(145) Acute cardiovascular adverse effects associated with the use of prescription medications: A Danish nationwide screening study

Background: An alternative approach to traditional pharmacovigilance is the use of large healthcare databases and epidemiological methodology to analyze drug-outcome associations. Though still in its infancy, hypothesis-generating epidemiological screening of such databases can be used to identify unsuspected adverse outcomes.

Objectives: This study aims to conduct a hypothesis-generating screening for acute cardiovascular effects of prescription medications.

Methods: This Danish nationwide screening study was conducted among incident cases of cardiovascular diseases using a case-crossover study design between January 2000 and December 2022. From the study cohort, we identified all newly diagnosed cases of myocardial infarction, ischemic stroke, heart failure, venous thromboembolism, myocarditis and arrhythmia using discharge diagnoses registered in Danish national patient register. We examined exposure to all drugs by using a fixed prescription duration of 90 days; except for antibiotics and drugs prescribed in smaller quantities, for which we used a fixed prescription duration of 30 days. We assessed exposure status on a focal date (day 0) and 3 reference dates corresponding to days -180, -270 and -360 prior to index date. We calculated odds ratios (ORs) with 95% credible intervals (CIs) for associations between exposure drug and risk of cardiovascular effects using Mantel Haenszel (MH) method. Bayesian shrinkage was applied using a prior OR of 1.0 (95% CI 0.25 - 4.00).

Results: After applying exclusion criteria, we identified 191,979 cases of arrhythmia, 145,148 cases of MI, 132,271 cases of ischemic stroke, 71,821 cases of HF, 16,127 cases of VTE and 10,045 cases of myocarditis. Our study identified more than 30 drug-outcome associations for each cardiovascular outcome. Few major drug classes, such as antibiotics, opioid analgesics, laxatives etc. consistently demonstrated associations across all cardiovascular outcomes. However, few more specific associations were apparent, e.g. pantoprazole use which, in contrast to use of other PPIs, was associated with ischemic stroke (OR 1.68; CI 1.55-1.83) as well as other cardiovascular outcomes. Similarly, oxazepam stood out among other benzodiazepines and demonstrated increased risk of HF (OR 2.12; CI 1.75-2.56) and most other cardiovascular outcomes.

Conclusions: The results highlight several important associations across various pharmacological drug classes that warrant further investigation in tailored pharmacoepidemiological analyses.