Lawrence E. Klein

Medications and diagnoses in relation to falls in a long-term care facility

The association between falls, drugs, and diagnoses in elderly residents of a long‐term care facility was explored using case‐control methodology. The odds of being a faller rather than a control were significant (P < .01) for those taking antidepressants, sedatives/hypnotics, or vasodilators, and for those with osteoarthritis or depression.

An experiment to change detection and management of mental morbidity in primary care.

A randomized clinical trial was conducted in a group practice for the primary care of adult patients to address the effect of feedback to providers of information from a psychiatric screening questionnaire, the General Health Questionnaire (GHQ). The practice is staffed by faculty, residents, and health care extenders of The Johns Hopkins University School of Medicines Division of Internal Medicine. The patient population was drawn mainly from the inner city community in Baltimore that surrounds the hospital, where the practice is physically based. The GHQ was administered at the time of a regular visit to the practice and results made available to the clinicians for randomly allocated subsamples of their patients. The study results showed that feedback of GHQ information led to only marginal effects on overall detection of mental health problems among the patients in general. However, marked increases in detection occurred among the elderly, blacks, and men, subgroups that ordinarily have relatively low rates of detection of mental morbidity by primary care practitioners. Feedback of GHQ information did not affect management.

Effect of physician tutorials on prescribing patterns of graduate physicians

Pessimism has been expressed in the medical literature as to the efficacy of educational interventions in modifying practice patterns of graduate physicians. As a result, a prospective controlled trial of a specific form of educational intervention, the physician tutorial, was designed to test this belief. Physicians in the experimental group were surveyed to assess their knowledge of the effectiveness, cost and side effects of antibiotics used in the treatment of a preselected index condition, and a tutorial was developed to modify suboptimal prescribing patterns. Antibiotic usage patterns were initially similar for the experimental and control physicians. Prescribing patterns afterwards were statistically different, with the experimental physicians increasing their prescription of the encouraged antibiotics and decreasing that of the discouraged antibiotics. This was accompanied by a statistically significant decrease in direct drug charges. These results suggest that educational programs can be effective in modifying graduate physician prescribing practices.