Drug-drug interactions in medical patients: Effects of in-hospital treatment and relation to multiple drug use

OBJECTIVES: Adverse drug reactions (ADRs) are a major cause of hospital admissions, thereby leading to significant medical and economical problems. Drug interactions contribute to a major part of ADRs, especially in elderly patients and in patients under polymedication. As a peculiarity of the German health care system the general practitioner is less involved in patient care during hospital stay than in other countries. Consequently, changes in medication at the transition point from out-patient to in-patient care and back may contribute to drug-related problems. PATIENTS: In the present study we investigated potential interactions in 169 consecutive patients with the diagnosis CHD (coronary heart disease) or COLD (chronic obstructive lung disease) who were admitted to the University Hospitals of Hannover and Magdeburg. METHODS: For each patient, potential interactions between prescribed drugs at admission, at discharge, and 3 months after discharge were assessed by using drug interaction data bases. RESULTS: We found that the number of drugs taken per patient as well as the number of interactions per patient are higher during hospitalization than before admission (pre-admission), and fall back after the hospital stay (post-discharge), but not to the pre-admission level. The number of potential interactions was significantly correlated in a polynomial manner to the number of drugs taken by each patient. The number of patients without a potential interaction was 44.4% pre-admission, 39.6% at discharge, and 39.1% post-discharge. Patients with potential interactions had a mean of 2.8, 2.7, and 2.4 interactions at each of the time points. Drug classes mainly involved with potential interactions were kaliuretic diuretics (recorded by 43.3% at discharge), ACE inhibitors (30.3%), anticoagulants and aggregation inhibitors (20.4%) and digitalis glycosides (14.7%). Regarding the frequency of the interaction categories, 68-70% of the potential interactions demand clinical attention, while 1-2% are life-threatening. 17-19% may result in therapeutic benefit; 10-12% are without clinical relevance. CONCLUSIONS: The risk of drug interactions increases exponentially with the number of drugs given to a patient. Although we have no data on the fraction of interactions that became clinically manifest, our study indicates that prescribing fewer drugs can reduce the risk of suffering from sickness secondary to drug interactions. Taking this risk into account may help to improve the quality of drug treatment and to save costs.