Assessment of direct oral anticoagulants administered as potentially inappropriate medications to elderly inpatients

Background There is evidence that direct oral anticoagulants administered as potentially inappropriate medications increase the risk of bleeding and thromboembolic complications, which represent serious threats to human health. Objective To identify direct oral anticoagulants administered as potentially inappropriate medications for hospitalized patients aged ≥ 65 years in our hospital, and to determine associated factors and the correlation between potentially inappropriate medications and adverse reactions. Setting Xi’an Central Hospital, China. Method A retrospective cross-sectional study of elderly hospitalized patients who received either dabigatran or rivaroxaban at Xi’an Central Hospital between June 1, 2018 and June 1, 2019. The evaluation criteria of direct oral anticoagulants were formulated based on drug labels, disease guidelines and the 2019 American Geriatrics Society Beers Criteria, and any non-compliance with the criteria was considered to be potentially inappropriate medications. The Pearson chi-square test and a binary logistic regression model were used. Main outcome measure Factors associated with potentially inappropriate medications and correlation between potentially inappropriate medications and adverse reactions. Results This study analysed 315 patients aged ≥ 65 years. The application of our evaluation criteria identified 155 (49.2%) instances of potentially inappropriate medications, comprising 5 different types of potentially inappropriate medications. Fifteen adverse drug reactions occurred in the study participants. The Pearson chi-square test revealed significant differences in number of medications (p = 0.021) and creatinine clearance rate (p = 0.002) between potentially inappropriate medications and non-potentially inappropriate medications groups. In the binary logistic regression model, potentially inappropriate medications use was associated with creatinine clearance (creatinine clearance < 30: OR = 3.590, 95% CI = 1.214–10.615, p = 0.021), and there was no significant correlation between potentially inappropriate medications and adverse drug reactions after eliminating the confounding factors (age, length of hospitalization, number of disease combined) with p values of less than 0.25 (adjusted OR = 0.372, 95% CI = 0.117–1.182, p = 0.094). Conclusion This study revealed that the incidence of potentially inappropriate medications was relatively high, number of medications and creatinine clearance differed significantly between potentially inappropriate medications and non-potentially inappropriate medications groups, and potentially inappropriate medications was associated with creatinine clearance (creatinine clearance < 30 mL/min), but there was no significant correlation between potentially inappropriate medications and adverse drug reactions after eliminating the confounding factors.