Drug-disease interaction in elderly patients in family practice.

OBJECTIVE To determine the frequency of potential drug-disease interaction in elderly patients in family practice. To assess which drugs and diagnoses are associated with a high risk related to drug-disease interaction and whether there are gender- or age-related differences. METHODS In routinely recorded electronic patient records, patients at least 65 years old with at least one diagnosis named in Beers list and one prescription were identified. Potential drug-disease interaction (PDDI) was presumed if within the same 3 months a "Beers" diagnosis and a potentially inappropriate prescription with respect to this diagnosis were documented for a patient. Multiple logistic regression analysis identified factors associated with a high risk of PDDI. RESULTS Of 24,619 patients (63.4% women) corresponding to our inclusion criteria, 10.4% were exposed to at least one PDDI. Almost no (0.0%) PDDI was associated with the most common Beers disorder hypertension (prevalence 49.2%). However, 23.4% of men suffering from bladder outflow obstruction (prevalence 17.6% in males) were exposed to at least one PDDI. PDDI was quite common in some rarer conditions, for example, indications for anticoagulation (prevalence 2.6%, 31.5% PDDI). PDDI was not influenced by gender, but associated with taking more than 4 drugs (OR 1.91 (1.83 - 2.00)), suffering from more than one Beers disorder (OR 1.24 (1.16 - 1.31)), and advanced age (OR 1.10 (1.05 - 1.15)). CONCLUSIONS High risk patient groups could be identified. Some disorders as well as some drugs are particularly prone to risky constellations; these should be reflected in systems assisting prescribing with regard to patient safety.