Frequency and determinants of potential drug‐drug interactions in an elderly population receiving regular home visits by GPs ‐ results of the home medication review in the AGnES‐studies

Abstract

There is limited knowledge about prevalence and determinants of potential drug-drug-interactions (pDDI) in the ambulatory health care setting. In this manuscript we analysed the prevalence and determinants of pDDI in the AGnES home visit population (GP-supporting, community-based, e-health-assisted, systemic intervention). 779 Home-dwelling patients received an IT-supported home medication review (HMR). The interaction monographs of the ABDA-database were used to identify pDDI. A binary multivariate logistic regression model was used to analyse determinants for occurrence of moderate and serious pDDI, respectively. Patients (mean age f: 79.8 yrs; m: 76.2 yrs) took a mean number of 6.8 active substances (SD = 3.3) regularly. 626 patients (80.4%) with an HMR had at least one pDDI (mean = 4.8; SD = 4.4). In 454 patients (58.3%) we found at least one pDDI with moderate or serious relevance (mean = 2.9; SD = 2.8). The most frequent moderate pDDI was the combination of potassium excretion diuretics (e.g. hydrochlorothiazide) with anti-inflammatory drugs (e.g. ibuprofen; n = 258 patients). Adjusted for age and gender, multiple binary logistic regression showed significant results for the number of taken active substances (continuous variable; OR = 1.48; 95%-CI 1.382-1.585), metabolic diseases (OR = 1.52; 95%-CI 1.039-2.223), a diagnose of a muscular-skeletal disease (OR = 1.741; 95%-CI 1.204-2.517), infectious disease (OR = 0.127; 95%-CI 0.021-0.783), and gastro-intestinal disease (OR = 0.538; 95%-CI 0.322-0.899). Using a comprehensive, computer-assisted HMR in an ambulatory care setting we have identified a high proportion of pDDI of moderate or serious clinical relevance. These pDDI require an intervention from intensified monitoring to a change in medication. Further investigations should focus on clinical outcome of pDDI.