Postmarketing surveillance of adverse drug reactions: A correlational study approach using multiple data sources

The authors performed a correlational study on nationwide spontaneous adverse drug reaction (ADR) reports related to the antioxidant thioctic acid (Thioctacid; ASTA Medical in Germany from April 1992 to March 1995. Thioctacid was predominantly utilized by general practitioners and internists for treatment of diabetic neuropathy. The total number of treated patients was estimated using a nationwide drug prescription database (MediPlus, IMS: 362 general practitioners and internists) and pharmacy drug sales data. All Thioctacid prescriptions in MediPlus were assessed and the mean cumulative dosage/patient/year was calculated. Then, the total number of Thioctacid patients in Germany was estimated: N = nationwide pharmacy sales (kg)/mean cumulative dosage/patient (kg). From April 1992 to March 1995, 78 patients with 112 ADRs were notified on spontaneous reports. There was a decreasing number of ADR reports per year (32 to 21 patients/year). In parallel, the estimated total numbers of treated patients/year also decreased from 426,658 to 304,155 (p < 0.05). Poisson regression models were fitted including incident cases with ADR and the number of patients per year for each of the 3 years. No significant secular trend of ADR (p = 0.91) and no impact of mean cumulative dosages was observed in the models. There were also no changes in the pattern of ADR (e.g. skin reactions, gastrointestinal complaints). This study suggests that a combination of spontaneous reporting systems with prescription drug databases and pharmacy sales data may be a useful tool to perform rapid postmarketing monitoring of ADR. (C) 1998 John Wiley & Sons, Ltd.