Diagnostic Test Ordering in the Evaluation of Febrile Children

Abstract

To assess the independent effects of physician and environmental factors on test ordering. We prospectively studied 6191 consecutive visits by nonadmitted febrile (rectal temperature > or = 38.0 degrees C) children less than 18 years of age to a children's hospital emergency department from March through November 1989. Multiple logistic regression analysis was used to control for the mutually confounding effects of patient, physician, and environmental factors and to assess attending staff-trainee interactions (effect modification). Patients evaluated by hospital-based subspecialists were significantly more likely to undergo tests than patients evaluated by community-based physicians (odds ratio [OR] for undergoing at least one test, 1.13; 95% confidence interval [CI], 1.04 to 1.23; OR for complete blood cell count, 1.28; 95% CI, 1.12 to 1.46). Children seen by physicians with more than 10 years of experience were significantly less likely to undergo tests than those seen by their more junior colleagues, but this effect was modified substantially by trainee presence and level. For example, when children were seen by a physician with more than 10 years of experience and no trainee was involved, the OR for undergoing at least one test was 0.81 (95% CI, 0.73 to 0.91). If the same physician saw the same patient with a junior trainee, the OR for undergoing at least one test was 1.08 (95% CI, 0.95 to 1.24). Patients seen between July and November were significantly more likely to undergo at least one test than those seen between March and June (OR, 1.28; 95% CI, 1.20 to 1.36). Attending staff, trainee, and seasonal effects on test ordering have important implications for febrile children and their families, for clinical training and supervision, and for health care costs.