Adverse drug reactions causing hospitalization can be monitored from computerized medical records and thereby indicate the quality of drug utilization.

Purpose: To use computerized medical records to study the frequency of adverse drug reactions (ADR) as a cause for acute admission to a university hospital.

Methods: Computerized medical records in 168 consecutively acutely admitted cases to a short-term internal medicine ward at a university hospital were retrospectively evaluated to see if an ADR could have caused the admission.

Results: In 18 cases (11%), an ADR was judged to be the reason for acute admission to the hospital. Augmented pharmacological effects (type A reactions) accounted for 89% of the cases. An additional eight cases (5%) were intentional intoxications with prescription drugs. ADR-associated hospital admissions were age-related (average age 72 compared to 65 in patients admitted for other reasons) and also associated with poly-pharmacy (8.3 drugs compared to 5.2). The ADR was considered to arise from a drug-drug interaction in four cases (22%). Although all ADRs fulfilled the criteria for mandatory reporting, none of them were reported to the Swedish Adverse Drug Reactions Advisory Committee (SADRAC).

Conclusions: The observed frequency of ADR-related acute hospital admissions in this pilot study is similar to earlier reported data from Sweden and abroad. Almost all of these reactions were known pharmacological (type A) reactions and many therefore theoretically preventable. In contrast to many prospective and costly studies, this retrospective analysis of computerized medical records offered a simple method that can be used in routine health care and also serve as an indicator of the quality of drug use.