Work satisfaction and career aspirations of internists working in teaching hospital group practices

This paper presents data on the characteristics, work activities, job-related stress, work satisfaction, and career aspirations of 150 faculty and 595 housestaff physicians who regularly provide continuous primary care in 15 teaching hospital-based group practices. The faculty were young, board-certified generalists; they had been recruited from local training programs and spent the majority of their time seeing patients and supervising housestaff. Job satisfaction among faculty and housestaff was generally high. Dissatisfaction occurred most often with aspects of work over which physicians had little control. Although work-related stress was common, it was not related to job satisfaction. Compared with housestaff in traditional residency programs, housestaff enrolled in special Primary Care Training Programs reported significantly greater job satisfaction. For all housestaff, satisfaction with work in the group practice was consistently associated with decreased interest in subspecialty training. assisted in preparing this report. Received from the Department of Medicine and the School of Public Health, UCLA Center for the Health Sciences. Los Angeles. California. Supported by Grant #59082 from the Robert Wood Johnson Foundation. The views expressed herein do not necessarily represent those of the Robert Wood Johnson Foundation.     

Health of homeless women with recent experience of rape

There is limited understanding of the physical health, mental health, and substance use or abuse correlates of sexual violence against homeless women. This study documents the association of rape with health and substance use or abuse characteristics reported by a probability sample of 974 homeless women in Los Angeles. Controlling for potential confounders, women who reported rape fared worse than those who did not on every physical and mental health measure and were also more likely to have used and abused drugs other than alcohol. Results should serve to alert clinicians about groups of homeless women who may benefit from rape screening and treatment interventions. Presented, in part, at the U.S. Substance Abuse and Mental Health Services Administration (SAMHSA) Second National Conference on Women in Los Angeles, Calif, June 1999. The research presented here was supported by a grant to Suzanne Wenzel from The Commonwealth Fund, and by a grant to Lillian Gelberg from the Agency for Health Care Policy and Research. Dr. Gelberg is a Robert Wood Johnson Foundation Generalist Physician Faculty Scholar.     

The relative risk of myocardial infarction in patients who have high blood pressure and non-cardiac pain

The authors conducted a population-based case-control study to determine the risk of myocardial infarction in patients who reported angina-like symptoms. The cases studied were those of patients who had high blood pressure and had sought treatment in 1984 with myocardial infarction as the first manifestation of coronary artery disease. Controls, a random sample of patients who had hypertension, were frequency-matched to cases by age and gender. Blind to case-control status, the authors reviewed the medical records of the 32 cases and 64 controls for reports of angina-like symptoms. While controls reported such symptoms at a constant rate, the events for the cases clustered near their infarctions. When a patient with hypertension sought medical advice for angina-like symptoms, the risk of infarction within 30 days was 14.2 (95% confidence interval, 2.8 to 71), and after 30 days it fell to 1.03. Among patients who have high blood pressure but no history of angina, presentations with prodromal symptoms in the primary care setting are so common that only about one in 100 such visits actually heralds myocardial infarction. Presented at the National Meeting of the Robert Wood Johnson Clinical Scholars Program, Scottsdale, AZ (October 1986) and the National Meeting of the American Federation for Clinical Research, San Diego, CA (May 1987). Supported in part by a grant from the Robert Wood Johnson Foundation, Princeton, NJ, and by the Health Services Research and Development Program, Veterans Administration Medical Center, Seattle, WA. 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 or the VA Medical Center.     

End-of-life decision making

OBJECTIVE: To identify the desired features of end-of-life medical decision making from the perspective of elderly individuals. DESIGN: Qualitative study using in-depth interviews and analysis from a phenomenologic perspective. SETTING: A senior center and a multilevel retirement community in Los Angeles. PARTICIPANTS: Twenty-one elderly informants (mean age 83 years) representing a spectrum of functional status and prior experiences with end-of-life decision making. MAIN RESULTS: Informants were concerned primarily with the outcomes of serious illness rather than the medical interventions that might be used, and defined treatments as desirable to the extent they could return the patient to his or her valued life activities. Advanced age was a relevant consideration in decision making, guided by concerns about personal losses and the meaning of having lived a “full life.” Decisionmaking authority was granted both to physicians (for their technical expertise) and family members (for their concern for the patient’s interests), and shifted from physician to family as the patient’s prognosis for functional recovery became grim. Expressions of care, both by patients and family members, were often important contributors to end-of-life treatment decisions. CONCLUSIONS: These findings suggest that advance directives and physician-patient discussions that focus on acceptable health states and valued life activities may be better suited to patients’ end-of-life care goals than those that focus on specific medical interventions, such as cardiopulmonary resuscitation. We propose a model of collaborative surrogate decision making by families and physicians that encourages physicians to assume responsibility for recommending treatment plans, including the provision or withholding of specific life-sustaining treatments, when such recommendations are consistent with patients’ and families’ goals for care. This research was partially funded by the Robert Wood Johnson Clinical Scholars Program. The views expressed above are those of the authors and do not necessarily reflect those of the Robert Wood Johnson Foundation.     

Measuring the Quality of Colorectal Cancer Screening: The Importance of Follow-Up

Purpose As evidence mounts for effectiveness, an increasing proportion of the United States population undergoes colorectal cancer screening. However, relatively little is known about rates of follow-up after abnormal results from initial screening tests. This study examines patterns of colorectal cancer screening and follow-up within the nation's largest integrated health care system: the Veterans Health Administration. Methods We obtained information about patients who received colorectal cancer screening in the Veterans Health Administration from an existing quality improvement program and from the Veterans Health Administration's electronic medical record. Linking these data, we analyzed receipt of screening and follow-up testing after a positive fecal occult blood test. Results A total of 39,870 patients met criteria for colorectal cancer screening; of these 61 percent were screened. Screening was more likely in patients aged 70 to 80 years than in those younger or older. Female gender (relative risk, 0.92; 95 percent confidence interval, 0.9–0.95), Black race (relative risk, 0.92; 95 percent confidence interval, 0.89–0.96), lower income, and infrequent primary care visits were associated with lower likelihood of screening. Of those patients with a positive fecal occult blood test (n = 313), 59 percent received a follow-up barium enema or colonoscopy. Patient-level factors did not predict receipt of a follow-up test. Conclusions The Veterans Health Administration rates for colorectal cancer screening are significantly higher than the national average. However, 41 percent of patients with positive fecal occult blood tests failed to receive follow-up testing. Efforts to measure the quality of colorectal cancer screening programs should focus on the entire diagnostic process. Supported by the Robert Wood Johnson Foundation Clinical Scholars Program (Etzioni), Department of Veterans Affairs (VA) Health Services Research and Development (HSR&D), and National Cancer Institute Colorectal Cancer (CRC) Quality Enhancement Research Initiative (QUERI) Service Directed Research (Project # CRS 02-163). The views expressed in this article are those of the authors and do not necessarily represent the views of the Robert Wood Johnson Foundation or the Department of Veterans Affairs.     

The association between hypertension treatment,control, and functional status

The authors examined the relationship between hypertension treatment, control, and functional status among 356 “uncomplicated” hypertensive patients receiving care in 16 teaching hospital group ractices. Antihypertensive drug therapy and blood pressure control were determined from a medical record review. Functional status (health perceptions, mental health, role, and physical functioning) was assessed with a questionnaire. After adjustment for potential confounders, hypertensive patients without drug therapy were less likely to have impairment in mental health functioning, compared with patients receiving one or more than one antihypertensive medication (9% versus 25% and 20%, respectively, p<0.05). However, uncontrolled hypertensive patients were more likely to have role limitations than patients controlled only at the end or throughout the record review period (51% versus 39% and 36%, respectively, p<0.05). Patients controlled throughout the review period had the least impairment for each measure of functional status. These preliminary findings suggest that pharmacologic therapy may have a negative influence on the mental health of “uncomplicated” hypertensive patients, but that the dual goals of blood pressure control and positive functional status are not incompatible. Presented in part at the American Federation for Clinical Research Annual Meeting, May 1986. Supported by a grant from the Robert Wood Johnson Foundation. The conclusions are those of the authors and do not necessarily reflect the opinion of the Robert Wood Johnson Foundation or the Rand Corporation. Dr. Siscovick was a Teaching and Research Scholar of the American College of Physicians and an NHLBI Preventive Cardiology Academic Awardee.     

A model for predicting the counseling practices of physicians

The authors present a conceptual model of the determinants of the counseling practices of physicians and an empirical test of the model. Seventy-six per cent of a 50% random sample of physicians in a western county medical society completed a questionnaire (n=151). This instrument measures the aggressiveness, the indications and techniques used by physicians in counseling patients about smoking, exercise, weight control and alcohol use. The independent variables assessed by this instrument are motivations, perceived skills and barriers, medical specialty, and personal health habits. Significant associations were found between the counseling practices reported and physicians’ personal health habits, attitudes and specialties. Non-surgeons counseled more patients, counseled more intensively, and used a greater variety of techniques than surgeons and obstetrician-gynecologists. In general, physicians who had poor health habits did not fully counsel patients about those habits; however, physicians attempting to improve poor habits counseled patients significantly more often than physicians who were not trying to change their own behavior. Health maintenance efforts among physicians may have a multiplier effect. Received from the Departments of Medicine and Psychiatry, UCLA School of Medicine, and the Rand Corporation, Santa Monica, California. Supported in part by a grant from the Robert Wood Johnson Foundation. The opinions and conclusions expressed in the text are those of the authors and do not necessarily represent the views of the Foundation.     

Chagas Disease Awareness among Latin American Immigrants Living in Los Angeles,California

Approximately 300,000 persons have Chagas disease in the United States, although almost all persons acquired the disease in Latin America. We examined awareness of Chagas disease among Latin American immigrants living in Los Angeles, California. We surveyed 2,677 persons (age range = 18–60 years) in Los Angeles who resided in Latin America for at least six months. A total of 62% of the participants recalled seeing triatomines in Latin America, and 27% of the participants reported triatomine bites at least once per year while living abroad. A total of 86% of the participants had never heard of Chagas disease. Of persons who had heard of Chagas disease, 81% believed that it was not serious. More than 95% of those who had heard of Chagas disease would want to be tested and treated. Most Latin American immigrants living in Los Angeles recalled exposure to vectors of Chagas disease. However, they have little knowledge of this disease. Increasing awareness of Chagas disease is needed in this high-risk population.     

The value of serum iron studies as a test for iron-deficiency anemia in a county hospital

To evaluate the performance of serum iron studies as a diagnostic test for iron-deficiency anemia in a county hospital, the authors identified retrospectively all general medicine patients who had had bone-marrow aspirates for the work-up of non-macrocytic anemias from 1978 through 1983. Re-reading a sample of aspirates from the 254 study patients (42 with iron deficiency) verified the presence of absence of iron. Analysis with logistic regression, likelihood ratios, and receiver operating characteristic curves demonstrated that the total iron-binding capacity (TIBC) performed markedly better as a diagnostic test than did the transferrin saturation test. While no single TIBC level was diagnostic, the TIBC provided a good estimate of the probability of iron-deficiency anemia. Presented at the national meeting of the American Federation for Clinical Research, Washington, D.C., May 4, 1985. Supported in part by a grant from the Robert Wood Johnson Foundation, Princeton, N.J.; and by the Health Services Research and Development Program, VA Medical Center, Seattle, and by the National Center for Health Services Research, DHHS, under research grant numbers HS-04080 HS-04996. 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 VA Medical Center, or the National Center for Health Services Research.     

Amebic abscess of the liver

Fifteen patients with amebic abscess of the liver were seen at two California county hospitals over a period of 3 1/2 years. Of the 15 patients, 9 diagnoses were proven and 6 were presumptive. Fourteen of the patients had made a recent trip or had ready access to Mexico. The clinical finding of particular note was the high incidence of right lower lung field abnormalities. Useful laboratory studies included the combination of an elevated BSP, alkaline phosphatase, and direct reacting bilirubin. The value of the liver scan is emphasized. The high incidence of secondary infection of the abscesses is stressed. The uncomplicated postoperative courses of those patients treated by surgical drainage and postoperative chemotherapy for amebiasis is of particular importance since this approach has been considered in the past to be most hazardous.The author is indebted to Robert A. O'Reilly, M.D. for his assistance and helpful suggestions. The indirect hemagglutination tests were performed through the courtesy of Dr. John F. Kessel, School of Public Health, University of California, Los Angeles, Calif.     

Can a specialty society educate its members to think differently about clinical decisions?

OBJECTIVE: Measure the effect of specialty society-developed continuing medical education (CME) on clinical decision making. DESIGN: Randomized controlled trial. SETTING: National sample of neurologists. PARTICIPANTS: Of 492 neurologists randomly selected from an ongoing American Academy of Neurology CME program, 248 were randomized to receive a mailed CME course, and 244 did not receive it. INTERVENTION: A mailed educational course on movement disorders, developed by the specialty society, containing information on diseases and practice recommendations with illustrative case presentations. MEASUREMENTS AND MAIN RESULTS: We assessed adherence to 16 practice recommendations on disease detection, diagnostic test use, and treatments by mailed survey sent to all subjects 4.5 months after the intervention group received the course (73% response rate). The survey contained detailed clinical scenarios to measure self-reported clinical decision making and short open-ended questions to measure factual knowledge. More intervention participants (up to 2.6 times more) than control subjects reported clinical decision making adherent to 9 of the 16 recommendations (p<.05). For 4 of the other 7 recommendations, adherence exceeded 85% in both groups. Within the intervention group, neurologists who read the educational course were 2 to 6 times more likely to be adherent than neurologists who did not. The intervention group had better factual knowledge than control subjects in six of seven areas (p<.01). CONCLUSIONS: This educational course improved neurologists’ reported decision making. Specialty society-developed CME that utilizes a similar format may enhance the effectiveness of mailed CME information to improve physicians’ approach to clinical decisions. Presented at the Robert Wood Johnson Clinical Scholars annual meeting, November 1995, and VA Health Services Research and Development annual meeting, February 1996. Supported by an unrestricted grant from the American Academy of Neurology and by the Office of Research and Development, Health Services Research & Development Program, Center for the Study of Healthcare Provider Behavior. Dr. Gifford received additional support from the Robert Wood Johnson Clinical Scholars Program and from the Bureau of Health Professions, Mid-Career Faculty Training Program in Geriatric Medicine and Dentistry, grant 5D31AH99000-08. Dr. Vickrey received additional support from a Clinical Investigator Development Award from NINDS (K08NS0 1669-02). Opinions are those of the authors and do not necessarily reflect the views of the sponsoring institutions, the Department of Veterans Affairs, the University of California, Los Angeles, or RAND.     

Clinical and laboratory findings on admission to an alcohol detoxification service.

A review of 260 consecutive admissions for alcoholic withdrawal to the Alcohol Detoxification Service of the Los Angeles County--University of Southern California Medical Center shows that 98% of the patients fulfill the criteria for alcoholism as stated by the National Council of Alcoholism Criteria Committee. Previous alcoholic history, current symptoms, physical findings, and laboratory tests indicate frequent multiple organ dysfunction in acute alcohol withdrawal. Hallucinations occurrred in more than 46% of patients with a blood alcohol over 100 mg%. More than 90% of the patients had one or more abnormal laboratory values, and the serum gamma glutamyl transpeptidase was the most commonly elevated laboratory test.     

Performance of cancer screening in a university general internal medicine practice

The authors evaluated use of seven cancer screening tests by 52 providers in a university general internal medicine practice, using 1980 American Cancer Society (ACS) recommendations as standards for comparison. Performance rates were determined by retrospective medical record reviews of a stratified random sample of 525 patients. In addition, the 48 physicians and four nurse-practitioners in the practice were interviewed to determine their opinions, knowledge and perceived use of the tests. Performance rates were low, significantly below the ACS quidelines for all tests except Pap smear. Providers used the tests significantly more often to evaluate patients with cancer risk factors or for new patients. They significantly overestimated their own performances of six tests. More than a fourth of the providers disagreed with the use of mammography, sigmoidoscopy, pelvic or rectal examinations for screening asymptomatic adults. Their knowledge about cancer screening and the ACS recommendations was highly variable, and frequently quite limited. Providers offered four major reasons for not performing the screening tests: provider forgetfulness, lack of time, inconvenience and logistical difficulties, and patient discomfort or refusal. Received from the Division of General Internal Medicine, Department of Medicine, and Institute for Health Policy Studies, University of California, San Francisco, San Francisco, California. Presented in part at the 111th annual meeting of the American Public Health Association, Dallas, Texas, November 1983, and at the 8th annual meeting of the Society for Research and Education in Primary Care Internal Medicine, Washington, D.C., May 1985. Supported by the Robert Wood Johnson Foundation and by PHS grant number 5 R01 CA37340, awarded by the National Cancer Institute, DHHS. Steven Schroeder, MD, Bernard Lo, MD, Steven Cummings, MD, Thomas Coates, PhD, Eliseo Perez-Stable, MD, Hal Luft, PhD, and Joyce Bird, PhD, provided helpful comments. The faculty and residents of the Division of General Internal Medicine and the General Internal Medicine Group Practice at the University of California, San Francisco, provided their interest and cooperation. Richard Meltzer, Matthew Boone, MD, and William Cunningham, MD, assisted in obtaining data.     

Lumbar spine films in primary care

Low back pain (LBP) often prompts radiography, although the diagnostic yield of lumbar spine films is low, and many radiographic abnormalities are unrelated to symptoms. Criteria have been proposed for selective x-ray use, but their value and safety are uncertain. To evaluate these criteria, the authors prospectively studied 621 walk-in patients with LBP. The yield of explanatory x-ray findings was over three times greater among patients with indications for radiography than among those without. Furthermore, an indication for x-rays existed for all patients found to have a malignancy, and for 13 of 14 patients with an identified fracture. Actual physician ordering, however, did not correspond well with the recommended indications. Application of selective criteria appears safe and may improve the yield of useful findings. It may not, however, reduce x-ray utilization from current levels without further refinement in the criteria. Received from the Division of General Internal Medicine, Department of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, Texas. Presented in part at the 48th Annual Scientific Meeting of the American Rheumatism Association, Minneapolis, June 8, 1984. Published in abstract form in Arthritis and Rheumatism 1984;27(No. 4. Supplement): p. S39. 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.     

Ability of primary care physicians to diagnose and managePneumocystis carinii pneumonia

This study assesses the ability of primary care physicians to diagnose and managePneumocystis carinii pneumonia (PCP) in a standardized patient (SP) with unidentified HIV infection. One hundred thirty-four primary care physicians from five Northwest states saw an SP with unidentified HIV infection who presented with symptoms, chest radiograph, and arterial blood gas results classic for PCP. Seventy-seven percent of the physicians included PCP in their differential diagnoses and 71% identified the SP’s HIV risk. However, only a minority of the physicians indicated that they would initiate an appropriate diagnostic evaluation or appropriate therapy: 47% ordered a diagnostic test for PCP, 31% initiated an antibiotic appropriate for PCP, and 12% initiated an adequate dose of trimethoprim— sulfamethoxazole. Only 6% of the physicians initiated adjunctive prednisone therapy, even though prednisone was indicated because of the blood gas result. These findings suggest significant delay in diagnosis and treatment had these physicians been treating an actual patient with PCP. Presented at the International Conference on AIDS, Berlin, Germany, June 6 –11, 1993. Supported by grant number HS 06454-03 from the Agency for Health Care Policy and Research. Dr. Curtis is funded by the Robert Wood Johnson Clinical Scholars Program. The views expressed herein are those of the authors and are not necessarily the views of the Agency for Health Care Policy and Research or the Robert Wood Johnson Foundation.     

Surveillance of listeriosis in Los Angeles County, 1985-1986. A first year's report

After a large food-borne outbreak of listeriosis in Los Angeles County, California, in 1985, the California State Department of Health Services instituted mandatory reporting of Listeria monocytogenes by clinical laboratories. From September 1, 1985, through August 31, 1986, 94 cases of listeriosis were reported in Los Angeles County for an annual crude incidence rate of 12 cases per million persons. Of the 94 cases, 37 (39%) were in neonates and/or their mothers and 57 (61%) were nonperinatal. The overall case fatality rate was 31%, with a perinatal case fatality of 16% (6 fetal and 23 nonperinatal); this compares with an epidemic perinatal case fatality rate of 32%. No significant differences were observed in age-adjusted, race-specific incidence rates among nonperinatal cases or race-specific incidence rates among perinatal cases. All but 2 of the nonperinatal patients had a known predisposing risk factor for the development of listeriosis, the most common of which was a prior history of steroid therapy. A clustering of cases was not identified. No common food sources were apparent. Patients presenting as perinatal cases were more likely to have ingested Mexican-style cheese, ice cream, and yogurt than those presenting as nonperinatal cases. Improved case ascertainment through mandatory reporting and laboratory-based surveillance will establish meaningful baseline levels of listeriosis.     

Transforming Clinical Practice to Eliminate Racial–Ethnic Disparities in Healthcare

Racial–ethnic minorities receive lower quality and intensity of health care compared with whites across a wide range of preventive, diagnostic, and therapeutic services and disease entities. These disparities in health care contribute to continuing racial–ethnic disparities in the burden of illness and death. Several national medical organizations and the Institute of Medicine have issued position papers and recommendations for the elimination of health care disparities. However, physicians in practice are often at a loss for how to translate these principles and recommendations into specific interventions in their own clinical practices. This paper serves as a blueprint for translating principles for the elimination of racial–ethnic disparities in health care into specific actions that are relevant for individual clinical practices. We describe what is known about reducing racial–ethnic disparities in clinical practice and make recommendations for how clinician leaders can apply this evidence to transform their own practices. Funding: Drs. Washington (#RCD-00-017), Saha (#RCD-00-028), and Moody (#RCD-03-183) are supported by grants from the Department of Veterans Affairs, Health Services Research and Development Service. Dr. Saha is supported by a Generalist Physician Faculty Scholar award from the Robert Wood Johnson Foundation. Drs. Horowitz (#P60 MD00270) and Brown (#P20MD00148) are supported by grants from the National Center on Minority Health and Health Disparities. Dr. Brown also received support from the University of California, Los Angeles, Resource Center in Minority Aging Research (#AG02004) and the Beeson Career Development Award (#AG26748). Dr. Cooper is supported by a grant from the National Heart, Lung, and Blood Institute (K24HL083113).     

Changing sources of care for HIV infection in California.

A telephone survey was conducted of random samples of primary care physicians in Los Angeles County and in the 26 rural counties of California to assess changes in medical practices that may have occurred with the development of new pharmacological agents and recent changes in the management of infection with the HIV. Seventy-two to seventy-three percent of the physicians selected participated in the survey in both areas. The results indicate a reduction by almost two thirds of the number of primary care physicians who plan to continue to provide care to HIV-infected patients, both in Los Angeles County and the nonmetropolitan counties of California. Although some were no longer caring for HIV patients because of deaths or patient relocation, the overwhelming majority had referred their patients to infectious disease specialists and HIV clinics because of the growing complexity of the management of these patients.     

Functional disability screening of ambulatory patients

The authors conducted a randomized controlled trial of functional disability screening in a hospital-based internal medicine group practice. They assigned 60 physicians and 497 of their patients to either an experimental or a control group. Every four months the patients in both groups completed a self-administered questionnaire measuring physical, psychological, and social function. The experimental group physicians received reports summarizing their patients’ responses; the control group physicians received no report. At the end of one year the authors found no significant difference between the patients of the experimental and control group physicians on any measure of functional status. Functional disability screening alone does not improve patient function. Supported by a grant from the Robert Wood Johnson Foundation, Princeton, New Jersey.     

Proficiency of internal medicine residents in outpatient care of patients with HIV infection

To determine internal medicine residents’ knowledge of HIV care, the authors conducted a survey of residents from four internal medicine programs in the San Francisco Bay area. On a knowledge test, the mean score was 42.4/55, 77% correct. The residents performed relatively worse on questions regarding didanosine and zalcitabine, tuberculosis prophylaxis, and risk of cervical neoplasia in HIV-infected women. Predictors of greater knowledge were specific residency program, higher postgraduate year, primary care residency track, and more extensive HIV experience. Primary care internal medicine residencies and programs with more exposure to HIV patients are most effective in producing knowledgeable residents. Supported by the AIDS Clinical Research Center, University of California San Francisco, VA Medical Center (141A), 4150 Clement Street, San Francisco, CA 94121; and National Institute of Mental Health Grant: MH44045 (Dr. Cooke). Dr. Bindman is a Robert Wood Johnson Generalist Physician Faculty Scholar. This research was conducted, in part, while Dr. Schultz was a fellow in general internal medicine and clinical epidemiology in the Division of General Internal Medicine, San Francisco General Hospital, San Francisco, California.