Cranial surgery serves as an essential therapeutic modality in neurosurgical practice for conditions including traumatic brain injury, cerebrovascular diseases, intracranial tumors, functional brain disorders, and intracranial infections, representing a critical intervention for these pathologies (Cagney et al., 2019; Ramakonar et al., 2018). Postoperative care is equally critical as surgical success, making brief ICU monitoring a common practice for patients undergoing craniocerebral procedures (Laan et al., 2020). In numerous neurosurgical centers across China, individuals receiving general anesthesia for such operations are routinely admitted to intensive care units for specialized observation and management. Individuals undergoing cranial surgery frequently experience nausea and vomiting of varying severity, with these manifestations being particularly prevalent among those sustaining mild to moderate cerebral injuries. Paradoxically, patients with severe neurological impairment [Glasgow Coma Scale (GCS) ≤ 8] often exhibit attenuated emetic responses due to suppressed brainstem reflex arcs (Brain Trauma Foundation et al., 2007b). These divergent clinical manifestations underscore the necessity for tailored postoperative management strategies. Effective control of vomiting not only enhances patient comfort but also prevents secondary complications including aspiration pneumonia and intracranial pressure elevation, which are critical determinants of neurological recovery (Andrews and Hawthorn, 1988; Eberhart et al., 2007).

Ondansetron, a therapeutic agent targeting 5-hydroxytryptamine subtype 3 (5-HT3) receptors through selective antagonism (Borner et al., 2020), has maintained a well-established safety profile through over two decades of clinical implementation (Garsed et al., 2014), solidifying its position as the pharmacological mainstay for perioperative emesis management in surgical care (Lee et al., 2020). Its mechanism involves competitive inhibition of serotonin binding at vagal afferent terminals in the chemoreceptor trigger zone and gastrointestinal tract (Hesketh, 2008). Clinical evidence supports its efficacy in reducing postoperative nausea and vomiting (PONV) across diverse surgical populations, including abdominal surgery patients (Zhang et al., 2019) and gynecological operation cohorts (Abdulhussain, 2022). Additionally, the administration of ondansetron can significantly reduce the incidence of PONV in patients undergoing craniotomy (Ryu et al., 2014).

While ondansetron prophylaxis demonstrates significant efficacy in preventing PONV following cranial neurosurgical interventions, its neuroprotective implications — specifically regarding 30- and 90 days mortality indices—remain inadequately substantiated in contemporary neurocritical care pharmacovigilance studies. This knowledge gap motivated our investigation into ondansetron’s impact on mortality outcomes in a targeted population of cranial surgery patients with GCS > 8. Our study employs propensity score matching (PSM) to address confounding factors inherent in retrospective analyses, aiming to provide evidence for optimized postoperative management protocols.

This retrospective propensity - matched study evaluated the impact of ondansetron on clinical outcomes in cranial surgery patients. The findings suggest that ondansetron use is associated with a significant reduction in 28 and 90 days mortality in this patient population.

The observed mortality reduction is a notable finding, considering the complex postoperative course of cranial surgery patients. The mechanism through which ondansetron exerts this effect might be multi - faceted. As a central nervous System 5 - HT3 receptor antagonist (Kirchhoff et al., 2019), its well-established antiemetic property could play a role (Basak et al., 2019; Mascarenhas et al., 2019; Vaid et al., 2020). Ondansetron exhibits high selectivity as a 5-HT3 receptor antagonist, demonstrating 250- to 500-fold greater binding affinity for 5-HT3 receptors compared to 5-HT1A and 5-HT2B subtypes (Wang et al., 2019). The 5-HT3 receptor subtype demonstrates anatomical localization within peripheral vagal afferent terminals and dense concentration in the nucleus tractus solitarii of the brainstem’s emetic circuitry (Tyers, 1992). The high selectivity for 5-HT3 receptors and lack of antagonistic activity toward dopamine and other non-5-HT3 receptors result in ondansetron’s minimal neurological side effects and low incidence of extrapyramidal symptoms (Jueckstock et al., 2010; Tyers, 1992; Yin et al., 2020). By reducing PONV, ondansetron may prevent secondary complications such as aspiration pneumonia and intracranial pressure elevation (Andrews and Hawthorn, 1988; Eberhart et al., 2007). Aspiration pneumonia is a common and potentially life - threatening complication in hospitalized patients (Ko et al., 2010; Thomson et al., 2016; Xu et al., 2015), and acute intracranial hypertension is a common and life-threatening neurological complication often directly related to poor prognosis of patients (Vaibhav et al., 2020). Clinical studies further demonstrate that aspiration pneumonia in neurosurgical and critically ill patients is associated with increased mortality (Thomson et al., 2016; Xu et al., 2015), while acute intracranial hypertension has long been recognized as a major determinant of poor neurological outcomes and death (Brain Trauma Foundation, 2007a). These findings strengthen the clinical plausibility of our proposed pathway, whereby ondansetron reduces PONV, thereby lowering the incidence of aspiration events and ICP crises, ultimately contributing to improved survival in cranial surgery patients. Moreover, recurrent nausea and vomiting may precipitate acute hypertensive episodes, thereby elevating the risk of intraoperative intracranial hemorrhage. Severe intracranial hemorrhagic events may precipitate cerebral herniation through mass effect displacement (Crash-3 trial collaborators, 2019), culminating in catastrophic neurological deterioration secondary to brainstem compression and irreversible parenchymal injury (Gallagher et al., 2018; Song et al., 2017). In cranial surgery patients, the administration of ondansetron may be associated with maintenance of intracranial pressure stability, prevention of aspiration pneumonia, and preservation of hemodynamic equilibrium constitute critical perioperative priorities. The systematic mitigation of these interrelated risk factors may underlie its beneficial effects on improving patient outcomes.

Moreover, emerging evidence of ondansetron’s potential neuroprotective effects through modulation of inflammatory cascades and oxidative stress pathways may also be relevant (Hassan et al., 2024; Tian et al., 2023). Although these mechanisms remain underexplored in the neurosurgical context, they could offer additional explanations for the reduced mortality observed in our study. Inflammation and oxidative stress are known to play important roles in the pathophysiology of brain injury and recovery (Chen et al., 2022; Magid-Bernstein et al., 2022). Ondansetron might be interfering with these processes, leading to better outcomes. Recent studies provide more direct evidence supporting these hypotheses. Network pharmacology and experimental models have shown that ondansetron suppresses neutrophil extracellular trap (NET) formation by downregulating neutrophil elastase, MPO, and PAD4, while reducing pro-inflammatory cytokines such as IL-6, IL-1β, and TNF-α (Tao et al., 2024). Animal studies further suggest that ondansetron mitigates blood–brain barrier disruption, brain edema, and astrocytic activation, and confers anti-apoptotic effects by decreasing Bax and calcineurin expression while increasing Bcl-2 levels (Sharma et al., 2019). Nevertheless, these explanations remain largely hypothetical, and future mechanistic experiments and prospective clinical studies are required to confirm whether these pathways play a causal role in the improved outcomes observed in cranial surgery patients.

In the subgroup analysis, a significant interaction was observed between MV status and the effect of ondansetron on both 28 and 90 days mortality. Specifically, ondansetron use was associated with a marked reduction in short-term mortality among patients who did not require MV (28 days HR = 0.38, 95% CI: 0.17–0.85; 90 days HR = 0.40, 95% CI: 0.22–0.72), whereas the survival benefit was attenuated and no longer statistically significant in patients receiving MV. This finding may reflect underlying differences in baseline severity and clinical trajectory. Patients requiring MV generally present with more severe neurological impairment, hemodynamic instability, and a higher burden of complications such as ventilator-associated pneumonia, prolonged sedation, and secondary infections, which may diminish or overshadow the potential protective effects of ondansetron. By contrast, in non-ventilated patients with relatively preserved physiological reserve, the ability of ondansetron to mitigate aspiration events, stabilize intracranial pressure, and reduce perioperative complications could translate into a more pronounced survival benefit. These observations suggest that the therapeutic impact of ondansetron may be more evident in less severely ill patients, and highlight the need for future prospective studies to clarify whether MV status is a true effect modifier or primarily a marker of confounding severity.

There are several limitations to this study. Firstly, as a retrospective study, it is subject to inherent biases despite the use of propensity score matching. Residual confounding may still exist, as it is difficult to account for all potential confounders in a retrospective design. Secondly, the data source is from a single database (MIMIC-IV), which may limit the generalizability of the results. Different patient populations in other regions or hospitals may have different characteristics and responses to ondansetron. Thirdly, the dosage, timing, and duration of ondansetron administration were not standardized in the dataset, which may have influenced the observed associations. Fourthly, important perioperative variables such as the type of surgery, duration of the procedure, and intraoperative blood loss were not recorded, and these unmeasured factors could potentially confound the outcomes. Finally, the study only evaluated short-term mortality (28 and 90 days), and long-term outcomes such as neurological function recovery and quality of life were not assessed.

This study provides evidence that ondansetron may be beneficial in reducing short - term mortality in cranial surgery patients with GCS > 8. However, future prospective studies with larger and more diverse patient populations are needed to confirm these findings, explore the underlying mechanisms further, and evaluate the long - term impact of ondansetron on patient outcomes. This could help in developing more optimized postoperative management protocols for cranial surgery patients.