Determinants of physician antibiotic prescribing behavior: a 3 year cohort study in Portugal.

Objectives Antibiotic misprescription is a major driver of resistance, which is a worldwide public health problem. Therefore, our aim is to assess the influence of the determinants of physician prescribing on the quality of antibiotic use. Methods A 3 year cohort study including all primary-care physicians working in Portugal's Central Regional Health Administration (n = 1094) was conducted. We assessed the determinants of prescribing using a pre-validated, personally addressed, reply-paid, self-administered questionnaire (sent four times to non-responders, between September 2011 and February 2012) designed to collect information on physicians' attitudes to and knowledge of antibiotic prescribing as well as their socio-demographic and professional data. To evaluate antibiotic prescribing, we've calculated ESAC 12 quality indicators per physician per year, allowing us to stratify them as good or poor prescribers according to their performance on those indicators. Associations between determinants and outcomes were fitted with generalized linear mixed models. Results The overall response rate was 46.1%. Emergency activity (OR [95% CI] = 0.29 [0.16-0.54]; p < 0.05) and workload (number of patients seen per day: OR [95% CI] = 0.97 [0.94-1.00]; p < 0.05; number of patients seen per week in emergencies: OR [95% CI] = 0.98 [0.97-0.99]; p < 0.05) were both related to poor quality of antibiotic prescribing. Statistically significant odds ratios were also obtained for ignorance (IqOR [95% CI] = 2.14 [1.31-3.52]), complacency (1/IqOR [95% CI] = 1.19 [1.01-1.41]) and responsibility of others (1/IqOR [95% CI] = 1.78 [1.10-3.06]). Conclusions The above results serve to emphasize workload, working at emergency departments and physicians' attitudes identified as critical factors affecting antibiotic prescribing. This provides new insights for clinicians, researchers and policy makers when it comes to developing and improving the clinical and economic outcomes of antibiotic use. Key limitations of the study included the difficulty of results extrapolation and the limitations of the stratification method based on the antibiotic prescribing quality indicators.