Is the "practice style" hypothesis relevant for general practitioners? An analysis of antibiotics prescription for acute rhinopharyngitis.

Much research in France or abroad has highlighted the medical practice variation (MPV) phenomenon. There is no consensus on the origin of MPV between preference-centered approaches versus opportunities and constraints approaches. This study's main purpose is to assess the relevance of hypotheses which assume that physicians adopt a uniform practice style for their patients for each similar clinical decision in a context of medical decision with low uncertainty and professional practice with weak regulation. Multilevel models are evaluated: first to measure variability of antibiotics prescription by French general practitioners (GPs) for acute rhinopharyngitis regarding clinical guidelines, and to test its significance in order to determine to what extent prescription differences are due to between or within GPs discrepancies; second, to prioritize its determinants, especially those relating to a GP or his/her practice setting environment, while controlling visit or patient confounders. The study was based on 2001 activity data, along with an ad hoc questionnaire, of a sample of 778 GP taken from a panel of 1006 computerized French GPs. We observed that a large part of the total variation was due to intra-physician variability (70%). It is patient characteristics that largely explain the prescription, even if GP or practice setting characteristics (location, level of activity, network participation, continuing medical education) and environmental factors (visit from pharmaceutical sales representatives) also exert considerable influence. This suggests that MPV are partly caused by differences in the type of dissemination of medical information and this may help policy makers to identify and develop facilitators for promoting better use of antibiotics in France and, more generally, for influencing GP practices when it is of interest. Copyright 2010 Elsevier Ltd. All rights reserved.