close

Вход

Забыли?

вход по аккаунту

?

The nature of u.s. rheumatology practice 1977

код для вставкиСкачать
1177
MANPOWER SERIES
THE NATURE OF
U.S. RHEUMATOLOGY PRACTICE, 1977
WALLACE V. EPSTEIN and CURTIS J. HENKE
A description of academic and community rheumatologist practice activities in the United States has
been assembled from 826 responses to a 1W7 survey of
American Rheumatism Association members. Reported
work hours and patient load are similar to other medical
subspecialists, and three-fourths of the clinical time is
devoted to persons with predominantly rheumatic disease problems. Significant differences in activities occur
between academic and community practices and among
rheumatologists with different training backgrounds.
Few characteristics vary significantly across geographic
regions containing divergent rheumatologist-population
ratios. Descriptive productivity measures such as these
are necessary for planning; they provide part of the empirical analysis required to give reality to manpower policies.
The discussion of physician supply has recently
shifted from concern about an overall shortage to predictions of a coming oversupply of doctors in general,
From the Social Sciences Research Unit, Multipurpose Arthritis Center and the Robert Wood Johnson Clinical Scholars Program, University of California, San Francisco, and the Committee on
Rheumatologic Practice of the American Rheumatism Association.
Preparation of this report was assisted by Multipurpose Arthritis Center Grant No. AM-20684 from the NIAMDD and the Robert Wood Johnson Foundation, Princeton, New Jersey. The opinions,
conclusions, and proposals in the text are those of the authors and do
not necessarily represent the views of the Robert Wood Johnson
Foundation, the American Rheumatism Association, or its Committee on Rheumatologic Practice.
Wallace V. Epstein, MD: Professor of Medicine, University
of California San Francisco; Curtis J. Henke, MA, PhD C a n d Assistant Research Economist, University of California San Francisco.
Address reprint requests to Wallace V. Epstein, MD, Department of Medicine, University of California, 350 Parnassus X47, San
Francisco, CA 941 17.
Submitted for publication July 30, 1980; accepted in revised
form January 13, 1981.
Arthritis and Rheumatism, Vol. 24, No. 9 (September 1981)
an overabundance of specialists relative to primary care
physicians, and the general geographic maldistribution
of physicians (14). Much of this debate has been conducted in generalities and with only limited information. We believe that this general type of discussion is
inappropriate and potentially misleading. For each specialty group, the assessments of oversupply and/or inappropriate distribution should be based on the professional activities of the physicians, the value of care by
this group compared to other physicians, the actual geographic distribution of members of the specialty, and
the ways in which local practice conditions influence
what they do. If the claims of specialty and geographic
maldistribution are valid, as well they might be, then an
analysis of such data should make this evident.
The work of assembling an empirical description
for the specialty of rheumatology has already begun; yet
the published accounts to date are mostly incomplete or
of limited scope (5-9). Only one systematic sampling
has been presented to date, and it contains less detail
than is desirable for a consideration of this single specialty (10). We present here univariate and bivariate
analyses of survey results that describe the practice
characteristics of American Rheumatism Association
(ARA) members with declared specialties of rheumatology, internal medicine, and pediatrics. We consider this
group to be the best available population for studying
practice characteristics and location of these physicians;
the lack of competing national rheumatology societies
suggests that most physicians with a serious interest in
rheumatology would be ARA members, and we believe
that the membership requirements screen out most physicians without a serious interest in rheumatology. We
have already reported the training characteristics of this
group of rheumatologists (1 1). In this report, we first
EPSTEIN AND HENKE
1178
present their geographic distribution and practice arrangements; then we describe their practice activities
and the diagnoses of the rheumatologic patients they
treat. We investigate differences among groups of rheumatologists according to their practice arrangement,
training background, and geographic location.
METHODS
In early 1977, the national office of the ARA
mailed 1700 questionnaires to all members who listed
their primary specialty in the ARA records as rheumatology, internal medicine, or pediatrics. Members with
primary specialties of orthopedics, physiatry, general
practice, and “other” were not surveyed. Respondents
were assured anonymity, and no identification of their
training program was required. The questionnaires
yielded 949 replies, a 56% response rate. Eight hundred
twenty-six of these were from rheumatologists actively
practicing in the United States. The responding physicians answered questions concerning their training, current practice, and the distribution of their professional
efforts. Because the survey was anonymous, no search
was made for nonresponders, nor was external validation of responses attempted. The response rate varied
for individual questions, and the means and percentages
reported in this paper exclude the missing responses.
Three practice arrangements (designated fulltime academic, private practice with academic affiliation, and private practice without academic affiliation)
are used most often in the analysis of the responses.
Small numbers of respondents who listed other equivalent designations have been distributed among these
three categories. Some practice characteristics are also
contrasted for private practice organization: rheumatologists in solo or partnership practice versus those prac-
Table 1. U.S. rheumatologists, January 1977: respondents, population, and population per capita, by census division and practice setting
Census division
Setting
h’ew
Middle
England Atlantic
~
East
Korth
Central
West
North
Central
South
Atlantic
East
South
Atlantic
West
South
Central
Mountain
Pacific
U.S.
.~
~
Rheumatologist respondents
Internal medicine
Academic
Salaried
Community
Total
Pediatrics
Academic
Salaried
Community
Total
Total responses
22
7
53
82
50
7
128
185
30
4
75
109
0
0
0
0
82
2
0
5
7
I92
1
0
1
2
Ill
48
14
79
141
93
29
275
397
1
0
0
1
I42
3
1
5
9
406
0.258
0.081
0.753
1.092
6
0
29
35
38
4
105
147
216
36
547
799
2
2
37
4
1
2
7
154
18
1
8
27
826
20
10
64
94
17
6
47
70
59
20
175
254
412
I20
1007
1539
2
0
0
2
72
6
0
55
4
0
0
4
98
265
28
3
15
46
I585
0.161
0.029
0.2 12
0.402
0.1 12
0.047
0.300
0.459
0.193
0.06 1
0.479
0.733
0.226
0.070
0.626
0.922
0.205
0.057
0.476
0.738
18
0
35
53
35
9
68
112
16
28
6
4
38
48
2
0
2
55
3
0
0
3
115
1
0
0
1
29
3
0
0
3
51
47
13
134
194
24
2
72
98
83
22
132
237
21
4
29
54
6
I
3
3
1
1
1
0
0
5
1
204
2
0
I
3
101
242
0.130
0.034
0.335
0.499
0.155
0.012
0.435
0.601
0.253
0.068
0.392
0.713
0
11
1
0
0
Rheumatologist population *
Internal medicine
Academic
Salaried
Community
Total
Pediatrics
Academic
Salaried
Community
Total
Total population
10
5
I1
Rheumarologists per 1OO.OOO people
Academic
Salaried
Community
Total
0.401
0.115
0.647
1.163
* Since the mailing for this study occurred between data collections for published directories, the information about practice settings in the population was incomplete. Hence, the population of practicing U.S. rheumatologists was estimated by: 1) counting from the 197475 ARA membership
directory the specialty, practice setting, and geographic location of the 1479 listed U.S. physicians with specialties of rheumatology, internal medicine, or pediatric$; 2) tallying from the responses those who had finished training or had begun an initial practice after the publication of the directory and adding these 106 to the directory count; and 3) finding in the responses rheumatologists who were in a practice at the time of the directory
but had moved prior to the survey, and adjusting the distribution for these 5 I movers.
NATURE OF U.S. RHEUMATOLOGY PRACTICE
ticing in groups (single specialty or multispecialty).
Throughout this paper, the designation “solo” means
“solo or partnership.” In addition, “community rheumatologist” is used synonymously with “private practice rheumatologist.” To classify rheumatologists by
training orientation (most research, average, most clinical), we ranked the respondents by the ratio of reported
share of training time in research to time in outpatient
clinics and inpatient consultations, and then we divided
the group into thirds. The geographic categorization
that we use is the standard Department of Commerce
census divisions.*
Practice characteristics are stratified by practice
setting, training duration and orientation, and practice
location to test the hypotheses that significant differences exist in practice characteristics related to these
variables. Differences in mean values between two categories are tested using a standard two-sample r-test. For
differences across three or more categories, we use an Ftest in an analysis of variance framework. In each case,
the reported P value is for rejection of the hypothesis of
equality of means in all categories. When testing for independence in distribution among categorical variables,
we use a standard contingency table and chi-square test.
The census divisions and states included are as follows:
New England-ME, NH, VT, RI, CT, MA; Middle Atlantic-NY,
NJ, PA; East North Central-OH, IN, IL, MI, WI; West North Central-MN, IA, MO, ND, SD, NE, KS; South Atlantic-DE, MD, DC,
VA, WV, NC, SC, GA, FL; East South Central-KY, TN, AL, MS;
West South Central-AR, LA, OK, TX; Mountain- MT, ID, WY,
CO, NM, AZ, UT, NV; Pacific-WA, OR, CA, AK, HI.
1179
RESULTS
Table 1 presents the number of responders to the
questionnaire and the population of U.S. rheumatologists. From the 826 completed responses from actively
practicing U.S. rheumatologists and the then-current
ARA membership directory, we project that there were
1539 internist rheumatologists and 46 pediatrician rheumatologists in January 1977. The combined internist
and pediatrician rheumatologist population self-categorized as full-time academic was 28% of the total; community (private practice) rheumatologists constituted
64%; the balance fit neither category, classifying themselves as salaried full-time or in some non-practice situation. The Middle Atlantic and New England states
combined contained 35% of all internist rheumatologists; by contrast, these divisions contained 23% of the
1976 U.S. population. The rheumatologist-to-population ratio varied from 0.40 per 100,OOO in the East
South Central census area to 1.16 in the New England
states. The density of academicians was greatest in New
England and of community practitioners in the Middle
Atlantic states. In contrast, the lowest concentrations
were in the West South Central and East South Central
areas, respectively.
Table 2 further divides the responding community rheumatologists according to academic affiliation
and the nature of practice organization. (Recall that the
designation “solo” includes partnership arrangements.)
Of those who were in community practice, 77% claimed
academic affiliation, and 44% of those whose arrange-
Table 2. Practice setting of responding community rheumatologists, by census division
-
Census division
Setting
Internal medicine
With academic affiliation
Solo
Group
Unknown
Without academic affiliation
Solo
Group
Total internal medicine
Pediatrics
With academic affiliation
Solo
Group
Unknown
Without academic affiliation
Group
Total pediatrics
Total community
East
South
Central
West
South
Central Mountain Pacific
East
North
Central
West
North
Central
20
32
5
13
19
I
24
18
16
4
75
19
I1
4
6
2
53
16
7
128
7
11
1
75
35
I5
7
68
0
0
0
3
0
2
I
0
0
0
0
0
0
5
0
0
1
53
I33
76
New
Middle
England Atlantic
30
11
1
South
Atlantic
U.S.
0
7
I2
1
41
36
3
227
167
29
6
7
38
2
7
29
9
16
105
60
547
9
64
0
0
0
0
0
0
0
0
0
0
0
0
0
I
0
4
1
2
0
0
0
0
0
0
0
0
1
0
2
I
8
35
68
16
38
29
107
555
0
0
EPSTEIN AND HENKE
1180
ment was reported were in a group practice. Clearly
there were regional differences in distribution among
practice organizations, with solo practice exceeding
group practice on the East Coast and in the South (New
England, Middle Atlantic, South Atlantic, and East
South Central), while group practice was the major pattern in the rest of the country.
The responding ARA members provided the
date of their training, its duration, and characteristics
that allow us to recognize relatively more research or
more direct clinical emphasis. From these data, we
tested whether the distribution of current practice settings is independent of training characteristics (Table
3). The association between decade of training and
practice setting is statistically significant (P = 0.003),
but the size of the effect is small among graduates since
1951. The share who go into community practice has
varied by only two percentage points during that time,
and the share who become academicians has stayed
within a range of five percentage points. The association
between training date and choice of academic affiliation
among community practitioners is similarly significant
but small among graduates of the last 30 years. When
all years are considered together, there is a significant
and important relationship between training duration
and distribution among practice settings (P< 0.0005). If
we look at duration within each training decade, the difference is not significant, however, which suggests that
this association for all rheumatologists may be an artifact of the correlation between duration and training
Table 4. Distribution of time among professional activities, by
practice setting (share of total professional time in each activity)
community
Activity'
Academic
With
academic
af6liation
Without
academic
affiliation
0.30
0.24
0.18
0.13
0.15
0.78
0.13
0.01
0.04
0.04
0.92
0.W
0.00
0.01
0.02
Patient care
Teaching
Basic research
Clinical research
Administration
For each activity the difference among the three practice settings is
significant at P < O.OOO1.
date. Training programs reported as having the largest
percent of time devoted to research clearly provided the
highest percent of their graduates for academic positions, and those with the largest numbers of clinics and
consultations provided the largest share of trainees to
community practice. These extremes of research versus
clinical experience did not have a consistent effect on
subsequent academic affiliation for those in community
practice, even though the results were statistically significant.
In characterizing practice activities, we first considered the use of professional time. Table 4 shows the
reported distribution of time among five categories of
practice activities. Academicians devote almost the
same percent of time to teaching as to patient care (24%
and 30%), while basic and clinical research take a
smaller share (I 8% and 13%). Only in share of time for
Table 3. Current practice setting by decade, duration, and orientation of training (population estimates)
Training
characteristic
Academic
All rheumatologists
Academic affiliation, community only
Salaried
With
Community
Without
No.
%
No.
%
No.
%
P
No.
%
No.
%
P*
Decade7
Through 1950
1951-60
1961-70
1971-77
57
50
181
152
20
26
31
30
27
21
31
203
125
379
315
71
64
64
62
0.003
139
101
309
243
68
81
82
77
64
24
70
72
32
l9
18
23
0.003
44
9
11
5
9
Duration
t l year
2 1 year
61
379
20
30
34
89
11
7
217
805
70
63
o.Oo0
;;;
65
81
75
155
35
19
0.000
Orientation
Most research
Average
Most clinical
236
I24
80
45
24
15
39
39
45
7
7
8
246
364
412
47
69
77
o.Oo0
;;;
54
54
122
22
15
30
0.000
290
78
85
70
Total
440
28
123
8
1022
64
792
77
230
23
* P value for testing the hypothesis that the distribution among the practice settings considered is independent of the training characteristic.
f Decade in which speciaIty training was completed; for those with no rheumatology training, the decade residency was completed.
NATURE OF U.S. RHEUMATOLOGY PRACTICE
teaching do regional differences of significance exist, the
extremes being 17% by New England academic rheumatologists and 29%in the South Atlantic states. Community rheumatologists with academic affiliations devote 78% of their time to patient care, significantly more
than the academicians and significantly less than the
community rheumatologists without such affiliations (P
< 0.0005 for each pair). This distinction of affiliation
seems operationally most striking in regard to teaching,
which takes 13% of the time of those with, and only 4%
of the time of those without, such affiliation. These community rheumatologists with academic affiliation most
likely constitute what are usually designated “clinical
faculty.”
Table 5 presents a statement of professional activities, stratified by practice setting. Academic rheumatologists and community rheumatologists with academic
affiliations total 59 hours per week, and community
rheumatologists who deny academic affiliations work
only slightly fewer (difference not statistically significant). Solo and group practice show the same total professional time per week.
In contrast to total professional time, there are
highly significant differences in the hours devoted to patient care among practice arrangements. Academicians
average 18 hours per week in patient care, while community rheumatologists report from 45 to 49 hours. This
4-hour difference between community practitioners
with and without academic affiliations is significant (P
= 0.003). Regional differences in patient care hours oc-
1181
cur only for community rheumatologists with academic
affiliation, the extremes being 52.7 hours in the East
South Central division and 42.5 hours in the New England and Middle Atlantic divisions. Again, practice organization led to no difference.
The fraction of all patient care time devoted specifically to persons with rheumatic diseases varies across
practice arrangements. The most striking differences exist in the comparison of community rheumatologists
with and without academic affiliation in different sections of the country. For example, in the Middle Atlantic states, those with affiliations devote 71% of their patient care time to persons with rheumatic diseases, while
the unaffiliated rheumatologists devote 5 1% of their patient care time to these patients. The solo versus group
arrangement has no consistent effect. Academicians allocated 83% of their patient care time to rheumatic disease patients, with no significant differences found by
region.
Further characteristics of rheumatologic practice
shown in Table 5 indicate that community rheumatologists see more rheumatic disease outpatients per week if
they maintain an academic affiliation than if they do
not, and both see more than twice as many as those in
full-time academic practices. In contrast, all see about
the same number of inpatients per week. As might be
expected, full-time academicians and community practitioners with academic affiliations report a significantly
higher percent of their patients to have rheumatic disease than community practitioners without such an af-
Table 5. Practice characteristics reported by U.S. rheumatologists, by practice setting
p
p
p
p
_
_
_
p
-
Community*
p
p
p
p
p
-
Characteristic
Full-time
academic
Without
P
Solo
58.6
17.9
58.5
45.4
56.3
49.3
0.065
0.003
58.4
47.3
83
75
64
0.000
23.0
55.2
46.5
7.3
84
7.8
72
21.4
17.4
~~
Time in all professional activities (hrs/wk)
Patient care time (hrs/wk)
% of clinical time with rheumatic disease
patients
Number of rheumatic disease outpatients
seen per week
Number of rheumatic disease inpatients
seen per week
% of all patients with a rheumatic disease
Length of a rheumatic disease patient visit
(minutes)?
Number of rheumatic disease consults
per month
% of each week’s rheumatic disease patients
seen for consultation
Academic affiliation
With
16
-
Practice organimtion
~Group
P
_-57.9
0.663
47.8
0.695
72
72
0.974
0.024
56.4
52.6
0.260
6.6
60
0.197
O.Oo0
7.0
68
8.4
69
0.669
0.796
27.2
30.0
0.180
29.2
26.8
0. I82
18.7
13.4
0.009
16.I
19.6
0.05 I
9
9
0.914
8
9
0.563
* Reported P values are from I-tests of the difference between community rheumatologists with academic affiliation and without academic affiliation, and the difference between community rheumatologists in solo and in group practices.
t Calculated as: (reported time per week in rheumatic diseases patient care) i(rheumatic disease outpatients seen per week + 5 times rheumatic
disease inpatients seen per week).
EPSTEIN AND HENKE
1182
Table 6. Rheumatology in community practice, by decade. duration,
and orientation of training
No. of
rheumatic
outpatients
seen
per week
% of clinical
time with
rheumatic
disease
patients
Training
characteristic
Decade*
Through 1950
1951-60
1961-70
1971-77
P
Duration
t l year
rl year
P
Orientation
Most research
Average
Most clinical
P
No. of
rheumatic
disease
consults
per month
67
65
76
74
0.004
47
54
60
47
0.007
16
19
21
0.001
65
74
0.001
49
54
0.185
13
19
0.003
72
76
73
0.405
47
55
58
15
19
I1
in
0.145
0.060
* Decade in which specialty training was completed; for those with no
rheumatology training, the decade residency was completed.
filiation. A reverse finding exists for the duration of a
visit for a person with a rheumatic disorder (not statistically significant). Academicians and those in the community with academic affiliations see 17 to 19 consultations for rheumatic disorders per month, and there is a
significantly greater amount of such activity in group as
opposed to solo practice. As a percent of all specialty
patient care, academicians devote almost twice as much
attention to consultative activities as do community
practitioners.
Analysis of these practice characteristics by census division reveals that consultation for persons with
rheumatic disease is more frequent in East North Central and West North Central practices and least common in practices of the New England and Middle Atlantic states. Note that East North Central and West
North Central states are among those with the lowest
concentration of rheumatologists per capita, and the
New England and Middle Atlantic states are the highest
(Table 1).
Using patient care activities of community rheumatologists as markers, we can examine the effect of duration, date, and emphasis of fellowship training. Table
6 reveals that among all community rheumatologists,
decade of training is significantly associated with the
mean share of patient care time devoted to rheumatic
disease patients, the number of rheumatic disease outpatients seen per week, and the number of specialty consultations per month. Share of time and number of consults also vary by the duration of the specialty training.
The effect of research versus direct clinical training em-
Table 7. Patient care activity of community rheumatologists, by census division and age of practice
Share of clinical time with
rheumatic disease patients
Total patient care hours
-
~~
Census
division
New England
Middle Atlantic
East North
Central
West North
Central
South Atlantic
East South
Central
West South
Central
Mountain
Pacific
us
Practice age
_______.
-
-
Practice age
-
5 1 yr.
2-3 yrs.
4--5yrs.
>5 yrs.
sl yr.
2-3 yrs.
4-5 yrs.
>5 yrs.
38.3
39.0
40. I
46.0
53.0
39.4
41.4
44.5
I .oo
0.7 I
0.77
0.66
0.68
0.77
0.76
0.66
49.0
48.5
40. I
46.5
0.84
0.6 1
0.94
0.72
42.5
50.7
53.3
49.8
45.8
41.5
45.7
44.5
0.58
0.53
0.67
0.74
0.58
0.79
0.67
0.8 I
45.0
60.0
-
55.5
0.89
0.92
-
0.67
58.3
41.5
43.8
59.0
54.6
51.1
48.2
49.0
43.9
47.5
50.3
46.9
0.74
I .OO
0.72
0.95
0.76
0.81
0.9 1
0.67
0.75
0.75
0.82
0.66
44.3
48.5
45.8
46.1
0.76
0.74
0.78
0.7 1
P for diff.
by prac.
age
0.3398
0. I966
NATURE OF U.S. RHEUMATOLOGY PRACTICE
phasis on these three practice characteristics is shown to
be of no statistical significance for rheumatologists who
choose community practice.
The speed with which community practices develop to their maximum of patient care hours and share
of time in rheumatology is examined in Table 7. Community rheumatologists with practices established for 1
year or less devote an average of 44 hours per week to
patient care, and those in practice longer than 5 years
spend 46 hours. It is surprising to learn that no significant difference is to be found between the fraction of
patient care time devoted to persons with rheumatic disease in practices of less than 1 year and of those that are
older. This almost immediate jump to a full schedule of
patient care, with three-fourths of the time devoted to
rheumatic disease patients, is worthy of attention.
The content of rheumatologic practices by major
diagnosis, expressed as a fraction of all rheumatic disease patients in the particular practice, is shown in
Table 8. Osteoarthritis and degenerative arthritis constitute 16% of all academicians’ rheumatic disease patients
but 34% for community practitioners, a highly significant difference. Rheumatologists without academic affiliation see relatively more such persons than those
with affiliation, and neither practice organization nor
fellowship orientation is associated with the prevalence
of this entity in their practices. In contrast, rheumatoid
arthritics constitute 24% to 25% of the practices of all
rheumatologists, including academicians. Gout is also
equally frequent in nearly all settings. Rheumatic diseases of low prevalence such as juvenile rheumatoid arthritis, systemic lupus erythematosus (SLE), scleroderma, polymyositis and dermatomyositis, ankylosing
spondylitis, Reiter’s syndrome, and overlap rheumatic
diseases are all significantly more common in the practice of academic-based rheumatologists. Fibrositis, bursitis, tendonitis, and peritendonitis, however, are two to
three times more common in the practice of community
than full-time academic rheumatologists.
Having or not having an established academic affiliation is also associated with statistically significant differences in practice content. For each significant difference, those with affiliation are more similar to full-time
academicians than are those without affiliation. In the
few significant differences that occur between solo and
group practice, the group practices are consistently
more like the practices of academicians. Using the emphasis on research as opposed to direct clinical activities
as a stratifying characteristic, we find that SLE is more
common in the practice of community rheumatologists
whose fellowship training most emphasized research.
1183
All other entities are distributed independently of training orientation.
DISCUSSION
The perceived problem concerning numbers,
types, and distribution of physicians has been discussed
primarily in the context of physician-to-population ratios (5-7,9). Holden (12), Thier and Berliner (l3), Vilter
(14), Hadley (15), Ramsay (16), and many others have
forcefully argued that the information required for a
complete manpower policy demands more than these
ratios and the prevalence of a particular set of disabilities in the population. In particular, issues of access,
outcome, case mix, opportunity cost (the value of the alternative use of the resources utilized), and activity differences must also be addressed. This paper deals with
the ratios, but primarily with some of these other issues.
Estimates of the size of the U.S. rheumatologist
population depend on how one chooses to define this
group. Since there is only one national organization of
rheumatologists, the ARA, we considered that source
the most appropriate to identify these specialists. A recent analysis of medical subspecialties conducted by the
University of Southern California Division of Research
in Medical Education (USC/DRME) based specialty
designation on the AMA master file (10). They reported
that there were 645 practicing nonfederal internist rheumatologists in the United States in 1976 devoting half
their practice to persons with rheumatic diseases. In
contrast, the present study based on ARA membership
projects that there were 1258 such rheumatologists in
that same year, with significant age and practice location differences found in a comparison of these two
study populations (17). Even though they represent a
somewhat different population, we shall frequently
compare the characteristics recorded by rheumatologists
in the USC/DRME study, which was based on a log
diary of their practice activities, to the responses in our
survey. The USC/DRME findings for other subspecialties, in which the population sampled may be
more complete (16,18,19), provides interesting comparison to rheumatology.
A projected national total of 1539 rheumatologists in internal medicine and 46 in pediatrics provide
the basis for this report. That number represents 0.47%
of all U.S. physicians and projects a national ratio of
0.74 rheumatologist per 100,000 population in 1976.
The importance of the rheumatologist--population ratio
differences in Table 1 depends on whether it can be
demonstrated that the course of the diseases and the
Table 8. Distribution of rheumatologic diagnoses, by practice setting and rheumatology training orientation (mean fraction of all rheumatic disease patients in the practic
* When respondents grouped bursitis and fibrositis, all were attributed to bursitis.
NATURE OF U.S. RHEUMATOLOGY PRACTICE
availability of rheumatologist services to persons developing rheumatic diseases in the central areas is poorer
than for persons developing a comparable illness on the
coasts, where the ratios are higher. We are aware of no
studies advancing data to support an outcome and access difference.
As shown in Table 3, 28% of all rheumatologists
categorize themselves as full-time academicians and
about 64% are in community private practice. These
shares have remained nearly constant among new graduates for many years, even as the annual number of
graduates has risen. This may indicate a growth of academic opportunities commensurate with the number of
graduates, although it could also reflect constant relative desirability of the two types of practices among new
graduates in the presence of excess numbers of openings. Regional differences in practice settings (Table 2)
are important for manpower planning because they are
associated with significant differences in the content of
rheumatology practice (Table 5). In a descriptive study
like this one, we have no evidence on which to choose
between the view that the practice content differences
result from the differences in practice arrangement and
the equally plausible explanations that either reverse
the causation or attribute both practice activity and arrangement variation to some third factor, such as training background.
Comparison of rheumatologist practice settings
from our survey to other medical subspecialties is hindered by definition differences. A study of specialists in
pulmonary diseases conducted by one of their own specialty groups found that 24% were in full-time academic
positions in 1976 (20). But most other reports have used
the AMA master file categorization for physician practice arrangements, which is different from ours. Even
the office-based patient care category, which contains
52% of USC/DRME nonfederal rheumatologists (lo), is
not exactly comparable to our community physician
category, which contains 63% of our total population.
AMA figures for all U.S. nonfederal physicians in 1976
estimate 67% primarily in office-based patient care and
7.4% with primary activities of research, teaching, and
administration (2 1). Definition of a desirable distribution among practice settings depends on not only a
knowledge of what each type of rheumatologist does
but also the value of each different activity, which is a
much more difficult question.
Our responses indicate that academic and community rheumatologists overall devote 65.8% of their
time to patient care (Table 4), an activity that must often coincide with teaching and research. The log diary
1185
records of the USC/DRME rheumatologists indicated a
national average of 53% of time devoted to patient care
(10). These figures are not quite comparable because we
have excluded salaried rheumatologists (8% of the population) who are often not in patient care situations, but
this exclusion cannot account for the entire discrepancy.
Cardiologists reported in 1970 that 61% of their time
was spent in direct patient care with another 19% in patient care combined with teaching or research (22). In
1976, patient care was reported to be 58% of cardiologists’ professional time, 61% of gastroenterologists’ time,
49% of the time of specialists in pulmonary diseases,
and 41% of hematologists’ time (10). Since teaching partially, and research entirely, are the domain of academicians, any desired increases in these activities could be
accomplished with loss of only a relatively small share
of overall patient care time.
The number of hours per week in all professional
activities (Table 5 ) is quite similar when academic affiliation or practice organizations are compared. These responses are also similar to the 52.8 hours per week (excluding on-call time) reported in the USC/DRME
study (10). Overall, U.S.office-based physicians in 1976
worked 52.2 hours per week, ranging from 55.7 hours
for internists to 48.1 hours for psychiatrists (21).
Much interest is concentrated on the time that
physicians spend in direct patient care. The national average in 1976 for office-based physicians was 46.5 hours
per week in direct patient care, with a range of 41.3 to
49.4 (21), comparable to our community rheumatologists. The study of USC/DRME rheumatologists found
4.7 hours per day for total patient care, or 28 hours per
week (10). This figure is considerably smaller than the
weighted average of 37.7 hours per week that we can
calculate from Table 5 . Differences in the categories
used to report activities make it difficult to know if the
discrepancy is due to the difference in population studied, difference in definition of activities, or to the greater
accuracy inherent in direct reporting with the objective
log diary method. Overall, we found community-based
rheumatologists are within the national averages in time
spent in direct patient care. Among the small regional
differences found for community rheumatologists with
academic affiliations, the smallest number of hours occurs in the two census divisions with the highest density
of rheumatologists per capita, and the area with the
highest hours is the census division with the lowest
number of rheumatologists per capita.
The share of patient care time devoted to persons
with rheumatic diseases (Table 5 ) and the time it takes
the practice to achieve this level (Table 7) may be sensi-
1186
tive to the demand for the rheumatologist’s services.
Only community rheumatologists without academic affiliation devote more than a quarter of their time to persons without a rheumatic disease. No available evidence
shows whether rheumatologists desire to treat more
rheumatic disease problems than they do now (as shown
in Table 5 ) . We do not know whether this is an optimal
content in terms of physician and patient satisfaction,
nor do we know its effect on the quality of the care delivered. The mechanism that determines the specialty
content of a practice is complex and depends on referral, individual patient behavior, and physician interest
and capability.
The actual numbers of rheumatic disease patients seen in an outpatient setting (Table 5), one measure of specialty productivity, depends on academic affiliation but not solo versus group organization. Since
we have no guide to case complexity, numbers alone
may fail to portray accurately the contribution of each
type of rheumatologist. In 1976, U.S. office-based internists saw an average of 73 outpatients per week (similar
to the number seen by community rheumatologists, as
calculated from Table 5, lines 4 and 6) and made 45 inpatient visits (21). The fact that rheumatologists see on
the average only 1 inpatient per day with a rheumatic
disease (Table 5) emphasizes that this is a specialty for
the ambulatory ill.
Community rheumatologists have reported that
60% to 72% of their patients have rheumatic diseases
(Table 5), while the academicians claim that 84% of
their patients have such conditions. The overall average
is slightly higher than that found for the USC/DRME
rheumatologists (65.4%) (lo), but they include more
physicians not in patient care. Both compare favorably
to the share of specialty encounters by pulmonary disease specialists (37%), gastroenterologists (47%), endocrinologists (49%), and nephrologists (43%) (10). Academic affiliation appears to influence this important
characteristic of U.S. rheumatologists.
A recent study of time spent with patients by internists compared to family and general practitioners revealed that the specialist spent more time for almost all
categories of patient services, although the disease severity in that study was not well delineated (23). We
found (Table 5) that rheumatic disease patients seen by
a community rheumatologist received more time per
visit than the 24-minute average calculated for patients
of office-based internists (21).
Consultations are most frequent in practices with
academic affiliations. Academicians do not seem most
active in the number of consultations per month, but, as
EPSTEIN AND HENKE
a proportion of time in clinical practice, they do see
more than those in the community. Again, absence of
knowledge of disease complexity makes us cautious in
the interpretation of the work output in consultations
performed by those in full-time academic careers in
contrast to those in the community.
Details of the diagnoses of rheumatic diseases by
different kinds of rheumatologists reveal that the academic rheumatologists devote much of their practice
time to the subset of conditions with the hghest mortality rate. If these are also the least common or most difficult to control and if medical schools are considered the
last stop in the referral process, then this distribution
may reflect an efficient triage. Even so, the largest numbers of such persons are still cared for in the community
because there are many more private practitioners, and
each cares for more rheumatic disease patients. It would
be very desirable to have case-complexity data for the
different practice arrangements to reflect further on the
efficiency of the triage system.
In examining factors that might influence what
rheumatologists do in their practices, one should also
consider specialty training background. As Table 3 illustrates, few differences have been observed since 195l
in positions assumed and academic affiliations maintained. Since duration of training of more than 1 year is
so much the norm at present, its influence for future
rheumatology trainees seems very slight. The general
emphasis of training programs does appear to have significant association with career path (Table 3), but
training activities for those who have chosen community practice do not have a similar association (Table 6).
Decade and duration of training are associated with
output differences (Table 6), but these are also not useful for influencing future training patterns.
Another potential source of variation in practice
activities is difference in practice location, with the corresponding variation in practice environment and market forces. We found surprisingly few activity differences associated with census division for each practice
setting. One reason may be that census divisions are too
aggregate a unit to reflect differences in practice environments. This explanation is supported by the statistical significance of particular Standard Metropolitan
Statistical Areas in a preliminary multivariate analysis
of rheumatologist productivity (24). This potential influence of practice environment on activity is another
essential part of a complete manpower analysis.
In this report, we have presented a description of
the distribution and activities of U.S. rheumatologists.
These results and similar data from another systematic
NATURE OF U.S. RHEUMATOLOGY PRACTICE
sampling (10) provide important inputs to formulation
of a useful rheumatologist manpower policy, but we
also need information concerning case complexity, access, outcome, and cost before development of rational
manpower proposals. In a subsequent paper, we shall
examine the practice activities of these rheumatologists
in a multivariate framework.
REFERENCES
1. Califano JA: The government-medical education partner-
ship. J Med Educ 54:19-24, 1979
2. Petersdorf RG: Some issues in medical education. Forum
Med 3:203-209, 1980
3. Health Professions Educational Assistance Act of 1976
4. Schwartz WB, Newhouse JP, Bennett BW, Williams A P
The changing geographic distribution of board-certified
physicians. N Engl J Med 303:1032-1038, 1980
5. Professional Manpower in Rheumatology. Prepared by
Cresap, McCormick and Paget, Inc., Management Consultants for the Arthritis Foundation. New York, Arthritis
Foundation, 1973
6. McCarty DJ: Geographic distribution of consultant rheumatologists in the United States-1975. Arthritis Rheum
1956-58, 1976
7. Yelin EH, Henke CJ, Epstein WV: Rheumatology manpower in California. Arthritis Rheum 205305-8 10, 1977
8. American Rheumatism Association Committee on Rheumatologic Practice: A description of rheumatology practice. Arthritis Rheum 201278-1281, 1977
9. McCarty DJ: Geographical distribution of consultant
rheumatologists in the United States-1979. Arthritis
Rheum 23:344346, 1980
10. Girard RA, Mendenhall RC, Tarlov AR, Radecki SE,
Abrahamson S: A national study of internal medicine and
its specialties. I. An overview of the practice of internal
medicine. Ann Intern Med 90:965-975, 1979
1 I. Henke CJ, Epstein WV: The training of United States
rheumatologists, 1927-1976. Arthritis Rheum 22:187-196,
I979
1187
12. Holden WD: A perspective of physician manpower. N
Engl J Med 300:493495, 1979
13. Thier so, Berliner RW: Manpower policy: base it on
facts, not opinions. N Engl J Med 299:1305-1307, 1978
14. Vitler RW: Medical manpower. Forum Med 2:752-754,
1979
15. Hadley J: Alternative methods of evaluating health manpower distribution. Med Care 17:1054-1060. 1979
16. Ramsay DL, Weary PE, Krasner M, O’Sullivan V, Sullivan B, Lee P, Lee A, Martin E: Manpower for dermatologic care, Report of the Dermatology Manpower Study,
Department of Dermatology, New York University Medical Center, 1977
17. Epstein WV, Henke CJ: Letter to the editor. Ann Intern
Med 91:649, 1979
18. Cherkin D, Lawrence D: An evaluation of the American
Medical Association’s physician master file as a data
source-one state’s experience. Med Care 15:767-779,
1977
19. Mendenhall RC, Ramsay DL, Girard RA, DeFlorio GP,
Weary PE, Lloyd JS: A study of the practice of dermatology in the United States. Arch Dermatol 114:1456-1462,
1978
20. Coates EO, Nadel JA, Petty, TL, Pierce AK, Snider GL,
Wiseman DH, Murray JF: Survey of professional manpower in pulmonary diseases. Prepared for National
Heart and Lung Institute by American Thoracic Society
and the American College of Chest Specialists, 1972
21. Gaffney JC: Profile of medical practice 1978. American
Medical Association, Center for Health Services Research
and Development, 1979
22. Swan HJC, Gifford RW: Current profile of the professional activities of the American cardiologist. Am J Cardiol 34:417428, 1974
23. Noren J, Frazier T, Aetman I, DeLozier J: Ambulatory
medical care: a comparison of internists and family-genera1 practitioners. N Engl J Med 302:ll-16, 1980
24. Henke CJ, Epstein WV: Variation in characteristics of academic and community rheumatologists. Paper presented
at the Manpower and Fellowship Education in Rheumatology Workshop, Glen Cove, New York, September 2729, 1980
Документ
Категория
Без категории
Просмотров
0
Размер файла
909 Кб
Теги
nature, practice, rheumatology, 1977
1/--страниц
Пожаловаться на содержимое документа