Association between ondansetron use and mortality of patients on mechanical ventilation in the intensive care unit: a retrospective cohort study

BACKGROUND: Basic studies show that selective 5-hydroxytryptamine type 3 (5-HT(3)) serotonin-receptor antagonists can protect organs from inflammatory injury and have shown lung protection. Whether 5-HT(3) receptor antagonists ondansetron benefits patients with mechanical ventilation is unclear in the intensive care unit (ICU). METHODS: The Medical Information Mart for Intensive Care-IV (MIMIC-IV) database was reviewed to identify patients on mechanical ventilation (aged >16 years) in the ICU, which was divided into two groups according to whether ondansetron is used. Demographic characteristics, medical history data, clinical parameters, diagnosis and treatment measures were included as covariates. Ondansetron use was defined as any kind of ondansetron administration regardless of the dose before the induction of mechanical ventilation. The primary outcome was in-hospital death. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated by multivariable Cox regression. Propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) were performed to further adjust for confounding factors. Kaplan-Meier (KM) curves with log-rank test were also performed. RESULTS: A total of 18,566 patients on mechanical ventilation were included (5,735 with ondansetron use). The overall in-hospital mortality rate of patients on mechanical ventilation was 18.9% (3,512/18,566). Approximately 13.0% (746/5,735) and 21.6% (2,766/12,831) in-hospital mortality rates occurred in the ondansetron and non-ondansetron use groups, respectively. Multivariable regression indicated that ondansetron usage was associated with a 33% and 32% lower risk of in-hospital and 60-day death (HR =0.77, 95% CI: 0.70–0.85, P<0.001; HR =0.68, 95% CI: 0.62–0.75, P<0.001) in the whole sample. Multivariable regression post-PSM indicated that ondansetron usage was associated with a 38% and 31% lower risk of in-hospital and 60-day death (HR =0.62, 95% CI: 0.56–0.68, P<0.001; HR =0.69, 95% CI: 0.62–0.77, P<0.001). Log-rank test for the KM curve of ondansetron and 60-day death was statistically significant (P<0.001). The duration of ventilator use pre- and post-PSM was statistically different (P<0.001 and P=0.007) in the two groups. CONCLUSIONS: Ondansetron usage was significantly associated with a lower mortality risk of ventilated patients in the ICU. The 5-HT3 receptor antagonist use is may be new potential adjunctive therapeutic strategy for patients on mechanical ventilation in the ICU.