Licensed Physicians Who Workin Prisons: a Profile

RICHARD L. LICHTENSTEIN, PhD, MPHANNETTE RYKWALDER, MPH

Dr. Lichtenstein is assistant professor, Department ofMedical Care Organization, School of Public Health, Uni-versity of Michigan, 109 Observatory St., Ann Arbor, Mich.48109. Ms. Rykwalder is patient education coordinator,University of Virginia Hospitals, Charlottesville, Va.

This research was supported by grant No. HS 04127-01from the National Center for Health Services Research,Public Health Service, and by grant No. LEAA 77-ED-99-0026 from the Law Enforcement Assistance Administration.Tearsheet requests to Dr. Lichtenstein.

SYNOPSIS ...............................

A profile of the personal and professional char-acteristics of the physicians who work in America'sprisons was obtained by analyzing data from a largerstudy of all licensed physicians in the United Stateswho worked in a prison at least 12 hours a monthduring the fall of 1979. Psychiatrists were not in-cluded, nor were physicians working in jails.

The population of 382 prison physicians com-prised two major groups-those who worked inprisons full time and those who worked in thempart time. Part-time physicians, who represented themajority of physicians involved in prison work (58percent), were found to resemble closely the typicalphysician in the United States; they were predom-inantly trained in America, specialized, and boardcertified. In contrast, full-time prison physicians,who accounted for 73 percent of the total hoursphysicians spent working in prisons, differed sig-nificantly from the typical U.S. physician. They wereolder, less specialized, less likely to be board certi-fied, and more likely to be graduates of non-U.S.medical schools.

The professional characteristics of the full-timeprison physicians raise serious questions about thequality of medical care they are likely to provide.It would seem, based on their professional attributes,that the part-time physicians are able to providebetter quality care than their full-time colleagues.Prison health system could thus assure higher qual-ity care to inmates by relying primarily on part-time rather than full-time practitioners.

PHYSICIANS WHO PROVIDE HEALTH CARE in prisonswork in one of the most difficult practice settingsa physician is likely to encounter in the UnitedStates. Prison health systems and the physicianswho work in them face many problems. A majorproblem has been financial. Prison health systems,like the correctional systems that usually operatethem, have been sorely underfunded for many years(1). In addition, the provision of health services toinmates has not been viewed historically as a highpriority by corrections officials, who typically havebeen more concerned about security, housing, andfood services (2). Underfunding and low prioritystatus resulted in prison health systems that, in mostcases, failed to meet community standards. Existingdescriptions portrayed many prison health programsas virtually primitive (3, 4).A major force for change in prison health care

has been the courts. In a landmark 1976 legal case(Estelle v. Gamble, 429 U.S. 97 [1976]) and inseveral subsequent cases, the courts ruled that in-mates have the right to receive medical care at theprevailing community standard (5). Facing the

threat of inmate lawsuits and Federal receivershipif they did not meet the community standard ofmedical care, many prison systems began upgradingthe quality of their health services.

Reallocated State funds and Federal funds pro-vided by the Department of Health and HumanServices and the Law Enforcement Assistance Ad-ministration were used to upgrade the scope, con-tent, quality, and management of health programsin many prisons (3). In addition, several profes-sional societies, including the American PublicHealth Association, the American Medical Associa-tion (AMA), and the American Correctional Asso-ciation, promulgated standards for correctionalhealth programs. Despite advances at some prisons,however, the state of prison health care remainsinadequate (3, 6). And for physicians, working ina prison is a very different experience than practic-ing in almost any other setting (7).

Physical facilities for providing medical care inprisons are frequently small, dark, noisy, old, andcrowded, and technical equipment and supplies areoften antiquated or unavailable (1, 3, 6). The num-

November-December 1983, Vol. 98, No. 6 589

ber of personnel, both professional and ancillary, isoften inadequate (1). Security rules and securitypersonnel may infringe on the physician's autonomy(7). Patients' attitudes have been portrayed as de-manding, disrespectful, uncooperative, and hostile,and patients are said to frequently feign illnessesto avoid work or other duties (4, 7). Clinically, in-mate populations have been found to have higherlevels of illness than the general population of simi-lar age (8). Pay, under civil service schedules, isusually extremely low for physicians working inprisons, and the status they are accorded by theirmedical colleagues suffers (9). Fear for one's safety,which must affect all physicians working in prisonsto some degree, further contributes to making thepractice of medicine in a prison a seemingly un-attractive endeavor (4).

Yet hundreds of licensed physicians in the UnitedStates do work in prisons, some for extended periodsof time. Who are these physicians? What is theirmotivation for working in prison health care? Howdo they compare with the general population ofU.S. physicians? To answer these questions, we haveanalyzed data from a larger survey, made inferencesabout the probable quality of care provided by phy-sicians who work full time in prisons, and offereda strategy to cope with the situation.

Methods

Data presented in this paper were obtained aspart of a larger study to identify the organizationalcharacteristics of prison health programs that wereassociated with physician satisfaction and retention(9). Since the objective of that study was to identifypossible approaches for retaining licensed physiciansin the prison setting, data were not collected on anyunlicensed (usually foreign-trained) physicians whomay have been working in prisons. Although somecontend that unlicensed physicians are a majorsource of health manpower in correctional facilities(10), others argue that omitting unlicensed physi-cians from the study is inconsequential. The immedi-

ate past president of the American CorrectionalHealth Services Association stated that, due tolegal pressures, the use of unlicensed physicians inprisons is now "rare" (Richard Kiel, personal com-munication, September 8, 1983). "In the last fewyears, there has not been a large enough number ofunlicensed physicians practicing as the responsiblephysician in prisons to have any impact on thisstudy's findings at all." Unlicensed physicians not-withstanding, the data presented do reflect ac-curately the licensed physicians who worked inprisons at the time of the study.

The study was conducted in fall 1979. The popu-lation included all licensed physicians in the UnitedStates who diagnosed or treated patients with non-psychiatric problems, on a regularly-scheduled basis,for at least 12 hours per month, inside a Federal orState prison for adults or adolescents. Besides omit-ting the unlicensed physicians, the study also ex-cluded most psychiatrists, physicians who treated in-mates in their own offices or in nonprison hospitals,and physicians who practiced inside prison wallsonly occasionally.

Physicians working in jails-as opposed to prisons-were also excluded. A jail generally houses per-sons who have not yet been convicted of a crime butare awaiting trial and, often, persons already con-victed but sentenced to less than 1 year. Jails areusually operated by local (county or municipal)jurisdictions, although some States and the U.S.Bureau of Prisons operate jails. In contrast, prisonshouse persons who have been convicted of crimesand sentenced to terms longer than 1 year. In theUnited States, prisons are operated by 52 jurisdic-tions-the 50 States, the District of Columbia, andthe Federal Government.Names of eligible physicians were obtained from

corrections officials in each of these 52 jurisdictions.All 588 physicians meeting study criteria were senta self-administered questionnaire eliciting informa-tion about personal and professional characteristics,job satisfaction with several aspects of work, andintentions of staying on or leaving the job. Followupprocedures consisted of three additional contacts bymail and one by telephone.A total of 382 usable questionnaires was received,

a response rate of 65 percent. The response rate bycorrectional system is shown in table 1. Some physi-cians did not answer all questions, which accountsfor totals not adding to 382 in some of the tables.To the extent possible, nonrespondents were com-pared with respondents in terms of personal andprofessional characteristics. This analysis showed

590 Public Health Reports

Table 1. Response rate among eligible physicians, bycorrectional system

Number ofNumber of physicians Percent

State respondents eligible responding

Alabama ..............Alaska ................Arizona ...............Arkansas ..............California ..............Colorado ..............Connecticut ............Delaware ..............Florida ................Georgia ...............Hawaii ................Idaho .................Illinois ................Indiana ................Iowa ..................Kansas ................Kentucky ..............Louisiana ..............Maine .................Maryland ..............Massachusetts .........Michigan ..............Minnesota .............Mississippi ............Missouri ..............Montana ...............Nebraska ..............Nevada ...............New Hampshire ........New Jersey ............New Mexico ...........New York ..............North Carolina .........North Dakota ..........Ohio ..................Oklahoma .............Oregon ................Pennsylvania ...........Rhode Island ...........South Carolina .........South Dakota ..........Tennessee .............Texas .................Utah ..................Vermont ...............Virginia ...............Washington ............West Virginia ..........Wisconsin .............Wyoming ..............District of Columbia .....U.S. Bureau of Prisons

1 2 501 4 254 6 672 2 100

29 49 594 4 1002 2 1000 2 020 30 675 9 563 4 751 2 50

17 28 615 10 50

05 6 833 5 60

11 21 521 2 50

12 20 609 30 3017 23 742 3 671 1 1001 3 334 5 801 3 332 4 501 2 506 14 432 3 67

29 37 7830 36 83

011 18 617 9 782 5 40

15 24 631 3 331 3 330 1 0

10 10 10014 16 881 1 1001 1 100

17 27 637 17 412 2 1006 11 551 2 504 6 67

51 69 74

only one important difference: foreign medical grad-uates were somewhat less likely to respond to thequestionnaire than graduates of American medicalschools. Of the respondents, 23 percent had beengraduated from non-U.S. schools, compared with 30percent of the nonrespondents. This difference

should be borne in mind when interpreting the datathat follow.

Results

Characteristics of licensed prison physicians. Dataon personal and professional characteristics of li-censed prison physicians are presented in tables 2-4.The mean age of the physicians was 48 years (table2). Seven percent were under 30 years old and 4percent over 70, with the remainder about equallydistributed in their 30s, 40s, 50s, and 60s. Nearlyall (96 percent) prison physicians were males, afigure that mirrors the proportion of male prisoners(96 percent) (6).

Seventy-seven percent of the respondents weregraduates of American medical schools, and 23 per-cent were graduates of foreign medical schools(table 3). If there had been a 100 percent responserate, about 25 percent would be foreign medicalgraduates, based on information we found aboutnonrespondents. Ninety-five percent of the respon-dents were fully licensed by the States in which theyworked. The other 5 percent consisted of physicianswith only Federal licenses, including National HealthService Corps personnel (about 1 percent), andphysicians with licenses restricted to the institutionemploying them (4 percent).

Thirty-seven percent of the respondents were cer-tified by one or more specialty boards, 24 percentwere board eligible, and 7 percent restricted theirpractice to a specialty although they were neitherboard eligible nor certified; 32 percent claimed nospecialty. (Family practice was considered a spe-cialty in this survey.) Forty-two percent of the re-spondents were employed full time in prisons (atleast 35 hours per week), and 58 percent worked inprisons only part time. The average number of yearsrespondents had worked in prisons was 5.7, mostlyat their current institution (an average of 5.2 years).

Questions concerning physicians' reasons for tak-ing jobs in prisons were also included (table 3).Twenty specific reasons in the questionnaire havebeen grouped into four categories-monetary rea-

November-December 1983, Vol. 98, No. 6 591

3119M.~ ~ ~ s pstvines,jo-related reasons, no al-

ternatives.

Monetary reasons. Supplementary income whileu 1 building a private practice; income while deciding

on future plans; supplementary income during re-*.~~~. tirement; good pay; supplementary income to mili-

tary pay.

~~~. ~~~~Positive interest reasons. Interest in correctional| | S i | health; inmates need care; interesting caseload; rep-

Table 2. Age distribution of prison physicians by employment status

Age groups (percentages)

Under 30 30-39 40-49 50-59 60-69 70 or more Mean ageGroup Number years years years years years years (years)

All prison physicians. ..................... 378 7 25 23 21 21 4 48Full-time prison physicians ..... ........... 160 8 18 16 31 26 22 50Part-time prison physicians ...... .......... 218 6 30 27 14 18 5 47

Sexual distribution: 96 percent male; 4 percent female (N = 380).2 Total does not add to 100 percent because of rounding.

Table 3. Specific characteristics of licensed prison physicians by employment status (percentages)

All prison Full-time Part-timephysicians physicians physicians

Characteristics (N = 382) (N = 160) (N =221)'

Training: 2American medical school graduate .77 68 84Foreign medical school graduate .23 333 16

Licensure status: 4Fully licensed .95 89 99.5Restricted to specific institutions .4 8 .5Federal license .1 3 0

Specialty status: 5Board certified .37 28 46Board eligible .24 22 23Self-proclaimed specialist .7 8 8Nonspecialist .32 42 23

Reasons for entering correctional work:Positive interest:

Yes .49 41 55No.51 5945

Monetary interest:Yes .49 35 59

No.51 6541Job related:

Yes .41 58 28No.59 4272

No alternative to correctional work:Yes .16 19 14

No ..................................................... 84 81 86

No data on 1 respondent.2 No data on 2 respondents.I Percentages do not add to 100 percent because of rounding.

4No data on 1 respondent.5 No data on 18 respondents.

592 Public Health Reports

resents a challenge; a source of patients in my spe-cialty area.

Job-related reasons. Provides job security; poorpersonal health or age made other work difficult;wanted regular hours; interested in an organizedpractice; it was an alternative to military service;wanted a change of pace.

No alternatives. Was having licensure problems;had limited alternatives; was having malpracticeproblems; and group has a contract to provide ser-vices.

Each physician could select his or her two mostsignificant reasons for entering correctional prac-tice. Their reasons are presented in table 3 in dichot-omous fashion, indicating the percentage of physi-cians who did or did not select each category as oneof their two choices. Only 16 percent indicated thatthey chose correctional work because they had noalternative to that field. The dominant reasons forentering prison health care were a positive interestin the field or a monetary interest (49 percent each).Slightly fewer physicians (41 percent) chose job-related factors such as regular hours or job security.

Differences between full- and part-time physicians.Although 42 percent of respondents worked inprisons full-time, they accounted for most of thehours spent by physicians practicing in prisons. Wefelt it imperative to describe the characteristics ofthese full-time prison physicians and to contrastthem with those of their part-time colleagues (tables2-4). Tests of statistical significance are inappro-priate and were not used here because the data rep-resent differences for a population rather than asample.The full-time physicians were an older group

than the part-time physicians. Their mean age was50 years, and 57 percent were over 50 years old,compared with the part-time physicians, whose meanage was 47 years and only 37 percent of whomwere over 50 years of age. Full-time physicians weresomewhat less likely to be fully licensed by a State,with 8 percent holding restricted State licenses. Vir-tually all part-time practitioners were fully licensed(table 3). Full-time physicians were also less likelythan their part-time counterparts to be board-certi-fied specialists (28 percent versus 46 percent) andmore likely to declare no specialty at all (42 versus13 percent). The two groups of physicians also dif-fered in the location of their medical school train-ing. Only 16 percent of the part-time physicians

Table 4. Income and time worked in prison

Full-time Part-timeCharacteristic physicians physicians

Average income per hour:Board certified ........ ..... $23 $40Board eligible ........ ...... 21 38Self-proclaimed specialist ... 21 36Nonspecialists ........ ..... 19 32

Overall .................. 20 38

Hours per week (average): 1In correctional work ........ 42 11In other settings ....... ..... 6 39

Total ................... 48 50

'160 full-time and 221 part-time physicians.

were foreign medical graduates, but twice that pro-portion, or 33 percent, of full-time practitionerswere graduates of foreign medical schools.

Levels of direct pay received by part- and full-time prison physicians differed markedly (table 4).Part-time physicians received an average hourlycompensation of $38 an hour, but their full-timecolleagues averaged only $20 an hour. The rates forfull-time physicians, however, do not reflect the valueof their fringe benefits. Those benefits, which arenot available to part-time prison physicians, may vir-tually equalize the overall costs of compensation tothe prison systems.

Pay differences between full- and part-time physi-cians reflected their specialty credentials to varyingdegrees. Direct pay of part-time physicians rangedfrom $32 an hour for nonspecialists to a high of$40 an hour for diplomates of specialty boards. Incontrast, full-time physician direct pay was less re-flective of credentials, ranging from $19 an hour fornonspecialists to $23 an hour for board-certifiedspecialists. Significantly, the full-time physicians withboard certification were paid at rates about 28 per-cent lower than part-time practitioners with no spe-cialty training and 43 percent lower than the part-time physicians with board certification. Even fringebenefits would not account for this skewed patternof compensation.

Full-time practitioners worked an average of 42hours per week in the prisons and the part-time phy-sicians averaged only 11 hours per week (table 4);the full-time providers accounted for 6,720 (73 per-cent) of 9,151 total physician hours. Full-time phy-sicians were clearly the mainstay of correctionalhealth care despite the larger number of part-timepractitioners.The two groups of physicians also revealed differ-

ences in their reasons for entering work in the field

November-December 1983, Vol. 98, No. 6 593

of corrections (table 3). Only a small percentage ineach group felt that they had no alternative toprison work, with little difference between the twogroups (19 and 14 percent of full- and part-timephysicians, respectively). There were large differ-ences, however, in the other three categories of rea-sons for entering prison health work. Full-time phy-sicians were most likely (58 percent) to have en-tered the field for reasons related to the job, such asregular hours and job security, rather than for rea-sons related to a positive interest in the field (41percent) or for monetary reasons (35 percent). Incontrast, only 28 percent of part-time physiciansentered the field for job-related reasons; 59 percenttook the position for monetary reasons, such as sup-plemental income, and 55 percent because of a posi-tive interest in the field and the patients.

Differences between the profiles of prison physiciansand of all U.S. physicians. To put the profile ofphysicians working in prisons in perspective, we

compared characteristics of the entire population ofprison physicians and the subgroups of part-timeand full-time prison physicians with a profile of allU.S. physicians (11) prepared in 1980 by the Amer-ican Medical Association (table 5).

The entire population of prison physicians did notdiffer statistically from the general physician popu-lation with respect to mean age, age distribution, or

percentage of foreign medical graduates. Statisticallysignificant differences were found, using a test ofproportions with the normal approximation (12),between these two populations in the percentages ofwomen, board-certified physicians, and physiciansdeclaring no area of specialization. Women ac-counted for 4 percent of the prison physicians and9 percent of all U.S. physicians. Board-certified spe-cialists represented 37 percent of the prison physi-cians and 49 percent of the all-U.S. group. Thirty-two percent of the prison physicians declared nospecialty, but only 16 percent of the general physi-cian population did so. All of these differences aresignificant at P < .001.

Part-time prison physicians were notably similarto the all-U.S. group of physicians. The two groupsshowed no statistically significant differences in age,sex distribution, proportion of foreign medical grad-uates, or percentage of board-certified specialists.Only in the proportion of physicians declaring nospecialty, 23 percent of the part-time prison groupand 16 percent of the all-U.S. physician group, didthe two groups reveal a statistically significant dif-ference (test of proportions, P < .01 ) .

Profiles of full-time prison physicians and all U.S.physicians were compared also, and the differenceswere striking. Statistically significant differenceswere found in all five bases of comparison-agedistribution and the percentages of women, foreign

Table 5. Comparison of prison physicians with all U.S. physicians (percentages)

All U.S. All prison Part-time prison Full-time prisonphysicians ' physlcians physicians physicians

Variables (N = 437,486) (N = 382) (N = 221) (N = 160)

Age.............2100.0 100.0 100.0 100.0Less than 35 years ........ ................... 27.4 319.3 322.0 415.635-44 years .................................. 25.0 23.5 26.1 19.445-54 years .................................. 19.5 23.0 22.5 23.855-64 years .................................. 14.5 23.0 17.9 30.065 or more years ......... .................... 13.5 11.3 11.5 11.3Mean (years) ................................. 47.0 48.0 47.0 50.0

Sex ........................................... 100.0 100.0 100.0 100.0Male. 5 91 6 96 394.5 6 97.5Female ...................................... 9 4 5.5 2.5

Training .......... ............................. 100 100 100 100.0American medical school graduate ..... ......... 580 377 384 667.5Foreign medical school graduate ..... ........... 20 23 16 32.5

Specialty status:Board certified ........... .................... 49 637 46 628No specialty ................................. 16 632 723 642

SOURCE: Reference 11, pp. 140, 148, 149.2 Totals do not equal 100 percent because of rounding.I Not significant.4 Chi square goodness of fit test, P < .005.

5Percentage of physicians involved in patient care.6 Test of proportions using the normal approximation, P < .001.'Test of proportions, P < .01.

594 Public Health Reports

medical graduates, board-certified physicians, andnonspecialists.

Full-time prison physicians, whose mean age was50 years, were significantly older (P < .005) as agroup than the AMA profile of all U.S. physicians,which had a mean age of 47 years, as determinedby the chi square goodness of fit test (13). In the55-64 year age grotup, the full-time physicians hada particularly high concentration when comparedwith the U.S. physician group.The full-time physician group had a lower per-

centage of women than the overall U.S. physicianpopulation, presumably due to the predominance ofmale inmates. Full-time prison practitioners alsodiffered in terms of training, specialization, andbackground. The percentage of foreign medicalgraduates was nearly 33 percent in the full-timeprison group, compared with 20 percent in theoverall U.S. physician population. In terms of boardcertification, only 28 percent of the full-time prisonphysicians were diplomates of specialty boards,while 49 percent of all U.S. physicians had attainedthat status. No specialty was declared by 42 percentof the full-time prison group and 16 percent of allU.S. physicians. These differences were all signifi-cant at P < .005.

Conclusion

Physicians who work in prisons are an extremelydiverse group, with a wide range of backgrounds,specialty training, and practice experience.Two distinct subgroups-part-time and full-time

physicians-have been noted. The majority are part-time physicians who also practice outside penal in-stitutions. They closely resemble the typical U.S.physician in the AMA profile. Their presence nodoubt decreases the isolation of prison medicine andpermits inmates access to providers whose personaland professional attributes are virtually indistin-guishable from health care providers available tomembers of the larger community.

Although fewer in number than their part-timecolleagues, full-time prison practitioners providenearly three-fourths of the physician-hours in thenation's prisons, and they can be considered thedominant providers of physician care to prison in-mates. Full-time physicians differ from the profileof all U.S. physicians and from that of their part-time colleagues in a number of important ways.Although no definitive conclusions can be drawnfrom a statistical profile such as the one we havepresented, the characteristics of the full-time prison

physicians do raise questions about the probablequality of the care they render in comparison withphysicians in the mainstream of medical care.

In the absence of data about the technical qualityof the care actually delivered by these full-time phy-sicians, inferences about the quality of care theyrender may be made based on their professionalattributes (14). One cannot be sanguine about thequality of care the full-time prison physicians arelikely to provide since their profile indicates that adisproportionate number have characteristics asso-ciated with lower quality care.

Licensure, longer periods of training, board certi-fication, and specialization have all been associatedwith higher quality performance (15-17). Amongfull-time prison physicians there is a high percentagewith restricted licenses, limited postgraduate medi-cal training, no area of specialization, or no boardcertification. The physician's age, although it is con-founded with the amount of training, has also beenfound to be associated with quality (15). Older phy-sicians, particularly those in general practice withno additional training, have been found to providelower quality care (18). Fifty-seven percent of thefull-time prison physicians were nonspecialists overthe age of 45.

Another striking characteristic of full-time prisonphysicians is the high proportion (33 percent) whoare graduates of foreign medical schools. Severalauthors (15-17, 19, 20) have compared the qualityof care provided by graduates of foreign and U.S.medical schools, and their results seem to indicatethat, in general, foreign medical graduates performas well as U.S. graduates. This is particularly truewhen the foreign medical graduate has received ad-ditional postgraduate medical training and is a spe-cialist (19, 20). Because these characteristics do notdescribe most foreign medical graduates who workfull time in prisons, we can infer that they probablyprovide care of lesser quality than the typical U.S.physician (17, 19). The vast majority of foreignmedical graduates working full time in prisons arenot board certified (84 percent). Most lack ad-vanced training (56 percent), and many practice nospecialty (47 percent) or have only a restrictedlicense (25 percent). The ability to speak Englishhas also been associated with the quality of careprovided by foreign medical graduates in the UnitedStates (19, 20). Although we do not know anythingabout the language skills of the foreign medicalgraduates working full time in prisons, any problemsthey have in providing high quality care would beexacerbated by a poor command of English.

November-December 1983, Vol. 98, No. 6 595

Certain structural attributes of practice settings,such as a high degree of staff organization, extensiveuse of peer review, affiliations with medical schools,and the group practice rather than solo practicemode of delivery, have also been associated withhigher quality care (17, 19, 20). Several studiesfound that even physicians with characteristics as-sociated with lower quality can provide high qualitycare in appropriately structured facilities and or-ganizations, while the absence of such structuralcharacteristics can make matters worse for the indi-vidual clinician (16, 17, 19, 21). Unfortunately, theprison health systems across the United States, withsome exceptions, have few of these positive struc-tural characteristics. Most are not organized tosupervise full-time providers properly, except, per-haps, where part-time physicians are active in suchendeavors (1, 3, 6). Thus, the characteristics offull-time providers and the nature of their work set-tings cause further concern about quality of care.

Since the part-time physicians who work in pris-ons seem, based on their professional characteristics,more likely to provide high quality care than currentfull-time practitioners, and since the fiscal and"environmental" conditions that make full-time pri-son practice unattractive to mainstream providers arenot likely to change soon, perhaps the staffing ofprison health programs should be reconsidered infavor of a much heavier reliance on part-time physi-cians. Although many bureaucratic and financial im-pediments are likely, it would seem, based on con-siderations of physician quality alone, that this is thebest way of assuring inmates of high quality care.Through the use of appropriate recordkeeping andfull-time nonphysician health personnel such asnurses, nurse practitioners, and physician assistants,continuity of care can be maintained without a full-time physician. Coupled with functioning medicalinformation systems, utilization reviews, medicalaudits, and pharmacy profile systems, reliance onpart-time physician staffing might further decreasethe isolation of prison health programs and at thesame time improve their clinical quality.

References ...............................

1. Comptroller General, General Accounting Office: AFederal strategy is needed to help improve medicaland dental care in prisons and jails. GAO No. GGD-78-96. General Accounting Office, Washington, D.C.,Dec. 22, 1978.

2. Lindenauer, M. R., and Harness, J. K.: Care as partof the cure: a historical overview of correctionalhealth care. J Prison Health 1: 56-66 (1981).

3. General Accounting Office: More than money is neededto solve problems faced by State and local correctionsagencies. GAO No. GGD-81-104. General AccountingOffice, Washington, D.C., Sept. 23, 1981.

4. Coste, C.: Prison health care: part of the punishment?New Physician 25: 29-35 (1976).

5. Winner, E. J.: An introduction to the constitutionallaw of prison medical care. J Prison Health 1: 67-84(1981).

6. Comptroller General; General Accounting Office:Women in prison: inequitable treatment requires ac-tion. GAO No. GGD-81-6. General Accounting Office,Washington, D.C., Dec. 10, 1980.

7. Thorburn, K. M.: Croucher's dilemma: should prisonphysicians serve prisons or prisoners? West J Med134: 457-461 (May 1981).

8. King, L. N., and Whitman, S.: Morbidity and mor-tality among prisoners: an epidemiologic review. JPrison Health 1: 7-29 (1981).

9. Lichtenstein, R. L.: Physician job satisfaction and re-tention in prison health programs. PB 82 146101. Na-tion Technical Information Service, Springfield, Va.,1982.

10. Weiss, R. J., et al.: Foreign medical graduates and themedical underground. N Engl J Med 290: 1408-1413(1974).

11. Glandon, G. L., and Shapiro, R. J.: Profile of medicalpractice, 1980. American Medical Assoc., Chicago, Ill.,1980, pp. 140, 148, 149.

12. Snedecor, G. W., and Cochrane, W. G.: Statisticalmethods, Ed. 6, Iowa State University Press, Ames,Iowa, 1967, pp. 209-216.

13. Hays, W. L.: Analyzing qualitative data: chi squaretests. In Statistics. Ed. 3, Holt, Rhinehart and Winston,New York, 1981, pp. 536-566.

14. Donabedian, A.: The definition of quality and ap-proaches to its assessment. Health Administration Press,Ann Arbor, Mich., 1980, pp. 80-85.

15. Palmer, H., and Reilly, M. C.: Individual and institu-tional variables which may serve as indicators of qual-ity of medical care. Med Care 17: 693-717 (1979).

16. Morehead, M. A.: Quality of medical care providedby family physicians as related to their education,training and methods of practice. Health InsurancePlan of New York, May 1958. Mimeographed.

17. Reidel, R. L., and Reidel, D. C.: Practice and per-formance: and assessment of ambulatory care. HealthAdministration Press, Ann Arbor, Mich., 1979, pp.183-198.

18. Peterson, 0. L., et al.: An analytical study of NorthCarolina General Practice 1953-1954. J Med Educ 31(Pt. 2): 49-74, December, 1956.

19. Williams, K. N., and Brook, R. H.: Foreign medicalgraduates and their impact on the quality of medicalcare in the United States. Milbank Mem Fund Q 53:549-581 (1975).

20. Rhee, S.: U.S. medical graduates versus foreign medi-cal graduates: are there performance differences inpractice? Med Care 15: 568-577 (1977).

21. Payne, B. C., and Lyons, T. F.: Method of evaluatingand improving personal medical care quality: episodeof illness study. University of Michigan MedicalSchool, Ann Arbor, 1972, pp. 1-59.

596 Public Health Reports