Effects of cannabis on in-patient diabetic patients with prior revascularized (PCI and/or CABG) myocardial infarction and health care service utilization

Abstract

The legalization of cannabis for recreational and medical purposes has contributed to its increased popularity and caused concern in the United States (US) (1). Cannabis users have been reported to have an increase in hospitalizations and emergency room visits for acute myocardial infarctions (MI) and other adverse effects of cannabis (2,3). The available literature supports the poor hospital outcomes following percutaneous coronary intervention (PCI), and increased risk for Major Adverse Cardiac and Cerebrovascular Events (MACCE), resulting in increased healthcare utilization and costs in patients having chronic cannabis use, and prior diabetes, revascularization procedures, chronic lung disease, concurrent tobacco use, hypertension, and dyslipidemia exhibited (4). We conducted propensity score-matched analysis to assess and better understand the implications of cannabis use disorder (CUD) on hospitalized patients with diabetes mellitus (DM) and prior revascularized MI, i.e., either PCI or coronary artery bypass graft (CABG), using a publicly available US population representative hospitalization sample database.We identified hospitalizations with type 1 or type 2 DM, prior revascularized MI (PCI or CABG), and CUD from the National Inpatient Sample (NIS) database between October 2015 and December 2017 using previously validated ICD-10 codes. Propensity score analysis with 1:1 matching was conducted using a caliper width of 0.2 and near neighbor match adjusting age, sex, race, type of admission (elective/non-elective), payer status, median household income quartile of patient's zip code, hospital bed size, location/teaching status and region to identify two cohorts, CUD positive (CUD+, n=190) and CUD negative (CUD−, n=190). The CUD+ and CUD− groups were compared for comorbidities and in-hospital outcomes. The age groups in this study were 18-44, 45-64, and ≥65 years. MACCE, consisting of all-cause mortality, acute MI, cardiac arrest and stroke, and subsequent revascularization, were the primary outcomes in the analysis. Multivariable logistic regression model to assess MACCE with vs. without CUD was adjusted for age at admission, sex, race, hypertension, hyperlipidemia, obesity, renal failure, coagulopathy, congestive heart failure, chronic pulmonary disease, depression, drug abuse, alcohol abuse, and tobacco use disorder. C-statistics was performed for the fitness of model. The median duration of hospital stays, and charges were the secondary outcomes. We used Mann-Whitney U test and Pearson's Chi square test for P value statistics for continuous and categorical variables respectively. P value of less than 0.05 was taken as significant statistical association.In this nationwide propensity-matched analysis (1:1), 380 DM patients with previous MI and revascularizations were stratified into CUD+ and CUD− cohorts with Letter to the Editor