Antibiotics in Dutch General Practice: electronic GP databases and national reimbursement data.

Background. A variety of databases such as data from registration forms, electronic patient records and claims data of health insurance companies, are used in evaluation studies on antimicrobial management in general practice. Aim. To assess to which degree prescribing figures for systemic antibiotics from a national electronic GP database correspond with nationwide reimbursement figures, and to investigate for which indications antibiotics are prescribed. Design and Methods. In a cross-sectional design the number of prescriptions from the Second Dutch National Survey of General Practice (DNSGP-2) (90 general practices) were compared with the claims data from the Dutch Drug Information System / Health Care Insurance Board (GIP/CVZ) by calculating the number of prescriptions per 1000 patients/inhabitants per year prescribed by GPs in both sources. The volumes were compared for the total number of antibiotic prescriptions and according to subgroups. Indications for which GPs prescribed antibiotics were described. Results. In reimbursement data, 347 antibiotic prescriptions per 1000 inhabitants were prescribed by GPs, while DNSGP-2 showed 263 prescriptions per 1000 patients with relatively high volumes of sulphonamides & trimethoprim, and small volumes of macrolides and quinolones. About half of the prescriptions were prescribed for respiratory diseases, a quarter for urinary diseases and 7% for ear diseases. Conclusions. GPs participating in research networks prescribe less antibiotics than Dutch GPs in general, and they are cautious in prescribing newer and more broad-spectrum antibiotics. The results support the value of prescription rates from GP patient records as outcome measures in intervention and correlational studies