Compliance of German pediatric patients with oral antibiotic therapy: results of a nationwide survey.

BACKGROUND: Noncompliance with a prescribed therapy is a common problem in ambulatory pediatrics.
OBJECTIVE: To establish a nationwide status quo of compliance of German ambulatory pediatric patients with oral antibiotics prescribed for various bacterial infections. PATIENTS AND METHODS: In this study, organized and financed by the German Society for Pediatric Infectious Diseases, 42 pediatricians in private practice who were selected to represent the 3 main regions of Germany and residence in large cities or small towns, respectively, enrolled consecutive patients who had bacterial infections that required therapy with oral antibiotics. Choice of agent and duration of treatment were left to the study physicians. Compliance was measured by a standardized telephone interview on the penultimate day and a urine bioassay for antibacterial activity on the last day of the planned treatment period. Parents did not know the true purpose of the study.
RESULTS: Five hundred eight-four patients were fully evaluable. The most frequent diagnoses included tonsillopharyngitis (n = 231), otitis media (n = 170) and lower respiratory tract infections (n = 114). Most frequently prescribed antibiotics included amoxicillin (n = 102), potassium penicillin V (n = 81) and clarithromycin (n = 67). Overall compliance (positive urine test) on the last day of therapy was 69.5% (406 of 584 patients). Compliance was not significantly influenced by the region of residence or the underlying bacterial infection. It was significantly associated with the antibiotic used (macrolides, 89.0%; penicillins, 62.2%; cephalosporins, 66.4%; P = 0.0001 for macrolides vs. the others). Best compliance was found with clarithromycin (94.0%) and erythromycin estolate (89.8%). Compliance was also significantly better in patients > or =6 years old (77.7%; P = 0.016); with a treatment duration of < or =7 days (77.6%; P = 0.014); when the drug package contained a dose-taking reminder (79.7%; P = 0.003); and when the pediatrician's behavior toward the patient was assessed by the parents as "very sympathetic" or "sympathetic" (72.6%; P = 0.017). Subjecting all variables to logistic regression analysis, we found 3 variables to be significant predictors of treatment compliance: choice of antibiotic (P = 0.0001); patient age (P = 0.0008); and residence in town or city (P = 0.03).
CONCLUSIONS: A noncompliance rate of >30% is unsatisfactory. Whereas some variables significantly associated with compliance cannot be influenced (patient age; place of residence in town or city), others are amenable to modifications. These include the physician-patient interaction and the choice of antibiotic. Agents should be preferred that are well-accepted by patients, that enable short-term therapy with few daily doses and with a package that contains a dose-taking reminder.