Specialty Training and Specialization Among Physicians Who Treat HIV/AIDS in the United States [Landon BE et al. J Gen Intern Med 2002;17:12] This is another report from HCSUS, which is an analysis of patients in care throughout the U.S. The present study is a survey of randomly selected physicians (identified by patients) who agreed to complete a questionnaire with a $25 incentive. There were 379 eligible physicians who completed the survey, representing 72% of those requested. The survey included questions regarding practice, self-assessment of HIV knowledge, and an 11-item test on competence in HIV care. The results showed that caseload correlated with test performance. Specifically, those with 50 or more patients were 9 times more likely to achieve a score of at least 80% on the knowledge scale. The average score was 6.4 for those with 0-19 patients, 8.4 with 20-49 patients, and 8.8 for those with 50 or more patients. About 80% of participants considered themselves "experts" in HIV care, and these physicians reported attendance at an average of 9.3 meetings and 2.3 national HIV CME meetings/year. With regard to the distinction between ID-trained physicians and generalists, the former accounted for 40% of the physician respondents, these physicians had an average HIV caseload of 150 and an average knowledge score of 9.0 on an 11-point scale. The generalists accounted for 56% of respondents, had a median caseload of 200 HIV-infected patients, and a knowledge score of 8.5. On the basis of the national sampling, the authors estimated that about 8.5% of patients under care for HIV infection (approximately 13,115) received care from physicians with less than 20 active cases. The authors conclude that generalists perform comparably to infectious disease trained physicians on an HIV knowledge quiz and that expertise in the HIV field correlates strongly with case load and participation in CME activities. Some of these results are summarized in the following table:

Comment: The results of this study from HCSUS clearly support the impression that most HIV care is provided by physicians with relatively high case loads, supporting the concept that this is a specialty. Care for approximately 92% of the patients sampled, which represents a cross section of HIV care in the U.S., is provided by physicians with at least 20 patients with HIV infection. The results also show that expertise is defined by caseload and CME education, at least in terms of performance in the knowledge-based quiz and self-perception of HIV knowledge. These findings will surprise no one. In fact, qualifications for expertise in HIV care have been defined by caseload and CME courses by DHHS, the IDSA HIV Medicine Association, and the American Academy of HIV Medicine. Perhaps the most controversial part of the study is the assumption that knowledge of HIV can be accurately assessed by 11 questions, and in fact, 11 questions that are not included in the manuscript. These are also questions that were not subject to scientific review and psychometric analysis as is customary with the American Board of Medical Specialists, which has the expertise for such testing. Thus, perhaps a safer conclusion from this study is that about 40% of HIV care in the U.S. is delivered by ID trained physicians, about 56% is delivered by "generalists," and more than 90% of patients receive care by physicians who have substantial case loads and are relatively rigorous in attending the multiple CME courses relevant to this subject.
posted 3/05/2002

back to news table of contents