Published in:
01-08-2003 | Original Research Article
Identification of Adverse Drug Reactions in Geriatric Inpatients Using a Computerised Drug Database
Published in: Drugs & Aging | Issue 10/2003
Login to get accessAbstract
Introduction and objective
Geriatric patients with multiple comorbidities are at high risk of experiencing an adverse drug reaction (ADR) during hospitalisation. The aim of the study was to compare the rate of ADRs as predicted by a computerised pharmacological database to the actual rate determined by direct observation in a sample of geriatric patients.
Study design
During a 4-month period, geriatric patients were monitored using prospective observation. Patients were intensively screened for ADRs by a pharmacoepidemiological team (PET), consisting of two pharmacists and a physician. Actual ADRs detected by the PET were compared with those predicted by a computerised drug database. Furthermore, the set of actual ADRs, which resulted from drug-drug interactions (DDIs), were contrasted with potential DDIs signalled by the database. The main outcome measures were the incidence of actual ADRs. For the detection rate of the database we focused on frequent ADRs (>1% according to product information and database) and all DDIs indicated automatically by the database.
Results
163 patients (121 female), mean age 79.8 ± 7.1 years (range 60–98), were included in the study which was conducted on a geriatric rehabilitation hospital ward. The mean duration of hospitalisation was 24.3 ± 8.4 days. Elderly patients received an average of 14.0 drugs (range 2–35) during their hospital stay.
Of all patients, 60.7% experienced at least one ADR. The PET detected a total of 153 ADRs, with a mean of 0.9 ADRs per patient (range 0–5). The computerised drug database predicted an average of 309 potential ADRs for each patient; however, only 21 ADRs per patient were of high frequency. In 48% of ADR-positive patients (defined by PET) at least one of these frequent ADRs occurred.
DDIs were detected by the PET in 14.7% of patients. Our database indicated a mean of 12 potential DDIs per patient. In 14 out of 24 DDI-positive patients, at least one signal indicated a real DDI. The database sensitivity was consequently 58.3%.
Conclusion
In geriatric patients the incidence of ADRs is high. Computerised drug databases are a useful tool for detecting and avoiding ADRs. Our software, however, also produced a large number of signals that did not relate to actual ADRs found by the PET. The sheer number of these ‘false’ signals shows the need for refinement and optimisation of databases for daily clinical use.