Ondansetron and the Risk of Sudden Cardiac Death among Individuals Receiving Maintenance Hemodialysis

Abstract

Background: Individuals receiving hemodialysis have high incidence of sudden cardiac death and are susceptible to QT interval prolonging medication-related cardiac complications. Ondansetron, an antiemetic with known QT prolonging potential, is associated with fatal arrythmias in the general population when administered intravenously. The cardiac safety of ondansetron in the hemodialysis population is unknown. Methods: We conducted a new-user, active-comparator, cohort study using US Renal Data System data (2012-2019) to examine the association between the initiation of oral ondansetron vs. antiemetics with lesser QT prolonging potential (promethazine, metoclopramide or prochlorperazine) and the 10-day risk of sudden cardiac death among individuals receiving hemodialysis. We used inverse probability of treatment weighted survival models to estimate adjusted hazard ratios (aHRs), risk differences (aRDs), and 95% confidence intervals (CIs). We used an intention-to-treat approach in which non-sudden cardiac death was considered a competing event. We examined additional cardiac outcomes in secondary analyses. Results: Of the 119,254 study patients, 64,978 (54.5%) initiated ondansetron, and 54,276 (45.5%) initiated a comparator antiemetic. Initiation of ondansetron versus a comparator antiemetic was associated with higher relative and absolute 10-day risks of sudden cardiac death, aHR (95% CI) = 1.44 (1.08, 1.93); aRD (95% CI) = 0.06% (0.01%, 0.11%). The number needed to harm was 1,688. Analyses of additional cardiac outcomes yielded similar findings. Conclusions: Compared with initiation of antiemetics with lesser QT prolonging potential, initiation of ondansetron was associated with higher short-term cardiac risks among people receiving hemodialysis.

Authors' Reply: Pharmacokinetic Issue May Influence the Perceived Cardiac Risk Posed by Ondansetron in Patients Undergoing Chronic Hemodialysis

“…We thank Hung and Chao 1 for their interest in our study and their laudatory comments regarding the importance of the subject and the comprehensiveness of the analysis. 2 In subgroup analyses, we found that patients using at least one other medication with known risk of Torsades de Pointes at the time of ondansetron initiation may have had a lower 10-day relative risk of sudden cardiac death compared with patients not using such medications at the time of ondansetron initiation. Small subgroup sample sizes limited a formal assessment of statistical interaction and also warrant caution when interpreting these findings.…”

Prescription and Dispensation of QT-Prolonging Medications in Individuals Receiving Hemodialysis

“…The focus on a limited number of high-risk medications allowed for more detailed, medication-specific descriptions of prescription and dispensation patterns of medications shown to be associated with a higher risk of sudden cardiac death among people receiving maintenance hemodialysis. 12 , 13 , 14 , 15 We defined new use as a prescription fill for a study medication after a 180-day washout period free of prescription fills for the study drug of interest. Studying new use supported reliable ascertainment of the origin of study prescription medications (ie, prescriber and dispenser) and captured the point in time when the prescriber made the prescribing decision and might have considered medications that were already prescribed to the patient in their decision to prescribe the study medication.…”

“…For example, compared with clinically relevant comparator medications, QT-prolonging medications with known risks of torsades de pointes (TdP), such as respiratory fluoroquinolones, azithromycin, ondansetron, and certain selective serotonin reuptake inhibitors, are associated with higher risks of sudden cardiac death in people receiving hemodialysis. 12 , 13 , 14 , 15 This risk is further elevated in the setting of concomitant use of other QT-prolonging medications with known risk of TdP than in the setting of stand-alone use. 12 , 14 Prescriber education programs, 16 , 17 medication reconciliation, 18 , 19 , 20 and electronic health record (EHR) supports 21 , 22 , 23 reduce unsafe prescribing practices in populations not undergoing dialysis.…”