gms | German Medical Science

Background and aim: Antibiotics are widely used for the treatment of infections in children and the best choice when treating acute bacterial infections. However, due to emerging resistance of bacteria against certain antibiotics a careful consideration of prescription is necessary, in particular, when the cause of infection is unclear. Antibiotic prescription may differ between physician groups. We therefore compared the prescription of antibacterials for systemic use in children between paediatricians and general practitioners.

Material and method: We used the statutory health insurance sample AOK Hesse/KV Hesse, comprising 20% of the whole AOK insurants in Hesse. Overall, 47,033 continuously insured children in 2006 between 2 and 17 years of age were included in the analysis. Prevalences of antibiotic use (at least one prescription) in children by age group and gender as well as antibiotic prescriptions by physician groups were calculated. In an analytic approach, odds ratios for antibiotic prescription were calculated by means of logistic regression adjusted for potential confounder, i.e. age, sex, number of contacts, diagnosis quarter in the year, further diagnoses, comparing general practitioners and paediatricians. In doing so, diagnosis specific models were run. Antibiotic use was defined as at least one prescription in a diagnosis quarter in the respective physician group when a certain diagnosis was given. Diagnoses were identified by ICD-10 code including respiratory (J00–06, J20–22) and urinary tract infections (N30, N39.0), pneumonia (J13–16, J18), otitis media (H65–66) and scarlet fever (A38). Systemic antibiotics were classified according to ATC code J01.

Results: Prevalence of systemic antibiotic use was 44.3% in girls and 41.2% in boys. Gender difference in prevalence was most prominent in the age groups 15–17 with 40.8% and 34.5% in girls and boys, respectively. Betalactams and penicillins (J01C) were the most frequently prescribed antibiotics followed by the group of macrolides, lincosamides and streptogramins (J01F). Mean number of prescriptions were 2 per year in both boys and girls. General practitioners and paediatricians prescribed 82.2% of all antibiotic prescriptions with 43.5% and 38.7%, respectively. In 12.2% of all practice contacts with general practitioners antibiotics were prescribed, in contrast to paediatricians with only 8.5% of all contacts. Significantly decreased odds ratios for antibiotic prescription were observed for paediatricians as compared to general practitioners with OR (95% CI) of 0.46 (0.43–0.50) for respiratory tract infections, 0.42 (0.32–0.56) for urinary tract infections and 0.38 (0.28–0.50) for nonsuppurative otitis media. No significant associations were observed when assessing scarlet fever, pneumonia and suppurative and unspecified otitis media.

Conclusions: In this retrospective analysis of a statutory health insurance sample, paediatricians were associated with a lower odds ratio of prescribing antibiotics in a diagnosis-specific analysis as compared to general practitioners when assessing diagnoses where discretion for antibiotic therapy is given. However, in diagnoses where antibiotic therapy is clearly indicated, i.e. scarlet fever, pneumonia and suppurative otitis media, no differences in prescription by physician group were observed. Reasons for antibiotic prescription differences are speculative and may be due to characteristics of the physician group itself or their patients. We tried to address differences in risk profiles between children treated by paediatricians and general practitioners by adjusting for potential confounders available in routine data. However, residual confounding, e.g. severity of disease, cannot be excluded. Further studies are warranted to verify the present findings and to gain more insight into potentially different treatment strategies between physician groups.