LIBRARY

Mlchigan State
Unlverslty

 

 

 

PLACE II RETURN BOX to roman this Mom hum your need.
To AVOID FINES mum on or baton dd. duo.

DATE DUE DATE DUE DATE DUE

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU II An mm mm Opportunity lmwon
Wm‘

TREATMENT DIFFERENCES FOR AMBULATORY HYPERTENSIVE
PATIENTS SEEN IN OFFICE-BASED SETTINGS

BY

Linda S. Hughes

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

Master of Science in Nursing
College of Nursing

1996

ABSTRACT
TREATMENT DIFFERENCES FOR AMBULATORY HYPERTENSIVE
PATIENTS SEEN IN OFFICE-BASED SETTINGS
BY

Linda S. Hughes

Cardiovascular disease is the leading cause of death
for Americans and hypertension is a major risk factor for
it. Despite this, fewer than one—fourth of all
hypertensive persons control their blood pressure. Two
purposes of the current study are: to determine whether
appropriate interventions are being ordered, and to
determine what conditions are associated with giving
adequate treatment for hypertension.

Data for this investigation came from the 1992
National Ambulatory Medical Care Survey of patients seen by
office-based physicians and practitioners.

A very significant finding was that only one-fifth of
newly diagnosed hypertensive patients got even one non-
pharmacological treatment. Only 85.5 percent of these
hypertensive patients got any treatment or follow—up,

including medications.

For non-pharmacological treatments, obese patients

were three times as likely as others to receive it,

Hispanics only one-third as likely. After combining all

treatments and follow-up into a single category of

"intervention", however, the differences disappeared.

ACKNOWLEDGEMENTS

This research project would not have been completed
without the valuable assistance of many people. I am very
grateful to Manfred Stommel, the chairperson of my thesis
committee, for guiding me through the research process and
for helping to clarify the statistical interpretations. I
would also like to thank Suzanne Budd and Brigid Warren for
their ongoing support and for the time they devoted to
reviewing drafts of this thesis.

I would also like to thank my family, who have
generously supported me in the completion of this project

with their love and patience.

iv

TABLE OF CONTENTS

LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . Vi
LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . Vii
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . l
THEORETICAL FRAMEWORK . . . . . . . . . . . . . . . . . .7
REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . .13
METHODS . . . . . . . . . . . . . . . . . . . . . . . . 29

Data Base Survey . . . . . . . . . . . . . . . . . 29

Variables . . . . . . . . . . . . . . . . . . . . .35
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . 45
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . .53
IMPLICATIONS . . . . . . . . . . . . . . . . . . . . . .58

General Implications . . . . . . . . . . . . . . . 58

Implications for Practice . . . . . . . . . . . . .61

Limitations . . . . . . . . . . . . . . . . . . . .65
REFERENCES . . . . . . . . . . . . . . . . . . . . . . .68
APPENDIX A . . . . . . . . . . . . . . . . . . . . . . .72
APPENDIX B . . . . . . . . . . . . . . . . . . . . . . .79

APPENDIX C - UCRIHS Approval Letter. . . . . . . . . . .79

Table
Table
Table

Table

Table

4a:

4b:

LIST OF TABLES

Frequencies of the Patient Characteristics
Treatment and Follow-up Frequencies
Bivariate Correlations Matrix

Logistic Regression to Predict the Odds of
Patient or Setting Characteristics to
Influence Behavioral Therapeutic Treatments
Logistic Regression to Predict the Odds of
Patient or Setting Characteristics to

Influence Behavioral or Pharmacoloogical
Treatment or Appropriate Follow-up

vi

46

.47

.49

51

.52

Figure 1:
Figure 2:

Figure 3:

LIST OF FIGURES

King's Model of Goal Attainment
King's Model as it applies to this study

Age-adjusted death rates (per 100,000)
for diseases of the Heart . . .

vii

.10

12

.22

TREATMENT DIFFERENCES FOR AMBULATORY HYPERTENSIVE

PATIENTS SEEN IN OFFICE-BASED SETTINGS

INTRODUCTION

In 1995, cardiovascular disease was the leading cause
of death for Americans, according to the Center for Disease
Control (1995), and hypertension is a major risk factor in
most cardiovascular diseases (editors, American Journal of
Public Health, 1994). Diseases caused by hypertension
include stroke, ischemic heart disease, congestive heart
failure and renal failure (MacMahon, Petro, and Cutler,
1990 and Collins, Petro, and MacMahon, 1990).

Hypertension remains the most prevalent, yet
medically treatable, chronic condition among adult
Americans. Nearly 50 million people, or more than twenty
five percent of the adult United States population, are
affected by hypertension (The Fifth report of the Joint
National Committee on Detection, Evaluation, and Treatment
of High Blood Pressure [JNC], 1993).

Hypertension is one of the commonest reasons for
visits to family physicians. Left untreated, it and its
sequelae can lead to serious morbidity and mortality,
according to Kligman (1993). Although usually silent,
hypertension may present with fatigue, occipital and
pulsating headaches in the early morning, lightheadedness,

1

flushing, epistaxis, chest pains, visual and speech
disturbances and dyspnea. Without treatment, 70 percent of
hypertensive patients die of congestive heart failure or
coronary heart disease, 15 percent die of cerebral
hemorrhage and 10 percent die of uremia (Blake, 1994).

Despite the risks of high blood pressure, national
data indicate that fewer than 25 percent of all
hypertensive persons currently meet the recommended blood
pressure goal of less than 140/90 mm Hg. (JNC, 1993)

The gap between the potential to control blood
pressure and its achievement was recognized soon after
effective hypotensive agents became readily available
(Creditor and Schoenberger, 1972). Since then, a
consistent pattern of high rates of patient dropout and
failed therapy has been widely documented (Stockwell,
1994). Despite great technical advances, specifically,
the development of a wide range of new antihypertensive
drugs and drug classes, the process of care available in
the community has not changed to any great extent, and
treatment itself remains highly ineffective, according to
Stockwell.

The Fifth Report of the Joint National Committee on
Detection, Evaluation, and Treatment of High Blood Pressure
(JNC) outlines currently accepted treatment protocols for
managing the care of hypertensive patients. This report

was derived from solid scientific evidence accumulated over

two decades through research under the auspices of the
National Institutes of Health (NIH). The JNC was
established in 1976 by the National High Blood Pressure
Education Program of the National Heart, Lung, and Blood
Institute of the NIH to provide concise, clinically
relevant consensus guidelines on the detection and
treatment of hypertension. The resulting documents are
authored by clinical hypertension experts and public health
specialists who develop a consensus on hypertension
management that integrates their differing viewpoints
(Roccella, 1992.)

Each JNC publication is triggered by the release of
significant hypertension research results and is,
therefore, built on a strong foundation of clinical
experience. Each publication is also endorsed by the
Coordinating Committee, a body of professional and
voluntary organizations that includes the American Academy
of Family Physicians, the American Academy of Cardiology,
American College of Chest Physicians, American College of
Physicians, American Diabetes Association, American
Dietetic Association, American Heart Association, American
Hospital Association, American Medical Association,
American Nurses' Association, American Osteopathic
Association, American Pharaceutical Association, American
Public Health Association, American Red Cross, American

Society of Hypertension, and many other prominent health

organizations. This coordinating committee also includes
eight federal agencies: The Agency for Health Care Policy
and Research, Health Care Financing Administration, Health
Resources and Services Administration, National Center for
Health Statistics, Centers for Disease Control, National
Heart, Lung and Blood Institute, National Institute of
Diabetes and Digestive and Kidney Diseases, and Veterans
Affairs.

The current study will concern itself with
"appropriate" treatment. Because it is nurse-practitioner-
oriented, not a medical or pharmacological study, it will
focus mainly on non-pharmacological treatments for
hypertension. According to the JNC, non-pharmacological
treatments of hypertension are always appropriate, whereas
the appropriateness of pharmacological treatments would
vary according to many factors which were not reported on
this survey (exact blood pressures, patient history, any
prior treatments and responses to such treatment, physical
condition, complications, comorbid conditions, etc.)

Data for this investigation came from the 1992
National Ambulatory Medical Care Survey (issued in October,
1995). This survey was conducted throughout the entire
year, 1992, by the National Center for Health Statistics
(NCHC). The survey provides data from samples of records
of medical care given to patients seen by office-based

physicians and practitioners. The basic sampling unit is

the physician—patient encounter or visit. The physicians
included in the study were sampled from a list of providers
classified by the American Medical Association or the
American Osteopathic Association as those giving office-
based care.

In this study, the types of treatment offered to the
newly-diagnosed hypertensive patients will be examined.

Given the problems of failure to implement known,
effective therapies, the current study is an attempt to
determine whether or not appropriate interventions are
being used with hypertensive patients in the United States,
and what factors are associated with the use of appropriate
interventions. The implication of this study is that if
patients are not receiving adequate treatment, and if the
optimum conditions were known, then care providers should
work to change conditions in order to improve the care of
hypertensive clients, to decrease morbidity and mortality

from cardiovascular disease.

Research questions:

1) Do the interventions which are being prescribed
for hypertensive patients meet the current standards of
care for hypertension, as outlined by the JNC?

2) What personal characteristics of patients such
as: age, race, ethnicity, gender or physique or their

healthcare environments, such as their insurance

arrangements, predict that appropriate, adequate treatment
is being prescribed by providers?

According to Hickey, Ouimette and Venegoni (1996),
nurse practitioners will‘increasingly be providing the type
of primary care that physicians have been giving in the
past, especially as capitation and managed care increase.
"Nursing and nurses are broadly focused on the healthcare
needs of all people, especially of underserved and
vulnerable populations." Nurses are exquisitely prepared
to manage chronic illnesses, such as hypertension, a major
focus in U.S. healthcare. So, advanced practice nurses can
take the implications from this physician study and apply
them to their practices, as they work with hypertensive

patients in primary care.

THEORETICAL FRAMEWORK

Imogene King's conceptual framework of

open systems

and her theory of goal attainment apply to this project's

central theme of differences in treatment of hypertensive

patients (King, 1971.)

First, King's open systems framework is composed of

three interacting systems;
the interpersonal systems,
Individuals comprise one type
called the personal systems.

dyads, triads,
another type of system called
Groups with special interests
which make up communities and

systems in King's theoretical

For the purposes of this

and small and large groups,

these are the personal systems,

and the social systems.

of system in the environment
Individuals interact to form
which comprise

interpersonal systems.

and needs for organizations,
societies are called social

framework.

project, the healthcare

practitioner and patient form an interpersonal system,

generally a dyad King (1971.)

interpersonal systems are interaction,

transaction, role,

and stress.

King's concepts for

communication,

Interaction means a dynamic reciprocity, defined as

the observable behaviors of two or more persons in mutual

presence.

current study.
Communication,

nonverbal,

7

as defined by King (1971),

situational, perceptual,

This is the provider-patient visit in the

is verbal or

irreversible

interactions which transmit ideas from one person to
another. For the current problem, this is the information
that is shared between provider and patient at the visit.

King defines transactions as "a process of
interactions in which human beings communicate with
environment to achieve goals that are valued ... goal—
directed human behaviors." For the current topic,
transactions are directed toward reducing blood pressure
and reducing the risk of cardiovascular disease.

The nurse's (or provider's) role is that of a
professional, using skills, knowledge and values to
identify goals and to help others to achieve goals. These
goals would be the health and well-being of hypertensive
patients, specifically the control of hypertension to avoid
cardiovascular disease.

For this project, the individual or social systems are
relevant, but are not the main focus. The main focus is on
the interpersonal relationship between patient and
caregiver. King's (1981) theory of goal attainment,
however, is of critical relevance to the current topic.

King explains her theory of goal attainment as "in the
interpersonal systems in which two people, who are usually
strangers, come together in a health care organization to
help and be helped to maintain a state of health that
permits functioning in roles." All of the above

characteristics of interpersonal relationships apply. King

states that caregivers and clients communicate information,
set goals mutually, and then act to attain those goals.

The steps are the basics of the nursing process:
assessment, diagnosis, planning, implementation and
evaluation. For the current project, since only a one-time
visit is studied, the assessment, diagnosis, planning are
done within the context of the single, first-time visit of
a hypertensive patient to a new provider. The patient
comes away from the visit with a diagnosis of hypertension,
and a plan of treatment. In some cases, the implementation
is done within the single visit (such treatments as
counseling.) In other cases, the implementation is only
ordered at the visit, to be done at a later time. The
evaluation is generally not done at this first visit, but
at a subsequent Visit, to discern the efficacy of treatment
in lowering the patient's blood pressure. Only if a
follow-up visit was not grdered, can we determine that
evaluation was probably not done.

King's theory of goal attainment suits the current
project in that we are looking at a way to interrelate
several concepts to arrive at a determination of whether or
not goals of lowering hypertensive patients' blood
pressures could be met. King's way of relating the

individual and the larger society to the specific

interaction are particularly relevant to this study's

1O

efforts at determining what factors help or hinder

appropriate treatment of hypertension.

Figure 1: King's model of goal attainment.

 

 

 

 

W Perceptionl 4— Feedback 0— ————————— '1

i I . |
I
l
l
I
I

1

:1 ,
2'

:3

Judgement

 

 

 

 

 

 

Interaction I—o Transactionl

/'

Action —0 Reaction

 

 

Judgement

 

I
I
I
I
|
t I

m Perception ! ‘— Feedback I‘— _________ .J

 

 

 

 

 

 

 

. ‘ , , ' ‘ Transacuon
Action -9 RCQCIIOD --o Interacuon I ... (Goal outcome)

 

 

11

This figure shows the relationships of the caregiver
and client, whose separate perceptions lead to judgment,
action, reaction, then together for interaction and
transaction, or goal. Interaction is the client and
provider meeting, where communication takes place, which
can lead to mutual goal-setting, and subsequent goal
attainment. The goal, in this case is control of the
client's blood pressure.

The current study examines not this final goal, but
the steps which lead to it. The outcome variables are the
giving or ordering of correct treatments and the ordering
of follow-up. Without knowing the specific content of the
interaction between caregiver and patient, or the patient's
perspective, the only part of this model which can be
determined from the current study is the caregiver's
perception, judgement and action. We can tell from the
medical care survey that the patients have high blood
pressures (perception) and that they have been diagnosed as
having hypertension (judgement). We can tell from the
survey what therapies or follow—up were ordered (action).
The survey, as it was designed and utilized, tells us
nothing else. King's model, as it applies to this survey,

is only the portion shown in Figure 2.

12

Figure 2: King's model as it applies to this study.

 

 

 

W I Perception I

1

lludgemcnt I

 

\

 

| Action I

 

 

High blood pressure.
Health problems

Hypertension
Medical and nursing diagnoses
Treatment Options and plans

Prescription of interventions
and, or follow-up

REVI EW OF LITERATURE

Hypertension is a major risk factor for coronary heart
disease (Houston, 1992). In the United States,
approximately 50 million persons have elevated blood
pressure or take antihypertensive medication (JNC, 1993).
Primary hypertension develops from unknown causes between
the ages of 25 and 55 years in persons without preexisting
renal disease. Nonfatal and fatal cardiovascular diseases,
including coronary heart disease and stroke, as well as
renal disease and all—cause mortality increase
progressively with higher levels of both systolic and
diastolic blood pressure. These relationships are strong,
continuous, graded, consistent, independent, predictive,
and etiologically significant, according to the JNC (1993).

Blake (1994) stated that through the provision of
continuous care and preventive medical education, family
physicians are well equipped to recognize, manage and,
possibly prevent hypertension in their patients.

Much is known about how to treat hypertension successfully.
Hundreds of studies have been done which examine both
pharmacological and non-pharmacological treatments for
hypertension. Over the past two decades, these studies

have been closely evaluated by the National

13

14

Institutes of Health. Many of the studies, in fact, have
been done by government researchers within the National
Institutes of Health. The Fifth Report of the JNC (1993)
is the Institute's recommendations based on all these
studies. The recommendations are first for life—style
modifications: tobacco avoidance, weight reduction,
moderation of alcohol intake, physical activity,
modification of dietary habits (moderation of sodium and
fat intake, adequate intake of potassium, calcium and
magnesium). If medications are ordered for the further
reduction of hypertension, the JNC V recommends that these
lifestyle modifications continue. (The 1988 Report of the
Joint National Committee on Detection, Evaluation, and
Treatment of High Blood Pressure had the same
recommendations for non—pharmacological treatment of
hypertension.)

A meta-analysis of non—pharmacological interventions
(Devine and Reifschneider, 1995) reviewed 102 studies
between 1961 and 1993 and found that across all studies, a
beneficial lowering of blood pressure was related to what
was termed "psychoeducational care" —— such as educational
programs for weight reduction, exercise, sodium
restriction, compliance with medications, compliance with

follow-up care, and relaxation.

15

Recommendations for non—pharmacological interventions
include:
Tobacco avoidance
Weight reduction
Moderation of alcohol intake
Physical activity
Modification of dietary habits
Limitation of sodium
Limitation of fat
Provision of adequate potassium
Provision of adequate calcium

Provision of adequate magnesium

The JNC did not find scientific evidence to support
recommendations for reducing caffeine, for varying the
proportions of carbohydrate or protein in the diet, for
increasing amounts of garlic or onion. The report did not
find sufficient evidence to support the use of relaxation
or biofeedback therapies for definitive treatment of
hypertension.

According to Alderman (1994), there is a direct
positive relationship between body mass index and blood
pressure. The Intersalt study of fifty-two communities

world wide showed that weight, among all measured

16

characteristics except age, had the strongest and most
consistent correlation with blood pressure (Intersalt
Cooperative Research Group, 1988). Alderman (1994) stated
that individuals with central or upper body obesity, as
indicated by a higher waist-to-hip ratio, are even more
likely to have an elevated blood pressure. Central obesity
also carries a greater risk of cardiovascular mortality.
Reisen, Abel, Modan, M., Silverberg, Eliahou and Modan, B.
(1978) found that weight reduction lowers blood pressure.
Even with sodium and potassium intakes constant, a 5 kg
weight loss was associated with a systolic blood pressure
reduction of 5.4 mm Hg and a diastolic reduction of 2.2 mm
Hg. The Reisen group found that the addition of sodium
restriction to weight loss enhances blood pressure decline.

A meta—analysis of seventy-eight sodium-reduction
studies (Law, Frost, and Wald, (1991) showed that a 100
mmol (about 6 grams) reduction in daily sodium intake would
lower systolic blood pressure by 14 mm Hg in hypertensives
and 10 mm Hg in those with normal blood pressures. Since
some of the studies in this analysis were not well
controlled, a more rigorous analysis was done by Cutler,
(1991) using only randomly-controlled trials. The Cutler
analysis showed that a 100 mmol/day reduction of sodium was
associated with a decline of 5.7 mm Hg systolic and 2.7 mm
Hg diastolic in hypertensive subjects and 2.23 mm Hg

systolic in those with normal blood pressures.

17

The Intersalt Cooperative Research Group (1988)
reported that a high dietary potassium intake may protect
against hypertension, and a potassium deficiency may
increase blood pressure and induce ventricular ectopy. The
group recommends that normal plasma concentrations of
potassium should be maintained, by diet alone or, if
necessary, by supplementation.

Hamet, and Lambert, (1991) found that calcium
deficiency is associated with hypertension, and a low
calcium intake may amplify the effects of a high sodium
intake on blood pressure. The JNC (1993) stated that many
but not all epidemiological studies showed an inverse
association between dietary calcium and blood pressure.

The JNC (1993) stated that although there is some
evidence of an association between lower dietary magnesium
intake and higher blood pressures, the associations are not
strong enough to recommend increasing magnesium intake for
control of blood pressure.

In a review study of available research on the
relationship between dietary fats and blood pressure,
Sacks, (1989) looked at randomly controlled studies and
found that diets varying in total fat and proportions of
saturated fats to unsaturated fats had little, if any,
effect on blood pressure. However, the JNC concludes that
since dyslipidemia is a major independent risk factor for

coronary artery disease, the limitation of dietary fat is

18

an important treatment to be included with treatments for
hypertension. The American Heart Association states that a
diet high in saturated fat and cholesterol is a main cause
of high blood cholesterol and a high low density
lipoprotein level. Both saturated fat and cholesterol
raise the blood cholesterol level. Studies of large groups
of people show that the risk of heart attack for a person
with a blood cholesterol level of 240 mg/dl or more is
twice that of a person whose cholesterol level is 200
mg/dl. (American Heart Association, 1993).

In the 1993 National High Blood Pressure Education
Program Working Group Report on Primary Prevention of
Hypertension the group stated that increased physical
activity, either alone or as part of a weight loss program,
has frequently been proposed as a means to lower blood
pressure. The group sites more than thirty studies of
various types. Some were observational studies of
participants' activity levels and blood pressures, some
were studies of exercise capacity and blood pressure, some
were studies comparing physically active or physically fit
participants with those age-matched subjects who were less
active. All of these studies showed significant inverse
relationships between levels of physical activity and blood
pressures.

The Multiple Risk Factor Intervention Trial Research

Group (Neaton, and Wentworth, 1992) showed a strong

19

positive relationship between smoking and death from
coronary heart disease, although not specifically from
hypertension. The National High Blood Pressure Education
Program, however, found a positive relationship between
smoking and hypertension, and recommends the avoidance of
smoking for the prevention of coronary heart disease and
stroke (Roccella, 1993). According to the National
Institutes of Health publication, the Physician's Guide
"How to help your hypertensive patients stop smoking"
(1984), smoking and hypertension are two major independent
risk factors for cardiovascular disease. The combination
of these two risk factors greatly increases the probability
of disease. The Framingham Heart Study data show that 40-
year-old males who smoked and had elevated blood pressure
(greater than 165 systolic) were three and one—half times
more likely to develop cardiovascular disease than males
without these risks. Males in this age group with elevated
blood pressure but not smoking were only twice as likely to
develop cardiovascular disease as those without these
risks.

The National High Blood Pressure Education Program
Working Group Report on Primary Prevention of Hypertension
(1993) also evaluated the effect of stress management on
hypertension. In several cross-sectional, case-control and
prospective studies in a variety of populations, long-term

stress appears to play an important role in the development

20

and maintenance of hypertension. The relationship between
stress and hypertension seems to persist even after
controlling for age, race, body mass index, type—A
behavior, alcohol consumption, family history of
hypertension, urinary sodium excretion, and physical
activity. Recent studies have emphasized the role of job
strain in the development of hypertension. Individuals who
are exposed to stressful situations, but lack the decision—
making ability to control their exposure and response, may
be at special risk for hypertension. The studies have also
indicated that lack of social support in the environment
may also increase the risk of hypertension. Although these
associations between stress and hypertension are well—
established from research studies, successful treatment of
stress for the reduction of blood pressure appears
difficult. Studies of treatments such as relaxation and
biofeedback show inconsistent results.

Shea (1994) stated in an editorial in the American
Journal of Public Health that national surveillance data
continue to show substantial gaps in detection, awareness,
initiation of treatment and control of hypertension, with
control rates even lower among minority and socially
disadvantaged groups. He said studies of patients with
hypertensive emergency and severe uncontrolled hypertension

have found that almost all of these hypertensives were

21

aware of having hypertension and had been previously
treated.

Since the literature states that hypertension is often
asymptomatic, patients may have hypertension and not be
aware of it. And even if hypertension is asymptomatic or
unknown to the hypertensive individual, it can be causing
cardiovascular problems.- This is why the Joint National
Committee for the Detection, Evaluation, and Treatment of
High Blood Pressure (1988, 1993) recommends screening for
hypertension and treatment when it is discovered. So, the
current study includes patients who were discovered to have
hypertension, regardless of the reason that brought them to
see the physician at the surveyed visit. This is a study
of what treatments are provided any time a patient presents
to a physician's office Visit with hypertension, even if
the finding of hypertension was unexpected or not the
reason for the visit. Finding and treating hypertension
before it becomes symptomatic and before serious morbidity
results is one of the main goals of the National Institutes
of Health JNC.

Some research documents a significant gender gap in
the treatment of cardiovascular disease. Of the 550,000
people who die each year of cardiovascular disease in the
United States, 250,000 are women, according to the U.S.
Public Health Service, National Center for Health

Statistics (1993). Men and women have different patterns

22

of morbidity and mortality. Men have consistently shown
higher rates of cardiovascular disease, although rates for
both men and women have declined in recent years. Black
men have higher death rates for cardiovascular disease than
white men, and black women have higher rates than white

women. See Figure 3:

Age-Adjusted Death Rates (per 100,000) for Diseases of Heart

 

 

 

 

 

 

 

an
Dvwmeuhbs
0 White Females
I Black Males
400 -------------------------------------------------------------------- 0 Black Females «
m ................... . --------------------------------------------------------------

Rate per 100,000

 

§

 

 

 

 

 

 

1950 1960 1970 1990
Year
Ageoadjusted death rates (per 111).”) for diseases of the heart. (From United States Public Health Service.

National Center for Health Statistics. Health. United Sara, ms. DHHS Pub. No. "85189-1232. Washin¢ton. D.C..
U.S. Government Printing Office. 1989.)

Eigu;§_1; Age-adjusted death rates (per 100,000) for

diseases of heart.

23

According to Douglas, (1993), although women utilize
the health care system more frequently than men, recent
studies indicate that they have less access than men to
certain significant diagnostic and therapeutic
interventions, especially those related to cardiovascular
disease and organ transplantation. Armitagge, and
Schneiderman, (1979) examined the extent of the workup
physicians ordered on patients with one of five complaints
common to both men and women: chest pain, back pain,
headache, dizziness, and fatigue. Across the board, men
received a more extensive workup than women for all
complaints studied. The researchers could not explain the
difference on the medical facts alone.

There are other patient characteristics that affect
their health. Black women have one and a half times the
death rate from heart disease as white women (National
Center for Health Statistics, 1988). Among the elderly,
blacks and the poor are less likely to possess private
insurance to supplement their Medicare health insurance
coverage (Rice, and McCall, 1985).

As a consequence of these disparities, those who live
below the poverty line, and a higher proportion of blacks
do, are less likely to receive preventive care services or
routine care for chronic disease (Kleinman, Gold, and

Makuc, 1981).

24

Shea, (1994) stated that the shortfall in
hypertension control reflects factors related to the
population as well as to the health care delivery system.
For example, educational level and social class are
powerful predictors of health outcome in many diseases and
undoubtedly work in multiple ways. Thus, one goal for
better control of hypertension in population groups would
be to improve the quality of primary care. According to
Shea, for hypertension control, active physician
involvement is needed in the process of promoting
compliance, which has long been recognized as a central
issue in achieving blood pressure control. To this end,
Shea suggests the use of periodic surveys of hypertensive
patients to monitor attainment of specific objectives for
hypertension control by different provider organizations.
This is a concept of "scorecard medicine" to assess if
physicians and healthcare groups are meeting certain
treatment criteria for hypertensive patients.

Hannan, Kilburn, Racz, Shields, and Chassin, (1994)
have studied this type of "scorecard medicine" for patients
who had coronary artery bypass graft surgery in New York,
and have found that it is helpful in improving outcomes
(decreased mortality and morbidity) for patients.

A large study done with Veteran's Administration (VA)
patients by Blue Cross and Blue Shield Association

(Shepard, 1995), however, showed that certain factors like

25

a greater number of laboratory tests, a greater time spent
in the waiting area and greater numbers of medications are
negatively associated with cost-effective control of
hypertension. This study reported that the factors
positively associated with cost-effective control of
hypertension are continuing care of the patient, greater
time spent with the patient during the visit, counseling
given to the patient, and no medications given at the first
visit. This study was longitudinal, and looked mainly at
future health-system costs such as hypertension-related
hospitalizations, and increased costs of clinic management
of these patients, including the costs of tests and
medications for these VA patients.

The studies cited above show that correct therapeutic
treatments can certainly lower patients' blood pressures
and decrease morbidity and mortality from cardiovascular
disease. Some of the studies show statistics about
treatment failures, and estimates of the numbers of
untreated hypertensives in the population. However, none
of these studies show what is being done in general primary
care settings to screen people for high blood pressure and
to treat those who are found to have hypertension. A link
needs to be made between what is known to be beneficial and
what is actually being done in routine patient care.

As stated in a special article, the National High

Blood Pressure Education Program Working Group report on

26

primary prevention of hypertension (1993): Despite its
undoubted benefits, treatment of established hypertension
is not a panacea. This approach requires an ongoing
commitment to the task of identifying and treating incident
cases. Each year approximately two million new
hypertensive patients are added to the pool of patients
requiring treatment for high blood pressure. The situation
is compounded by the fact that hypertensive patients may be
unaware of their condition. In addition, many sustain
vascular damage to their heart, brain, eyes or kidneys
before they come to the attention of a health care
provider, and many more are inadequately treated.
Provisional estimates suggest that approximately one third
(35 percent) of the 1988 through 1991 National Health And
Nutrition Examination Survey III (NHANES III) survey
participants with a systolic blood pressure (SBP) of 140 mm
Hg or more or a diastolic blood pressure (DBP) of 90 mm Hg
or more reported that they were unaware of a diagnosis of
hypertension. Even more troublesome, only forty-nine
percent of those with hypertension (those on
antihypertensive drug therapy or those with SBP greater
than or equal to 140 or DBP greater than or equal to 90)
were receiving antihypertensive medications, and only 21
percent of those being treated with antihypertensive
medications had a blood pressure less than 140/90 mm Hg.

These NHANES III findings are based on BP measurements

27

obtained at a single visit and may exaggerate the extent of
undertreatment in the community. Despite this limitation,
they suggest that large numbers of hypertensive patients
are unaware of their diagnosis and that many who are being
treated for hypertension probably have suboptimal blood
pressure control. Kligman, (1993), stated that without
treatment, 70 percent of hypertensive patients die of
congestive heart failure or coronary heart disease, 15
percent die of cerebral hemorrhage and 10 percent die of
uremia.

This indicates a need to look closely at the
physician's office visits of these hypertensive patients
and to determine what characteristics of patient or setting
that may be associated with an increase in likelihood of
adequate treatment. There is clearly a need to attempt
treatment of hypertension at any hypertensive patient's
entry into the healthcare system. It would greatly help if
it was understood what variables are significant in this
process. The current study might help to disclose factors
which can improve the likelihood of treatment and control
of hypertension.

The current study uses data collected from a large
national stratified random sample of patient visits to
physician offices. The data set, which included
physicians' office visits from many areas, metropolitan to

rural, and from many types of office settings, and from

28

many types of patients, provides a very broad sampling of
typical patients. The instructions given to the physicians
completing the surveys were thorough and explicit. The
broadness and representativeness of the survey help to make
it an accurate picture of what is done during actual
patient visits.

The two purposes of this study are: 1) To determine
if hypertensive patients are getting adequate treatment
when first found to have hypertension; and 2) To determine
what personal characteristics of patients or their
environments, such as insurance arrangements, might predict

the giving of adequate treatment or follow-up.

METHODS

I. THE DATA BASE SURVEY.

A. NATIONAL AMBULATORY MEDICAL CARE SURVEY.

This study uses 1992 data from the National Center for
Health Statistics (NCHS), as outlined above. The data
consist of samples of records of medical care given to
patients seen by office-based physicians. The basic
sampling unit is the physician-patient encounter or visit.
The physicians included in the study were from lists of
providers classified by the American Medical Association or
the American Osteopathic Association as those giving
office—based care. Excluded were federally-employed
physicians, and specialists in anesthesiology, pathology
and radiology. Also excluded were ambulatory encounters
made by phone, outside of physicians' office, and those
made in hospital or institutional settings.

In 1992 there were 34,606 patient records provided by
1558 doctors that participated in the survey. The 1992
NAMCS utilized a multistage probability design that
involved probability samples of primary sampling units--
geographical areas--, physician practices within the
primary sampling units (PSU) and patient visits within

practices. A PSU is a county, a group of adjacent

29

30

counties, or a standard metropolitan statistical area. The
first stage sample included 112 PSUs.

The second stage consisted of a probability sample of
practicing physicians selected from the master files
maintained by the American Medical Association and the
American Osteopathic Association. Within each PSU, all
eligible physicians were stratified by fifteen specialty
groups: general and family practice, osteopathy, internal
medicine, pediatrics, general surgery, obstetrics and
gynecology, orthopedic surgery, cardiovascular diseases,
dermatology, urological surgery, psychiatry, neurology,
ophthamology, otolaryngology, and all other specialties.

The final stage was the selection of patient visits
within the annual practices of sample physicians. This
involved two steps. First, the total physician sample was
divided into 52 random subsamples of approximately equal
size, and each subsample was randomly assigned to one of
the 52 weeks in the survey year. Second, a systematic
random sample of visits was selected by the physician
during the assigned week. The sampling rate varied for
this final step from a 100 percent sample for very small
practices to a 20 percent sample for very large practices,
as determined in a presurvey interview.

The first contact with the sample physician was
through a letter from the director of the National Center

for Health Statistics. Next, a representative phoned to

31

set up an appointment to discuss the survey and instruct
the doctor on how to complete the forms.

The actual data collection for the NAMCS was carried
out by the physician, aided by his or her office staff when
possible. Two data collection forms were employed by the
physician: The patient log and the patient record. The
patient log is used to sequentially list patients seen in
the physician's office during his assigned reporting week.
This list served as the sampling frame from which to sample
the visits for which data were to be recorded. A
perforation between the patient names and patient visit
characteristics permitted the physician to remove patient
names and protect confidentiality.

Based on the physician's estimate of the expected
number of office visits, each physician was assigned a
patient-sampling ratio. These ratios were designed so that
about 30 patient record forms were completed for each
physician office during the assigned reporting week.
Physicians expecting ten or fewer visits each day recorded
data for all of them, while those expecting more than ten
visits per day recorded data for every second, third, or
fifth visit based on the predetermined sampling interval.
These procedures minimized the data collection workload and
maintained approximately equal reporting levels among
sample physicians regardless of practice size. For

physicians assigned a patient sampling ratio, a random

32

start was provided on the first page of the log, so that
predesignated sample visits on each succeeding page of the
log provided a systematic random sample of patient visits
during the reporting period. For more information about

the data, and for definitions, see Appendix A.

Reliability and Validity Checks:

For each visit, information on many variables is
collected and coded, including patient characteristics,
conditions, tests, treatments given and disposition of each
visit. For data processing, completeness checks were made
by the field staff, then clerical checks were made on
receipt of the data for central processing. Item non-
response rates were less than five percent for most data
items. The only exceptions were visit duration (seven
percent), patient ethnicity and whether patient had been
seen before for the same condition (both eleven percent).
Information from the induction interviews and patient
record forms was keypunched with 100 percent verification
of the correctness of clerical data entry, according to
information provided by NCHS, and converted to electronic
data file. Extensive computer consistency and edit checks
were also performed. Data items still unanswered at this
point were imputed by randomly assigning a value from a
patient record form with similar characteristics;

imputations were based on physician specialty and broad

33

diagnostic categories. Any data field imputed in this
manner is indicated by a code in one of the columns on the
data file.

This survey has been administered over many prior
years, however, with extensive testing of the instrument
for reliability and validity since 1972.

The National Center for Health Statistics has supplied
information about the reliability and validity of the data.
For more detailed information on reliability, see

Appendix B.

Since this was an administrative data collection
procedure and the data were collected from regular office
visit records, the survey caused no physical harm to the
subjects. This was not an experimental study. The
treatment of the patients did not change due to the
physician's participation in this study. Concerning the
present researcher, the secondary analysis is based on data
published under Federal guidelines, which protect the
rights of human subjects and make it impossible to trace or
identify any of the individuals or practice groups
involved. The University Committee on Research Involving

Human Subjects has approved this project. See Appendix C.

 

Sc‘

34

B. DATA FOR THE CURRENT STUDY (SAMPLE).

The current study is concerned mainly with non-
pharmacological treatments given to hypertensive patients
at physician's office visits. From the 34,606 visit
records, only those who had been diagnosed as having
hypertension were selected for study, because the current
study relates to hypertension. Also, because the study
deals with initial treatments, only first visits for
hypertension at the office or clinic are included in the
data subset.

Among patients with hypertension, only those who had
not been seen before by the current physician were
selected. This selection was done in order to look at
initial treatments given to hypertensive patients.
Although some of these patients may have been diagnosed
with hypertension at some other time by a different
physician, they were new to the current physician at the
time of the visit, as defined by the survey questionnaire.
The reason for looking only at new patients is to rule out
the possibility that the prescribed treatments had already
been ordered for each patient at a prior visit to the
current physician. No medical histories of the patients
were supplied by this survey, and no information about
later visits to any physician was available because of the
way the study was constructed. This is a stratified random

sample of patient visits, an attempt at a cross—section of

35

all physician-office visits during 1992. The subset of
patient visits meeting all selection criteria involves a
sample of 573 subjects for study.

Finally, a frequency distribution showed that six
patients were children. These were excluded from the
subset, because the focus of the current study is on adult
hypertensive patients, not on pediatric patients. The
final subset total was 567 patients.

Several of the survey variables related to diagnoses.
These were in code numbers, using the International
Classification of Diseases and Assessments. The subset was
searched for codes which related to secondary, or
complicated hypertension. The intention was to eliminate
such complicated cases from the study. There were no cases
of secondary hypertension in the subset, however. The

final subset total remained at 567 patients.

II. VARIABLES.

For this study, the following variables have been
selected for examination: age, gender, race, ethnicity,
classification of obesity, hypercholesterolemia, smoking
status, chief complaint or reason for the visit, diagnosis,
type of insurance or payment arrangement. Then, the
disposition of each visit was noted: whether or not the
patient was referred to another physician, referred back to

a referring physician, referred to a specialist, told to

36

return at a specified time or date for follow—up, or
admitted to a hospital.

For purposes of examination, several variables were
recoded. The following section will describe all the
variables and explain any recodings.

Age: Ages of the patients were derived from exact
birthdates provided by the medical record data. For easier
statistical calculations, the age variable was split into
quartiles, with group one the youngest one-fourth, to group
four, the oldest one-fourth.

Gender: Female was coded (1), male was coded (2).

Reee: The four-category race variable (white, black,

Asian or American Indian) was subdivided and recoded into

two separate dichotomous variables: 1) black (coded 1) or
non-black (coded 0), and 2) Asian (coded 1) or not Asian
(coded 0). None of the patients in the study were American

Indian, so this category from the larger sample was
dropped. White was the default variable, if not black or
Asian. It was necessary to recode race into these
dichotomous variables because a three-category race
variable is meaningless for correlations and regression.

Eehnieiey; This was coded (1) for Hispanic origin, or
(2) for non-Hispanic.

Qbeeiey; This quality came from a part of the
original survey where the physician was asked to state if

the patient had one of several conditions at the time of

37

the visit. No definition of obesity was given to determine
if a patient qualified for this description or not. The
physician was simply asked to state yes or no if the
patient was obese. The physician was left to use his or
her own clinical judgement to determine this category. It
was coded (0) for no, (1) for yes.

Hypereholesterelemia: This was another condition as
obesity, above, in which the physician was asked to make a
clinical judgement and categorize each patient according to
cholesterol levels. Again, no exact definition or
laboratory levels were given for determining this category.
The responses were coded (0) for no, (1) for yes if the
patient had hypercholesterolemia.

Smgking stetes: The "Smoke cigarettes" variable was
recoded from a three-category variable (smoking, not
smoking or unknown smoking status) to a dichotomous
variable: Smoking (1), or Not smoking (0). If smoking
status was unknown, the patient was coded "System-missing"
for this variable. When smoking status was used for
statistical analysis, only those with known smoking status
were used.

Type gf ineurence or peymen; errengemengs: These were
separate dichotomous variables in the original study. Each
was coded (1) for yes, meaning the patient had a particular
type of insurance or payment arrangement or (0) for not

having it. The separate types were: Commercial insurance,

38

health maintenance organization, Medicaid, Medicare, and

patient or private payment. The physician was instructed
to check the source of payment to the best of his or her

knowledge, and to check all that applied. It is possible
to have more than one choice for each patient.

Therapeutig servicee: The survey respondent was
requested to check all appropriate boxes for the listed
services if provided or ordered during the office visit.
All forms of medication were to be excluded from this area
of the survey, as medications were noted in a different
section. This section was to include only counseling or
education services. The listed services which were
included as variables in this study are: Diet therapy,
exercise therapy, cholesterol-reduction therapy, weight—
reduction therapy and smoking cessation. Each of these
variables was coded (1) for yes, the therapy was given or
ordered, or (2) for no.

All the data analysis is done on a microcomputer using

SPSS software. Details of the analysis plan follow.

yerieblee yhigh meesure epprgpriateness gf eere
r e ' ns

Appropriate treatments for these hypertensive patients
fall into three categories:
1) Non-pharmacological: counseling or

education

 

 

— _ v ' .e..---—- H. u

39

2) Pharmacological: antihypertensive
medication

3) Follow-up or referral

According to the literature, especially the Joint
National Committee, Fifth Report (1993), non—
pharmacological counseling or education for lifestyle
changes (like diet or exercise) is always appropriate for
hypertensive patients. Pharmacological treatment is
appropriate if blood pressure cannot be controlled with
lifestyle changes. Since we do not know the details of the
lives of these patients, it is remotely possible that they
may not have any major lifestyle problems which contribute
to their hypertension. So any pharmacological treatment of
hypertension could be appropriate for a patient as a
beginning therapy. The JNC states that follow-up visits
for hypertensive patients are essential. The 1993 report
says, (the) "Frequency of visits and intensity of
evaluation and treatment should be the minimum sufficient
to achieve and maintain control of blood pressure and other
major risk factors and to contain progression of vascular
disease." Failure to follow-up with a hypertensive patient
would always be inappropriate.

Each of these three categories will be explained in
more detail:

1) The non-pharmacological antihypertensive

treatments given to these hypertensive patients include

4O

counseling or education: diet, exercise, cholesterol
reduction (if cholesterol levels are high), weight
reduction (if obese), smoking cessation (if smoking). The
content of the teaching interventions are not known from
the survey data. The survey instructions given to the
physicians reads, "Item 16, Therapeutic services: Check
all appropriate boxes for the listed services if provided
or ordered during the visit. All forms of medication
should be excluded as this will be collected in item 17."

All the non—pharmacological, or behavioral (counseling
or education) treatments are first counted. Since each
separate service is coded (1) for yes, (0) for no, the
total number of treatments can be tallied. The number of
services which each patient received is quantified. Next,
this variable is recoded for each patient into a new
variable called "behavior". This is coded (1) if eey
counseling or education is ordered or given, or (0) if none
is provided.

Thus, a measure of appropriateness is whether or not
at least gee of these treatments were given or prescribed
at the visit. Since we already know from the survey that
the patient has hypertension, we would expect, from being
familiar with the literature and the standards of care that
have been outlined over several decades by the Joint
National Committee, that some type of non-pharmacological

treatment is appropriate. The exact treatment given would

41

vary according to the specifics of each case. But the only
inappropriate care from the standpoint of the current study
would be no treatment.

The first outcome variable for this study then is
"Behavior", a dichotomous category of whether or not a non—
pharmacological behavioral intervention is provided to a
hypertensive patient. This variable will be used for
frequencies, correlations and a regression. This will
indicate how many the patients received this therapy, what
other factors the variable is related to, and what might
predict the likelihood of a patient to receive this
therapy.

2) In addition to these non-pharmacological
treatments, the giving or ordering of an antihypertensive
medication was also deemed appropriate because a physician
may determine during the visit that the patient has no
lifestyle risks that needed to be addressed or that the
hypertension is so severe that pharmacological treatment is
indicated immediately.

Because detailed pharmacological information was included
in the original survey, it was possible to determine which
patients were given any antihypertensive medications at the
surveyed visit (or given a prescription for any). In the
original survey, it was possible to indicate up to three
separate medications given or prescribed at the visit. The

medications were indicated in codes which showed their

42

function. Through a process of recoding, the
antihypertensive medications were quantified for each
patient into a new variable called "vascular medications".
This variable was coded "yes" (1) for being given or
prescribed any antihypertensive medication, or "no" (0) for
no antihypertensive medication.

3) A further consideration of appropriateness of
care must take account of the disposition of the patient
visit (follow-up or referral). Several possible
dispositions were given, including referral to another
physician, scheduling an appointment for return at a
specified date or a certain time interval or admission to a
hospital, or no follow—up. According to the literature,
any of the preceding follow-up measures constitute
appropriate care, except no follow-up. Each of the above
variables was coded (0) for no, or (1) for yes if the
disposition was planned or ordered. These were grouped and
recoded into a new variable called "follow-up". A measure
of no (0), or (1) yes, is used to code this new variable.
Any type of follow-up means that the patient remains in the
healthcare system and has the possibility of maintaining
control over blood pressure. No follow-up would leave the
patient at risk for uncontrolled hypertension and a

progression of vascular disease, according to the JNC.

43

Combining the three above treatment and follow-up
variables yields a single measure of appropriateness, the
second outcome variable for this study. The three
variables of non—pharmacological, medications and follow-up
are recoded into a new variable called "intervention".
Appropriate care is defined for this study as at least one
of the three treatment or follow—up interventions.
Inappropriate care is defined as: no treatment without
follow-up (0). All other care represents various degrees
of appropriateness lumped together into a single category
of "intervention" (1). This outcome variable will be used
in the final logistic regression in an attempt to determine
what factors might predict appropriate treatment for
patients with hypertension.

The definition of appropriate care employed here is
extremely liberal. It assumes that all the behavioral,
counseling or educational treatments are relevant and
effective. It assumes that the antihypertensive
medications are correct for each patient and are given in
the right dosages and times. It assumes that the follow—up
care will be therapeutic and proper. The survey does not
provide any details about the quality of any of these
measures, so that assuming they are correct is very

optimistic.

44

Potentiel Predictors of Appropriage Care:

The study looks at any differences in the above
treatments for different groups of patients. The groupings
are by (1) patient characteristics: age, gender, race,
ethnicity; (2) payment type: HMO, Medicare, Medicaid,
commercial or private insurance or patient payment.
Correlations will be computed to determine the effects of
individual variables on the dependent variables, behavioral
and overall treatment. Two logistic regression analyses
are performed to determine the effects of all of the
independent variables simultaneously on the dependent
variable. One analysis looks at the dependent variable
behavioral therapy, the second looks at the dependent
variable of any appropriate treatment or follow-up.
Logistic regression is used because the dependent variable
is dichotomous instead of continuous. The logistic model
is set up to ensure that the resulting probability will be
between zero and one.

The purposes, again, are to see first how often
appropriate treatments are given, and secondly, to see if
any of the above characteristics of the patient or the
setting influence the types of treatment given to

hypertensive patients.

RESULTS

A. CHARACTERISTICS OF THE DATA SUBSET (of 567 total
patients).

General gomments aboe; Lhe subset.

The data subset is roughly similar to a cross—section
of the United States population
with some exceptions. These hypertensive patients are
significantly older than is the general population. And,
of course, the sample included only adults. Females in the
study sample outnumbered males, 308 to 259. The proportion
of females to males in the subset is greater than in the
general population (51.2 percent female to 48.8 percent
male), but the sample represents only adults and includes
many older adults which are more likely to be women in the
population because women have a greater life expectancy.

So the study sample is actually very close to the general
population in gender distribution.

In the study sample, the percentage of blacks are
almost identical to that of the general population. Asians
and Hispanics, however, are underrepresented in the study
sample compared to the percentages in the general
population. There are no native American Indians in the
sample; again an underrepresentation.

The percentages in the subset of people who were

classified as obese or identified as smoking cigarettes are

45

46

slightly higher than in the general population. This is
probably an indication that these are risk factors for

hypertension.

Table l. EBEQflENQIES_QE_IHE_EAIIENI_QHABACIEBISIICSI

 

 

a. Patient Characteristics.
1. Age (in years). Mean. Median. £11.2eyl Range
61.5 64 14.48 18-91
Patient Characteristics Member Bergen;
Female 308 54.3
Male 259 45.0
3. Race.
White 486 85.7
Black 69 12.2
Asian or Indian 12 2.1
4. Ethnicity.
Non-Hispanic 545 96.1
Hispanic 22 3.9
5. Obesity.
Obese 109 19.0
Not obese 464 81.0
6. Smoking cigarettes.
Smoking cigarettes 58 10.1
Not smoking cigarettes 362 63.2
Smoking status unknown 153 26.7
7. Hypercholesterolemia.
Elevated cholesterol levels 68 11.9
Normal cholesterol levels 505 88.1

47

 

b. Type of Insurance or Payment Arrangements.

 

 

Commercial insurance 192 33.9

Health maintenance organization 93 16.4

Medicaid 45 7.9

Medicare 253 44.6

Patient, private payment 66 11.6
B. TREATMENT AND FOLLOW—UP FREQUENCIES. (The survey allowed multiple
responses.)

The following section lists the frequencies of the ordering of
behavioral treatments, antihypertensive medications, follow-up or

referral, and various combinations of these.

 

 

Table 2. T AT E OLL — P FRE E
a. Treatments Yes No
Number, percent Number, percent
1. Diet therapy. 87 (15.3) 480 (84.7)
2. Exercise therapy. 72 (12.7) 495 (87.3)
3. Cholesterol therapy. 33 (5.8) 534 (94.2)
4. Weight reduction therapy. 51 (9.0) 516 (91.0)
5. Smoking cessation. 14 (2.5) 553 (97.5)
(Of the 58 smokers) 14 (24) 44 (76)

6. Therapy for the obese.

No. of Treatments Number, percent
None 68 (63.6)
One 6 (1.1)
Two 16 (15.0)
Three 17 (15.9)

7. Therapy for those with elevated cholesterol levels.
For the 68 patients identified as having high cholesterol
levels, 33 (48.5 percent) were given therapy for reducing
cholesterol.

48

8. Total behavioral treatments (counseling, education) for
lowering blood pressure.

445 patients (78.5 percent) got no treatment.
122 patients (21.5 percent) got at least one treatment, as

follows:
No. of Treatments Number, percent
One 48 patients (8.5 percent)
Two 34 patients (6 percent)
Three 22 patients (3.9 percent)
Four 15 patients (2.6 percent)
Five 3 patients (0.5 percent)

9. Antihypertensive medications.
417 patients (73.5 percent) got no medications.
150 patients (26.5 percent) got at least one medication.

 

b. Disposition of the visit Yes No

(Multiple responses allowed) Number,percent Number,percent

 

Follow-up or referral.

119 patients (21 percent) got no follow—up, no referral.
448 patients (79 percent) got follow—up or referral.

 

 

1. Follow-up: (Given appointments
or told to return at
a specified time). 379 (66.8) 188 (33.2)
2. Referral:
To another physician: 22 (3.9)
Back to referring physician: 33 (5.8)
To another physician 16 (2.8)
To another specialist: 15 (2.6)
For hospital admission 15 (2.6)
c. Treatment combinations.
1. Total treatments for hypertension, including antihypertensive
medications.

348 patients (61.4 percent) got no treatments for

hypertension.
219 patients (38.6 percent) got at least one treatment, as
follows:
No. of Treatments Number, percent

One 122 (21.5)
Two 38 (6.7)
Three 35 (6.2)
Four 15 (2.6)
Five 8 (1.4)

Six 1 (0.2)

49

2. Any treatment or follow—up: (1) behavioral, (2) medications,
or (3) follow-up or referral.

82 patients (14.5 percent) got no treatment or follow—up.

485 patients (85.5 percent) got at least one, including:
292 patients (51.5 percent) got one.
151 patients (26.6 percent) got two.
42 patients (7.4 percent) got all three.

C. CORRELATIONS BETWEEN VARIABLES.

Correlations were computed to assess the associations
between various characteristics and treatments. Table 3
contains only the correlations between the outcome
variables of interest and characteristics of patients or
the payment arrangements. As can be seen, with the
exception of obesity, which correlated with getting a
"behavioral" treatment at 0.27, these correlations are
generally very small, though many are statistically
significant. The correlations among the potential
predictor variables (patient characteristics and insurance
arrangements) are not shown here, but they too are
uniformly low, most less than 0.3.

Table 3: BI RIATE C RRELATIONS MATRIX n = 567

 

 

Characteristic "Intervention" "Behavior"
1. "Intervention" .56*

2. "Behav" .56*
3. Age —.1* -.l9*
4. Gender 0.0 .04
5. Black/not .l* .06
6. Asian/not .04 -.02
7. Ethnicity —.04 -.05
8. Obesity .12* .27*
9. Hyperchol. .03 .16*
10. Smoker .04 .11*
11. Com.Insur. .02 .11*
12. H.M.O. -.05 —.01
13. Medicaid .04 -.04
14. Medicare -.12 -.2*
15. Priv.Pay. .14* .12*

* Indicates the correlation coefficients which are
statistically significant at p=0.05 or less.

50

D. LOGISTIC REGRESSION TO DETERMINE PREDICTORS OF

TREATMENT.

Two separate logistic regressions were performed, to
estimate the odds of predicting appropriate treatment on
the basis of the identified set of independent variables.
One regression was done for the outcome variable "Behav",
to predict the odds of a patient getting behavioral therapy
for the reduction of hypertension. The other regression
was done for the outcome variable "Intervention", to
predict the odds of a patient getting any of the
therapeutic treatments or follow-up.

For the first regression, only two of all the
independent variables were significant predictors of a
patient getting the counseling or education therapies to
control blood pressure (dependent variable: "Behav").

These two predictor variables were obesity and ethnicity.
Obese patients were 2.8 times more likely to get behavioral
therapeutic treatments than non-obese patients. This odds
ratio was significant to p=.000. Hispanics were less than
one—third as likely as non-Hispanics to get behavioral
treatments. This odds ratio was significant at p=.016 (see
Table 4a).

The second regression tested the likelihood of getting
"Intervention", any appropriate behavioral or

pharmacological treatment or follow-up. This regression

51

showed that there were some differences in odds ratios for

some of the variables in the sample. The strongest

difference was that Medicaid patients appeared 3.5 times
However,

more likely to get any of the interventions. none

of the differences was statistically significant. They

could have occurred by chance or sampling irregularity.

None of the independent variables,

patient

characteristics or payment arrangements showed any

significant effect on the dependent variable

"Intervention".

appropriate care or follow-up was not significantly

A patient's likelihood of getting

affected by the independent variables

Table 4a.

(see Table 4b).

Logistic Regression to Predict the Qdds of Patient or

Setting Characteristics to Influence Behavioral Therapeutic
Treatments ("Behav").

 

 

Variable B S.E. Wald Sig Odds

Age -.1304 .1363 .9156 .3386 .8777
Gender .1679 .2189 .5886 .4430 1.1829
Black/not .3711 .3153 1.3852 .2392 1.4494
Asian/not .3956 .7146 .3065 .5799 1.4853
Ethnicity -1.1733 .4887 5.7635 .0164 .3093
Obesity 1.0147 .2477 16.7816 .0000 2.7586
Comm. Insur. .3358 .2644 1.6135 .2040 1.3991
HMO Pay. -.3302 .3668 .8101 .3681 .7188
Medicaid -.0285 .4425 .0042 .9486 .9719
Medicare -.6332 .3417 3.4332 .0639 .5309
Patient Pay .4362 .3262 1.7882 .1811 1.5468

 

Table 4b.

ngistic Regtession to Predict the Odds of Patient or

Setti g characteristics to Influence Behavioral or
Ehermacglogical Treatment or Appropriate Follow-up

("Intetvention").

 

 

Variable B S.E. Wald Sig Odds
Age .1181 .1506 .6150 .4329 1.1254
Gender —.0418 .2445 .0293 .8641 .9590
Black/not .3184 .4311 .5453 .4602 1.3749
Asian/not -.1678 .8034 .0436 .8346 .8455
Ethnicity .0616 .6511 .0090 .9246 1.0636
Obesity .0448 .3217 .0194 .8893 1.0458
Commpay .0187 .2992 .0039 .9503 1.0188
HMOpay -.5291 .3765 .9751 .1599 .5891
Medicaid 1.2622 .7628 .7378 .0980 3.5332
Medicare —.2244 .3845 .3408 .5594 .7990
Patient pay 1.3463 .4445 .6071 .4359 1.4139

 

DISCUSSION

The frequency of giving therapy to reduce hypertension
in these hypertensive patients was remarkably low. This
clearly disregards the guidelines outlined by the Joint
National Committee on Detection, Evaluation and Treatment
of High Blood Pressure (1993). Perhaps even more
surprising is that a full 14.5 percent of these
hypertensive patients were given no treatment and no
follow-up. And 78.5 percent of the hypertensive patients
were given no counseling or educational therapy for
reducing their blood pressure. These findings are perhaps
the most significant of this study. Many patients clearly
are not being given appropriate treatment for hypertension.

When being surveyed for a National Institutes of
Health study, physicians would probably be more likely
rather than less likely to give the correct and appropriate
treatments. So these low treatment rates are difficult to
understand.

The percent of patients receiving at least one
appropriate treatment also was very low. Only 60.7 percent
of all these hypertensive patients got any treatments for
hypertension, including medications. This is a very
significant finding. And only 85.5 percent of these

patients got any treatment or follow—up. It is highly

53

54

significant that 21 percent of the patients newly
identified as having hypertension at an office visit with a
physician did not get any type follow—up for the

hypertension.

Three of the patient characteristics, age, race, and
obesity were significantly correlated with the dependent
variable, "Intervention". One payment arrangement, private
payment was significantly correlated with "Intervention".
It is a note-worthy finding of this study that most of
these patient characteristics or payment arrangements did
not significantly affect the appropriateness of their care.
The patients identified as having high cholesterol levels
or identified as smoking were not significantly likely to
be given the "Intervention", appropriate treatment or
follow-up. These are significant findings from this study.
This lack of appropriate treatment clearly violates the
acceptable standards of care, outlined in the literature.

The patient characteristics of age, obesity, elevated
cholesterol levels and smoking were significantly
correlated with getting the counseling or educational
therapy "Behav". Age was negatively correlated, meaning
that the oldest patients were less likely to get a
counseling or behavioral intervention. The other

characteristics were positively correlated. Specific

55

therapies were generally directed toward obvious patient
needs.

The obesity status of patients (comparing obese to
non—obese hypertensive patients) correlates only 0.12 with
getting or not getting epy therapeutic treatments to reduce
their health risks, as measured by the outcome variable
"Intervention". Obesity status is, however, somewhat more
strongly related to behavioral intervention.

Compared to non-smokers, hypertensive smokers were not
significantly more likely to be given smoking cessation
therapy. This treatment would always be appropriate. This
is especially surprising because these patients were
identified at the visit as being hypertensive and smoking,
a high—risk combination.

In the logistic regression predicting the behavioral
interventions alone, the only predictors that were
significant were obesity and ethnicity. Obesity greatly
increased the odds of getting care, and Hispanic ethnicity
greatly decreased these odds. It is not possible to
determine from this study what factors caused the
significant influence of these two variables. This is an
area that would warrant further study. There may be other
factors which are not clear from this survey which have
caused these results. For instance, the majority of the
Hispanic patients were from metropolitan regions in the

Southern and Western United States. These include some

56

very large cities such as Miami, Los Angeles, Houston and
Dallas. The offices may have been crowded or understaffed.
The offices may not have had the resources to offer the
correct therapeutic interventions. For the obese patients,
their risk factor of obesity was probably difficult to
ignore. The stereotypes and biases that obese people are
at risk for cardiovascular problems probably worked to get
therapy for these people. This raises another
consideration that people with equal, but less obvious
risks may have been overlooked.

In the logistic regression of "Intervention", no
characteristics of the patient, the setting or payment
arrangements were significantly associated with a patient's
getting appropriate treatment or follow-up. This is a
noteworthy finding. No bias or discrimination appears to

be influencing patient care.

Stu ur oses.

To address the study purposes, the first purpose, to
assess whether treatments were meeting established
standards of care was met. A clear picture of inadequate
treatment emerged. The outcome of the second purpose, to
discover what factors affected the ordering of treatments
or follow—up, was not as obvious. No definitive factors
were found which influenced general patterns of care, but

some very significant associations with the likelihood of

57

getting behavioral interventions appeared. Obese patients
were much more likely to get treatment, and Hispanics were
much less likely. Further study of this important issue

would be warranted.

IMPLICATIONS
W

This study had a significant finding of poor treatment
of hypertensive patients. Many hypertensive patients were
not given appropriate treatment as recommended by the
National Institutes of Health, the Joint National Committee
on Detection, Evaluation, and Treatment of High Blood
Pressure. The time and energy of very knowledgeable people
and important health agencies went into the preparation of
the JNV-V, and excellent—quality research was used as its
basis. Putting its recommendations into practice should be
a priority for all physicians and practitioners. It was
not clear from this study what the causes of failure were.
There is a need for further study of these issues. There
is a need to know what takes place during these physician
office visits and what rationale the physicians have for
not ordering the indicated treatments. There is a need to
know what conditions would increase the likelihood of their
ordering appropriate treatments.

Another implication of this study is to educate health
care providers to recognize random opportunities when
hypertensive people present themselves to the health care
system and to be prepared to offer appropriate
interventions, even if this is not the reason for the

visit.

58

59

This study was an interesting way to look at
treatments and follow-up for hypertension because the study
sample was part of a very large, general survey. The
practitioners who responded to the survey were not paying
particular attention to any certain type of patient. This
probably gave a more accurate look at typical care than if
the survey was only looking at care of hypertensive
patients.

The consequences of uncontrolled hypertension are
severe, leading to serious disease and death. Health care
providers should feel obligated to treat hypertension any
time they discover it. The current study seems to support
Stockwell's (1994) conclusion that "despite great technical
advances with new medications, the process of care
available in the community has not changed to any great

extent, and treatment itself remains highly ineffective."

Implications of the theoretical model.

Imogene King's model of open systems and theory of
goal attainment was an appropriate guide for the current
study, except that the model focuses on the process, but
not final outcomes.

For the open systems model, there were certainly
characteristics of each patient, the health-care setting
and the larger society which impacted the interactions

between physician and patient. Obesity and smoking status,

60

race and ethnicity were examples of individual patient
characteristics which seemed to affect care, at least
according to the frequency distributions and correlations.
Insurance and payment arrangements were characteristics of
the interpersonal systems which seems to have affected care
in some way. Then the larger social system considerations
of National health coverage through Medicare and Medicaid
as well as the work of many government agencies which went
into the study of hypertension treatments also had some
impact on care.

According to King's theory, the end product of the
interaction is the final attainment of the mutual goal
between patient and provider. In this study, the goal
would be control of blood pressure. This study does not
provide any follow—up data or any mechanism to determine if
this goal was met. However, we can use the model to the
point of the interaction between patient and provider. We
must assume that if a correct treatment was never given or
ordered for a patient, that it would be most difficult for
that patient to achieve the goal of control of
hypertension. And we must assume that if follow-up is not
ordered or provided, it would not be possible to assess
goal attainment. Perhaps a further study, a cohort study,
could provide the necessary information about goal
attainment. This limitation of the current study should

not preclude using the King theoretical model to the point

61

of ordering or providing appropriate treatment and follow-
up. King's theory should certainly be used to influence
the perceptions and judgements of caregivers to be more
aware of hypertensive patients in their caseloads.
Providers should be routinely screening for hypertension
and treating it when found.

Nothing about the interaction is known from the
survey. The Health Belief Model (Rosenstock, 1966, 1974)
could be applied to the interaction, to involve the patient
in the care process and increase the likelihood that
treatments would be followed. The important concepts of
the Health Belief Model are that if a person believes that
he or she is susceptible to a disease, that the disease is
serious, and that the perceived benefits of action outweigh
the perceived barriers or costs (such things as time,
money, effort), the person will take action to avoid the

disease.

Impiigations for practice.

This section is intended for nurse practitioners, as
this is the focus of the study. However, the implications
would also apply to physicians or other practitioners who
see patients in office-based settings.

1. This study should help providers be aware that any
patient may have hypertension. The patients involved in

the study came to the physicians' offices for a variety of

62

reasons, not necessarily problems associated with
hypertension. During these first visits to a physician,
these patients were discovered to have hypertension.

2. The provider should be ready with a pre-planned
screening mechanism to check all patients for hypertension,
regardless of their presenting complaints or reason for the
visit. Taking blood pressure readings on all patients is
very quick and easy, and need not be done by the physician.
it should be an automatic part of every office visit and
can be done by a nurse or medical assistant. The JNC
recommends routine blood pressure screenings.

3. The provider should be ready with an automatic protocol
to follow for any patient discovered to have hypertension.
This should include appropriate treatment and follow-up
even if hypertension was not the reason for the visit. If
the provider does not have extra time to initiate the
treatments at the initial visit, he or she should order
them and should certainly have the patient come back for
follow-up. Anything less should be considered malpractice,
given the risks to the patient of uncontrolled
hypertension.

4. The provider should communicate to each patient the
serious effects of uncontrolled hypertension. And the
provider should educate each hypertensive patient about the

condition and its chronic nature. This should increase

63

each patient's likelihood of expecting and seeking
adequate, appropriate treatment and follow-up.

5. This study should help reduce biases that only certain
types of patients are referred for appropriate care. The
final regression analysis shows that no patient
characteristics of race, ethnicity, gender or age affected
their getting appropriate care, and no type of insurance or
payment arrangement affected appropriate care.

6. Although the final regression showed no biases, there
was a very significant finding that Hispanics were far less
likely to receive appropriate behavioral counseling to
reduce hypertension. Although it may not have been overt
bias, it shows a disturbingly significant problem. Perhaps
the healthcare settings which Hispanic people have access
to are somehow deficient. Regardless, this study shows
that Hispanic people are at increased risk of not having
their hypertension adequately treated. This places them at
higher risk for cardiovascular disease and death, a
significant national public health problem. Although the
numbers of Hispanic people in this study were small, the
regression analysis shows that this effect is statistically
significant and could not have occurred by chance or
sampling abnormalities. This problem should certainly be
addressed by further study and definitive action.

7. Obese patients were three times as likely to get

behavioral treatments. This triggers two considerations.

64

First, obese patients are the most likely to get
appropriate treatment. Obesity is a very obvious risk
factor for cardiovascular disease. It fits providers'
stereotypes of people with high risks. Obesity is easy to
recognize and difficult to ignore. It induces physician
action. The second consideration is that hypertensive
patients who have risk factors which are not so obvious may
be overlooked or forgotten. They may have risks which are
equally high, but not so obvious —- perhaps a sedentary
lifestyle or a diet high in saturated fats. The
implication is that providers should continue to treat
those with obvious risks, but they should also evaluate
hypertensive patients for any risk factors, and give
appropriate treatment as needed.

8. The main limitation of appropriate care seems to be a
provider's willingness to give or order correct treatments
and follow-up. Providers should give careful thought to
providing appropriate antihypertensive care to all of their
patients in general. Each provider could establish a
personal "scorecard” method of increasing his or her
percentage of appropriate treatments and follow-up for
hypertensive patients. Perhaps provider characteristics
which affect appropriate treatment should be studied. It
would be helpful to know what causes a provider to order
treatments or not. Any provider deficits could be

addressed through training or incentives.

65

9. Smoking status was listed as "unknown" for greater
than 25 percent of these hypertensive patients. Providers
should certainly be asking their hypertensive patients if
they smoke, as it is a major risk factor for cardiovascular
disease. It is difficult for providers to order the
appropriate smoking cessation therapy if smoking status is
unknown.

10. Nurse practitioners are trained to look more
wholistically at their patients than physicians are trained
to do. If the patients in this survey had been viewed more
wholistically, perhaps their likelihood of getting
appropriate treatment would have been greater. In general,
physicians are trained to treat the presenting problems
only. Nurse practitioners are trained to look at the whole
person, screen for likely risks at each age group, and

offer more education and counseling types of interventions.

Limitatigns of the study.

One major limitation of the National Ambulatory
Medical Care Survey is that it only provides a one—visit
look at each patient. And the information about each
patient is limited. Patients were not followed over time.
The survey did not give physical data about each patient,
blood pressure readings, history or current health status.
The survey did not measure the adequacy of treatments or

report any follow-up or eventual control of blood

66

pressures. It did, however, allow the quantification of
treatments and the ordering of follow-up.

Another major limitation is that the quality or
content of any counseling or educational programs was not
stated. The effectiveness of these therapies was unknown.
The current study simply assumed that all these therapies
were correct and appropriate, although they may not have
been. The study only measured whether or not these things
were ordered or given. And when antihypertensive
medications were ordered, the current study assumed that
they, also, were correct and appropriate, though they may
not have been. The codings and allowances were extremely
generous, always assuming that correct things were being
done. This may have distorted the study findings, counting
care as appropriate when it was not.

Another major limitation of the current study is that
the results only apply to the providers and patients who
were surveyed in 1992. The verifications, data entry and
computer work took almost three years. The results were
not released until October, 1995. The results may no
longer be an accurate representation of current practice.
Although the recommendations for lifestyle modifications
had been made by the Joint National Committee on Detection,
Evaluation and Treatment of High Blood Pressure were made
over twenty years ago, they have been receiving more

attention in recent years. Patients and physicians in 1992

67

may not have been as aware of appropriate treatment of
hypertension as they are currently. The results of the
original survey and the current study should not be
automatically extended to the current general population.

A limitation of both the original survey and the
current study is that most variables are dichotomous. For
both of the outcome variables, various treatments and
follow-ups were coded, counted and recoded into dichotomous
variables of yes or no for getting appropriate care and, or
follow-up. It would perhaps have been more accurate to
place these outcome variables on scales of appropriateness
if more information about the patients and the treatments
were available. The dichotomous coding certainly provides

only a rough measure of appropriate treatment.

REFERENCES

Alderman, M. (1994). Non-pharmacological treatment of
hypertension. Lancet, 344, 307-311.

American Heart Association, (1993). Controlling your

risk fectors for heart attack. American Heart Association
National Center: Dallas, TX.

Armitagge, K., Schneiderman, L., & Bass, R. (1979).
Response of physicians to medical complaints in men and

women. Journal of the Americen Medical Association, 241,
2186—2187.

Blake, G. (1994). Primary hypertension: The role of

individualized therapy. American Family Physician, 50,
138—146.

Collins, R., Petro, R., MacMahon, S, (1990). Blood
pressure, stroke, and coronary heart disease, part 2:
Short-term reductions in blood pressure: Overview of
randomised drug trials in their epidemiological context.

Lancet, 335, 827-838.

Creditor, M., & Schoenberger, J., (1972). Control of
hypertension: An unmet challenge. Illinois Medical
Journal 140, 521.

 

Cutler, J., Follmann, D., Elliott, P., & Suh, I.
(1991). An overview of randomized trials of sodium

reduction and blood pressure. Hypertension, 17 (suppl. 1),
27-33.

Devine, E. & Reifschneider, E. (1995). A meta—
analysis of the effects of psychoeducational care in adults
with hypertension. Nursing Research, 44, 237-245.

Douglas, P. (Ed.). (1993). Cardiovascglet health and
dieease in women. Philadelphia: W. B. Saunders Company.

Editors, American Journal of Public Health, (1994).

Hypertension Control, 1994. Amerigan Jogrpai gf Pubiig
Heelth, 84, 1725-1727.

Hamet, P., Lambert, J., et al. (1991). Interactions
among calcium, sodium, and alcohol intake as determinants
of blood pressure. Hyperteneion, 17 (suppl. 1), 150-154.

Hannon, E., Kilburn, H., Racz, M., Shields, E., &
Chassin, M. (1994). Improving the outcomes of coronary

68

69

artery bypass surgery in New York State. Qournal of the
Ametigep Megicel Association, 271, 761- 766.

Hickey, J., Ouimette, R., & Venegoni, S. (1996).
Advangeg practice nursing. New York: Lippincott.

Intersalt Cooperative Research Group. (1988).
Intersalt: An international study of electrolyte excretion
and blood pressure: Results for 24 hour urinary sodium and

potassium excretion. British Medical Journal, 297, 319—
328.

Joint National Committee on Detection, Evaluation, and
Treatment of High Blood Pressure (1988). 1988 report.

Arghives of Internal Medicine, 148, 1023-1038.

Joint National Committee on Detection, Evaluation, and
Treatment of High Blood Pressure (1993). The fifth report.

Archives of Internal Medicine, 153, 154-183.

King, I. (1971). Toward a theory for nursing:
Genera; concepts of human behavior. New York: John Wiley

& Sons, Inc.

King, I. (1981). A theory for nursing: Systems,
concepts, process. New York: Wiley.

Kleinbaum, D. (1994). Statistics in the health

eciengee: Logistic regression, a self-learning text. New
York: Springer-Verlag.

Kleinman, J., Gold, M., & Makuc, D. (1981). Use of
ambulatory medical care by the poor: Another look at

equity. Medical care 19, 1011-1128.

Kligman, E. (1993). Hypertension. In: Mengel, M., &

Schweibert, L., (Eds.) Ambulatory medicine: The primary
gere gf families. Norwalk, CT: Appleton & Lange.

Law, M., Frost, C., & Wald, N. (1991). By how much
does dietary salt reduction lower blood pressure?: 3:
Analysis of data from trials of salt reduction. British
Medigei Jgurnal, 302, 819-824.

MacMahon, S., Petro, R., & Cutler, J., (1990). Blood

pressure, stroke, and coronary heart disease. Part 1,
prolonged differences in blood pressure: Prospective
observational studies corrected from the regression bias.

Lepcet, 335, 765-774.

70

National Ambulatory Medical Care Survey (1992).
National Center for Health Statistics, U.S. Public Health
Service, Department of Health and Human Services.
Washington, D.C.

National Center for Health Statistics. (1995).
Health, United States. U.S. Public Health Service, DHHS
Pub. Washington, D.C.: U.S. Government Printing Office.

National High Blood Pressure Education Program Working
Group, Report on primary prevention of hypertension (1993).

Archives gf Internal Medicine, 153, 186-203.

Neaton, J., & Wentworth, D. for the Multiple Risk
Factor Intervention Trial Research Group. Serum
cholesterol, blood pressure, cigarette smoking, and death
from coronary heart disease: Overall findings and
differences by age for 316,099 white men. Atghives of
Internal Medicine 152 56-64.

 

Reisin, E, Abel, R., Modan, M., Silverberg, D,
Eliahou, H., & Modan, B. (1978). Effect of weight loss
without salt restriction on the reduction of blood pressure
in overweight hypertensive patients. New Englahd Journal
of Medicine 298 1—6.

 

Rice, T., & McCall, N. (1985). The extent of
ownership and the characteristics of Medicare:
Supplemental policies. Ingpiry, 22, 188-200.

Roccella, E. (1993). National high blood pressure
education program working group report on primary
prevention of hypertension. Archives of Internal Medicine.
153, 186-203.

Sacks, F. (1989). Dietary fats and blood pressure: A
critical review of the evidence. Nutrition Review, 47,
291-300.

Shea, S. (1994). Hypertension control. American

Journal of Puhlic Health, 84, 1725—1727.

Shepard, D., Stason, W., Perry, H., Carmen, B &

Nagurney, J. (1995). Multivariate cost-effectiveness
analysis: An application to optimizing ambulatory care for
hypertension. Ingpity, 32, 320-331.

Stockwell, D., Madhavan, S., Cohen, H., Gibson, G., &
Alderman, M. (1994). The determinants of hypertension

awareness, treatment and control in an insured population.

Ametigan Joprhel gf Pphiic Health, 84, 1768-1774.

71

Verbrugge, L. (1989). The twain meet: Empirical
explanations of sex differences in health and mortality.

Jogtnal of Health and Social Behavior, 30, 282—304.

Working Group on Physician Behaviors to Reduce Smoking
Among Hypertensive Patients. (1984). Physician's Suide;
How tg help your hypertensiye patients stop smoking. U.S.
Department of Health and Human Services, Public Health
Service, National Institutes of Health. Washington, D.C.:
U.S. Government Printing Office.

APPENDIX A

De ' ' i n of e ain terms used in this s ud :

Office(s): Premises that the physician identifies as
locations for his or her ambulatory practice.
Responsibility over time for patient care and professional
services rendered there generally resides with the
individual physician rather than any institution.

Visit: A direct, personal exchange between ambulatory
patient and the physician (or members of his or her staff)
for the purpose of seeking care and rendering health
services.

Ambulatory patient: An individual presenting for
personal health services, neither bedridden nor currently
admitted to any health care institution on the premises.

Patients (in-scope): All patients seen by the
physician or member of physician's staff in physician's
Office(s).

Physician (in-scope): All duly licensed doctors of
medicine and doctors of osteopathy currently in practice
who devote most of their practice to caring for ambulatory

patients at an office location.

The following section gives a brief description of
each variable in the data survey:

72

73

Date of visit: The actual date of the patient visit
to physician's office.

Age: Age of the patient in years.

Sex: Gender of the patient.

Race: Race of the patient, Black, white,
Asian/Pacific Islander or American Indian/Eskimo/Aleut.

Ethnicity: Hispanic or not Hispanic.

Expected source of payment: Health maintenance or
other prepaid medical plan, Medicare, Medicaid, other
government payment, private or commercial payment, patient
payment, no charge, other or unknown payment.

Referred: Was patient referred by another physician-—
yes or no.

Injury: Was the visit related to an injury——yes or
no.

Smoking: Does the patient smoke cigarettes--yes, no
or unknown.

Visit reason: Patient reasons for the visit. Three
separate reasons could be coded for each visit. The codes
are from a long list of possible reasons.

Diagnosis: The physician's diagnosis of the patient's
condition, coded according to established medical coding
from the International Classification of Diseases, 9th
Revision, Clinical Modification. Up to three separate

diagnoses could be coded for each patient visit. A

74

separate variable "Does this patient currently have
hypertension?" was included in the data collection process,
independent of the official diagnosis. All patients who
were identified as having hypertension were included in
this study, even if their first diagnosis was something
other than hypertension.

Seen before: A coding of whether the patient had been
seen before by the current physician--yes or no.

If yes: If the patient had been seen before by the
current physician, was it for the same condition--yes or
no. .

Several codings applied to conditions that the patient
currently had at the time of the visit: Depression,
hypertension, hypercholesterolemia or obesity--yes or no
for each of these. These coding of these conditions was
separate from and/or in addition to the diagnoses
previously coded.

Several variables concerned ambulatory surgical
procedures, which were not relevant to the current study.

Several codings pertained to diagnostic screenings
which were done during the office visit: Blood pressure
taken, urinalysis, EKG at rest or during exercise,
mammogram, chest x-ray, other radiological study, allergy
testing, spirometry, Pap test, strep throat test, HIV
serology, cholesterol measurement, hearing test, visual

test, mental status exam. These were coded either yes or

75

no for these tests. The instructions were to check
appropriate boxes for all services ordered or provided
during this visit for the purpose of screening or
diagnosis. The instructions went on to say that during a
visit for a complete physical examination, several of the
services may be ordered or provided. Each service should
be marked. The physician was instructed to check "other"
for any diagnostic or screening service ordered or provided
but not included in the list, then the physician was
instructed to describe the service on the line provided.
The possible choices listed were: 1) None, 2) Blood
pressure, 3) Urinalysis, 4) EKG - Resting, 5) EKG —
Exercise, 6) Mammogram, 7) Chest x—ray, 8) Other radiology,
9) Allergy testing, 10) Spirometry, 11) Pap test, 12) Strep
throat test, 13) HIV serology, 14) Cholesterol measure.
Many therapeutic services were coded if done or not
done. These were therapy for: diet, exercise, cholesterol
reduction, weight reduction, drug abuse, alcohol abuse,
smoking cessation, family or social, growth or development,
family planning, other counseling, psychotherapy,
corrective lenses, hearing aid, physical therapy, other
therapy or no therapy. These were coded yes or no. The
physician was instructed to check all appropriate boxes for
the listed services if provided or ordered during the
Visit. The physician was instructed that all forms of

medication should be excluded as they are included in a

76

different section. It is very clear from the instructions
that any and all non-pharmacological treatments be
included, whether they were actually done at this visit or
only ordered at the visit and done at a different time.
This is important to the current study -— this means if
treatments were not marked in this area, they were not done
and were not ordered to be done. The possible choices for
therapeutic services were listed by type. The first choice
is for no therapy. Next, a list of counseling or
educational services in focused areas was given: 1) Diet,
2) Exercise, 3) Cholesterol reduction, 4) Weight reduction,
5) Drug abuse, 6) Alcohol abuse, 7) Smoking cessation, 8)
Family/social, 9) Growth/development, 10) Family planning,
11) Other counseling. Then a list of other therapeutic
services was given: 1) psychotherapy, 2) Corrective
lenses, 3) Hearing aid, 4) Physiotherapy or 5) Other
therapy. If "Other therapy" was checked, the physician was
instructed to specify the type.

Many codings were done for medications, if ordered
during the visit. These were not relevant to the current
study.

Disposition of the visit was coded for the following
choices: No follow-up planned, return at a specified date
or interval, return if needed, telephone follow-up,
:referred to other physician, returned to referring

physician, admit to hospital or other disposition. These

77

were coded yes or no. The instructions to the physician in
this area were: "Eight codes are provided to describe the
physician's disposition of the case on this visit. The
physician should mark as many categories as apply." Then
the choices were: 1) No follow—up planned: No return visit
or telephone contact is scheduled for the patient's problem
on this visit; 2) Return at a specified time: The patient
is told to schedule an appointment or is instructed to
return at a particular time.

3) Return if needed, PRN: No future appointment is made,
but the patient is instructed to make an appointment with
the physician if the patient considers it necessary (PRN,
pro re nata, as necessary, as the occasion arises.) 4)
Telephone follow-up planned: The patient is instructed to
telephone the physician on a particular day (to report on
this progress, or if the need arises, if he has any trouble
or wishes further consultation). 5) Referred to other
physician: The patient is instructed to consult of seek
care from another physician. The patient may or may not
return to this physician at a later date. 6) Return to
referring physician: The patient was referred to this
physician and is now instructed to consult again with the
physician who referred him. 7) Admit to hospital: Patient
is instructed that further care or treatment will be
provided in a hospital. No further office visits are

expected prior to that admission. 8) Other: Any other

78

disposition of the case not included in the above
categories. The physician was instructed to specify what
other disposition was given.

The duration of each visit was given in minutes of
time spent actually with the physician. The physician was
instructed to not include the time the patient spent
waiting to see the doctor and not to include the time the
patient spent receiving care from someone other than the
doctor.

Patient weight was given for each case, as an integer,
to produce national estimates from sample estimates.

Geographic region of the location of the physician's
practice was coded for northeast, midwest, south and west
U.S., and whether the location was in a metropolitan area
or not.

Then, data was given about each physician: type of
specialty, type of doctor (M.D.

or D.C.).

APPENDIX B
Marginel Data.

For visits to aggregated specialties in 1992, any
estimate less than 676,000 has a relative standard error
that is greater than 30 percent. Such an estimate is
considered unreliable accordint to the standards of the
National Center for Health Statistics. For individual
specialties the 30 percent relative standard error level
varies. Separate listings of all the data categories, with
numbers and percents of visits, was supplied with the NAMCS
data. The instructions with this listing state that if any
figure did not meet standards of reliability or precision,
a star (*) would be placed with the figure. There were no
stars in this listing, meaning that all the data used for

this study met standards of reliability and precision.

79

APPENDIX C

gnivereity Committee on Reeearch Involving Human Subjects

Wall

80

"I11111111111111.1111“