparation of the manuscript.

Friends and colleagues at the University Centers
for International Rehabilitation with their daily interest
and assistance made this work more enjoyable and sustained
my spirits. I shall especially remember the wit and
patience of Jim Mullin, who taught me to speak without
expletives to the computer, and the personal warmth and
camaraderie of Madan Kundu, Denise Tate, Bill Frey, Ron
Wolthuis, Paul Schneider, Bob Jarvis, and Owen Dailey.

I remain indebeted to Don Melcer, friend and
mentor, who valued the differentness of the educational
paths I have taken, and whose steadiness and example
helped me negotiate the path through graduate school.
During this time Don's wife, Eleanor, watched over my
family's welfare and contributed often to our happiness.

Our growing friendship with David and Peggy Rolfe

added a stimulating, much appreciated dimension to my

iv

family's life during the past year of labor, and banished
often the lurking specters of obsession and depression.
David also spurred my professional development in many
ways that have helped me anticipate an active professional
life.

Most of all I wish to acknowledge the devotion and
understanding of my wife, Janneke, and our children,
Niels and Fiona. We have come this far together, and I
intend to share with you more completely, joyfully, the

adventures still before us.

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . . . .

Chapter

I.

II.

III.

IV.

THE PROBLEM . . . . . . . .

Purpose . . . . . . . . .
Importance of the Research
Definition of Terms . . . .
Summary and Overview . . .

REVIEW OF LITERATURE . . .

The Human Experience of Pain

Personality Inventories and

of Chronic Low Back Pain
Treatment Studies . . . . .
Implications and Summary .

METHOD OF RESEARCH . . . .

Experimental Sample . . . .
Instrumentation . . . . . .
MMPI . . . . . . . . . .
Pain Survey . . . . . .
Procedure . . . . . . . .
Statistical Analysis . . .

ANALYSIS OF RESULTS . . . .

Pain Survey . . . . . . . .

Back Treatment Success Scale

Sufferers

Re-analysis of Data for Total Sample

Population . . . . . . .
Summary of Results . . . .

DISCUSSION OF RESULTS . . .
Summary . . . . . . . . . .

Discussion . . . . .
Limitations of the Study .

Implication for Future Research . . .

Conclusions . . . . . . . .

vi

Page

viii

10
11

13
14

21
30
38

41

42
44
44
47
49
57

60

61
62

67
82

84

84
86
95
98
101

APPENDICES

REFERENCES

Pain Survey . . . . . . . .

Cover Letter Accompanying Pain

Endorsement Percentages for MMPI
Form—R Items by Successful vs.

Loadings on Eight Factors of 4S
MMPI Form-R Items Discriminating
Between 76 Successful and 109
Unsuccessful Patients Treated

Survey . . . . . . . . .

Unsuccessful Subjects (N = 185)

for Chronic Low Back Pain

vii

Page
103

103

105

106

113

114

LIST OF TABLES
Table Page
3.1 Pilot Study of Pain Survey; Ability of

Individual Foils to Discriminate
Between Successful and Unsuccessful

Former Patients . . . . . . . . . . . . 52
3.2 Comparison of Demographic Data Obtained

from Pilot and Research Samples . . . . 55
3.3 Endorsement of Criterion Questions by

Pilot vs. Research Samples . . . . . . 56
4.1 Response to Pain Survey of Treatment

Success and Treatment Failure Groups . 63
4.2 Back Treatment Success Scale; Direction

of Item Response Endorsed More
Frequently by Successful vs. Un-

successful Subjects (N = 185) . . . . . 65
4.3 Analysis of Variance of Back Treatment

Success Scale Scores of 26 Successful

and 36 Unsuccessful Subjects . . . . . 66
4.4 Scale Categorization of Items Dis-

criminating at .10 Level Between 76
Successful and 109 Unsuccessful Back

Clinic Patients (N = 185) . . . . . . . 68
4.5 MMPI-Item Loadings on Factors Related to

Successful/Unsuccessful Treatment

Outcome . . . . . . . . . . . . . . . . 76
4.5.1 Factor 1: Hypochondriasis . . . . . . . 76
4.5.2 Factor 2: Organic . . . . . . . . . . . 77
4.5.3 Factor 3: Manipulative . . . . . . . . . 77
4.5.4 Factor 4: Reactive Depression . . . . . 78
4.5.5 Factor 5: Familial Conflict . . . . . . 78

viii

Table Page
4.5.6 Factor 6: Somatization . . . . . . . . . . 79
4.5.7 Factor 7: Hypochondriasis . . . . . . . . 79

4.5.8 Factor 8: General Neurotic . . . . . . . . 80

ix

CHAPTER I

THE PROBLEM

Americans spend more than 100 billion dollars per
year on health care. Of this amount, nearly 13 billion
dollars is spent by persons seeking relief from chronic
low back pain (Fordyce, 1976a), while the national yearly
cost of related compensation, lost wages, and lost potential
tax revenue conservatively may be estimated at an additional
25 billion dollars (Leroux, 1979). It is evident that in
the United States, the reported frequency of low back
injuries is increasing decade by decade (Sternbach, Wolf,
Murphy, and Akeson, 1973b). There is an evergrowing popula-
tion of chronic low back pain patients who have failed to
respond to medical, physical, and psychophysiological treat-
ments for relief of pain (Bonica, 1976; Kraus, 1976;
Melzack, 1973; Shealy, 1976), and disproportional payments
for this one syndrome are seriously jeopardizing the
functioning of employment compensation programs (Finneson,
1976).

Russek (1955) reported that in 1955 chronic low
back pain sufferers accounted for 12.4% of all industrial

injuries and 16% of all compensation payments in the State

of New York. Statistics from the State of Washington
Department of Labor and Industries (McGill, 1968) revealed
that back injuries constituted 5% of industrial claims,
12% of contested settlements, and 24% of days lost. In the
State of California in 1970, 37.8% of all newly filed
industrial claims were based on low back pain, and this
percentage of back settlements had steadily increased from
29% in 1961 to 39.3% in 1969 (Osterloh, 1971, cited in
Sternbach et a1., 1973b). Comparatively, from 1971-1973
in the Commonwealth of Pennsylvania only 9% to 11% of
industrial injuries were classified as low back injuries,
but 30% to 40% of all compensation payments were for low
back pain (Bureau of Vocational Rehabilitation, Common-
wealth of Pennsylvania, cited in Finneson, 1976). More

recent statistics cited in the Rehab Brief (University of

 

Florida Rehabilitation Research Institute, 1978) showed
that claims related to low back pain account for 85% of
California's workmen's compensation medical budget.
According to Department of Labor Statistics of the State
of Michigan (Pinto, 1979), lower and upper back injuries
comprised the largest single class (24.5%) of industrial
injuries.

Such findings also point to the significance of
chronic low back pain as a rehabilitation problem.
Finneson (1976) reported that 281 low back pain patients

receiving employment compensation and successfully treated

by nonsurgical means averaged 36 days of total disability
per episode. White (1969) found that 4 years after surgical
treatment, only 39.5% of low back pain patients had re-
turned to work comparable to that performed prior to
surgery.

Despite the frequency and persistence with which
this syndrome has been presented to the medical profession,
and despite its enormous consequences to the individual
sufferer and to society, chronic low back pain "...remains
a baffling, frustrating, and elusive problem to clinical
practitioners" (Blumetti and Modesti, 1976). A recent
important response to this problem has been the upsurge of
interest in a multidisciplinary pain clinic approach to
complex chronic pain problems (Bonica, 1976). These
clinics combine in an inpatient setting several treatment
modalities such as orthopedic surgery and neurosurgery,
chemotherapy, relaxation therapy, individual and group
counseling, withdrawal of analgesic medication, exercise,
physical therapy, and regulation of diet (e.g., Sternbach,
1974; Shealy, 1976; Pheasant, 1972).

Increasingly, as part of this development, the
discipline of psychology has been called upon to explain
chronic low back pain as a psychophysiological condition
with substantial psychological components. Treatment pro-
grams have incorporated such diverse approaches as operant

conditioning (Fordyce, 1976b), psychoanalytic counseling

(Sarno, 1976), biofeedback and relaxation techniques
(Shealy, 1977; Bullock, Jerome and Pool, 1975), life—
situational counseling (Sternbach, 1974), and family-
oriented treatment (Hudgens, 1979) in the attempt to
ameliorate chronic pain. Reported success rates of 60 to
75% have confirmed the efficacy of these interventions.
Nevertheless, even with intensive screening to eliminate
the especially poor candidates, 25 to 40% of patients

seen in these clinics failed to achieve significant relief
from pain (Shealy, 1977; Sarno, 1976; Bullock, 1977).
These "low back losers" as one clinical team (Sternbach
et al., 1973b) has termed them, continued to seek treat-
ment, but consistently failed to find relief from their
pain; moreover, each subsequently unsuccessful treatment
further complicated their problem (Shealy, 1976; White,
1966) and further reduced the odds for recovery (Wilfling,
Klonoff, and Kokan, 1973). Clearly, there is a growing
pool of chronic low back pain sufferers who have failed and
who continue to fail to respond to the available resources
of the healing profession. Further investigation of
factors related to failure, and to success, appeared

warranted.

Purpose

The purpose of this study was to investigate the
psychological (personality) characteristics of those

persons who succeed, and those who fail, in multidisciplinary

inpatient treatment for chronic low back pain. Patients'
pre-treatment responses to The Minnesota Multiphasic
Personality Inventory (hereafter: MMPI) were statistically
analyzed to determine which items significantly differen-
tiated between patients with successful and those with
unsuccessful outcomes. These significantly differentiat-
ing items comprised a Back Treatment Success Scale, a
subscale of MMPI items designed to predict success or failure
in treatment for chronic low back pain. Additionally, the
items selected were examined on the basis of content

to facilitate identification and discussion of factors

related to success or failure in treatment.

Importance of the Research

 

The past four decades have witnessed a growing
awareness by health care professionals that the onset of,
reaction to, and outcome of treatment for chronic low back
pain are in many cases highly dependent upon psychological
factors (e.g. Fetterman, 1937; Sargent, 1946; Sullivan,
1955; Yochelson, 1966). The need for an empirically de-
veloped psychological test predictive of individual out-
come of treatment for this syndrome has been recently dis—
cussed (Waring, Weisz, and Bailey, 1976). These authors
reviewed and attempted to validate extant treatment studies
in which outcomes were correlated with patients' psycho-
logical characteristics, and stated that results were in-

conclusive and of doubtful clinical validity. Indeed, few

such studies have been published. One major study in this
review (Wiltse and Rocchio, 1975) and one not included
(Lippincott, 1976) were unfortunately conducted on a
sample of patients whose low back pain had been treated by
chemonucleolysis, a chemo-surgical procedure later proven
ineffective except as a placebo (Martins et a1., 1978).
The purpose of Lippincott's (1976) study, the development
of an MMPI subscale predictive of treatment outcome, was
synonymous with that of the present research. The fact
that the treatment enjoyed by her sample later proved to
be a placebo, is definitely of interest, and her study is
worthy of reinterpretation in light of this finding.
Nevertheless, an important need remained to be met through
the development of such a scale based on treatments of
proven effectiveness. The medical and psychological treat-
ments administered to patients in this study (EMG bio-
feedback with or without facet rhizotomy) are of proven
effectiveness (Shealy, 1977; Bullock, Jerome, and Pool,
1975; Bullock, 1977; Jerome, 1978).

Several potential clinical applications of such a
scale have been suggested. First, multidisciplinary teams
could utilize such an assessment tool in their evaluation
of patients for treatment. McGill (1968, p. 176) has
stated that improper evaluation of the "subjective com-
plaints of compensation patients with low back pain creates

great adverse psychological effect on the already anxious,

apprehensive patient, with the result being prolonged
absenteeism." Hoover (1968) of the Mayo Clinic has stated
that psychiatric illness may prevent a patient from obtain-
ing the intended relief from surgery.

"Evaluation of the contribution of emotional

instability to the patient's pain is by far

the most difficult problem and the source of

greatest error in making a decision to

fusion." (p. 192)
White (1966) has maintained and Wilfling et a1. (1973)
systematically have demonstrated that each successive un-
successful surgery reduces the probability that a patient
will ultimately experience relief from chronic low back
pain. This correlation may be due to somatic complica-
tions (irritating scar tissue, post-operative pain) as
mentioned by Shealy (1976), or to the psychological con-
sequences of failure (learned helplessness, somatic pre-
occupation) as suggested by Sternbach (1974), or to a
learned life-style (White, 1966), or simply to the
intractability of the physical condition, a suggestion
which no one has made except in certain categories of
clearly identifiable organic pathology (e.g. infection,
neoplasm, rheumatoid diseases); the fact remains that
practical identification of crucial psychological factors
would assist a pain specialist in the decision to recommend
surgery or not.

Second, the identification of psychological factors

associated with treatment failure may lead to the

recommendation that a patient undergo intensive counseling
prior to re-application for surgical procedures (Sternbach,
1974). Either the counseling component of the regular
treatment program may be expanded to accommodate certain
individuals, or extramural referrals can be made. Such a
procedure emphasizes the best matching of patient with
treatment and facilitates the optimum use of available
rehabilitation resources.

Third, the collection of psychological data may
provide an opportunity for realistic discussion with the
patient of factors which could impede efforts to seek
relief from pain (Sternbach, 1974). Many chronic pain
patients have little or no conception of psychosomatic
dysfunctions and easily assume that any mention of emo-
tional factors is an attempt to convince them that their
pain is "imaginary," "all in their heads," or "just nerves."
Reference to patients' scores on an objectively scored,
normed personality inventory may provide a bridge to the
discussion of cognitive, emotional, and motivational
factors affecting the experience of pain. When this occurs,
patients can begin to accept responsibility for their role
in the maintenance or eradication of pain, an aspect of
treatment deemed highly important under the holistic or
"whole person" approach to treatment espoused by numerous

pain clinics (Shealy, 1976; Jerome, 1978).

Fourth, an expeditious evaluation may enhance the
likelihood that patients will return to productive activity
before becoming massively conditioned to pain. Lamaze
(1970) has described the work of Pchonick and of Rogov,
who in separate experiments were able to condition subjects
to experience a neutral stimulus as painful, and a painful
stimulus as pleasant. In sufferers of chronic pain, such
conditioning eventually contributes not only to the per-
vasiveness of the experience of pain, but also to the in-
creasing attempts to withdraw from activities associated
with pain (Fordyce, 1976b). Effective reinforcement in
such cases is forthcoming from secondary gains either in
the form of compensation payments (Finneson, 1976) or in
the form of sympathy or other considerations shown a
"sick" person (Foster, 1964). A quickly administered test
allowing timely diagnosis to interrupt or prevent the
adoption of such a lifestyle may be considered a worthwhile
contribution to the fields of pain treatment and research.

The challenge and the importance of devising such
an instrument has been well summarized by Thomas and Lyttle
(1976). These authors have noted that

"...certain patient personality profiles on

the MMPI appear closely associated with

chronic complaints of low back pain [and

that] predictions made from psychological

data were more accurate than predictions

based solely on the physical conditions

of the patients...[0]ne would expect

psychological assessment to be well inte-

grated into the diagnostic procedures of
medical therapy programs that involve

10

patients with complaints of low back pain.
However, psychological evaluation is not
a routine part of diagnostic procedures,
probably because there is no single
reliable and easily identifiable psycho-
logical variable that predicts prognosis
with as much accuracy as the clinical judge-
ment of a psychologist based on multiple
psychological tests and an interview.

The problem then is the fact that many
orthopedists in general practice do not
have easy access to a psychologist's
evaluation of their patients." (p. 125)

The attempt was made in this study to develop a diagnostic
scale based on reliable psychological variables. Whether
or not this attempt was entirely successful, the empirical
derivation of factors related to treatment outcome may be
viewed as a worthwhile contribution toward the development

of such a diagnostic tool.

Definition of Terms

 

Chronic low back pain - Operationally is pain described by
the patient as moderate to severe, that has been
persistent for six months or longer, and has failed
to respond to medical, physical or psychological
treatment (Jerome, 1978). Nosologically is pain
situated in the lumbosacral region at the level of
the third, fourth and fifth lumbar vertebrae and
the sacrum, which is not the result of fractures,
neoplasms, infective lesions, congenital anomalies,
or diseases referred from other organic systems

(Thomas and Lyttle, 1976).

11

Facet Rhizotomy - Is a technique which involves thermo-
cautery to produce denervation at the facet joint.
Facet denervation attempts to block neural trans-

mission at the origin of the pain.

Psychological variables - Refers to cognitive, affective/

 

evaluative, and motivational factors residing
within the individual that can be perceived, ob-
jectively described, and measured, the organization
of which modulates the individual's experience of

pain and differentiates him/her from other persons.

Summary and Overview

 

This study has been designed to explore the rela-
tionship between relief from chronic pain and individual
psychological variables. Clinical research and practice
suggest that such variables play an important role in
modulating the human experience of pain and in affecting
the outcome of treatment for relief of chronic pain. A
more thorough understanding of this relationship is
warranted. The procedures followed in this study have been
intended not only to lead to an identification of such
critical modulating variables, but also to result in the
development of an MMPI subscale capable of practical
application in the prediction of treatment outcomes for

chronic low back pain.

12

This chapter has presented evidence of the potential
viability and practical utility of such a study. Chapter
II reviews current theory ascribing a decisive role to
psychological variables in the human experience of pain.
Findings resulting from the previous use of personality
inventories with chronic low back pain patients are reviewed,
with primary attention to those studies incorporating the
MMPI, the inventory used in the present study. The
majority of such studies have demonstrated the validity and
reliability of the MMPI as an indicator of psychophysio-
logical involvement and treatment prognosis of patients
having chronic low back pain.

The sample population, research procedures and
design of this study are the tOpic of Chapter III, the
statistical results of the research the topic of Chapter IV.
These results are then discussed in Chapter V, with

particular attention to their implications for treatment

and future research.

CHAPTER II

REVIEW OF LITERATURE

That psychological variables play a significant
role in the origin and longevity of the syndrome known as
chronic low back pain has been reported with increasing
frequency in the literature of the past four decades. The
particular action and interaction of these variables has
been the subject of considerable recent research. A more
complete understanding of these processes and the applica-
tion of this knowledge to clinical practice is the subject
of the present study.

To better understand the role of psychological
factors in the human experience of pain, it is necessary
to review the various theories of what pain really is.

The review leads to the Melzack and Wall (1965) formulation
of the Gate Control Theory of Pain, which described a
physiological basis for the role of psychological (cognitive,
affective, and motivational) variables in the pain process.

The first section of this chapter therefore re-
capitulates the Melzack and Wall conceptualization as a
basis for understanding pain as a multidimensional ex-

perience, involving both physical and psychological domains.

13

14

The second section summarizes the research relating chronic
pain in general and chronic low back pain specifically to
individual personality factors. The final section reviews
previous studies which have utilized standardized per-
sonality inventories to investigate the interaction be-
tween personality variables and the outcome of treatment

for chronic low back pain.

The Human Experience of Pain

 

The common view of pain is that it is a biologically
useful sensation informing the organism of potential or
actual damage or harm. The neural links between an
organism's sensory capacity and its motor capacity enable
the transmission of pain signals to activate the organism
to respond, either by defensively removing itself from
the noxious stimulus or by aggressively dealing with its
source.

Pain has traditionally been conceptualized by
scientific investigators as an objective stimulus related
to noxious levels of warmth, pressure, etc., which is
transmitted by neural pathways to the subjective potential
(brain) of the organism. In fact, the traditional approach
of science has sought to explain all human perception in
terms of direct linear causality and to investigate it by
the increasingly refined dissection of the neurological
systems responsible for the transmission of sensation from

sensing organ to perceiving brain (Steiner, 1962). Man

15

has thus been considered a passive recipient of impressions
from an objective world external to him. Within such a
framework, pain has been viewed as a strictly physiological
sensory experience.

Typical of such conceptualizations of pain were
the specificity theories, which defined pain as a primary
sensation with special peripheral receptors, neuronal trans-
mitters, and receivers in the central nervous system. The
pathway from periphery to center was seen as an uninterrupted
transmission system, with the intensity of perceived pain
in direct proportion to the intensity of the stimulus
applied. The relevant peripheral receptors were believed
to be specialized for the sensation of pain and distinct
from other main groups of sensory receptors, for instance
those for mechanical or thermal stimuli.

The specificity theories cannot account for the
vast individual differences in subjective perception of a
constant noxious stimulus. The subjective report of pain
is a notoriously unreliable measure of objective stimula-
tion (Sternbach, 1974). Even though these theories have
proved to be of heuristic value in research on the mechanics
of pain transmission, the deficiencies of the theories are
serious. Melzack and Wall (1965) have placed specificity
theory in the following perspective:

Physiological specialization is a fact that

can be recognized without acceptance of the

psychologic assumption that pain is deter-
mined entirely by impulses in a straight-

16

through transmission system from the skin

to a pain center in the brain. (p. 972)

Pattern theories arose in response to the defi-
ciencies of specificity theory. These theories stated that
information generated by peripheral receptors is coded in
the form of patterns of nerve impulses. The peripheral
receptors are sensitive to pain, a separate system of
neuronal fibers transmits the information to the brain,
and the brain interprets the patterns of impulses as pain
(Head, 1920; Lewis, 1942; Noordenboos, 1959). Essentially,
pattern theories attempted to account for the complexity
of the pain experience by referring to the encoding and de-
coding of neural impulses. Despite the complexity of the
theory, man in this model remained a passive recipient of
the pain sensation. With both the pattern and the
specificity theories, the psychology of experience was re-
duced to the physiology of stimulation and transmission.

Medical practice has long been based on these models.
Treatment has attempted to relieve pain by eradicating the
"painful" stimulus, either by treating directly the injured
or diseased organ, or by blocking the transmission of
sensation by means of analgesic medication. An example of
‘ the compelling power of such models in the interpretation
of basic data has occurred in conjunction with the practice
of administering morphine to persons suffering from severe
pain. For decades it was assumed that morphine acted

directly upon the neural capacity by inhibiting the

1?

transmission of painful stimuli. Not until the late 1950's
did Beecher (1959) conclusively demonstrate that morphine
inhibited the subjective reaction related to anxiety about
pain, and not the function of neural transmission.

This and other research contributed to the growing
recognition that, psychologically, man is not a passive
but an active participant in the experience of pain.
Anticipation of pain, anxiety and attention (Hill, 1952),
cultural background (Chapman, Finesinger, Jones, and Cobb,
1947), early experience (Melzack, 1973) and prior condi-
tioning (Pavlov, 1927) were all shown to have a profound
effect on both pain experience and response. From a
strictly phenomenological viewpoint, it would seem obvious
that human beings play an active role in responding to
pain. Common responses include the cognitive effort to
identify the noxious stimulus, emotional reactions such as
anguish or indifference, and the motivational response of
fight or flight. On the other hand, basic evidence exists
that man plays an active role not only in the response to
pain but also in the perception of pain. Much of this
evidence was incorporated by Melzack and Wall (1965) in
their formulation of the Gate Control Theory of Pain.

The Gate Control Theory of Pain proposed by Melzack
and Wall (1965) provided a basis for considering the active
role played by cognitive, affective, and motivational

factors in the actual perception of pain. In so doing, it

 

18

made an important distinction between the perception of
pain and the sensation of pain, by insisting that all per-
ception is an active response to sensation. 'Sensation is
an available "given," while the act of perception is
interpretive, creative.

The perception of pain, according to Melzack and
Wall (1965), has three active dimensions, each of which is
associated with a particular neural organization or

system: Sensory-discriminative activity enables a person

 

to locate the painful stimulus in time and space. The

motivational-affective system relates to the responsiveness

 

to noxious stimulation and to emotional input such as fear

and anxiety. The cognitive-evaluative system has the

 

capacity to act very rapidly in identifying, evaluating,
and selectively modifying the sensory input; through this
system past experience, disposition, and attention all
exert their influence in evaluating the input as threaten-
ing or not. This analysis of input rapidly interacts with
the motivational-affective and sensory-discriminative
functions to compare the stimulus with other input and
with memory, and to bring into action response strategies.
As a result, the cognitive-evaluative system has the capacity
directly to modulate sensory input before it is transmitted
to the sensory-discriminative and the motivational—
affective systems. The temporal priority of this system

is an important aspect of the Gate Control Theory, as it

19

implies that cognitive activity can intercept and either
minimize or exaggerate sensory input before a person has
an opportunity to make an affective/motivational response
to the input as "painful."

The probability that "higher central nervous system
activity" plays a critical role in the perceived intensity,
duration, and significance of pain, a position now firmly
supported by experimental evidence (Wall, 1976), may
justifiably be said to have been anticipated by the
phenomenological movement in twentieth-century philosophy.
The phenomenologists have insisted that all human percep-
tion is determined by each individual's total disposition
toward the object of perception. This disposition involves
a set of cognitive, affective and conative functions sub-
sumed in the concept of "intentionality" (Husserl, 1962),
which may be characterized as an active orientation toward
sensory experience. All perception is an act of not only
focusing and identifying but also interpreting the nature
of the world in individually relevant terms which reveal
personal intent toward the objects of perception.

Stated most simply, perception is an act of problem-
solving (Gregory, 1970, cited in Brady, 1976) of active
attempts to cope with life and find it meaningful. This
concept of perception as active coping has been given

expression in psychological terms by Sternbach (1968):

20

By "perceptual" is not meant a passive
reception of stimuli, if indeed that
ever occurs, but an active process of
searching, discriminating, and distort-
ing that reflects an adaptive, need-
satisfying, motivated perception. By
"coping" is meant the comparable pro-
cess in overt behavior. We hyphenate
perceptual-coping styles to point up
the obvious interaction: coping is in
part a function of perception, which in
turn is a function of the same motives
which underlie overt behavior. (p. 157)

This conceptualization of perceptual-coping as an
act of problem-solving adds vital dimensions to the study
and treatment of human pain. Pain-as-an-answer-to-a-prob-
lem-in-living becomes as important a concept as pain-as-
response-to-a-noxious-stimulus.* What a patient says or
does about his pain may no longer be viewed as mere sub-
jective reactions to an objectively quantifiable pain
stimulus, but must be appreciated as an integral feature
of the pain itself (Fordyce, 1976b) so that stimulus, per-
ception, and response are viewed as interrelated aspects

of the entity-in-pain to be treated. As Merskey has noted:

 

*
To View pain as a consciously contrived attempt to

solve problems in living is to assume far too simplistic

a perspective. Merskey (1976) has also noted that the
interaction between psyche and soma must lie beyond conscious
control. It is actually rather difficult to imagine a pain,
or even to recollect vividly the experience of severe pain.

 

If patients have pain for psychological

reasons, and many do, this must usually

occur because of the operation of mental
mechanisms that produce it independently
of the patient's conscious wishes.

(p. 711)

21

It is easier to honor the patient's ex-
perience if we keep in mind that what-
ever the physical basis for pain it can
be known to an individual only through
his consciousness. Thus pain is al-
ways and only a psychological experience.
(p. 712)

Personality Inventories and Sufferers of
Chronic Low Back Pain

The measurement of personality characteristics
associated with low back pain has most often been under-
taken with the MMPI. MMPI profiles have been compared to
discriminate between patients with physical findings
(organic) versus without physical findings (functional),
between sufferers of acute versus chronic low back pain,
between back pain patients versus other orthopedic patients,
between those receiving and those not receiving employment
compensation, and between patients successful in treatment
and those unsuccessful.

Investigating clinical findings that many cases of
chronic low back pain could not be related to physical
deficits, Hanvik (1951) attempted to differentiate the
MMPI profiles of back patients designated functional (no
physical findings) versus organic (positive physical find-
ings). His sample consisted of two groups of thirty
veterans equated for age, socioeconomic class, marital
status, and intelligence. The functional group's composite
profile demonstrated a higher "neurotic triad," i.e., the

Hypochrondriasis, Depression, and Hysteria scales of the

22

MMPI. Additionally, the Depression scale was significantly
lower than the two adjacent scales, thus yielding a V-
shaped profile known as the psychosomatic-V or "conversion-
V," so called to indicate that patients with this profile
are likely to repress or deny their emotions (high hysteria,
low depression) and instead focus on somatic concerns
(hypochondriasis). These functional patients have been
described (Lachar, 1974) as having a strong need to inter-
pret their circumstances in a logically and socially
acceptable manner, and as resisting suggestions of any weak-
ness or unconventionality in their character. In general
they are described as egocentric, immature, and dependent.
Their complaints of pain appear to allow them to avoid
awareness of anxiety and conflict, albeit at considerable
cost in emotional control and repression.

The organic group's composite profile, by contrast,
had non-significantly elevated and approximately equal
standard scores (i.e., no conversion—V) on the scales of
the neurotic triad, and the profile in its entirety was
essentially normal. The functional group, in addition to
the conversion-V, also recorded significant elevations on
the Psychopathic Deviate, Psychasthenia, and Schizophrenia
scales, further indicating the presence of psychological
factors involved in these patient's experience of pain.

Apparently Hanvik's (1951) results, suggesting con-

version hysteria as an etiological factor in functional

23

low back pain, were of sufficient face validity as to
generate little controversy. It was not until 1964 that
another MMPI study of low back pain was reported, this
time comparing 58 low back pain patients with a group of 72
patients with limb fractures (Phillips, 1964). As the low
back patients in this study were not given functional/
organic diagnoses, it may be assumed that they repre-
sented a mixed group in this respect. The profile of this
mixed group of back patients had a significantly elevated
neurotic triad, both above the mean and above the group
with fractures; however, there was no evidence of a con-
version-V. Such a profile indicates neurotic involvement
with an existing physical condition, rather than a con-
version of emotional conflict into physical symptoms al-
though, from the perspective of Hanvik's (1951) earlier
findings, it may be argued that this non-V composite pro-
file was a direct result of Phillip's failure to discrim-
inate between functional and organic conditions among his
subjects. An additional noteworthy finding of Phillip's
(1964) study was that the amount of neuroticism, as
measured by the MMPI, was negatively correlated both with
prompt completion of a rehabilitation program and with
symptomatic improvement in the medical condition, in-
dicating that affective and motivational variables may
play a decisive role not only in task achievement but also

in the chronicity of pain.

24

Further research utilizing the functional/organic
dichotomy was undertaken by Haven and Cole (1972).
Composite MMPI profiles revealed no significant differences

among organic, functional, and malingering male veterans

(N 44) with chronic low back pain.

Gentry, Shows, and Thomas (1974) studied 56 male
and female patients whose chronic low back pain had per-
sisted despite at least one surgical intervention. Both
males and females had significant elevations on the
neurotic triad, with the males scoring slightly higher.
This between-sex difference is consistent with the report
of Sternbach et al. (1973a).

Beals and Hickman (1972) studied 180 industrially
injured patients treated in a physical rehabilitation center,
and found that the group of back-injured patients evidenced
an elevated neurotic triad on the MMPI, with acute patients
tending to peak on the Depression scale and chronic
patients tending toward a conversion-V. The chronic patients
also had elevations on the Psychasthenia and Schizophrenia
scales. In general Beals and Hickman found greater psycho-
pathology in back-injured patients than in extremity-
injured patients andichhronic, multiply-operated than in
acute patients. In addition, patients with higher eleva-
tions on the Hypochondriasis and Hysteria scales were less

likely to return to work. In this study the chronic

patients most closely resembled Hanvik's (1951) functional

25

group, indicating that chronicity of pain may have been a
confounding variable in Hanvik's study.

Recent studies by Sternbach and associates (1973a,b)
have contradicted Hanvik's findings of MMPI differences
between functional and organic patients. In a sample of
68 patients, 44 of whom had physical findings and 24 of
whom did not, no significant group differences were found
(Sternbach et a1., 1973b). The same researchers (1973a)
reviewed the MMPI profiles of another sample of pain
clinic patients, 81 with positive findings and 36 without.
Again, no significant differences were found. In both
studies the composite profile for all patients revealed
neurotic triad elevations that were two standard devia-
tions above normal, or higher than would be obtained by
96% of the normal population. There was no evidence of
a conversion-V. The authors therefore discounted the
diagnosis of conversion hysteria and gave preference to
the diagnosis "psychophysiological reaction with depres-
sion." They concluded that the organic/functional dichotomy
is of questionable value in the evaluation and treatment
of patients with chronic low back pain, and pointed out
that once a patient's pain has passed from the acute to
the chronic stage, and thus begun to dominate the patient's
emotional and social life, chances are slight that the
patient will get well, benefit from surgery, or success-

fully adapt to permanent disability, without appropriate

26

intervention. Successful intervention dependtheavily on
treating the depression and helping the patient to meet
"those needs which, unmet, have resulted in excessive
somatic concern and bodily pre-occupation" (1973b, p. 229).
This concept of an adopted life-style based on
pain has been supported by research (Sternbach et al.,
1973b) into the differences between acute and chronic low
back patients. Pain of more than six months' duration
was defined as chronic, of less than six months, as acute.
Acute patients had neurotic triad elevations one standard
deviation above normal, chronic patients had neurotic
triad elevations two standard deviations above normal.
Additionally, acute patients had slightly higher eleva-
tions on the Paranoia and Hypomania scales, indicating a
greater sense of urgency and apprehension about their pain.
In the transition from the acute to the chronic state,
this anxiety evidently tends to be replaced by depression.
Finally, in attempting to identify differences be-
tween low back patients with compensation action pending
(n = 36) and those with such action settled or never
initiated (n = 81), Sternbach et al. (1973b) discovered
that both groups had significant elevations on the neurotic
triad, with the litigants significantly higher than the
non-litigants. Also, the former group scored significantly
higher (T-score = 70 vs. T-score = 61) on the PsychOpathic

Deviate scale, which typically reflects anger, rebelliousness,

27

and resentment against authority figures. The authors
suggested that this profile of "compensation neurosis" was
inauspicious for improvement in condition, at least until
the litigation was settled.

While the majority of evidence supported Sternbach
and his co-workers in their suggestion that all chronic
low back pain patients had significant neurotic involve-
ment, their findings that the MMPI did not significantly
differentiate between functional and organic patients, as
earlier indicated by Hanvik (1951), has in turn been con-
tradicted by two more recent studies. Clinicians at the
Seattle Veterans' Administration Hospital (Freeman, Calsyn,
and Louks, 1976) classified 36 patients as either organic,
functional, or "mixed" on the basis of physical findings,
the "mixed" group comprising those patients who had some
organic basis for their pain, but insufficient to
explain the full degree of their reported pain. The patients
were matched for age and educational level, 12 to a group.
Each patient was administered the MMPI as part of a
routine evaluation at admission. All three groups obtained
composite profiles with significant elevation on the
neurotic triad, but the organic group scored significantly
lower on all three of the scales, and the functional group
slightly higher than the mixed. The functional and mixed
groups showed clear psychosomatic-V patterns, the organic

group had none. Consistently, the Psychasthenia and

28

Schizophrenia scales were significantly elevated, but only
for the mixed and functional groups. These results were
clearly supportive of Hanvik's (1951) findings that the
MMPI did in fact differentiate between low back patients
with and without positive physical findings. However,
these results also supported the conclusions of Sternbach
et al. (1973b) that even patients with a clear organic
basis for their reported pain were significantly neurot-
ically reactive to their pain. Hanvik had found minimal or
no neuroticism among organic patients in his sample.

In a further study the Seattle authors (Calsyn,
Louks, and Freeman, 1976) contrasted the MMPI profiles of
31 organic with 31 mixed patients. Virtually identical
comparisons prevailed between these groups as on the pre-
vious study. Again the conversion-V was present for the
mixed but not for the organic group, with the organic group
nevertheless evidencing a neurotic profile. Here, as
before, the mixed group scored significantly higher on the
Psychasthenia scale. Unlike the previous organic group the
patients here classified as organic obtained a significant
elevation on the Schizophrenia scale, as did the mixed
group.

It is important to ask why some research has deter-
mined that the elevation and configuration of the MMPI-
neurotic triad does differentiate functional from organic

pain patients, while other research has not found this

29

result. Two possibilities should be considered. Swartz
and Krupp (1971) have found that older medical patients
tend to have elevations on the Hypochondriasis and Hysteria
scales, and that these scales are therefore less useful in
making a functional vs. organic diagnosis among older
patients. The research of Sternbach et al. (1973b) con-
cerned patients in their early forties while Hanvik (1951)
did not report the ages of his patients. One might sus-
pect that Sternbach's patients were older, based on the
fact that Hanvik's sample consisted exclusively of
veterans likely to have served in World War II, his research
having been conducted shortly after the war (1949-1950).
The average age of his subjects was likely to have been
closer to 30 than to 40 years. The second reason for the
contradictory evidence is likely to involve the criterion
for evidence of positive physical findings. Hanvik's
criterion - protruded intervertebral disc - was especially
stringent, and was confirmed both by X-ray and upon re-
moval of the disc at time of surgery. Sternbach's criterion
included such manifest behavior as impairment in gait,
reflexes, and range of movement, findings which some in-
vestigators would consider to be of potentially psychogenic
etiology, and not necessarily indicative of disc pathology.
In summary, the majority of MMPI studies of chronic
low back pain patients have provided evidence that such

patients are neurotic. Elevations on the Hypochondriasis,

30

Depression, and Hysteria scales consistently are found in
group profiles. These elevations tend to be greater among
patients who have minimal or no detectable organic basis

for their pain, indicating the possibility of psychosomatic
pathologyf They are also likely to have an elevated score
on the Psychasthenia and Paranoia scales, and occasionally
on the Schizophrenia scale. It has been suggested by
Sternbach et al. (1973b) that regardless of the extent of
physical findings, chronic low back pain patients are likely
to be neurotically involved with their pain, and that treat-
ment of the neurosis is important for any recovery from or

adjustment to back-related disability.

Treatment Studies

 

A smaller body of research has attempted to identify
personality factors associated with outCome of treatment
for chronic low back pain. In an analysis of the relation-
ship between demographic, medical, and psychological
factors and successful outcome in a rehabilitation program
for patients with low back disabilities, Nagi, Burk, and
Potter (1965) studied the case records of 125 admissions
to the Ohio Rehabilitation Center. The outcome criteria
consisted of achieved improvement in activities of work and
daily living, as compared with the expectations of the
clinical team. Psychological factors were measured by

psychiatric ratings and unspecified psychological testing.

31

These researchers found that emotional and personal prob-
lems were more prevalent among non-achievers, and concluded
that emotional and motivational factors must be assessed
and addressed in physical rehabilitation programs.

Wilfling, Klonoff, and Kokan (1973) administered
the MMPI to 26 male veterans who were to have spinal fusion
for low back pain relief. Post-operatively (no indication
is given of elapsed time), a success/failure rating of good,
fair, or poor was assigned to each patient on the basis of
a combined score for employment status, presence of pain,
range of movement, the patient's rating of the value of
surgery, and the thsician's rating of the degree of dis-
ability. The three outcome groups were successfully dif-
ferentiated by the MMPI: The poor and fair groups both
had high elevations on Hypochondriasis, and both were
significantly higher than the good group. Both the poor
and fair groups scored high on the Depression scale, with
the poor group nearly three, the fair group nearly two, and
the good group less than one standard deviation(s) above
normal. On the Hysteria scale, both fair and poor groups
showed significant elevations. Additionally, the poor and
fair groups scored significantly lower than the good group
on the Ego Strength scale.

MMPI comparisons were also reported for a grouping
based on number of previous surgeries. The multiply -

Operated group showed significant elevations on all three

32

scales of the neurotic triad, with Hysteria and Hypo-
chondriasis scores significantly higher than the singly
operated group. Also, no previously multiply-operated
patient achieved a good outcome.

These results were partially supported by another
treatment outcome study conducted by Wolkind and Forrest
(1972). The Middlesex Hospital Questionnaire, a self-
rating measure of neurotic behavior and symptoms, was
administered to 50 male patients prior to treatment for
low back pain. After six sessions patients evaluated their
own outcomes by means of another self-rating questionnaire.
The good versus poor outcome groups showed significant dif-
ferences on the depression, somatic concomitants of
anxiety, and obsessionality scales, with the poor outcome
group in each case scoring higher in the neurotic direction.

Pheasant and Holt (1973) attempted to determine
personality correlates of response to treatment for low
back pain at an orthopedic hospital. An initial sample of
95 and a second, cross-validational sample of 94 patients
were administered a battery of psychological tests and then
assessed daily for response to treatment, the outcome
criterion being improved functioning in the activities of
daily living. In evaluating their data the authors found
that poor response to treatment was positively correlated
with neurotic symptomatology as interpreted from the MMPI.

Support for this relationship, however, was minimal, and

33

another inventory of health concern, the Cornell Medical
Index, provided non—confirmatory data.

Blumetti and Modesti (1976) compared 42 patients'
pre-treatment scores on the MMPI with the outcomes of their
neurosurgical treatment for intractable back pain. The
unsuccessful group, as determined by their report of non-
significant relief of pain at six-month follow-up, scored
significantly higher than the successful group (n = 19) on
both the Hypochondriasis and Hysteria scales. Viewing
these results together with patients' performance on the
Rorschach, the authors concluded that

those patients who responded favorably to

neurosurgical intervention for chronic

low back pain are relatively less

pathologically pre-occupied with overall

bodily concerns, less dependent, and

more capable of a higher level of in-

dividuation. They are also less rigid

and constricted in terms of defense

mechanisms and can rely on more than

just somatic complaints to deal and

interact effectively with the world

around them. (p. 324)

A similar study Of 34 consecutive admissions to a
general hospital orthopedic surgery department was conducted
by Waring et a1. (1976). Pre-treatment administration of
the MMPI, however, was unable to uncover any differences
between the poor and good outcome groups, which were deter-
mined on the basis of physicians' post-operative ratings
of operative success. The discrepancy between these and

previous findings, suggested the authors, may have been a

result of the small sample size. More likely, they believed,

34

it was due to the thorough pre-admission screening by the
surgeon, who had excluded many patients with poor surgical
prognoses. Those who had been excluded were most likely
those with disproportionate neurotic involvement, i.e.,
precisely those who in other studies (Calsyn et a1., 1976;
Freeman et a1., 1976) have been designated "functional"
or "mixed" and who obtained significantly higher elevations
on the neurotic scales of the MMPI and achieved poor treat-
ment outcome (Wilfling et a1., 1973).

Another treatment outcome study by Wiltse and
Rocchio (1975) attempted to identify "good surgical risks."
Their sample comprised 130 surgical candidates with low
back pain, who had not had previous surgery but had also
not achieved significant relief of pain from extensive
conservative treatment. Prior to surgery each patient was
interviewed and administered three psychological tests, the
MMPI, the Cornell Medical Index, and the Quick Test of
intelligence. The determination of surgical outcome was
based on the surgeon's rating, one year after treatment, of
improvement in physical condition and symptomatology.
Analysis of data revealed that the single best predictor of
the outcome criterion were the scores on the MMPI Hypo-
chondriasis and Hysteria scales. Of patients with very low
scores (54 and below), for example, 90% showed good or
excellent improvement, while only 10%<mfthe patients scoring

above 85 were able to attain this degree of improvement.

35

Using another pair of cut-off points, scores of 75 and
above portended only a 25% chance of achieving good or ex-
cellent results. A full 32% of the 130 patients scored 75
and above, while 44 percent scored 64 and below.*

A different use of the MMPI in predicting treatment
outcome has been proposed by Sternbach (1974). Based on
his experience in pre-treatment evaluation, inpatient
counseling, and post-treatment follow-up at a pain rehabil-
itation clinic, he has suggested that the usual dichotomous
classification of MMPI profiles in terms of high and low

elevation on the neurotic triad be expanded to a four-fold

 

IAlthough this study confirms earlier findings that the
degree of neurotic involvement is significantly correlated
with treatment outcome, its findings are subject to question.
The effectiveness of chemonucleolysis, the surgical treat-
ment employed in this study, was not confirmed by the re-
sults of a recent double-blind study (Martins et a1., 1978)
conducted under the auspices of the federal Food and Drug
Administration. In this study a control group injected with
a placebo had a 46% recovery rate, while 55% of those re-
ceiving the normal procedure (intervertebral injection with
chymopapain) showed significant improvement. This difference
in improvement rates is not significant, but because of
criticisms of their study on methodological grounds, Martins
and co-workers deemed their research inconclusive and
advocated conducting a more definitive double-blind study.
This has not occurred to date (Stromberg, 1979).

The 46% of patients achieving good or excellent results
in the Wiltse and Rocchio (1975) study, however, was within
the same range of improvement recorded in the double-blind
study. If chemonucleolysis as administered by Wiltse and
Rocchio was an active, effective procedure, then their MMPI
results clearly are in line with other findings that
neuoticism is correlated with outcome of treatment for low
back pain. If, on the other hand, chemonucleolysis has no
value except as a placebo, then one must infer that among low
back patients, neuroticism plays the same role in response to
a genuine as in response to a placebo treatment, i.e., re-
gardless of treatment, less neurotic persons get better and
more neurotic persons do not.

36

approach. It will be remembered that Sternbach and
associates (1973a,b) consistently have found that various
types of low back pain patients (organic and functional,
chronic and acute) all have significant elevations on the
neurotic triad. Consequently, Sternbach (1974) advocated

a more configural approach to interpretation and evaluation,
based on four typical clinical profiles representing
hypochondriasis, reactive depression, somatization reaction,
and manipulative reaction, respectively. Of these, the
hypochondriacal profile characteristically has hypo-
chondriasis as the peak score, with an elevation more than
two standard deviations above normal. The Depression,
Hysteria, Psychasthenia, and Schizophrenia scales are also
significantly elevated. This pattern can only be achieved
by endorsement of a great number and variety of pain
complaints, including many that would be totally unrelated
to chronic low back pain. Prognosis for this patient group
is poor, as they continue to seek medical help after dis-
charge, their pain remaining the central factor in their
lives.

The "reactive depression" profile consists of an
elevated neurotic triad, with the Depression score as peak.
The Psychasthenia scale is also elevated. This configura-
tion reflects the subjective experience of depression,
which the patient acknowledges and views as a reaction to

living with pain. Premorbid adjustment was likely good, as

37

are response to treatment and post-treatment adjustment.
For many of these patients, treatment results in signif—
icantly reduced pain, and they subsequently cease to seek
medical help.

The "somatization reaction" is equated with the
classical conversion-V or psychosomatic-V profile, where
the Hypochondriasis and Hysteria scales are significantly
elevated, and the Depression score forms a valley between
them. The depressive component of this syndrome is not
absent, but latent, obscured because the patient focuses
on physical symptoms. This pattern may reflect such a
psychophysiological condition, or it may also indicate
among medical and surgical pOpulations that the patient
has learned to live with pain by deriving satisfaction
from the invalid role. These patients can usually be
assisted in treatment to enrich their lives and become less
symptom-centered.

The profile for "manipulative reaction," a term
coined by Sternbach, resembles closely that for hypo—
chondriasis, with the addition of a high score on the
Psychopathic Deviate scale. This scale reflects anger,
manipulativeness, and acting out, and its elevation with
the neurotic triad is a sign that the patient is likely to
use his symptoms to get what he wants, whether this be
satisfaction in frustrating the physician or a pay-off

through favorable settlement of a compensation claim. These

38

patients are not malingerers in the usual sense, but are
consummate game players, and Sternbach recommended exclud-
ing them from treatment programs. Sternbach (1974)
cautioned that these results were based only on a small-
scale retrospective study and have not been systematically

confirmed.

Implications and Summary

 

Previous research repeatedly has demonstrated the
effectiveness of the MMPI in evaluating the psychological
characteristics of chronic low back pain patients. Such
patients consistently have recorded pathologically high
scores on the Hypochondriasis, Depression, and Hysteria
scales of the MMPI, indicating that their condition may
best be viewed as psychophysiological in nature. In addi-
tion to elevations on this "neurotic triad" of scales, low
back patients often record significant elevations on the
Psychopathic Deviate, Psychasthenia, and/or Schizophrenia
scales, further indicating the presence of factors such as
secondary gain, acute anxiety, and thought disorders.

Less frequent use of the MMPI has been made in
attempting to correlate patient psychological character-
istics with treatment outcomes. Researchers have most often
found an inverse relationship between increasing elevations
on the neurotic triad and favorable prospects for treatment.

Sternbach (1974) has suggested that greater accuracy in

39

predicting treatment outcome may result from a configural
approach to interpretation. Standard interpretation of the
MMPI combines the elevational and configural approaches.

The present research has attempted to further im-
prove the accuracy of the MMPI as a diagnostic instrument
for use with chronic low back pain patients. Although such
patients have shown a consistent pattern of responding to
the MMPI in terms of both configuration and elevation of.
scales, the standard clinical and validity scales of the
MMPI were constructed without reference to chronic pain
patients as a nosological group. A need exists to deter-
mine whether the personality characteristics of low back
pain patients, in this case patients successful and those
unsuccessful in treatment, can be more precisely defined by
means of the MMPI.

To accomplish this, the current study has addressed
the hypothesis that analysis of responses of these patients
to individual items on the MMPI may be more revealing of
patient personality variables than the analysis of response
to scales per se. By using empirical techniques essentially
similar to those employed in the original development of
the MMPI scales, the attempt has here been made to develop
a new scale of MMPI items with maximum potential to dis-
criminate between successful and unsuccessful chronic low
back pain patients. Examination of the content of these

items should result in a more specific identification of

40

factors necessary to consider in the pre-treatment screen-
ing of patients, in determining their prognosis, and in
devising treatment programs which may aid in their im-

provement.

CHAPTER III

METHOD OF RESEARCH

To test the major hypothesis of this study -- that
patients successful in treatment for chronic low back pain
responded differently than unsuccessful patients to items
on the MMPI -- it was necessary to determine the outcome
of treatment for a sample of low back patients and to
correlate this outcome with patients' pre-treatment
responses to the MMPI. To accomplish this a self-rating
questionnaire, the Pain Survey, was developed and mailed
to a large number of former low-back patients. The
questions measuring outcome status concerned degree of
pain relief obtained, ability to return to normal work and
activities, (dis)continued search for medical assistance
for pain relief, and overall subjective rating of satisfac-
tion with treatment. Responses to these questions formed
the basis for assignment of subjects to successful vs. un-
successful treatment outcome groups.

Treatment outcome was then correlated with pre-
treatment MMPI records taken from the respondents' (N = 185)
medical files. MMPI responses of a randomly selected sub—
sample of two-thirds (N = 51) of the successful and two-
thirds (N = 72) of the unsuccessful respondents were

41

42

analyzed to identify items which significantly discriminated
between the successful and the unsuccessful patients.
Those items significant at the .10 level comprised a Back
Treatment Success Scale, which was then compared in a cross-
validational procedure to the MMPI responses of the re-
maining one-third (N = 62) of the total sample, in the
attempt to determine the reliability and potential use-
fulness of the scale.

This chapter begins with a description of the
major medical and other demographic characteristics of
the sample population. Salient features of the measure-
ment devices used in the study, the MMPI and the Pain
Survey, are discussed. The survey procedures and response
data for the piloting and administration of the Pain Sur-
vey are presented. The chapter concludes with the state-
ment of the research hypothesis and an outline of the
statistical analyses by which the null hypothesis was

tested.

Experimental Sample

 

Subjects of this study were former patients of the
Ingham Low Back and Pain Clinic, Lansing, Michigan, all
of whom had terminated treatment between six and forty-
two months prior to the inception of the study. Their
pain had been defined as chronic, i.e., of more than six
months' duration, prior to treatment. All of the subjects

had experienced failure with one or more types of

43

conservative treatment (massage, manipulation, thermal
packs, bed rest, etc.) and many had experienced failure
with one or more surgical treatments. All were deter-
mined by their referring physicians and by the Clinic's
orthopedic surgeon to have pain and disability severe
enough to warrant intensive, inpatient treatment.

Patients who had been pregnant at time of treat-
ment were excluded from the study, on the premise that
their pain had been caused or exacerbated by their
temporary condition. Patients who had been discovered
during or after treatment to have cancer or multiple
sclerosis as a primary or complicating cause of pain were
also excluded. Illiteracy and senility, previously found
by Jerome (1979) to reduce the likelihood of treatment
success among patients at this clinic, were also a basis
for exclusion.

For the total sample of 185 subjects, average age
at the time of the present study was 44.4 years with a
range of 22 to 72 years. There were 48.1% males and 51.9%
females in the sample. A total of 19.5% of the subjects
were employed in semi-skilled or unskilled labor, and 24.9%
in skilled occupations, while 39.5% of the subjects were
unemployed. Housewives constituted 9.7% of the samples
students 3.4%, and non-disabled retirees 2.2%. High
school education had been completed by 65.8% of all sub-

jects, while 11.4% had completed four years of college.

44

With respect to age, sex, and employment these
figures are very similar to data obtained by Jerome (1978)
in a previous study of patients from the same clinic.
For his sample, mean age was 41.5, with a range of 19 to
77 years; 45% were females and 55% males; and the number
of skilled workers in the sample equalled 27%. Supple-
mentary data from that study are therefore also likely
to be highly representative of the current sample: In
Jerome's study the average number of prior surgeries per
patient was .98 and the typical patient had been
hospitalized twice previously for back pain treatment at
other facilities. Upon admission 93% of Jerome's sample
were taking narcotic analgesics, 28% major tranquilizers,
and 56% muscle relaxants. On a five-point pain intensity
scale (0 = no pain, 5 = horrible/unbearable) the average
pain intensity reported by patients was 4.03. Jerome
(1978) further noted that at time of referral, 80% of
his sample were not working due to pain. Such data appear
characteristic of the typical low-back patient involved in
in-patient treatment at other clinics (cf. Sternbach,

1974; Wilfling et a1., 1973).

Instrumentation

 

MMPI. This study utilized the Minnesota Multiphasic

 

Personality Inventory (MMPI) and a self-rating question-

naire entitled "Pain Survey" (see Appendix A), which was

45

developed specifically for the study. These instruments

will be discussed individually.

MMPI. The availability of MMPI records for the majority
of former patients of the Clinic was a primary factor in
its selection as a measure which would allow a broad basis
for comparison of patient personality factors. Originally
designed to identify and diagnose persons with psychiatric
disorders, it is currently used to provide a personality
description of patients in a wide variety of counseling,
medical, and employment settings. It has also emerged as
a preferred instrument in evaluating candidates for low
back treatment, so that the body of extant research into
the personality characteristics of low back pain patients
refers far more often to the MMPI than to any other single
personality measure.

The scales of the MMPI were empirically developed
by contrasting normal groups with carefully studied
clinical cases. Scale construct validity was based on
the usefulness of the diagnostic groups represented by
the clinical cases. The greater the number of items a
person answered the same as a given criterion group, the
more likely he/she was to share other characteristics of
the group (Dahlstrom and Welsh, 1960).

Each item consisted of a statement to which
respondents answered either "True" or ”False", a format

which allowed item-for-item comparison of response between

46

two groups, as was intended in the present study. The
566 items comprising the ten clinical and four validity
scales cover

...areas such as health, psychosomatic

symptoms, neurological disorders, and motor

disturbances; sexual, religious, political,

and social attitudes; educational, occupa-

tional, family, and marital questions; and

many well-known neurotic or psychotic be-

havior manifestations, such as obsessive

and compulsive states, delusions, hallu-

cinations, ideas of reference, phobias, and

sadistic and masochistic trends.

(Anastasi, 1976, p. 497)

In constructing the test, the literal meaning of the item
did not necessarily determine its inclusion in a specific
scale. The primary criterion for including an item was
its ability to discriminate between a sample of the
normal population and a sample of patients already diag-
nosed as belonging to a criterion group.

Eight of the clinical scales were developed in
this manner and consist of items differentiating between
a control group of nearly 700 persons and a specific
clinical group, usually numbering 50 persons. Of the
other two clinical scales, the Masculinity-Femininity Scale
items were selected for their ability to discriminate be-
tween normal women and normal men, a high score on this
scale typifying interests of the opposite sex. The Social
Introversion Scale was developed subsequent to the other

scales, based on item responses of two contrasting groups

of college students. These scales are supplemented by

47

four validity scales, which monitor the protocol for care-
lessness, malingering, misunderstanding, and test-taking
attitudes (Anastasi, 1976).

For this study the Form-R of the MMPI was used,
which permits all 14 basic scales to be scored from the
first 399 items. Only these 399 items were scored and
analyzed. The test was administered individually in the
standard manner described by Hathaway and McKinley (1967)

prior to or at admission to the Low Back and Pain Clinic.

Pain Survey. The patients' self-report Pain Survey

 

(Appendix A) was developed specifically for the purpose

of evaluating long-term treatment outcomes among subjects
in this study. It consisted of two groups of questions
which subjects were requested to answer at least six months
after termination of treatment. The first group of four
questions was based on criteria suggested by the literature
and by the clinic staff as representing desirable outcomes
of treatment. These criteria included relief from low

back pain, return to work and normal activities, cessation
of medical treatment for back pain, and subjective rating
of benefit from treatment. Each of the four forced-choice
questions had two foils (in each case labeled (a) and (b))
that were considered indicative of successful outcome and
between one and four foils intended to denote an unsuccess-

ful outcome.

48

Patients were requested to affirm only the one
best answer to each question. It was thus possible to
score from a maximum of four, to a minimum of zero,
answers indicative of success on the four questions. For
this study, subjects marking either three or four success-
ful foils were deemed to have achieved a successful treat-
ment outcome. Patients marking either two, one or zero
answers indicative of success were considered to have
achieved an unsuccessful treatment outcome.

Part II of the Pain Survey consisted of a group of
ten questions, designed to measure the impact of pain on
patients' lives; these were included as a validity control
for the first four questions. It was hypothesized that
patients achieving a successful treatment outcome, as
indicated by a total score of three or four on the first
group of questions, would be significantly less affected
by pain in their activities of daily living, and that this
would be reflected by their answers to Part II. The
activities represented were evident in the questions re-
produced on the second page of Appendix A.

Two additional items, not intended for analysis in
this study, were included to provide feedback for the
clinic staff. Under each of the headings "SUGGESTIONS FOR
IMPROVING THE CLINIC" and "OTHER COMMENTS" several blank

lines were provided to allow for unstructured response.

49

Procedure

 

Four hundred ninety former patients of the Ingham
Low Back and Pain Clinic who had been administered the
MMPI prior to or at outset of their treatment and who had
terminated treatment at least six months prior to this
study were identified as potential respondents. Of these
an initial sample of 79 were mailed questionnaires
(Appendix A) accompanied by cover letters (Appendix B) ex-
plaining the purpose of the study and soliciting their
cooperation. Potential respondents were informed that
the purpose of the questionnaire was "to obtain new know—
ledge that will be of help in treating patients coming
through the program." They were assured that their
responses were confidential, would not become part of
their medical records, and would in no way influence their
future treatment. A stamped, self-addressed envelope was
enclosed for the return of the questionnaire.

Of this sample of 79, 33 respondents returned
valid questionnaires and met the medical criteria for in-
clusion in the study. Valid MMPI's were on file for 28 of
these respondents, who as a group comprised the total
pilot sample. Of these 28 respondents, 15 answered either
three or four of the four criterion questions in the
direction intended to indicate successful outcome of treat-
ment, while 13 answered either three or four of the

criterion questions in the opposite direction. No

50

respondent marked two of the four questions in one direc-
tion and the other two in the opposite direction. It was
thus apparent that the questions as piloted differentiated
quite decisively between two groups of respondents.

Confirmation that these two groups did in fact
represent the successful and the unsuccessful patients,
respectively, was provided in part by the orthopedic
surgeon and the psychologist of the clinic, who to the
extent possible verified that individual responses and
implied treatment outcome were congruent with respondents'
treatment experiences at the clinic. In these delibera-
tions the unstructured responses and comments by patients
were also consulted, as many of the respondents were very
definite in expressing their satisfaction or dissatisfac-
tion with their treatment and its outcome.

A statistical comparison of the two identified
groups provided additional evidence of the validity of the
Pain Survey and the criteria for assigning outcome status.
Patients of the two groups were compared on the basis of
their responses to the ten questions measuring impact of
pain on their life activities. For each of these questions
a "YES" response, acknowledging the presence and delete-
rious impact of pain, was scored 1 (one), while a "NO"
response was scored 0 (zero). At one point per question,

higher total scores for the ten questions would indicate

51

considerable impact of pain on the respondent's life,
while lower total scores would indicate minimal impact.

Individual and mean scores were computed for each
of the two groups. Among the 15 respondents belonging to
the group tentatively identified as successful, individual
scores ranged from 0 to 7, with a mean of 3.7. Among the
group tentatively designated unsuccessful, individual
scores ranged from 0 to 9, with a mean of 6.2. A t-test
(two-tailed, pooled variance, 26 df) revealed the dif-
ference in means to be significant at the .005 level. This
significant difference in the impact of pain on former
patients' lives permitted positive identification of the
two groups as successful and unsuccessful beneficiaries
of treatment.

A further statistical analysis of the response by
pilot subjects to the four criterion questions was con-
ducted to determine the ability of each foil to dis-
criminate among successful and unsuccessful respondents.
Results of a 2 x 2 chi-square, appropriate for this non-
interval data (Nie, Hull, Jenkins, Steinbrenner, and
Bent , 1975), comparing individual vs. group-appropriate
response to each foil, are shown in Table 3.1. Given the
relatively small number of subjects in the pilot sample,
the majority of the foils successfully discriminated be-
tween the two groups of respondents. It was considered

likely that those foils which failed to discriminate to

52

 

TABLE 3.1. Pilot Study of Pain Survey; Ability of Individual Foils to
Discriminate between Successful and Unsuccessful Former
Patients
S (%) U (%) Significance
n = 15 n 13
Question 1. (affect on pain)
a. major pain relief 7 (46.7) 0 (0.0) .016
b. some pain relief 8 (53.3) 2 (15.4) .090
c. no lasting pain relief 0 (0.0) 11 (84.6) .000
Question 2. (work and activities)
a. return to normal 7 (46.7) 0 (0.0) .016
b. return to less strenuous 8 (53.3) 2 (15.4) .090
c. unable to return 0 (0.0) 11 (84.6) .000
Question 3. (continued medical
help)
a. no, no more pain 2 (13.3) 0 (0.0) .528
b. no, less bothered by pain 8 (53.3) 0 (0.0) .007
c. want to, can't afford 2 (13.3) 0 (0.0) .528
d. no, lost faith in doctors 0 (0.0) 2 (15.4) .401
e. currently in care 3 (20.0) 9 (69.2) .024
f. searching for help 0 (0.0) 2 (15.4) .401
Question 4. (benefit from treat-
ment)
a. helped very much 5 (33.3) 0 (0.0) .072
b. helped enough 10 (66.7) 3 (23.1) .054
c. helped very little 0 (0.0) 7 (53.8) .005
d. no help 0 (0.0) 3 (23.1) .175

 

 

53

a statistically significant degree (e.g., 3a, 3c, 3d, 3f,
4d) evidently failed to do so because of the small sample
size. This was obviously the case with 3a and 4d. The
three remaining foils in question (3c, 3d, 3f) were all
intended to indicate an unsuccessful outcome of treatment,
and appeared to provide necessary response alternatives
for potential unsuccessful subjects, i.e., alternatives
which might reflect most accurately actual post-treatment
status and activities with respect to continued use of the
health care system. All foils were therefore retained,
and the questionnaire was used without modification for
the study proper.

The Pain Survey with accompanying letter was then
sent to the remaining 411 former patients. After two weeks,
when it appeared that fewer than the desired 40% of
potential subjects would respond, 24 non-respondents were
contacted and administered the questionnaire by telephone.
A total of 182 valid questionnaires were obtained from
this combined second mailing and telephoning, of which 157
could be matched with valid MMPI's. Of those responding
by mail, 39.1% had achieved a successful treatment outcome,
and of those responding by telephone, 37.5%. Response by
telephone was therefore considered not to have been a con-
founding factor, and the data gathered in this manner were

included in all subsequent analyses.

54

It was considered desirable to include the pilot
data with the research data for the purpose of developing
the Back Treatment Success Scale, as the addition of 28
subjects would increase the stability and statistical
power of the item-for-item analysis. To determine the

appropriateness of including the pilot data, a comparison

was made of the questionnaire data obtained from the re-
search sample and the pilot sample (Table 3.2). This com-
parison indicated that the demographic Characteristics of
the two samples were essentially similar, with the excep-
tion of sex and employment. In examining the possible
biasing effect of sex on treatment outcome, however, a

2 x 2 chi-square analysis revealed no significant inter-
action (3 = .393, research sample; 2 = .829, pilot sample).
In examining the significant difference in the rate of
unemployment between the pilot sample (25.0%) and the
research sample (42.0%), a 2 x 6 chi-square crosstabulating
the six employment categories with treatment outcome re-
vealed that unemployment was significantly related

(9 = .001) to failure in treatment. This indicated that
the research sample likely contained a greater percentage
of treatment failures. The data presented in Table 3.3
document this difference in percentage but also show

that this difference was not significant (p = .146). In
other words, although the research sample contained a

greater percentage of unemployed and correspondingly un-

55

TABLE 3.2. Comparison of Demographic Data Obtained from Pilot and
Research Samples

 

 

Significance
Demographic Pilot Research (T-TEST:
Variables N = 28 N = 157 Pooled Variance)
Mean Age in Years 43.89 44.51 ns
Sex % Males 25.0 52.2 .01
% Females 75.0 47.8 .01

Mean Years Education 11.25 11.76 ns
Employment

Unemployed % 25.0 42.0 .05

Unskilled, Semi-skilled 25.0 24.8 ns

Skilled, Managerial, Prof. 17.9 19.7 ns

Housewife 25.0 15.9 ns

Student 3.6 4.5 ns

Non-disabled retiree 3.6 1.9 ns

 

56

TABLE 3.3. Endorsement of Criterion Questions by Pilot vs. Research

Samples

 

 

Questionnaire Response % Endorsing Significance
Pilot Research
(N = 28) (N = 157)
Q1. Pain Relief
a. major 25.0 19.1 .475
b. some 35.7 28.7 .455
c. none 39.3 52.2 .209
Q2. Employment, Recreation
a. return to normal 25.0 15.9 .244
b. partial return 35.7 40.8 .618
c. minimal or no return 39.3 43.3 .693
Q3. Return to Medical Care
a. no, no more pain 7.1 3.8 .429
b. no, less pain 28.6 22.9 .521
c. no time or money 7.1 10.2 .168
d. no faith in doctors 7.1 8.3 .840
e. currently in care 42.9 43.3 .965
f. looking for help 7.1 8.9 .760
Q4. Satisfaction with
Treatment
a. very satisfied 17.9 18.0 .991
b. mostly satisfied 46.4 34.0 .208
c. not very satisfied 25.0 23.1 .826
d. disappointed 10.7 23.7 .126
Success in Treatment 53.6 38.9 .146

(3 or 4) X (a or b)

 

57

successful former patients, this difference was not
great enough to justify the conclusion that the pilot and
the research samples were non-homogeneous.

Based on the assumption of homogeneity of the
research subjects and the pilot subjects, the two samples

were pooled to constitute a total sample population of 185

subjects. It was with respect to this total sample popula-
tion that all subsequent statistical analyses were per-

formed.

Statistical Analysis

 

The null hypothesis under consideration in this
study was as follows:

Items on the MMPI do not differentiate to a

statistically significant degree between

patients who achieve success and patients

who experience failure in multidisciplinary

treatment for chronic low back pain.
To test this hypothesis two-thirds of those subjects
identified by means of the Pain Survey as successful were
assigned to a Treatment Success Group (n = 50) and two-
thirds of those identified as unsuccessful were assigned
to a Treatment Failure Group (n = 73), while the remainder
of the subjects (n = 62), consisting of one—third of all
successful and one-third of all unsuccessful subjects,
were retained as a cross-validational subsample.

The MMPI responses of the Treatment Success Group

and the Treatment Failure Group were compared on an item-

for-item basis. This comparison of the two groups on the

58

true/false dimension of item response was accomplished
through a 2 x 2 chi-square analysis of each of the first
399 items on the MMPI Form-R. Items significant at the .10
level were selected for inclusion in the Back Treatment
Success Scale.
The MMPI records of the cross-validation subjects
(N = 62) were then examined for response to items on the
Back Treatment Success Scale. To obtain a scale score for
each subject, any scale item answered in the "failure"
direction (i.e., the direction -- TRUE or FALSE 1- which
had been endorsed with greater relative frequency by the
Treatment Failure Group) received a score of zero; those
scale items answered by cross-validational subjects in the
opposite direction received a score of one. The total
scale score (sum of 0's + 1's) for each subject was then
divided by the number of items which the subject had
answered either TRUE or FALSE, a procedure which prevented
non-response (CANNOT SAY) to any scale item from being a
source of measurement error.‘ Individual scores thus
ranged from 0 to 1, with relatively higher scores indicative

of success as measured by the scale.

 

*

“The standard procedure for administration of the MMPI
permits a testee to leave blank up to 30 items without in-
va11dating the profile. This presents a potential source
of measurement error, capable of vitiating the significance
test in the present cross-validation procedure, since any
subject may have failed to respond to a number of items on
the scale. For this reason the above-described proportional

scoring procedure was adopted to control for error intro-
duced by blank items.

59

Mean scores were then generated for two subgroups
of the cross-validational subjects, consisting of those
subjects previously defined by the Pain Survey as treatment
successes and treatment failures, respectively. The mean
scores were compared with a one-way analysis of variance
to determine the degree to which the Back Treatment Success
Scale differentiated between the two groups. The results

of these analyses are presented in Chapter IV.

CHAPTER IV

ANALYSIS OF RESULTS

Subjects' responses to the Pain Survey indicated
that the questionnaire accurately discriminated between
successful and unsuccessful former back clinic patients.
These responses provided the basis for assignment of sub-
jects to successful or unsuccessful treatment outcome
status.

A tentative Back Treatment Success Scale was de-
veloped, based on the MMPI-items which discriminated be-
tween members of a randomly selected subsample of two-
thirds of all successful and unsuccessful subjects. The
attempt to cross-validate this scale on the remaining one-
third of the sample population proved statistically non-
significant. The null hypothesis, that MMPI items do not
discriminate to a statistically significant degree between
successful and unsuccessful back clinic patients, was not
rejected.

In order to identify discriminating items with
greater statistical power, the scale development and cross-
validational subsamples were pooled and the test of item-

discrimination was applied to the MMPI records of the

60

61

total sample. By this procedure 45 MMPI-items were
identified as best discriminating between successful and
unsuccessful subjects. These items were subjected to
several analyses: they were categorized according to
traditional MMPI-scale membership, grouped into seven
content-related factors, and also statistically factor-
analyzed. These analyses yielded a number of personality
characteristics and profiles which appeared to be posi-
tively correlated with treatment outcome. On this basis
it was concluded that these 45 MMPI-items did discriminate
meaningfully between successful and unsuccessful former
patients of the back clinic, even though it was not
possible to establish the items' statistical significance
through cross-validation on another sample.

This chapter will present an analysis of the re-
search data. First, subjects' responses to the Pain
Survey will be summarized. Second, the attempt to develop
the Back Treatment Success Scale will be discussed, in-
cluding selection of items and cross-validation of the
scale. The third and concluding section of this chapter
contains the several analyses of the 45 MMPI-items best
discriminating between successful and unsuccessful sub-

jects of the total sample.

Pain Survey. As piloted, the Pain Survey had effectively

 

discriminated between successful and unsuccessful subjects

(Table 3.1). Analysis of the response to the Pain Survey

62

by the combined Treatment Success and Treatment Failure
groups (N = 123), which together comprised two-thirds of
the total 185 subjects in the study, indicated the
questionnaire's continued effectiveness in differentiating
between the two outcome groups. A 2 x 2 chi-square analysis
was used to crosstabulate outcome of treatment with
response to each foil oftflmasurvey's four criterion ques-
tions (Table 4.1). All but three of the foils dis-
criminated between successful and unsuccessful respondents
at the .002 level. These foils (3c, 3d, 3f), however, had
the three lowest response frequencies of all fifteen foils.
Because the three foils were all intended to indicate un-
successful treatment outcome, the sum of successful
respondents (BC + 3d + 3f = 4) was compared with a 2 x 2
chi-square to the sum of unsuccessful respondents (BC +

3d + 3f = 29) in the attempt to more closely approximate
the true statistical power of Question 3. This comparison

attained significance at the .001 level.

Back Treatment Success Scale. A 2 x 2 chi-square analysis

 

was performed comparing MMPI-item response (TRUE/FALSE)
with treatment outcome (successful/unsuccessful) for the
combined population (N = 123) of the Treatment Success and
Treatment Failure groups. By means of this analysis, 10
items were identified as significant at the .05 level and
an additional 18 at the .10 level. In a preliminary

attempt to eliminate items with the least stability,

63

TABLE 4.1. Response to Pain Survey of Treatment Success and Treatment
Failure Groups

 

 

Success Failure
Questiona (N = 50) (N = 73) Significance
n % n %
Ql. Pain Relief
a. major 28 56.0 1 1.4 .000
b. some 22 44.0 12 16.4 .002
c. none lasting 0 0.0 60 82.2 .000
Q2. Wbrk and Activities
a. return to normal 21 42.0 2 2.7 .000
b. some restrictions 29 58.0 16 21.9 .000
c. minimal return 0 0.0 55 75.3 .000
Q3. Return to Medical Care
a. no, no more pain 9 18.0 0 0.0 .002
b. no, less pain 28 56.0 1 1.4 .000
c. no, can't afford 3 6.0 11 15.1 .196
d. lost faith in
medical help 1 2.0 8 11.0 .123
e. currently in care 10 20.0 43 58.9 .000
f. still looking for
help 0 0.0 10 13.6 .020
Q4. Satisfaction with
Outcome
a. very satisfied 22 44.0 4 5.5 .000
b. mostly satisfied 27 54.0 17 23.3 .001
c. not very satisfied 1 2.0 25 34.2 .000
d. disappointed 0 0.0 27 37.0 .000

 

a
Full Text in Appendix A.

64

significance levels for these 28 items were computed for
the 185 subjects in the total sample population. Upon
comparing the item significance levels for the two groups
(N = 185, N = 123), the decision was made to exclude from
the scale four of the 28 items which had a greatly re-
duced significance (p > .30) with respect to the total
sample population. The 24 items retained for the Back
Treatment Success Scale, with the direction of item
response endorsed most frequently by successful and un-
successful subjects, is listed in Table 4.2.

To test the potential usefulness of the 24-item
scale in predicting outcome of treatment for a similar
group of chronic low back pain patients, a cross-valida-
tional procedure was conducted. This procedure involved
computation of Back Treatment Success Scale scores for
the 62 subjects whose MMPI records had not been included
in the chi-square item-analysis. To compute individual
scale scores, each scale item answered in the "failure"
direction (i.e., the direction which had previously
been endorsed by a greater percentage of unsuccessful than
successful subjects) was scored 0; each scale item answered
in the direction previously endorsed with greater relative
frequence by successful subjects received a score of 1.

By way of example, for Item 2 in Table 4.2 a subject
received a score of l for a TRUE response, because this

response had been endorsed relatively more often by

65

TABLE 4.2. Back Treatment Success Scale: Direction of Item Response
Endorsed More Frequently by Successful vs. Unsuccessful
Subjects (N = 185)

 

MMPI Direction of Successful S's Direction of Unsuccessful S's

 

Item Response % Endorsing Response % Endorsing
2 T 94.0 T 81.9

3 F 60.0 F 79.2
23 F 91.7 F 76.4
60 T 98.0 T 74.0
61 F 95.7 F 69.0
72 F 84.8 F 67.2
106 F 93.8 F 79.1
125 F 87.8 F 70.0
142 T 52.1 T 71.8
175 T 71.4 T 54.9
184 F 100.0 F 88.9
185 T 92.0 T 73.6
187 T 89.6 T 66.7
211 F 97.9 F 86.4
213 F 100.0 F 89.6
215 F 83.7 F 66.7
247 F 97.8 F 86.8
248 F 54.3 F 72.7
266 T 51.1 F 69.1
268 T 77.3 T 58.8
272 T 87.5 T 71.8
281 T 83.0 T 64.3
323 F 83.7 F 63.3
392 T 53.7 T 71.7

 

66

successful than by unsuccessful subjects. A proportional
scale score for each subject was then obtained by summing
the item scores (0's + 1's) and dividing the sum by the
number of items to which the subject had responded. This
procedure prevented items left blank from being scored as
0 or "failure." The resulting scale scores for each sub-
ject therefore ranged from 0 to 1 at the hypothetical
limits, with relatively higher scores indicative of success.
I Mean scale scores were then computed for two sub-
groups of the cross-validational subjects, one consisting
of those subjects identified by the Pain Survey as
successful (N = 26) and the other consisting of subjects
likewise identified as unsuccessful (N = 36). The com-

parison of mean scores is presented in Table 4.3.

TABLE 4.3. Analysis of Variance of Back Treatment Success
Scale Scores of 26 Successful and 36 Unsuccess-
ful Subjects

 

Source SS df MS F p
Between Groups .0345 l .0345 2.90 .094
Within Groups .7143 60 .0119

Total .7489 61

 

The difference in mean scores was statistically
significant at the .094 level. This comparison indicated

some stability of the scale-items across groups, but the

67

significance level obtained did not warrant rejection of

the null hypothesis.

Re-Analysis of Data for Total Sample Population

 

Despite the failure to reject the null hypothesis,
the moderate level of significance achieved in the cross-
validational procedure provided some evidence for the
homogeneity of the total sample and for the stability of
items included in the Back Treatment Success Scale.
Consequently, it was decided to increase the statistical
power of the chi-square computation by which scale items
had been identified; the scale-development sample and the
cross-validational sample were pooled in order to re-test
the significance of items discriminating between success-
ful and unsuccessful subjects. With the increased cell
frequencies for the 2 x 2 chi-square analysis, it was
assumed that significantly discriminating items would have
greater stability, and would thus provide a more
appropriate basis for the attempt to identify personality
factors associated with success or failure in treatment.

Recomputation of the chi-square analysis for all
185 subjects yielded 45 items significant at the .10 level.
The MMPI scale categorization of these 45 items, with
direction of response endorsed relatively more often by
unsuccessful than successful subjects, is presented in

Table 4.4. (Results of the chi-sqaure analysis for all

68

TABLE 4.4. Scale Categorization of Items Discriminating at .10 Level
between 76 Successful and 109 Unsuccessful Back Clinic
Patients (N = 185)

 

Scales
MMPI Diagnostic Validity
Item Hs D Hy Pd Mf Pa Pt Sc Ma Si L F K

 

23 T T T T
46

60 F

61 T

69 T

82 F f

91 t t

95
104 T T
106 . T T
113

115 F F
131 F
134 F f f F
145 t

156 T T T
159 T T
185 F
186 T

187 f F
211
215 F
220 F
225 F
230 F
243
247 T
248 F F

a . . .
Lower case f or t denote responses not scored dev1ant for spec1f1c
MMPI scale.

69

Table 4.4 (continued)

Item Hs D Hy Pd Mf Pa Pt Sc Ma

Si

 

249 F

251 T T
260 f

266 t t t
268 F t
272 F

278 T

282 t t

284 T T

292 F

295 t

323 T T

343 T

367

375

391

392

Total b 2 8 4 8 3 2 3 7 2

 

Does not include lower case responses.

70

399 items, with endorsement percentages of successful and
unsuccessful subjects, are contained in Appendix C). Each
of the ten diagnostic and three validity scales of the
MMPI is represented by at least two of the 45 items, but
the Depression and Psychopathic Deviate scales with 8
items each, the Schizophrenia scale with 7 items, and the
F scale with 9, together account for nearly 60% of the
item-scale categorizations. Surprisingly, 10 of the 45
items were endorsed with relatively greater frequency by
the unsuccessful subjects in a direction which indicated
no deviant response on any of the diagnostic or validity
scales. This does not indicate that the majority of
successful subjects had responded to these items in the
opposite, deviant manner; but it does indicate that
relatively more successful than unsuccessful subjects
answered in the deviant direction with respect to
established norms of a specific scale. This, in turn, does
not preclude the possibility that unsuccessful subjects
were in fact responding with greater relative frequency
in the deviant direction with respect to the distinct
nosological category here being investigated, namely, back
treatment failure.

An attempt was made to identify, by content
analysis of the 45 items from the MMPI, clusters of items
representative of personality characteristics common to

the unsuccessful subjects. A variety of such item clusters

71

emerged: (1) denial of social non-conformity (items #60,
225, 243, 292, 391 and 392; (2) self-depreciation tending
to guilt and paranoia (46, 61, 91, 104, 106, 159, 260,
278 and 284); (3) health complaints and disease phobia
(23, 131, 185, 230); (4) impaired faculties of concentra-
tion, coordination and awareness (156, 159, 186, 187, 251);
(5) depressed affect and behavior (134, 211, 248, 266,
268, 272, 343); (6) repressed hostility and authority
problems (82, 145, 375); and non-affirmation of funda-
mentalist religious beliefs with non-regular church
attendance (95, 115, 249). Significant single items in-
cluded acknowledgement of strong attraction by members

of the same sex (item 69) and of excessive use of alcohol
(item 215).

To determine whether and how these item-clusters
contributed to typical MMPI response patterns of this sample
of chronic low back pain patients, a statistical factor-
analysis of all 45 items was conducted. The SPSS image
factor program computed and retained all factors with
eigenvalues greater than 1.0 and listed values for the
loading of all items on each factor. With this program
the number of factors is usually equal to one-half the
number of variables in the set. It is assumed in each
case that a number of these factors will be unamenable to
interpretation, and will not be retained (Nie et a1.,

1975).

72

Image factoring in the present case generated 30
factors with eigenvalues greater than or equal to 1.0.

The first ten factors, with eigenvalues ranging from 9.82
to 3.25 and accounting for 58.5% of the variance, were
examined for feasibility of interpretation. Subsequently,
the first eight factors, with eigenvalues 9.82 - 3.60 and
accounting for 51.1% of the variance, were retained for
further interpretation. A complete tabulation of the
loadings of individual items on each of the eight factors
is presented in Appendix D.

In interpreting each factor, between 10 and 16
items were considered. For each factor, interpreted items
included all those with a loading greater than or
approximately equal to one-half the absolute value of the
highest weighted item. These items are arranged in eight
tables (4.5.1 - 4.5.8) representing the eight factors,
with items arrayed from top to bottom according to decreas-
ing absolute value of loading, without regard to sign.

Item responses are labeled T or F according to the direc-
tion more frequently affirmed by unsuccessful than success-
ful subjects. Loadings are designated "same" or "opposite"
with respect to the response direction. Thus, in Table
4.5.1, item 251 had the highest loading on Factor 1, with
an absolute value of .57. The response listed for this
item is T, indicating that a higher percentage of un-

successful than successful subjects responded TRUE to this

73

item. The loading of .57 is designated "same," to show
that the direction of loading coincides with T in the
Response column.

Each opposite or same item-loading must also be
interpreted in light of the entire response pattern repre-
sented by the factor on which that item loads. That is,
each factor represents the response pattern of a subset
of either successful or unsuccessful subjects, whereby it
is extremely unlikely that all items in a given pattern
will have been endorsed in either the same or in the
opposite direction as the Response-value. It is reason-
able to assume, for example, that Factor 1 represents
the largest subset of unsuccessful subjects because, first,
this factor accounts for a higher percentage of the vari-
ance than any other factor and, second, all but one of
this factor's thirteen highest weighted itms is weighted
in the same direction indicated under Response. However,
in interpreting the opposite weighting of this one item
(#282, Table 4.5.1), it is necessary to bear in mind
that only a 53% majority of unsuccessful subjects (see
Appendix C) responded FALSE to this item, leaving 47% of
the unsuccessful subjects who responded in a direction
opposite to "Response." For this reason loadings desig-
nated "opposite" were frequently found on factors

apparently representative of unsuccessful subjects.

74

Interpretation of factors with approximately equal
numbers of same and opposite item-loadings is complicated
by the fact that for all but 2 of the 45 items, a majority
of both successful and unsuccessful subjects responded to
a specific item in the same (T or F) direction (see
Appendix C). Item 249, for example, discriminated between
successful and unsuccessful subjects at the .05 level of
Significance; to this item 63% of unsuccessful and 78% of
successful subjects responded TRUE. Thus, where this
item occurs in Factor 4 as the most highly weighted item
(Table 4.5.4), and with an "opposite" loading, one cannot
at first glance determine whether, in this factor, this
particular item-loading is characteristic of successful
or unsuccessful treatment outcome, or whether this factor
as a whole profiles successful subjects.

Interpreting the significance of individual items
on a specific factor is further complicated by the fact
that the factor as a whole comprises a response-set.
Within this set any given item is therefore part of a
configural whole, and must be considered as such with
respect to its prognostic value. This phenomenon is con-
sistent with traditional interpretation of item values
on the clinical and validity scales of the MMPI, where for
a number of items a FALSE response may signify deviance on
a particular scale or scales, while a TRUE response connotes

deviance on another scale. Similarly, within this context

75

of the test asta whole, a significant elevation on a
particular scale connotes varying degrees and kinds of
psychopathology, according to the presence or absence of
other significantly elevated scales. An example of this
ambiguity in the present research may be seen in the load-
ing of item 248 (sometimes feel happy without reason) on
both Factor 3 (manipulative) and Factor 4 (Reactive
Depression), which are indicative of opposite treatment
outcomes.

A tentative attempt was made to identify each of
the eight factors as representative of either successful
or unsuccessful subjects, and to characterize the content
of the factor. As mentioned above, each succeeding factor
proved more difficult to interpret:

Factor 1. Hypochondriasis (unsuccessful)

Characteristics: 1. extreme self-alienation, poor
awareness of self

2. poor health and coordination
3. conflicted family relation-
ships
4. repressed hostility
5. attraction by same sex
Factor 2. Organic (successful)

Characteristics: 1. affirmation of fundamentalist
religious beliefs with regular
church attendance

2. belief in social order (law
enforcement)

3. less rigid defense system and
more realistic self-appraisal
than Factor 1

4. more healthy family and social

relationships based on give and
take

TABLE 4.5.

76

MMPI-Item Loadings on Factors Related to Successful/
Unsuccessful Treatment Outcome

 

 

 

 

successful subjects.

Table 4.5.1. Factor 1: Hypochondriasis
Item Loadingb Content
251 same (.57) no recall of own actions
156 same (.53) blank spells
247 same (.49) jealousy of family members
104 same (.46) lack of concern for what happens to self
23 same (.38) attacks of nausea and vomiting
187 same (.38) hands not clumsy or awkward
230 same (.38) normal circulatory-respiratory activity
186 same (.38) poor hand coordination
282 oppos. (.37) occasionally feel hatred for family members
61 same (.36) have not lived right kind of life
278 same (.35) feel strangers look critically
145 same (.34) feel like picking fist fight
69 same (.33) attracted by same sex
106 same (.33) feel wrong, evil
284 same (.32) sure I'm being talked about
aR = response endorsed by higher percentage of unsuccessful than

b . .
same or opp051te Wlth respect to R.

77

‘Table 4.5.2. Factor 2: Organic

 

Item R Loading Content

249 T oppos. (.54) believe there is Devil and Hell
278 T same (.46) feel strangers look critical

113 T same (.44) belief in law enforcement

115 F oppos. (.41) belief in life hereafter

134 F oppos. (.36) thoughts sometimes outrace speech
220 F oppos. (.31) loved my mother

91 T oppos. (.31) don't mind being made fun of

391 F oppos. (.29) remember playing sick

185 F oppos. (.28) can hear as well as most people
282 F oppos. (.28) occasionally feel hate for family member
82 F oppos. (.28) easily downed in arguments

225 F oppos. (.28) occasionally gossip a little

95 F oppos. (.27) regular church attendance

 

 

 

Table 4.5.3. Factor 3: Manipulative

Item R Loading Content

323 T same (.39) had peculiar, strange experiences
113 T same (.37) belief in law enforcement

392 T same (.36) talk to strangers in public places
215 T same (.35) excessive alcohol use

248 F oppos. (.32) sometimes happy with no reason
375 T same (.31) experts no better than I

82 F same (.29) easily downed in arguments

106 T oppos. (.29) often feel wrong or evil

134 F oppos. (.28) thoughts sometimes outrace speech
145 T same (.28) feel like picking fist fight

284 T same (.27) sure I'm being talked about

104 T oppos. (.27) don't care what happens to me

131 F same (.26) don't worry about diseases

Table 4.5.4. Factor 4:

Item R Loading

249 F oppos. (.44)
248 F oppos. (.40)
292 F same (.39)
367 T same (.33)
115 F oppos. (.32)
106 T same (.31)
113 T oppos. (.30)
46 F oppos. (.29)
392 T same (.24)
211 T same (.24)
95 F oppos. (.23)
272 F oppos. (.22)
61 T same (.22)

78

Reactive Depression

Content

believe Devil and Hell exist
feel happy without reason
don't speak till spoken to
not unusually self-conscious
belief in life hereafter
feel wrong, evil

belief in law enforcement
trusts own judgement

talk to strangers in public
can sleep in day, not at night
weekly church attendance
sometimes full of energy

haven't lived right life

 

Table 4.5.5. Factor 5:

Item
272
266
186
247
46 oppos. (.27)

R Loading
F
F
T
T
F
260 T same (.26)
F
F
T
T
F

oppos. (.43)
oppos. (.37)
oppos. (.28)

same (.28)

268
230
215

69
220

oppos. (.25)
oppos. (.25)
oppos. (.24)
same (.24)

oppos. (.24)

Familial Conflict

Content

sometimes full of energy

very excited at least once a week
poor hand coordination

jealous of family member
judgement better than ever

slow learner in school

excitement ends depression

normal circulatory-respiratory activity
excessive alcohol use

attracted by same sex

loved my mother

 

Table 4.5.6.

79

Factor 6: Somatization

 

Item R Loading Content
243 T same (.38) few or no pains

95 F oppos. (.36) regular church attendance

225 F same (.29) sometimes gossip a little

156 T same (.28) poor recall of own activity

23 T same (.28) attacks of nausea and vomiting
60 F oppos. (.26) don't read all editorials daily
343 T same (.23) stop and think before acting
106 T same (.21) feel wrong or evil

113 T same (.20) belief in law enforcement

104 T oppos. (.20) don't care what happens to me
367 T same (.20) not unusually self-conscious
247 T same (.18) jealousy of family member

248 F same (.18) sometimes feel happy - no reason
284 T .oppos. (.18) sure I'm being talked about

220 F same (.17) loved my mother

 

Table 4.5.7.

Factor 7: Hypochondriasis

 

Item R Loading Content
131 F same (.36) no worry about catching disease
106 T same (.31) often feel wrong, evil
251 T oppos. (.30) blank spells

392 T same (.28) talk to strangers in public places
82 T same (.27) easily downed in argument

323 T same (.24) peculiar, strange experiences
343 T same (.24) stop and think before acting
156 T oppos. (.23) poor recall of own activity

95 F same (.20) regular church attendance

61 T same (.18) have not led right life

 

80

 

 

Table 4.5.8. Factor 8: General Neurotic

Item R Loading Content
220 F oppos. (.34) loved my mother

46 F same (.26) judgement better than ever
134 T oppos. (.25) thoughts may outrace speech

145 T same (.25) at times feel like picking fight

23 T same (.24) attacks of nausea and vomiting

295 F same (.24) liked "Alice in wonderland"

104 T same (.21) don't care what happens to me

248 F same (.22) can feel happy without reason

325 F oppos. (.19) gossip a little

284 T oppos. (.16) sure I'm being talked about

159 T same (.15) can't understand so well what I read
272 F oppos. (.15) sometimes full of energy

282 F oppos. (.15) at times feel hatred for family member

 

81

Factor 3. Manipulative (unsuccessful)

Characteristics: 1.

2.
3.
4.

self viewed as in control,
without guilt, aggressively
mastering situations

world viewed as hostile,
threatening

strange experiences, excessive
alcohol use

manic mood swings

Factor 4. Reactive Depression (successful)

Characteristics: 1.
2.

3.

religious beliefs with regular
church attendance

self experienced as wrong, evil,
guilty

trust in own judgment and
ability to interact

Factor 5. Family Conflict(equivocal)

Characteristics: 1.
2.

3.
4.

ability to overcome depression
non-acknowledgement of poor
health or excessive alcohol use
conflicted family relationships
attraction by same sex

Factor 6. Somatization (unsuccessful)

Characteristics: 1.
2.
3.

4.
5.

depression, poor recall, guilt
regular church attendance

denial of social non—conformity,
denial of pain

nausea, anxiety

conflicted familial relationships,
including maternal

Factor 7. Hypochondriasis (unsuccessful)

Characteristics: 1.
2.
3.

4.
5.

disease phobia

self-alienation, low self-esteem
compulsive, good recall, no
blank spells

non-regular church attendance
peculiar experiences

82

Factor 8. General Neurosis (unsuccessful)

Characteristics: 1. love mother, hate some other
family member
2. depression, repressed hostility
3. declining powers of judgement
and comprehension
4. nausea, vomiting
5. disliked "Alice in Wonderland"

Summary of Results

 

The analyses presented in this chapter pertained
to the major research hypothesis, which was as follows:

Items on the MMPI do not differentiate to

a statistically significant degree between

patients who achieve success and patients

who experience failure in multidisciplinary

treatment for chronic low back pain.
Significant findings relevant to this hypothesis included:
1. With a chi-square analysis of the MMPI-item responses
of 123 former back pain patients, it was possible to
identify only 28 items discriminating at the .10 level, 10
of which were significant at the .05 level.
2. The best 24 of these items were incorporated into a
Back Treatment Success Scale, which did not have statis-
tically significant predictive validity in a cross-
validational procedure with an additional sample of 62
former back pain patients.
3. The cross-validational procedure, however, did indicate
considerable similarity of response among successful versus
unsuccessful subjects, respectively, of the research and

cross-validational samples. These samples were pooled for

further data analysis.

83

4. The statistical power of the chi-square item analysis
was enhanced by this pooling of subjects. For the total
185 subjects, 18 items were significant at the .05 level
and an additional 27 at the .10 level.
5. By examining the content of these 45 items, it was
possible to identify several personality characteristics
apparently related to outcome of treatment for chronic
low back pain.
6. A statistical factor analysis confirmed the previous
content analysis, and indicated the possibility of con-
figural relationships among identified personality char-
acteristics.

The implications and limitations of these findings

will be discussed in Chapter V.

CHAPTER V

DISCUSSION OF RESULTS

Summary

The past three decades have witnessed an accelerat-
ing rate of recognition and investigation of the psycho-
logical dimensions of pain. Most recently, these activ-
ities have culminated in the establishment of multi-
disciplinary clinics as an approach to treatment of per-
sons suffering from chronic low back pain. In such
clinics psychologists have assumed a major role on the
treatment team, providing a variety of counseling, be-
havioral management, and stress reduction and muscular
relaxation techniques, all of which aim to cure by
impacting upon cognitive, affective, and motivational
variables associated with the personal experience of pain.

Personality testing conducted in these and
similar settings has repeatedly confirmed the presence
of psychoneurotic disturbance among a majority of sufferers
of chronic low back pain. Moreover, the preponderance of
evidence gathered in such situations indicates that higher

levels of pre-treatment psychoneurotic involvement are posi-

tively correlated with unsuccessful treatment outcomes.

84

85

The MMPI consistently has detected such involvement and
has therefore emerged as a preferred instrument in the
evaluation of chronic low back pain patients.

The primary intent of this study was to explore
the feasibility of further refining the diagnostic power
of the MMPI by developing a subscale of MMPI-items for use
in predicting outcome of treatment for chronic low back
pain. A scale facilitating the identification of patients
with poor prognosis could provide a valuable diagnostic
tool and also improve the matching of treatment resources
and patient needs.

In working toward this end, the first phase of
research required the development of a measure to assess
long-term treatment outcome. The Pain Survey was con-
structed expressly for this purpose, then successfully
piloted. According to their response to this ques-
tionnaire, a total of 185 former pain clinic patients
were assigned either successful or unsuccessful treatment
outcome status.

From this sample population, two-thirds of the
subjects were randomly selected to comprise a scale-de-
velopment subsample, and the remaining one-third became
the cross-validational subsample. The former subjects'
MMPI records were analyzed by chi-square to determine
which items significantly discriminated between successful

and unsuccessful subjects. The items thus identified

86

constituted a tentative Back Treatment Success Scale,
whose ability to discriminate between successful and un-
successful subjects in the cross-validational sample was
then tested. This test proved statistically non-signifi-
cant.

It was determined, however, that the small size
of the scale-development subsample was likely to have been
a major factor in the non-significant outcome of the test.
Therefore, to provide as comprehensive a basis as possible
for future research, the two subsamples were pooled and
the chi-square test of item-discrimination was applied to
the total sample population. A total of 45 items, all
those discriminating at the .05 and .10 levels, were
examined for standard MMPI-scale membership, then
categorized on the basis of content, and finally factor-

analyzed.

Discussion

 

This section will discuss conclusions drawn from
the results of the statistical analysis, and limitations
of the study.

A major conclusion tentatively drawn from this
research is that the use of the MMPI in diagnosing and
screening chronic low back pain patients may be consider-
ably enhanced by an analysis of specific item responses.
By means of traditional MMPI analysis based on interpre-

tation of scale scores and their configural

87

interrelationships, previous research has established
several guidelines for the prediction_of treatment out-
come among such patients. Two investigative teams
(Wilfling, Klonoff, and Kokan, 1973; Pheasant and Holt,
1973) found poor response to treatment to be positively
correlated with significant elevations on all scales of
the neurotic triad -- Hypochondriasis, Depression, and
Hysteria. Two other teams (Wiltse and Rocchio, 1975;
Blumetti and Modesti, 1976) identified the Hypochondriasis
and Hysteria scales as so correlated.' Sternbach (1974)
obtained similar findings, and also identified elevations
on the Psychopathic Deviate, Psychasthenia, and
Schizophrenia scales as important predictors for back pain
patients with particular MMPI profiles. Partial corrobo-
ration of these findings had previously been obtained by
Wolkind and Forrest (1972) with the Middlesex Hospital
Questionnaire; they found that poor outcome was positively
correlated with neurotic scores on the depression, somatic
concomitants of anxiety, and obsessionality scales.

A direct comparison of these findings with the
present results would require the computation of mean
scale scores from the records of subjects participating in
this study. However, since these subjects were not atypical
of back pain patients participating in previous studies,
the assumption can be made that mean MMPI scale scores

for successful and unsuccessful subjects are essentially

88

equivalent for this and other studies. The more pertinent
analysis has to deal with the content and factor loadings
of the particular items selected as discriminating.

It should be noted here that the two types of
analysis (traditional scale configuration versus item
analysis) complement one another, but are not directly
comparable. For example, the fact that in the present
study two items on the Hypochondriasis scale were sig-
nificant discriminators does not mean that unsuccessful
subjects as a group scored an average of 2 points (raw
score) higher than successful subjects on this scale. The
true difference in scale elevation would be a function of
the differential in successful and unsuccessful subjects'
percentage of endorsement of all items on this scale, not
merely these discriminating at a pre-set level. A gen-
erally valid rule of thumb, however, would indicate a
positive correlation between mean elevation of a given
scale and the number of significant items belonging to
that scale.

With this in mind, results of the item analysis
do suggest both similarities and differences with respect
to previous outcome studies -- similarities, insofar as
numerous items from the Depression, Psychopathic Deviate
and Schizophrenia scales proved to be significant pre-
dictors, and differences insofar as numerous items from

the F scale proved significant. Why this particular

89

difference should exist is at least partially explicable
by referring to previous publications concerning the F
-scale. This is a validity scale monitoring test-taking
attitude, and high scores on this scale may indicate
either a deliberately distorted self-description which
claims fictitious mental symptoms in the attempt to escape
responsibilities or "an exaggeration of existing dif-
ficulties to gain attention and assistance" (Lachar, 1974,
p. 2). These interpretations can be reasonably discounted
among chronic pain patients, the great majority of whom
are notorious for their denial of psychological problems.
Attention must therefore be directed to Blumberg's (1967)
finding that the degree of F scale elevation is a good
indicator of the severity of psychiatric disturbance.
Where this finding applies, according to Lachar (1974),
the testee is likely to be extremely self-deprecatory and
suffering from severe stress. This appears to be the best
interpretation of the data obtained here, as these are
often-documented elements of the typical pain-patient pro-
file. However, since no previous research has cited the
F scale as a clinical predictor for this population,
further interpretation of this finding should await in-
dependent confirmation by future research.

Another apparent anomaly in the present findings
is the relative paucity of discriminating items belonging

to the Hypochondriasis (two items) and Hysteria (four

90

items) scales. In fact, the two items belonging to the
first scale also comprise two of the four items on the
second. This appears contradictory to previous findings
that high elevations on these scales are the most sig-
nificant and consistent predictors of poor treatment out-
come. Assuming that results of the current research are
valid and that they are congruent with previous findings,
the only possible explanation is that, even though a con-
siderable number of items on these scales may have been
endorsed more frequently by unsuccessful than successful
subjects, thus yielding higher mean scores for unsuccessful
subjects, for only a very few items did difference in rate
of endorsement reach statistical significance. This
explanation is completely plausible and, as stated above,
simply illustrates that scale analysis and item analysis
provide complementary results.

In this respect suggestions by Sternbach (1974)
regarding a more refined approach to configural interpre-
tation are worthy of some attention. He suggested that
analysis of personality differences between successful
and unsuccessful subjects could best proceed from an ex-
amination of four MMPI profiles frequently obtained by low
back patients, which he designated as typical of hypo-
chondriasis, reactive depression, somatization, and
manipulative reaction. These four profiles were reviewed

at length in Chapter III. They are<mfrelevance to this

91

study because their main features bear some correspondence
to the profiles obtained by statistical analysis of the

45 discriminating items. Direct comparison of these pro-
files and the factors obtained in this study is difficult
because, as mentioned above, the presence of significant
items does not directly translate into elevations on a
particular scale. Nevertheless, each factor does suggest
a response set, whose equivalent is the configuration of
several scale scores.

The most obvious profile-to-factor relationship is
between the manipulative reaction profile and Factor 3.
The manipulativeness, aggression, guile, and hostility
of the sociopathic personality typify the two response
sets. There is strong reason to suspect that these are,
in fact, health-system sociopaths, the work-weary, legally
involved, secondary-gainers, and/or those who simply
derive satisfaction from frustrating the clinical team's
attempts to relieve their pains.

The hypochondriasis profile appears most closely
related to Factor 1. Health complaints not typically
associated with low back pain are present. Dependency
conflicts with the family and the health care system
appear likely with each. Schizophrenic elements of self-
alienation indicate the severity of the condition.
Sternbach noted that these patients have a poor prognosis

and continue to seek medical help after discharge.

92

Factor 4 most closely approximates the char-
acteristics of reactive depression. Such a depression
has likely resulted from the repeated failure of attempts
to cope with and find relief from pain. Patients' pre-
pain lives were not symptom-centered. There are indica-
tions of social and religious interests and involvement,
together with a basic trust in one's own judgment, all
of which bode well for such patients developing a life
apart from pain. Sternbach contends that prognosis is
more favorable for this than for any other group.

The somatization or conversion-V profile bears a
striking resemblance to Factor 6. The characteristic
elements appear to be a rigid defense system based to a
great extent on physical health complaints, denial of any
social non-conformity, and elements of dependency.
Especially interesting in Factor 6 is the occurrence of
Item 243 as the highest loading item: though admitted to
a clinic for chronic pain, patients affirming this item
steadfastly maintain that they have few or no pains. This
appears to be an absurd and contradictory assertion to
deny pathology and to protect precarious self-esteem.

The correspondence of these four factors with gen-
erally recognized profiles characteristic of low back pain
is an encouraging aspect of this study, and lends a cer-
tain credibility to the results. It also indicates that

other factor-profiles may be typical of this population

93

and worth attempting to interpret. The items loading on~
Factor 2 do not seem evident of psychoneurotic involvement.
There are few, if any, indications of an abnormal profile.
In this respect one is reminded foremost of the "organic"
classification of back pain patients first established with
the MMPI by Hanvik (1951) and recently partially supported
by Freeman, Calsyn, and Louks (1976). Hanvik found no' '
evidence of neuroticism among these patients, while Freeman
et al. found minor elevations on the neurotic triad scales.
Factor 5 indicates conflicts in familial relation-
ships and sexual identification. Whether or not for the
patients profiled in this factor such conflicts are
indicative of poor treatment outcome is unclear. Never-
theless, the literature on chronic low back pain frequently
indicates difficulties within the immediate family, where
pain is used as means of manipulating conflict and of
meeting unfulfilled dependency needs. One recent study
has strongly emphasized the importance of directly treat-
ing such pathology within the family in order to best in-
sure treatment progress and a favorable outcome (Hudgens,
1979). Pathology detected by Hudgens in families with
either spouse suffering from chronic pain included extreme
dependency, indirect communications, narrow social con-
tacts, inability to handle anger appropriately, sexual
conflicts, a general power struggle between spouses, and

conflict over male-female roles. Such observations support

94

the plausibility of the item-complex represented by
Factor 5.

The most heavily weighted item on Factor 7 con-
cerns disease phobia, even though there are no significant
loadings of items dealing with poor health or health
complaints. Such a profile strongly suggests hypo—
chondriasis, linking low self-esteem and compulsive
attention to one's physical condition. Precisely this
aspect of hypochondriasis, in contrast to recital of
varied physical complaints, has been delineated in a re-
cent article by Meister (1980). Obviously, even though
no major health concerns are evident on the item-loadings,
these patients have focused hypochondriacal attention on
their musculo-skeletal system, as indicated by their ad-
mission to the clinic.

Dependency, depression, and passive agression are
the most salient characteristics included in Factor 8.
Again one is reminded of the possibility of interpersonal
conflict and manipulation. These are general characteristics
of low back pain patients often cited in the literature,
and they typify the personality functioning of patients
scoring high on the neurotic triad of MMPI scales. More
Specific indications are difficult to interpret from this
factor.

These comparisons provide a general confirmation

of the results obtained in this and previous research.

95

Even though no opportunity was available to crossvalidate
the 45 items significant for the total sample population,
the relevance of the factor analysis to previous research
findings supported the validity of the items as predictors
of treatment outcome. Whether or not these eventually
prove to be the best 45 items for inclusion in a predictive
subscale, and how such a subscale can best be used in
screening applicants for treatment, remains to be deter-
mined by future research. ‘It is apparent, however, that

a scale consisting of approximately 45 items would allow
not only the computation of a total scale score but also
the accurate identification of factors to be considered

in the design of individual treatment programs.

Limitations of the Study. Pain is a subjective experience.

 

There are no standardized methods to evaluate or measure
pain relief. Lacking such methods, and given the debat-
able validity of the patient's verbal report, such reports
were supplemented with behavioral criteria as measures of
treatment outcomes in this study. These, also, were not
standardized, but did conform to general clinical con-
sensus.

All subjects in the study were selected from former
patients of the Ingham Medical Center Low Back and Pain
Clinic, Lansing, Michigan. No attempt was made to con-
trol for the sex, age, marital status, educational dif-

ferences or other demographic variables in the sample

96

population. Findings of this research are therefore
limited to the subjects who took part in the study but
can be generalized by the Cornfield, Tukey, Bridge Argu-
ment (1956) to populations similar to the group of sub-
jects studied.

Although potential respondents identified for
this study were not atypical of chronic low back pain
patients participating in treatment programs throughout
North America, those actually participating in the study
selected themselves by their decision to respond to the
mailed questionnaire or telephone inquiry. Given this
fact of self-selection, it is not inconceivable that the
185 subjects in the sample population represented a sub-
set of former patients with certain personality char-
acteristics not normally distributed among originally
identified potential respondents. For example, successful
respondents may have had greater needs for dependency or

affiliation than successful non-respondents, while

unsuccessful respondents may have been more hostile or
more passive-aggressive than unsuccessful non-respondents.
Similarly, several former patients still pursuing legal
claims expressly declined to reSpond at risk of jeopar-
dizing their cases, and approximately seven percent of the
mailed questionnaires were returned by the postal service
as undeliverable, so that legal involvement and mobility

also affected response rates. Either of these could have

97

been associated with personality factors that would have
been reflected in patients' MMPI responses and, therefore,
in the item analysis.

Another sampling limitation is the number of sub-
jects who participated in the study. The considerable
differences in number, content, and significance level of
discriminating MMPI items selected from the full sample
population's records as opposed to the records of a
randomly selected two-thirds of this population, indicate
insufficient stability of the chi-square with this sample
size and nearly 400 variables. The stability of the 45
items significant for the total sample population re-
mains to be determined by future crossvalidation.

Attention should also be directed to the varying
time intervals per subject from completion of treatment
to follow-up. For the subjects responding, this interval
ranged from six to forty-two months. In the course of
three and one-half years many personal and environmental
variables may have emerged to mitigate the once positive
effects of treatment. WOrk lay-offs caused by the state
of the economy, death or injury to significant others,
familial break-up, could all have caused, directly or
indirectly, the changes in vocational or health status
associated with treatment outcome as measured by the Pain

Survey.

98

A final possible source of invalidity is instru-
mentation. A major goal of the clinic program is to
assist patients to cope with pain that may not be relieved
by treatment. Several respondents expressly noted their
appreciation of this treatment component when responding
to the questionnaire, most often as a rationale for their
decision not to seek further medical care. This dimen-
sion of treatment, however, was not measured systemat-
ically by the questionnaire. Presumably, it could have
been incorporated in the third criterion question, con-
cerned with continued efforts to find medical help, or
it could have been measured by a fifth criterion question.
With favorable response to this dimension considered
indicative of successful outcomes such an adjustment might
then have resulted in a shift of certain subjects from

the unsuccessful to the successful category.

Implications for Future Research

 

Of major importance in research correlating patient

personality characteristics with treatment outcome for

chronic low back pain is the establishment of basic
criteria for determining success in treatment. Relevant
research published in the 1970's, the decade during which
the multidisciplinary approach to pain has emerged as a
viable treatment concept, has not been based on stan-
dardized measures of treatment outcome. The measure de-

veloped and employed in the present study, the Pain

99

Survey, was meant to provide a standardized instrument
based on self-reported behavioral and affective/evalua-
tive rating criteria most commonly cited in the litera-
ture. In research concerned with the personal experience
of pain and the impact of pain on patient lifestyles, it
is clear that both the more objective/behavioral and the
more subjective/evaluative dimensions of treatment must
be measured. Future inclusion of a question concerned
with treatment impact on the ability to c0pe with pain
would appear to increase the validity of the outcome-
measure.

A major issue with respect to standardization of
outcome measures concerned the impact of time elapsed
after treatment on such factors as pain relief and con--
tinuation of vocational and everyday life activities.
Assessment of outcome at "one year or more after treat-
ment" appears to be the modal interval among related
studies, though selection of this time interval is no-
where defended or explained. It would appear important
to determine some norms detailing long-term effect of
multidisciplinary treatment which could serve as guidelines
in conducting post-treatment studies, in order more
sharply to differentiate between treatment effects due to
personality characteristics and effects related to the

passage of time.

100

Based on the potential contribution and the
limitations of the present study, the following indications
for future research are offered for consideration:

1) Further development and standardization of the Pain
Survey as an instrument measuring outcome of treatment for
chronic low back pain.

2) A study of the effects of time on the outcome of
multidisciplinary treatment for chronic low back pain,
with a view to clarifying possible interaction between
personality factors and the passage of time.

3) Sampling of a larger percentage of identified potential
respondents, either through personal interview or repeated
mail and telephone follow-up, to control for the possible
error introduced by self-selection of respondents.

4) Use of a larger sample to improve robustness of the
chi-square test of item discrimination, and comparison of
results with the 45 items selected here or, in lieu of
this, cross-validation of the 45 items on an available
sample. Either procedure would constitute a next step in

the develOpment of a scale with sufficient power to

accurately predict outcome of treatment for chronic low
back pain.

5) A thorough examination and consideration of the
resulting content-related item-response factors to deter-
mine the extent to which these should be specifically
addressed by therapeutic programs attempting to re—

habilitate the chronic low back pain patient.

101

6) An examination of the relationship between socio-
economic status/level of employment and the personality
profiles of chronic low back pain sufferers. It seems
reasonable to hypothesize, for example, that unemployed,
unskilled, and semi-skilled members of the work force would
be more likely to manifest the manipulative reaction
associated with anger, rebelliousness, acting-out, and
secondary gain, whereas skilled and managerial employees
may typically manifest other profiles (perhaps somatiza-
tion).

7) Comparison of factors here identified as related to
failure in chronic low back pain treatment with factors
related to failure in treatment for other chronic
syndromes, e.g., migraine and tension headaches, colitis,
dysmenorrhea. Findings of similarity among these factors
may lead to identification of a general "chronicity pro-
file" associated with resistance to medical and psycho—

logical treatment.

Conclusions

 

The research conducted here may best be considered
exploratory. Limitations with respect to sampling pro-
cedures and especially sample size did not allow final
development of a diagnostic instrument and the ascertain-
ing of its validity, reliability, and practical utility.

The design of research and analyses of the data

did, however, permit some progress to be made toward this

102

goal. An apparently valid treatment outcome measure was
developed. A 24-item Back Treatment Success Scale was
constructed which proved moderately significant in cross-
validation. This result supported the assumption that
patients who benefit from chronic low back pain treatment
differ from those who do not in their response to specific
items on the MMPI, an obvious condition for the develop-
ment of a predictive subscale.

The test of item-discrimination applied to the
MMPI records of the total sample and subsequent analyses
of the 45 best discriminating items provided further
support for the feasibility of eventual scale-construction.
The content of these items to a large extent coincided
with psychological characteristics previously identified
as critical in determining outcome of treatment for chronic
low back pain. Additionally, a statistical factor analysis
revealed patterns of response to these items which were
congruent with previously identified pain-patient profiles.
These results suggest that the 45 MMPI items here selected
and analyzed possess considerable validity as indicators
of treatment outcome, and may serve as a basis for the
construction of a scale as was here intended. A task
remains for future research to test the extent of their
reliability as a next step in the develOpment of a scale

with practical utility.

APPENDICES

APPENDIX A

PAIN SURVEY

NAME

APPENDIX A

PAIN SURVEY

 

DATE OF BIRTH

EDUCATION (highest grade completed)

 

 

CURRENT EMPLOYMENT

 

(CHECK ONLY THE ONE BEST ANSWER TO EACH QUESTION)

1.

How did the treatment at the Ingham Low Back and Pain Clinic
affect your pain?

a. relieved all or most of my pain.
b. relieved some pain.

c. relieved pain only for a while, or not at all.

 

2.

How did the treatment affect your ability to work and carry on
normal activities?

a. I am able to do my normal work and be just about as active
as I like.

b. I have been able to return to less strenuous work and activities.

c. Because of continued pain I have not been able to return to

 

3.

work or to many of my normal activities.

Since your treatment, have you continued to seek medical help for
your back pain?

a. No, because I have no more pain.

b. No, because my pain doesn't bother me as much any more.

c. I want to, but I can't afford the time or money for more

 

treatment.

103 ~

104

d. No, because I have lost all faith in doctors and specialists
to cure my pain.

e. Yes, I am still seeing a doctor or specialist.

 

f. I am still looking for a doctor or specialist who can relieve
my pain.

4. What effect has the treatment had on your life in general?

a. helped very much, I'm very satisfied.

 

b. helped enough that I'm mostly satisfied.

c. helped so little that I'm not very satisfied.

 

d. did not help at all, I'm disappointed.

Part II
YES NO

1. Does pain keep you from falling asleep at night?

2. Are you awakened by pain during the night?

3. Do you often drink to deaden the pain, or feel like it?
4. Do you have to lie down and rest often because of pain?
5. Are you afraid to be without your pain medicines?

6. Does pain interfere with your marriage or family life?

7. Does pain interfere with your sex life?

8. Are you unable to do the things you want because of pain?
9. Do you find that all you can think about is your pain?
10. Do you have fewer friends or go out with friends less

often because of pain?

SUGGESTIONS FOR IMPROVING THE CLINIC

 

 

 

OTHER COMMENTS

 

 

 

APPENDIX B

COVER LETTER ACCOMPANYING PAIN SURVEY

APPENDIX B

COVER LETTER ACCOMPANYING PAIN SURVEY

Dear Patient:

We are sending out this questionnaire to a small number of
patients treated in the past at the Ingham Medical Center,
Low Back and Pain Clinic. We are hoping to obtain new know-
ledge that will be of help in treating patients coming
through the program today.

Please complete these questions to the best of your ability
and return the questionnaire today in the self-addressed
envelope which is provided. Your responses are, of course,
confidential and will not be part of your medical record,
nor will they, in any way, influence your future treatment.

The goal of this survey is to better understand and effec-
tively treat the chronic pain patient coming through the
Clinic.

Thank you very much for your prompt attention to this
questionnaire. If you have any questions, please do not
hesitate to contact the Low Back and Pain Clinic at Ingham
Medical Center. The phone number is:

Once again, thank you for helping us with this questionnaire.

105

APPENDIX C

ENDORSEMENT PERCENTAGES FOR MMPI FORM-R ITEMS
BY SUCCESSFUL VS. UNSUCCESSFUL SUBJECTS (N = 185)

APPENDIX C

ENDORSEMENT PERCENTAGES FOR MMPI FORM-R ITEMS
BY SUCCESSFUL VS. UNSUCCESSFUL SUBJECTS (N = 185)

Item More Frequent Response Signi- Item More Frequent Response Signi-

 

 

# Ua (s) 5C (a) ficance # U (s) s (s) ficance
1 F (53) F (59) .530 22 F (81) F (78) .769
2 T (83) T (92) .129 23 F (80) F (90) .090
3 F (77) F (67) .209 24 F (84) F (88) .608
4 F (86) F (78) .205 25 F (76) F (69) .372
5 T (69) T (78) .201 26 F (52) F (52) .942
6 F (73) F (67) .454 27 F (88) F (94) .295
7 T (56) (50:50) .503 28 F (83) F (89) .351
8 F (62) T (72) .178 29 F (64) F (72) .397
9 F (87) F (92) .385 30 T (79) T (82) .794
10 F (86) F (88) .832 31 F (76) F (84) .266
11 F (75) F (77) .965 32 F (72) F (80) .304
12 T (58) T (53) .670 33 F (63) F (69) .429
13 F (52) F (62) .256 34 F (83) F (89) .365
14 F (75) F (69) .440 35 F (95) F (93) .843
15 F (55) F (60) .617 36 F (64) F (60) .664
16 F (86) F (92) .396 37 T (86) T (83) .769
17 T (92) T (89) .798 38 F (59) F (59) .897
18 T (54) T (62) .380 39 (50:50) (50:50) .951
19 F (86) F (88) .870 40 F (85) F (86) .883
20 T (64) T (59) .618 41 T (51) F (54) .573
21 F (66) F (72) .489 42 F (90) F (94) .466
aUnsuccessful subjects bPercentage of unsuccessful subjects

endorsing this response
c .
Successful SUbJeCts dPercentage of successful subjects
endorsing this response

106

Item More Frequent Response

#

43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75

U

emememmmmmemeemeeeameemmmmmmmmmme

(%)

(57)
(72)
(58)
(56)
(72)
(90)
(95)
(91)
(81)
(85)
(85)
(92)
(60)
(85)
(70)
(65)
(56)
(80)
(70)
(76)
(55)
(64)
(89)
(84)
(54)
(68)
(88)
(67)
(65)
(74)
(67)
(60)
(93)

S

8"er"10-3'11'11"!WWHWHHWHHHBWHHWMMWWWWHW'UH

(%)

(52)
(76)
(54)
(59)
(75)
(92)
(100)
(93)
(79)
(83)
(88)
(91)
(69)
(81)
(68)
(71)
(56)
(95)
(84)
(77)
(60)
(61)
(88)
(84)
(56)
(69)
(96)
(53)
(70)
(82)
(74)
(60)
(99)

1107

Signi-
ficance
.584
.692
.684
.080
.784
.807
.147
.854
.920
.975
.726
.998
.273
.606
.858
.604
.892
.009
.044
.994
.707
.802
.932
.918
.871
.929
.103
.149
.532
.264
.441
.905
.189

 

Item More Frequent Response

#

76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108

U

mememmeeemememmmmmememmmeememeamw

(%)

(74)
(55)
(52)
(53)
(85)
(51)
(76)
(96)
(54)
(97)
(73)
(76)
(92)
(63)
(90)
(51)
(77)
(54)
(58)
(71)
(77)
(76)
(72)
(59)
(57)
(77)
(74)
(53)
(87)
(88)
(80)
(80)
(78)

S

MHWH'GMHHHMHWHW'QMWWHWH"1'13'11'90-3'68'11'110-5'1'4

(%)

(77)
(56)
(51)
(56)
(84)
(53)
(62)
(93)
(54)
(100)
(62)
(78)
(92)
(66)
(91)
(68)
(77)
(65)
(65)
(56)
(80)
(87)
(81)
(66)
(54)
(79)
(75)
(51)
(96)
(90)
(92)
(84)
(85)

Signi-
ficance
.742
.903
.983
.304
.942
.885
.057
.583
.364
.377
.155
.319
.862
.812
.839
.034
.945
.191
.367
.064
.856
.110
.245
.427
.876
.890
.945
.896
.069
.917
.057
.633
.328

Item More Frequent Response

#

109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141

U

mmememmemmmemaemmm

(%)

(64)
(89)
(55)
(79)
(94)
(76)
(69)
(75)
(57)
(67)
(73)
(57)
(91)
(84)
(93)
(60)
(75)
(59)

50:50

"18'1188'31888680-3'118

(69)
(53)
(72)
(64)
(66)
(86)
(69)
(76)
(68)
(82)
(51)
(89)
(74)
(59)

S

"18"!86'60-388886'60-3HMMWNHWWHWWWHWHHW'UW

(%)

(68)
(91)
(65)
(69)
(100)
(79)
(85)
(81)
(64)
(67)
(80)
(55)
(97)
(86)
(99)
(51)
(83)
(55)
(60)
(66)
(55)
(78)
(78)
(72)
(78)
(85)
(85)
(73)
(85)
(53)
(96)
(78)
(55)

1108

Signi-
ficance
.483
.973
.234
.215
.098
.700
.023
.445
.426
.841
.380
.884
.142
.817
.130
.160
.313
.692
.256
.793
.818
.470
.068
.529
.203
.025
.179
.589
.687
.720
.198
.686
.667

 

Item More Frequent Response Signi-

#

142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174

U (%)

(73)
(70)
(88)
(77)
(85)
(59)
(61)
(67)
(69)
(97)
(52)
(57)
(91)
(62)
(86)
(81)
(68)
(65)
(96)
(74)
(54)
(73)
(84)
(71)
(58)
(84)
(95)
(85)
(59)
(54)
(57)
(60)
(51)

HHWBWH'UMHWHMWMMMM'U'HWt-J'd'd'flv-J'fi'd'd'd'fl'fl'flri

S

HB'QHWB'QWBWHMWRJ'UWW'UW"!H'll'fl'fli-J'fl'fl'fi'd'q'd'fit-J

(%)

(62)
(72)
(88)
(88)
(88)
(62)
(72)
(66)
(74)
(100)
(51)
(53)
(92)
(60)
(95)
(86)
(68)
(79)
(95)
(71)
(54)
(74)
(84)
(67)
(56)
(74)
(94)
(89)
(58)
(64)
(58)
(73)
(58)

ficance

.153
.878
.825
.099
.780
.827
.204
.921
.605
.373
.985
.715
.910
.926
.094
.563
.877
.062
.881
.766
.887
.928
.846
.684
.962
.152
.904
.568
.998
.248
.919
.120
.482

.109

Item More Frequent Response Signi-I Item More Frequent Response Signi-
# U (%) S (%) ficance # U (%) s (%) ficance
175 . T (56) T (67) .199 208 T (52) F (52) .755
176 (50:50) T (51) .999 209 F (95) F (96) .942
177 T (97) T (99) .829 210 r (96) F (100) .227
178 T (82) T (90) .251 211 F (85) F (94) .089
179 IF (74) F (79) .534 212 F (91), F (96) .361
180 F (55) F (62) .442 213 F (91) r (97) .187
181 P (63) F (63) .930 214 T (56) T (53) .822
182 F.(86) r (89) .699 215 F (64) F (83) .008
183 F (51) T (56) .442 216 F (85) F (84) .987
184 F (91) F (97) .136 217 T (69) T (64) .499
185' T (76) T (90) .030 218 F (91) F (97) .148
186 F (63) F (77) .064 219 F (76) F (69) .370
187 T (66) T (87) .004 220 T (92) T (99) .086
188 F (57) F (51) .476 221 T (63) T (63) .911
189 F (74) F (72) .990 222 ' r (67) F (68) .985
190 T (53) T (53) .943 223 F (62) F (74) .123
191 F (69) F (63) .512 224 F (69) F (76) .416
192 T (52) T (58) .533 225 T (75) T (91) .013
193 T (75) T (85) .140 226 T (62) T (66) .680
194 F (85) r (91) .382 227 F (95) F (92) .581
195 T (80) T (84) .638 228 T (80) T (83) .699
196 T (95) T (91) .369 229 F (79) r (78) .983
197 F (93) F (99) .173 230 F (52) T (61) .096
198 T (67) T (66) .920 231 F (64) F (60) .664
199 T (76) T (81) .499 232 F (64) F (59) .629
200 F (92) F (91) .923 233 F (77) F (67) .190
201 F (57) F (55) .907 234 T (56) T (51) .599
202 r (89) F (93) .510 235 T (55) T (59) .706
203 r (80) F (82) .917 236 F (84) F (77) .354
204 r (81) r (84) .807 237 F (56) T (51) .502
205 F (96) F (100) .228 238 T (65) T (64) .952
206 F (81) F (81) .856 239 F (60) F (65) .560
207 T (78) T (77) .975 240 F (72) F (78) .448

 

1110

Item. More Frequent Response Signi-

#

241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273

U

MBBWHWHW'IJHJHWW'U'UHHJH’U

(%)

(73)
(60)
(91)
(51)
(88)
(94)
(89)
(73)
(63)
(63)
(89)
(92)
(79)
(75)
(63)
(93)
(89)
(97)
(65)

F (58)

(50:50)

WBWMWH'UW'H'UWH

(76)
(60)
(55)
(79)
(71)
(60)
(60)
(96)
(67)
(80)
(72)
(52)

S

tit-3'11”:"18'4'6'0'11'1168'6'1168'110-3"JH'H'U'Ht-J'd'd’fl'd'd’dfil'd

(%)

(65)
(56)
(99)
(56)
(90)
(92)
(97)
(51)
(78)
(67)
(78)
(87)
(75)
(71)
(64)
(90)
(84)
(100)
(70)
(75)
(54)
(73)
(63)
(56)
(84)
(53)
(58)
(77)
(97)
(54)
(85)
(87)
(55)

ficance

.336
.766
.084
.528
.792
.770
.092
.005
.052
.654
.072
.313
.727
.728
.950
.703
.498
.377
.514
.035
.703
.742
.782
.940
.527
.024
.921
.037
.968
.110
.501
.033
.466

 

Item More Frequent Response‘ Signi-

#

274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306

U

HWHWHWWWWWHHBWWWBHWHWHMWWBMWW'dti'fl'fl

(%)

(51)
(90)
(90)
(70)
(76)
(88)
(71)
(68)
(69)
(61)
(65)
(93)
(92)
(52)
(84)
(82)
(52)
(94)
(78)
(91)
(68)
(57)
(69)
(80)
(74)
(68)
(67)
(69)
(86)
(73)
(64)
(78)
(92)

S

Quilt-3’11H'fl'fl'fl'fl'flflefl'fl'fi'fl'fla'fle'fifl'fl'fl’flHM'URJ'Alt-l'di-S

(%)

(56)
(95)
(97)
(81)
(61)
(92)
(77)
(77)
(53)
(68)
(78)
(93)
(93)
(55)
(88)
(89)
(59)
(99)
(65)
(92)
(74)
(71)
(74)
(86)
(83)
(60)
(64)
(78)
(85)
(75)
(71)
(75)
(93)

' ficance

.480
.338
'.129
.134
.037
.499
.473
.250
.048
.500
.065
.947
.938
.751
.578
.252
.222
.261
.074
.967
.471
.099
.616
.484
.230
.351
.804
.294
.958
.898
.374
.792
.924

Item More Frequent Response

#

307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339

U (%)

(67)
(64)
(83)
(65)
(60)
(91)
(65)
(66)
(94)
(56)
(62)
(72)
(S8)
(83)
(55)
(60)
(66)
(96)
(59)
(85)
(74)
(75)
(68)
(66)
(95)
(74)
(95)
(83)
(82)
(63)
(73)
(56)
(93)

"18'0'5'11'11"1'11"!WWWBMWMWHMM'QB'GBWNH’UNHHWW

S

mmmmmmmmmmmmemmmmemmmemmmmemmeamw

(%)

(73)
(63)
(78)
(68)
(60)
(96)
(70)
(66)
(96)
(56)
(63)
(76)
(69)
(84)
(54)
(52)
(81)
(99)
(64)
(77)
(81)
(86)
(66)
(59)
(96)
(81)
(91)
(80)
(76)
(65)
(73)
(51)
(97)

11].

Signi-
ficance
.487
.964
.546
.881
.874
.369
.650
.923
.833
.180
.913
.745
.221
.952
.962
.373
.055
.576
.574
.263
.383
.144
.971
.431
.910
.390
.528
.719
.402
.885
.862
.515
.401

 

Item. More Frequent Response Signi-
U (%)

#

340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372

eemmeammmmmmmmmmmmmemmmmmemmmmmma

(54)
(89)
(82)
(59)
(93)
(84)
(81)
(79)
(57)
(85)
(91)
(85)
(83)
(57)
(94)
(98)
(63)
(71)
(87)
(57)
(89)
(52)
(63)
(95)
(86)
(88)
(81)
(74)
(58)
(56)
(62)
(58)
(56)

S

aemmeemmmmmammwmmmmemmmmmammmmmme

(%)

(62)
(88)
(83)
(74)
(88)
(80)
(71)
(88)
(57)
(86)
(90)
(87)
(87)
(54)
(97)
(93)
(71)
(77)
(84)
(60)
(94)
(51)
(65)
(100)
(90)
(91)
(76)
(60)
(60)
(57)
(65)
(58)
(52)

ficance

.345
.852
.969
.076
.474
.661
.188
.183
.930
.962
.953
.831
.602
.823
.534
.218
.415
.474
.775
.840
.397
.844
.897
.161
.690
.664
.518
.081
.990
.978
.890
.920
.646

1312

Item More Frequent Response Signi-

#

373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399

U

WIMP-3HWHWBWHWWBHHBWHMHHWWWHMW

(%)

(77)
(54)
(74)
(81)
(59)
(63)
(72)
(74)
(66)
(62)
(57)
(57)
(54)
(58)
(71)
(84)
(63)
(84)
(69)
(71)
(78)
(77)
(78)
(65)
(57)
(74)
(69)

S

HWBHMW’UHGWJ

(50:50)

WWHHWH'GHWHWWHHBB

(%)

(81)
(51)
(60)
(85)
(59)
(66)
(68)
(83)
(63)
(58)

(64)
(53)
(54)
(69)
(79)
(67)
(77)
(54)
(55)
(78)
(68)
(84)
(56)
(59)
(76)
(58)

ficance

.661
.644
.104
.692
.928
.862
.699
.237
.757
.740
.453
.483
.933
.764
.935
.540
.720
.364
.089
.050
.882
.244
.460
.344
.960
.957
.198

APPENDIX D

LOADINGS ON EIGHT FACTORS OF 45 MMPI FORM-R

ITEMS DISCRIMINATING BETWEEN 76 SUCCESSFUL

AND 109 UNSUCCESSFUL PATIENTS TREATED FOR
CHRONIC LOW BACK PAIN

113
APPENDIX D
LOADINGS ON EIGHT FACTORS OF 45 MMPI FORM-R
ITEMS DISCRIMINATING BETWEEN 76 SUCCESSFUL

AND 109 UNSUCCESSFUL PATIENTS TREATED FOR
CHRONIC LOW BACK PAIN

6am0¢¢o~c~t owc-ohn cw.mcmo~r~o~¢mocm

IACIOI I
3
3°
S

=38
2
I
I
z.

7

gnaw-“Mos... uscmucv—tmwsmvcoawm NOD-0PM
ENAWN (QC NQC wt #01" NP|NON5\M¢a¢O¢fl51¢3~~OOM¢QEMNKOQ
cued-«.000. NGQOM mm cmm~~amom nannquKOv-M
PW'WU“ 3NNQMO°~MW1WNOOMO~IPWWM
ecu-venou-moom OmgmoWOOOMQOcv-OWOWOOGN
cocoon-000000.000 00.000.000.00...0.0.0.0000.

[ICIOR

6

c—vc-acsw. can: ”wwwuwwu‘dw ~08 we: O-IKQ “WNW
Ohvcmv Obit-05:061.- o Mute c‘uowoammc-tv Oven Guano-MN.-
~c coma 0&0th $9vpc~t< mcht‘MV-GMN uvvocen
ham-cone aO-FGU‘MOFN'CNQMMW flan-Mfumcab‘h OGQMOW- c
.‘JC f- :G’C (AWN-(J! .4069?va oo—c .v-fitofl-pQ-c c—ch goo-p" GNU-coo

0.0.0.0000...cocoa-00000000000000.0000-00coco

flCIOI

5

w—nmnommm mmcmmwmncm
"a. .OHw-NN-t'oommhoch-QM- :- Mum-«c VOU'M’.~&U‘.P¢" 03:50
”We? '0 ONW¢~O~O¢QVQO F'O—mswor Q awn-P00 0036001“
WOmc Wm mmnomcmmnawommc
GNOWCOccOmfl'GGGOM6WG¢NO~CFGMAWP¢POGP¢O
000.00IOOOIOOOOOOOOOOIOOOOOOIOOOIIOOOIOOOJOOO
ll IlIlllO‘l III ltllt

VICIOI

WM OONWQWWC'N commute"
mg 01—: WW ment-Shad”. mac-Mme: CNcNNv-«m
M000“) MOT-fine omcmhc- ococmxocmsoOv-oo-tsmc—
mcmcomhoOqu—t wwoanw—owovm (\me

64670.3 4
5
§:
0
O
O
‘1’;

- 0
S
3%
0
1
I
0
0
a
0
2
0
i’
3
1
C
3
b
l

3

MW“ oOlflW (ucmwwfifllw SON P‘OthGG-‘a5'9-‘M'MQ‘O

MQNOtno sooocowmvm 3!! WW GW'OhMMO chem

55‘0" Whl' P'VoQ ONOM‘OQ'C ””41" ~30 fi‘W-‘OOPOW m
M pct.

ratio:
0
0
Io-g
{5
.,
o
l
I
i
1
o
o
2
1
b
3

~P¢Q¢Qeuhon¢9mgm'vwcwa«nooo¢6¢Chceomhvm~hmdo¢
mmo-Oremw-o omens-nor. och 00¢ 02.21er- ~39 are 5M €

IACIOI ?
O
?
0
0
4
4
o
3
06
3
g
i
2
0
2
o
E
2
?

OMB P. M: WNCQOM” ”Wench o c-OM 66W”
envosccemm omwomcw-an Hm «WM
mnthmcocmoon¢omso~co~cw~ Mun-l econ
momma—ow ocommmm~tm ommonm~womom
tho-Fm" MP ~010- omrr- crmmgcc. Qw'lkwénuflh MNONOq-C-M

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOIOOOOOOOOOO

34618. 1

Lowmmoomehmov orifice v-OOCNQN-cmlnmvnh Ont-N
no om‘nOO—uuflmcmn “We «accent... coowewoo

‘34-'34..- ‘- --A- -A.AAA.A:B mmmn A.Mnn_.u‘ -.

 

REFERENCES

REFERENCES

Anastasi, A. Psychological testing (4th ed). New York:
Macmillan, 1976.

Beals, R.K., & Hickman, N.W. Industrial injuries of the
back and extremities. Journal of Bone and Joint
Surgery, 1972, S4-A, 1593-1611.

Beecher, H.K. Measurement of subjective responses:
Qualitative effects of drugs. New York: Oxford
University Press,_l959.

Blumberg, S. MMPI F scale as an indicator of psycho-
pathology. Journal of Clinical Psychology, 1967,
£3, 96-99.

Blumetti, A.E., & Modesti, L. Psychological predictors of
success or failure of surgical intervention for
intractable back pain, in J.J. Bonica and D.G.
Albe-Fessard (Eds.), Advances in pain research and
therapy (Vol. 1). New York: Raven Press, 1976.

Bonica, J.J. Management of chronic pain: Medicine's new
growth industry. Medical World News, October
1976.

Brady, R. Goethe's natural science: Some non-Cartesian
meditations, in K.E. Schaefer, H. Hensel, & R.
Brady (Eds.), Toward a man-centered medical science
(Vol. 1). New York: Futura Publishing Co., 1977.

Bullock, J.M. Facet injectra and facet rhizotomy 2-year
follow-up. Presentation to American Academy of
Orthopedic Surgeons, New Orleans, 1977.

 

Bullock, J.M., Jerome, J.A., & Pool, A.P. Back pain clinic:
Design, method and results. Presentation to
American Academy of Orthopedic Surgeons, San
Francisco, 1975.

Calsyn, D.A., Louks, J., & Freeman, C.W. The use of the
MMPI with chronic low back pain patients with a
mixed diagnosis. J. of Clinical Psychology, 1976,
fig, 532-536.

114

115

Chapman, W.P., Finesinger, J.E., Jones, C.M., & Cobb, S.
Measurements of pain sensitivity in patients with
psychoneurosis. Archives of Neurology and
Psychiatry, 1947, El, 321-331.

 

 

Cornfield, J., & Tukey, J.W. Average values of mean squares
in factorials. Annals of Mathematical Statistics,
1956, 21, 907-949.

 

Dahlstrom, W.G., Welsh, G.S., & Dahlstrom, L.E. An MMPI
handbook (Vol. 1): Clinical interpretation.
Minneapolis: University of’Minnesota Press, 1972.

 

Fetterman, J.L. Back disorders of psychic origin. The
Ohio State Medical Journal, 1937, 31(b), 777-781.

 

Finneson, B.E. Modulating effect of secondary gain on the
low back pain syndrome,:h1J.J. Bonica & D.G. Albe-
Fessard (Eds.), Advances in pain research and
therapy (Vol. 1). New York, Raven Press, 1976.

 

Fordyce, W.E. The treatment of chronic loco back pain.
Presentation to Rehabilitation Institute of
Chicago, March 17, 1976a.

 

Fordyce, W.E. Behavioral method for chronic pain and
illness. St. Louis: The C.V. Mosby Co., 1976b.

 

‘Foster, M.W. Neurosis and trauma. J. of Clinical
Orthopaedics, 1964, g2, 54-59.

 

 

Freeman, C., Calsyn, D., & Louks, J. The use of the
Minnesota Multiphasic Personality Inventory with
low back pain patients. J. of Clinical Psychology,
1976, 22, 294—298.

 

Gentry, W., Shows, W., & Thomas, M. Chronic low back
pain: A psychological profile. Psychosomatics,
1974, 15(4), 174-177.

 

Hanvik, L.J. MMPI profiles in patients with low-back pain.
J. Cons. Clin. Psych., 1951, 15, 350-353.

 

Haven, G.A., Jr., & Cole, K.M. Psychological correlates
of low back pain syndrome and their relationship
to the effectiveness of a low back pain program.
Newsletter for Research in Psychology, 1972, 11(2),
31-33.

 

Head, H. Studies in neurology. London: Kegan Paul, 1920.

 

116

Hill, H.E. Studies of anxiety associated with anticipa-
tion of pain. Archives of Neurological Psychiatry,
1952, 61, 612.

 

Hoover, N.A. Indications for fusion at time of removal of
intervertebral disc. J. Bone and Joint Surgery,
1968, 50-A(1), 189-193.

Hudgens, A.J. Family—oriented treatment of chronic pain.
Journal of Marital and Family Therapy, 1979, 5(4),
67-78.

 

Husserl, E. Ideas. New York: Collier Books, 1962.

Jerome, J.A. Decreasing musculoskeletal tension via bio-
feedbaEk and relaxation training_procedures for
the treatment of chronic low back pain. Unpub-
Iished doctoral dissertation, Michigan State
University, 1978.

 

 

 

Jerome, J.A. Personal communication. October 2, 1979.

Kraus, H. Management of chronic pain: Medicine's new
growth industry. Medical World News, October
1976.

 

Lachar, D. The MMPI: Clinical assessment and automated
interpretation. Los Angeles: Western Psychological
Services, 1974.

 

 

Lamaze, F. Painless childbirth. New York: Henry Regnery
Co., 1970.

 

Leroux, C. Chronic pain: When all traditional cures fail
us. Chicago Tribune, March 29, 1979, Sec. 2, 1-6.

 

Lewis, T. Pain. New York: Macmillan, 1942.

Lippincott, A.J. Development of a back surgery success
scale of the MMPI to help predict chemonucleolysis
surgical success (Doctoral dissertation, United
States International University, 1976). Disserta-
tion Abstracts International, 1976, 31, 1958B.
(University Microfiims No. 76-22, 392)

 

Martins, A.N., Ramirez, A., Johnston, J., & Schwetschenau,
P.R. Double-blind evaluation of chemonucleolysis
for herniated lumbar discs. J. Neurosurgery,
1978, 49, 816-827.

 

McGill, C.M. Industrial back problems. J. Occup. Med.,
1968, 19(4), 174-178.

 

117

Meister, R. Closet hypochondriacs. Psychology Today, 1980,
11(8), 28-37.

 

Melzack, R. The puzzle of pain. United Kingdom: Penguin
Books, 1973.

 

Melzack, R., & Wall, P.D. Pain mechanism -- A new theory.
Science, 1965, 150, 971-979.

Merskey, H. Psychiatric aspects of the control of pain,
in J.J. Bonica & D.G. Albe-Fessard (Eds.),
Advances in pain research and therapy (V01. 1).
New York: Raven Press, 1976.

 

Nagi, S.Z., Burk, R.D., & Potter, H.R. Back disorders and
rehabilitation achievement. J. Chronic Diseases,
1965, 18, 181-197.

 

Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K., &
Bent, D.H. Statistical package for the social
sciences (2nd ed.). New York: McGraw-Hill, 1975.

Noordenboos, W. Pain. Amsterdam: Elsevier, 1959.
Pavlov, I.P. Conditioned reflexes: An investigation of

the physiolo ical activity of the cerebral cortex.
CTV. Anrep Edi). Dover PuinEhers, 1927.

 

 

Pheasant, H.C. Adult back clinic. South California
Orthopedic Group. Los Angeles, Calif., 1972.

 

 

Pheasant, H.C. & Holt, J.M. Low-back disorder: Psycho-
logical, demographic, and organic correlates of
response to treatment. Orthopedic Hospital Report,
Los Angeles, Calif.: 1973.

 

Phillips, E.L. Some psychological characteristics
associated with orthopaedic complaints. Current
Practice of Orthopaedic Surgery, 1964, 1, 165-176.

 

Pinto, M. (State of Michigan Dept. of Labor). Personal
communication. April 13, 1979.

Russek, A.S. Medical and economic factors related to the
compensable back injury. Archives of Phys. Med.
and Rehab., 1955, 16, 316-323.

 

 

Sargent, M. Psychosomatic backache. New England J.
Medicine, 1946, 234(13), 427—436}

 

Sarno, J.E. Chronic back pain and psychic conflict.
Scandinavian J. of Rehab. Med., 1976, 8(3,4), 145-
153.

 

118

Shealy, C.N. Percutaneous radio frequency denervation of
spinal facet. Treatment for chronic back pain and
sciatica. J. of Neuro-surgery, 1975, 11, 448-451.

 

Shealy, C.N. Behavioral techniques in the control of pain,
in M. Weisenberg & B. Tursky (Eds.), Pain: New
perspectives in therapy and research. New York:
Plenum Press, 1976.

 

 

Shealy, C.N. Holistic techniques in pain management.
Presentation to Heritage Hospital Pain and Stress
Control Center, Muskegon Heights, MI, September
1977. ,

 

Steiner, R. Goethes Naturwissenschaftliche Schriften.
Stuttgart: Veflag Freies Geisteslében,’l962.

 

Sternbach, R.A. Pain -- A psychophysiolpgical ana1ysis.
New York: Academic Press, 1968. ‘

 

Sternbach, R.A. Pain patients: Traits and treatment.
New York: Academic Press, 1974.

 

Sternbach, R.A., Wolf, S.R., Murphy, R.W., & Akeson, W.H.
Aspects of chronic low back pain. Psychosomatics,
1973, 11, 52-56. (a)

 

Sternbach, R.A., Wolf, S.R., Murphy, R.W., & Akeson, W.H.
Traits of pain patients: The low-back "loser."
Psychosomatics, 1973, 1311), 226-229. (b)

 

Stromberg, L.R. (Medical Director, Travenol Laboratories,
Deerfield, 111.). Personal communication. May 11,
1979.

Sullivan, J.D. Psychiatric factors in low-back pain. New
York State Journal of Medicine, 1947, 41, 2683-2687.

 

Swartz, M.S. & Krupp, N.E. The MMPI "conversion-V" among
medical patients: A study of incidence, criteria,
and profile elevation. Journal of Clinical
Psychology, 1971, 11, 89-95.

 

 

Thomas, M.R. & Lyttle, D. Development of a diagnostic
checklist for low back pain patients. J. of
Clinical Psychology, 1976, 11, 125-129.

 

University of Florida Rehabilitation Research Institute.
The hurt that doesn't go away -- Treating chronic
pain through behavior modification. Rehab Brief,
1978, 1(9).

 

119

Wall, P.D. Modulation of pain by non-painful events, in
J.J. Bonica & D.G. Albe-Fessard (Eds.), Advances
in pain reseapgp and therapy_(Vol. 1). New York:
Raven Press, 1976.

Waring, E.M., Weisz, G., & Bailey, S. Predictive factors
in the treatment of low back pain, in J.J. Bonica &
D.G. Albe-Fessard (Eds.), Advances in pain research
and thera (Vol. 1). New York: Raven Press,

 

White, A.W.M. The compensation back. Applied Therapeutics,
1966, 1, 971-974.

White, A.W.M. Low back pain in men receiving workmen's
compensation: A follow-up study. Canadian
Medical Association Journal, 1969, 161(2), 61-67.

Wilfling, F., Klonoff, H., & Kokan, P. Psychological,‘
demographic and orthopaedic factors associated
with prediction of outcome of spinal fusion.
Clin. Orthop., 1973, 19, 153-160.

 

Wiltse, L. & Rocchio, P. PreOperative psychological tests
as predictors of success of chemonucleolysis in
the treatment of the low-back syndrome. J. Bone
and Joint Surgery, 1975, 57-A(4), 478-483.

Wolkind, S.N. & Forrest, G.J. Low back pain: A psychiatric
investigation. Postgraduate Medical Journal, 1972,
11, 76-79.

 

Yochelson, L. Psychiatric aspects of backache. Cur. Prac.
Orth. Surgepy, 1966, 1, 253-272.

 

 

 

        

    

 

HICHIGRN STRT

HI (I)

312 31

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E UNIV. LIBRRRIES
HMINIHFIIIHIWIWHIIHHHI
00643869