Trends in prescribing antibiotics for children in Dutch general practice.

OBJECTIVE To assess changes in antibiotic prescribing patterns for children between 1987 and 2001, and to identify general practice characteristics associated with higher antibiotic prescribing rates. METHODS Cross-sectional national survey of Dutch general practice in 1987 and 2001. Data were used for all children aged 0-17 years; 86 577 children in 103 participating practices in 1987, and 76 010 children in 90 participating practices in 2001. Population-based, contact-based and disease-based antibiotic prescription rates were evaluated by age, gender and diagnosis. Practice characteristics associated with inappropriate broad-spectrum antibiotic prescription were identified. RESULTS Population-based prescription rates decreased from 300/1000 children (95% CI, 292-307) in 1987 to 232/1000 children in 2001 (95% CI, 228-235). In 1987, the contact-based prescription rate was 108/1000 contacts (95% CI, 106-111) and this was somewhat similar in 2001: 103/1000 contacts (95% CI, 101-105). In 2001, increased disease-based prescription rates were observed for acute otitis media, acute bronchitis, acute upper airway infections, acute tonsillitis and cough. Overall, non-recommended broad-spectrum antibiotics were prescribed more often in 2001 than in 1987 (87% in 1987 versus 90% in 2001, P < 0.001). Adjusted for other practice characteristics, general practitioners in single-handed practices prescribed 58% more broad-spectrum antibiotics inappropriately for upper airway infections than general practitioners in group practices. CONCLUSION Antibiotic prescribing in children is still relatively low in the Netherlands. However, the prescription of broad-spectrum antibiotics for inappropriate diagnoses has increased, an unfavourable trend given the emerging bacterial resistance. Single-handed practices should especially be targeted to improve antibiotic prescribing in children.