The relationship between physician antibiotic prescribing patterns and medical training in the treatment of respiratory infections

Background. The widespread and often inappropriate use of antibiotics in the outpatient setting for treating respiratory infections is recognized as a significant health issue. This problem is one of the contributing factors to the spread of bacterial resistance and can add to the economic costs and adverse, health of the patient. Physician training characteristics may affect antibiotic prescribing practices for treating' respiratory infections. This study analyzed three factors related to physician training: type of medical school (allopathic or osteopathic), specialty (family practice or internal medicine) and years since medical school graduation in relation to antibiotic prescribing patterns in a Medicaid population.;Methods. A random sample of all Medicaid adult ambulatory patient visits for respiratory infections in primary care in the Health Partners of Philadelphia managed care network between May 1999 and June 2000 were reviewed (525 patients seen by 111 physicians). Comparisons of antibiotic prescribing were made based on physician type, specialty, and years since medical school graduation. Logistic regression was used to examine the magnitude of the effect of each factor.;Results. In the logistic regression analyses, the following physician variables were found to be related to the likelihood of inappropriately prescribing (either providing antibiotics when not indicated or not providing antibiotics when indicated) antibiotics at a statistically significant level: (1) M.D.s were less likely to provide the correct treatment than D.O.s (exp(b)=.48, Wald=9.33, p<.0005), (2) family practitioners were less likely to provide the correct treatment than general internists (exp(b)=.52, Wald=7.39, p=.01), and (3) each additional year since graduation reduced the likelihood of prescribing the correct treatment (exp(b)=.94, Wald=21.27, p<.0005).;Conclusions. Many clinicians and researchers have recognized that prescribing an antibiotic to a patient is a complex process that can be influenced by numerous factors. While many articles in the medical literature have suggested that physician specialty training and years of experience influence the prescribing patterns for drugs, they have been inconclusive in identifying consistent differences. This study's findings provide further support for the idea that medical training can influence physician prescribing for antibiotics. While this study was conducted in a Medicaid population and in a managed care plan, the results suggest that greater attention to this relationship may be warranted throughout health care.