Compatibility of medication with PRISCUS criteria and identification of drug interactions in a large cohort of patients with COPD.

Patients with COPD are often of advanced age and have a high number of medications due to their lung disease and comorbidities. Thus they are at risk for unwanted effects from drugs, either due to age or due to interactions between drugs. These issues are not clarified. We therefore assessed the number of medications and potential adverse effects in a large cohort of patients with COPD. The analysis was performed in 2741 patients of the German COPD cohort COSYCONET, using baseline data (visit 1) and follow-up data after about 1.5 years (visit 3). Spirometric GOLD grades 1-4 were found in 8/35/32/9% of patients and GOLD groups ABCD in 7/25/4/48% of patients, while the remaining patients (n = 450, 16.4%) could not be classified according to GOLD criteria. The compatibility of medication with age was evaluated via the PRISCUS list, drug interactions via the AiD clinic system, whereby only drug combinations occurring in at least 10 patients were considered (nine unwanted interactions, one wanted interaction). The median numbers of medications were 5 or more in all patient categories, among them 3 or more non-respiratory medications. In the total population there were 153 patients (10.2%) aged ≥65 years who had any medication of the PRISCUS list with intermediate or low risk. Serious adverse combinations of drugs according to AiD occurred in 114 patients (4.2%), while the number of unwanted but only potentially clinically relevant combinations was 175 (6.4%). The number of wanted combinations was 219 (8.0%). These numbers did not markedly change when restricting the analysis to patients of GOLD grades 1-4. Moreover, the results were similar for visit 1 and visit 3. We conclude that in a large cohort of COPD patients about 10% of patients aged at least 65 years had medications that could interfere with their age and that the proportions of patients with either unwanted or wanted drug interactions were both in the range of 8-10%. These results suggest that problems arising from the high number of medications were not very frequent in the COPD cohort analysed.