RHEUMATOLOGY TRAINING AT INTERNAL MEDICINE AND FAMILY PRACTICE RESIDENCY PROGRAMS

47 1 SPECIAL ARTICLE RHEUMATOLOGY TRAINING AT INTERNAL MEDICINE AND FAMILY PRACTICE RESIDENCY PROGRAMS DON L. GOLDENBERG, RAPHAEL J. DEHORATIUS, STEPHEN R. KAPLAN, JOHN MASON, ROBERT MEENAN, SUSAN G. PERLMAN, and JOHN B. WINFIELD The Medical Research Education Subcommittee of the American Rheumatism Association surveyed a random selection of large and small programs in internal medicine and family practice residency programs in order to evaluate their rheumatology training. Formal rheumatology training is offered in 90% of these residency programs, but many available positions are not being filled. A full-time staff rheumatologist was present at 69% of large internal medicine programs, 32% of small internal medicine programs, and 11% of family practice programs. The methods of rheumatology training are similar in most programs, although small internal medicine programs and family practice programs more often utilize physicians offices or outside medical centers for the rheumatology elective training. A majority of the directors of these residency programs thought Don L. Goldenberg, MD: Chairman, American Rheumatism Association Medical Research Education Subcommittee, Associate Professor of Medicine, Clinical Program Director, Arthritis Center, Boston University School of Medicine, Boston, Massachusetts; Raphael J. DeHoratius, MD: Professor of Medicine, Director, Division of Clinical Immunology/Rheumatology, Hahnemann University, Philadelphia, Pennsylvania; Stephen R. Kaplan, MD: Professor of Medicine, Brown University Program in Medicine, Providence, Rhode Island; John Mason, MA: Medical Sociologist, Research Associate, Arthritis Center, Boston University School of Medicine; Robert Meenan, MD: Associate Professor of Medicine, Associate Director, Multipurpose Arthritis Center, Boston University School of Medicine; Susan G. Perlman, MD: Assistant Professor of Clinical Medicine, Section of Arthritis and Connective Tissue Diseases, Department of Medicine, Northwestern University, Chicago, Illinois; John B. Winfield, MD: Chairman, American Rheumatism Association Education Committee, Professor of Medicine, Director, Multipurpose Arthritis Center, University of North Caroha School of Medicine, Chapel Hill. Address reprint requests to Don L. Goldenberg, MD, Boston University School of Medicine, Arthritis Center, K5, 71 East Concord Street, Boston, MA 021 18. Submitted for publication June 21, 1984; accepted in revised form October 23. 1984. that many basic skills and techniques were not taught adequately and that the training of their rheumatology residents was not equal to that of residents in cardiology or gastroenterology. Rheumatic diseases affect about 10% of the population (1). Five to ten percent of office visits to general physicians are related to musculoskeletal problems (2), and a small minority of these patients will be referred to physicians with specialized training in these disorders (3). The National Arthritis Plan suggested that every primary and family physician be trained in the diagnosis and treatment of the common musculoskeletal conditions (4). Rheumatology training of generalists occurs to a limited extent in medical school (9, but it is likely that the most intensive training in rheumatology occurs during residencies in internal medicine and family practice. Little is known, however, about the content or process of training in rheumatic diseases at this critical stage of medical education (6-8). Thus, the Medical Education Research Subcommittee of the American Rheumatism Association (ARA) designed and conducted a survey to evaluate the current status of rheumatology education in internal medicine and family practice programs in the United States. The results of this survey provide an overview of house staff training in rheumatology and support conclusions about the general physician s education in rheumatology in the United States. METHODS Seven committee members designed the questionnaire to determine basic information regarding rheumatology education. Rheumatology was not defined in any descriptive fashion, although aspects of rheumatology training that Arthritis and Rheumatism, Vol. 28, No. 4 (April 1985)

472 GOLDENBERG ET AL the committee believed to be important were indirectly conveyed to the respondents. These were specified in a question about the level of training in the following areas of rheumatology: musculoskeletal examination, differential diagnosis of acute arthritis, differential diagnosis of rheumatoid arthritis, management of rheumatoid arthritis, laboratory tests in the rheumatic diseases, synovial fluid analysis, joint aspiration technique, diagnosis and management of tendinitis, general principles of occupational and physical therapy, and psychosocial aspects of chronic disease. The respondents were asked to rate the level of training (good, fair, poor) in each of these areas, and were asked to compare the training of their residents in rheumatology with the training of residents in other medical subspecialty areas such as cardiology or gastroenterology. Program variables were surveyed by questions that could be answered by yes, no, or a numeric value. These questions included: the total number of residents in the program, the total number of residents in a rheumatology elective each year, the percentage of available rheumatology elective slots filled each year, the duration of rheumatology training, specific assignments during the rheumatology training, including inpatient and outpatient consultations and longitudinal care, private office experiences, rheumatology attending rounds, teaching conferences, and orthopedic training. Furthermore, the availability of other electives in orthopedics, physical and rehabilitation medicine, gerontology, cardiology, pulmonary medicine, gastroenterology, nephrology, oncology, hematology, endocrinology, and infectious diseases was surveyed. A description of the training program, including medical school affiliation and faculty rheumatologists, was provided. The specific weekly activities of rheumatology faculty in the following areas were also determined: house officer teaching conferences, inpatient consultations, hospital-based clinic, hospital-based clinical and basic research, and general medical attending responsibilities. The 1,125 internal medicine and family practice programs listed in the 1982-1983 Directory of Residency Programs (9) were initially grouped by the following variables: (a) internal medicine or family practice; (b) large or small (small was defined as fewer than 30 residents; large was defined as equal to or greater than 30 residents total in each residency program); (c) general geographic location. Since there were not many large family practice programs, we later grouped all family practice programs into 1 category and compared them with large internal medicine and small internal medicine programs. Two hundred twenty-four programs were randomly selected to receive the questionnaire. Completed questionnaires were received from 177 programs, a 79% response rate. The questionnaires were answered by the internal medicine or family practice director at 80% of the programs, and the other questionnaires were completed by the rheumatology program director or by an associate director of the residency training program. The number of respondents from the 3 general categories included: 61 large internal medicine programs (35%), 44 small internal medicine programs (25%), and 72 family practice programs (40%). The survey data were analyzed by applying chisquare tests for nominal data or counts and analysis of variance for continuous variables, to look for overall differences across the 3 program types. If a significant difference was found, initial analysis was followed by explicit paired comparisons using chi-square for counts and t-tests for continuous data. All differences reported as significant displayed P values 5 0.01. This conservative significance level was chosen rather than the more traditional 0.05 level, in order to adjust for multiple comparisons. RESULTS General characteristics and faculty of the residency programs. The mean number of house officers in all of the programs surveyed was 36, with a range of 3-225. The mean number of internal medicine house officers was 37, whereas the mean number of family practice house officers was 18. Seventy-three percent of the respondents reported that their program was a major affiliate of a medical school, and another 23% reported that their residency program had some affiliation with a medical school. Eighty-three percent of the large internal medicine programs, 70% of the small internal medicine programs, but only 13% of the family practice programs had a rheumatology department affiliated with their residency program (Table 1). A full-time staff rheumatologist was present at 69% of the large internal medicine programs, 32% of the small internal medicine programs, and 11% of the family practice programs. Each family practice program with a full-time rheumatologist was directly affiliated with a major medical center and an internal medicine training program. Large internal medicine programs were more likely to be affiliated with a rheumatology fellowship program and have more than 2 full-time staff rheumatologists. Overall, 65 residency programs (37% of those surveyed) did not have a full-time staff rheumatologist and 20 (14%) did not have a full-time or a part-time staff rheumatologist. Table 1. Rheumatology faculty manpower % of programs Large Small internal internal Family medicine medicine practice (n = 61) (n = 44) (n = 72) Rheumatology 83 70 13* department Full-time staff 69* 32 I1 rheumatologist >2 full-time staff 38* 9 6 rheumatologists * Significant difference compared with other 2 programs (P < 0.01).

RHEUMATOLOGY TRAINING IN THE U.S. 473 The family practice and small internal medicine programs without a full-time staff rheumatologist provided formal rheumatology education by employing a part-time rheumatologist or by sending their residents to training sites outside their own medical center. Eighteen of the 20 programs that did not have a fulltime or part-time faculty rheumatologist were contacted by telephone in order to determine who provided rheumatology education to their residents. Ten of these programs sent their residents to an affiliated medical center that did have a staff rheumatologist and an existing rheumatology training program. Five of the 18 programs sent their residents to the private practice of a rheumatologist or an ipternist. Three of the 18 programs reported that they provided no formal staffsupervised rheumatology training. The weekly teaching activities of the staff rheumatologists at large internal medicine programs more often included inpatient consultations, didactic teaching conferences, and arthritis clinics, when compared with the teaching activities of rheumatologists at small internal medicine and at family practice programs (Table 2). Less than 20% of the rheumatologists affiliated with family practice and small internal medicine programs participated in weekly clinical or basic research, whereas approximately 70% of rheumatologists at large internal medicine programs were involved in these investigative activities (Table 2). Content of rheumatology residency education. Formal training in rheumatology was available at 92% of all programs surveyed, including 93% of the internal medicine programs and 90% of the family practice programs. Three-fourths of the trainees received their formal rheumatology education during a 1- or 2-month elective. Most respondents reported that their programs provided a balance of inpatiept and ambulatory Table 2. Weekly activities of staff rheumatologists* Large Small internal internal Family medicine medicine practice programs programs programs (n = 61) (n = 44) (n = 72) Teaching conferences 78t 49 52 inpatient consultations 95 88 68t Arthritis clinic 89t 58 38 (hospital-based) Clinical research 76t 20 16 Basic research 66t 17 16 Attending rounds 84 63 49t * % who participate in each of these activities weekly. t Significant difference compared with other 2 programs (P < 0.01). patient training experience, although the family practice programs were more likely to provide only ambulatory training. During any single academic year, approximately one-quarter of residents at large internal medicine. small internal medicine, and family practice programs received some formal rheumatology education. In contrast, the mean percentage of house officers who received an orthopedic elective each year was 45% of residents at family practice program$, 9% at small internal medicine programs, and 4% at large internal medicine programs. Whereas more than 75% of available rheumatology elective positions were filled by residents at more than three-quarters of the large internal medicine programs, only about one-third of the small internal medicine and one-quarter of the family practice programs had filled more than 75% of their available elective positions (Table 3). In contrast, fewer than onequarter of available elective positions were filled at 8% of large internal medicine programs, and at 42% of family practice programs. The specific content of formal rheumatology training at most programs included teaching conferences, inpatient consultations, and attending rounds and ambulatory clinics (Table 4). Compared with the large internal medicine programs, family practice residency programs and small internal medicine programs more often provided some training in a private practice office setting. Family practice programs more often provided specific orthopedic training within their rheumatology elective in comparison with the internal medicine programs, while large internal medicine programs more commonly presented weekly teaching conferences. Perceived training quality. Quality of training in 10 specific educational variables was assessed by the program directors (Table 5). Approximately two-thirds of program directors at all residency programs surveyed considered the training of their residents in the Table 3. filled Percentage of available rheumatology elective positions % of programs Large internal Small internal medicine medicine Family practice (n = 61) (n = 44) (n = 72) <25% 8 18 25-5096 2 24 50-75% 11 20 >75% 79* 38 * Significant difference compared with other 2 programs (P < 0.01). 42 22 9 27

474 GOLDENBERG ET AL Table 4. Rheumatology elective curriculum % of programs offering training variable Large Small internal internal Family medicine medicine practice Training variable (n = 61) (n = 44) (n = 72) Teaching conferences 98* 81 83 Inpatient consultations 97 97 89 Attending rounds 93 86 80 Outpatient clinics 95 86 90 Emergency room 48 50 35 Private office 42* 67 78 Orthopedics 42 19 54 * Significant difference compared with other 2 programs (P < 0.01). general principles of occupational and physical therapy and the psychosocial aspects of chronic disease to be inadequate. However, there were significant differences in the inadequacies perceived by the program directors relative to the type and size of the program. For example, a significantly greater number of directors at family practice and small internal medicine programs, in comparison with those at large internal medicine programs, considered their residents training inadequate in the interpretation of laboratory tests, performance and interpretation of synovial fluid analysis, and the technique of joint aspiration (Table 5). In contrast, the family practice program directors, in comparison with the directors of the large and small internal medicine programs, more often believed that their residents training in the diagnosis and management of tendinitis and the psychosocial aspects of chronic disease was adequate. In general, however, there were significant inadequacies perceived with regard to each important training variable assessed. This perception of inadequacy was especially striking with regard to the basic techniques of a musculoskeletal examination and joint aspiration. The program directors also were asked to assess the perceived adequacy of their rheumatology training programs relative to that of cardiology and gastroenterology. Only 10% of respondents reported that the training received in rheumatology was slightly or much better than that received in cardiology and gastroenterology, whereas 39% of the respondents believed the rheumatology training was slightly or much worse than cardiology and gastroenterology training. When program directors were asked what they considered to be the single most important change in current rheumatology training that would help to improve the rheumatology skills of their residents, 40% of the respondents stated that exposure to more patients with arthritis was most important, 25% believed more general interest in rheumatology throughout the residency program was most important, 19% thought that more emphasis on ambulatory rheumatology was most important, and only 15% stated that a larger rheumatology faculty was most important. Table 5. Perceived inadequacies of rheumatolow house officer training* % of respondents who rated skill inadequate Large internal Small internal Family practice medicine programs medicine programs programs Specific training variable (n = 61) (n = 44) (n = 72) Musculoskeletal examination Differential diagnosis of acute arthritis Differential diagnosis of rheumatoid arthritis Management of rheumatoid arthritis Laboratory tests in the rheumatic diseases Synovial fluid analysis Joint aspiration technique Diagnosis and management of tendinitis General priniciples of occupational and physical therapy Psychosocial aspects of chronic disease 33 12 12 25 15t 25t 33t 53 76 61 * For each training variable listed, the program directors rated their residents education as good, fair, or poor. Inadequate represents the sum of fair and poor responses for each variable. t Significant difference compared with other 2 programs (P < 0.01). 50 23 23 30 34 55 52 57 82 73 39 31 31 43 46 70 61 33t 68 41t

RHEUMATOLOGY TRAINING IN THE U.S. 475 DISCUSSION After evaluating rheumatology education in U.S. medical schools, the Undergraduate Education Subcommittee of the Education Committee of the ARA reported that 10% of medical schools provided no formal rheumatology education and that 13% did not have a rheumatologist on their full-time faculty (5). Furthermore, 22% of the schools did not have a rheumatology fellowship program and 38% had 2 or fewer full-time rheumatologists. Only 62% of the schools provided specific training in musculoskeletal examination. An intense 1- or 2-month rheumatology elective was available to students at most medical schools, but fewer than 20% of the medical students participated in such electives. Three-fourths of the rheumatology program directors at the medical schools believed the educational impact of rheumatology on students was inadequate, primarily because of a lack of significant exposure to rheumatology patients. The current study, the first to evaluate the rheumatology training of residents in internal medicine and family practice programs throughout the United States, found similar educational deficiencies. While some formal rheumatology education is offered in 90% of these residency programs, the availability, utilization, and content of this education varies significantly with specific program characteristics. Those family practice and small internal medicine programs without a full-time staff rheumatologist (Table 1) utilized parttime rheumatologists or sent their residents to other programs for formal rheumatology education. Since formal training in rheumatology is not required, only about one-quarter of residents each year participate, a situation analogous to that in medical student rheumatology education (5). Whereas there are far too few elective positions available in rheumatology at U. S. medical schools (3, there is an underutilization of available electives in rheumatology at the level of residency training, especially in family practice and small internal medicine programs (Table 3). Furthermore, residents were often considered by program directors to be inadequately trained in the techniques of musculoskeletal examination, synovial fluid analysis, and joint aspiration. Ambulatory-oriented skills, such as the diagnosis and management of tendinitis and the psychosocial aspects of chronic disease, were more often judged to be inadequately taught in the internal medicine programs, whereas the interpretation of laboratory tests and synovial fluid analysis were more often judged to be inadequately taught in the family practice programs. In general, the program directors believed their residents were not as well trained in rheumatology as they were in cardiology or gastroenterology. These perceived inadequacies may correlate with the inefficient utilization of available elective training positions in rheumatology. Recent studies have concluded that a formal educational program in rheuniatology is important if basic skills are to be acquired during residency training (6-8). In contrast, significant education may be acquired by osmosis in medical subspecialties such as cardiology and gastroenterology, which are emphasized during general inpatient medical training. Rheumatology training skills, including a familiarity with many common musculoskeletal conditions such as soft tissue rheumatism and osteoarthritis, can best be provided in ambulatory sites over a long time frame. Therefore, residency programs should focus greater emphasis on ambulatory rheumatology training. The Internal Medicine Residency Review Committee recommends that most rotations on subspecialties should be at least six weeks duration (9). However, a 6-week rotation may not be optimal or practical in the provision of basic rheumatology training to most internal medicine residents. The Family Practice Residency Review Committee recommends that the curriculum in internal medicine should include both inpatient and outpatient experiences in medical subspecialties and that the faculty should include both general internists and subspecialists (9). In contrast, 160-200 hours of orthopedic training is required. Thus, formal training in orthopedics is provided for all family practice residents, and in a much greater degree than comparable rheumatology training. The family practice residents may receive substantial education in certain rheumatic conditions during their orthopedic training, although the content of this training is not known. The validity of the information gathered by this survey is subject to the accuracy of the program directors responses. Formal validity testing is not possible in such a survey. However, certain features of the questionnaire design and analysis should have served to increase validity. For example, most questions offered clear-cut dichotomous choices. The fact that our analyses focused on comparisons of program types, e.g., family practice versus internal medicine, tends to balance inaccuracies and reduce threats to validity. Furthermore, we are now in the process of followup site visits to a sample of the responding programs. These site visits will validate the reliability of the program directors responses and, more importantly, will provide an instrument to evaluate the

476 GOLDENBERG ET AL impact of these rheumatology training variables on residents performance. In conclusion, our survey of residency training at family practice and internal medicine programs found that formal training in rheumatology is offered at the great majority of programs throughout the United States, although currently the available educational positions are not fully utilized. Program directors thought that the rheumatology education of their trainees was not adequate in regard to a number of specific variables, and that they would benefit from greater exposure to rheumatology and a greater mastery of basic rheumatology skills. 1. 2. REFERENCES Bennett PH, Wood PHN: Population Studies of the Rheumatic Diseases. Amsterdam, Excerpta Medica, 1968 DeLozier JE, Gagnon RO: National Ambulatory Medical Care Survey. DHEW Publication No. 761772. Rockville, Maryland, US Department of Health, Education and Welfare, 1975 3. Ogryzlo MA: Editorial: specialty of rheumatology-manpower requirements. J Rheumatol 2: 1-4, 1975 4. Engleman EP: The National Arthritis Plan: an overview. Arthritis Rheum 20: 1-6, 1977 5. Goldenberg DL, Mason JH, DeHoratius R, Goldberg V, Kaplan SR, Keiser H, Lockshin MD, Rynes R, Sandson JI, Schumacher HR, Skosey J: Rheumatology education in United States medical schools. Arthritis Rheum 24:1561-1566, 1981 6. Goldenberg DL, Meenan RF, Allaire S, Cohen AS: The educational impact of a rheumatology elective. Arthritis Rheum 26:658-663, 1983 7. Eyanson S, Brandt KD: Some effects on houseofficers of an elective rheumatology rotation. J Rheumatol 7:25 1-257, 1980 8. Strosberg MA, Strosberg JM: Do family practice residents in the United States learn clinical rheumatology? An important question. J Rheumatol7:923-926, 1980 9. 1982-1983 Directory of Residency Training Programs. Chicago, American Medical Association, 1982