B-PO04-059 MEDICATION NONADHERENCE: A PREVIOUSLY UNIDENTIFIED CONFOUNDER UNDERLYING ICD THERAPY

Abstract

ICD shocks are associated with an increased risk of mortality. However, prospective analyses have failed to establish a correlation between ICD shocks and direct myocardial injury. Therefore, the reasons for worse outcomes associated with ICD shocks remain unclear. The majority of data for ICD shocks and worse outcomes is derived from observational studies and subgroup analysis which often have unidentified confounding factors. The goal of this study is to highlight one potential confounder not usually accounted for in observational trials and case controlled matched cohorts, namely guideline directed medication compliance. We hypothesized that patients who receive ICD shocks are more likely to be non-adherent with medications when compared with patients who do not receive shocks. In a single center retrospective analysis we compiled a database of all patients who underwent ICD implantation. We matched patients who received ICD shocks to a group of patients who did not receive shocks. Patients were matched by co-morbidities, NYHA class heart failure and ejection fraction. Medication adherence was expressed as doses dispensed in relation to the number of doses required in the dispensing period. We used a two-way mixed ANOVA using PROC GLM to see if there was a significant difference in adherence of betablockers between those who received an ICD shock and those who did not. We found 88 matched patients (27 with shocks and 61 without shocks). Patients who received ICD shocks were significantly less likely to be adherent to betablockers (60% versus 72%, P < 0.025) at three different time points. Non adherence with GDMT is associated with an increase in ICD shocks. This data suggests that the increased ICD shocks may not be responsible for the reported worsened outcome, but instead are a marker for medical non-adherence, which is associated with worsened outcomes.