Aim: The authors wanted to identify the factors associated with prescription or non-prescription of antibiotics by general practitioners (GPs), for three frequent infectious diseases: tonsillitis, acute bronchitis (AB) and exacerbations of chronic bronchitis (ECB).
Patients and methods: A cross-sectional study included 574 patients followed by 58 randomly selected GPs in Eastern France. Data on physicians, patients (clinical, social and environmental) and antibiotic treatment (type, motivations) was collected during the consultation via a standardized questionnaire.
Results: Official recommendations for tonsillitis, AB and ECB were known by GPs in 100, 96 and 91% of the cases, respectively. An antibiotic was administered in 61% of tonsillitis cases, 64% of AB and 84% of ECB. The correlation between prescribing or not antibiotics and official recommendations was variable (tonsillitis: 81.3%, ECB 89.6%, AB: 36.5%). Clinical criteria were the main factors influencing antibiotic prescription. A rapid antigen detection test for streptococcus A was performed in 65% of tonsillitis cases (positive in 55%) increased conformity with recommendations to 93.7%. A second consultation, poor tolerance and patient demand were also significantly associated with a higher rate of antibiotic prescription.
Conclusions: Recommendations were known, but semiologic definitions and nosologic limits varied from one physician to another. Recommendations must be based on routine medical practice surveys to include factors influencing prescription, even if non significant. Paraclinical tests help in therapeutic decision making and in the non-prescription of antibiotics but such tests need to be more commonly used.