The Association of Guideline-Directed Prophylaxis with Postoperative Nausea and Vomiting in Pediatric Patients: A Single-Center, Retrospective Cohort Study

BACKGROUND: Guidelines for postoperative nausea and vomiting (PONV) prophylaxis in pediatric patients recommend escalation of the number of antiemetics based on a preoperative estimate of PONV risk. These recommendations have been translated into performance metrics, most notably by the Multicenter Perioperative Outcomes Group (MPOG), used at over 25 children’s hospitals. The impact of this approach on clinical outcomes is not known. METHODS: We performed a single-center, retrospective study of pediatric general anesthetic cases from 2018 to 2021. PONV risk factors were defined using MPOG definitions: age ≥3 years, volatile use ≥30 minutes, PONV history, long-acting opioids, female ≥12 years, and high-risk procedure. Adequate prophylaxis was defined using the MPOG PONV-04 metric: 1 agent for 1 risk factor, 2 agents for 2 risk factors, and 3 agents for 3+ risk factors. PONV was defined as documented postoperative nausea/emesis or administration of a rescue antiemetic. Given the nonrandomized allocation of adequate prophylaxis, we used Bayesian binomial models with propensity score weighting. RESULTS: A total of 14,747 cases were included, with PONV in 11% (9% adequate prophylaxis versus 12% inadequate). Overall, there was evidence for reduced incidence of PONV with adequate prophylaxis: weighted median odds ratio 0.82 (95% credible interval, 0.66–1.02; probability of benefit, 0.97) and weighted marginal absolute risk reduction 1.3% (−0.1% to 3.1%). In unweighted estimates, there was an interaction between sum of risk factors and the association of adequate prophylaxis with PONV, with reduced incidence in patients with 1 to 2 risk factors (probability of benefit 0.96 and 0.95) but increased incidence in patients with 3+ risk factors receiving adequate prophylaxis (probability of benefit 0, 0.01, and 0.03 for 3, 4, and 5 risk factors). This was attenuated by weighting, with persistent benefit in 1 to 2 risk factors (probability of benefit 0.90 and 0.94) but equalization of risk in 3+ risk factors. CONCLUSIONS: Guideline-directed PONV prophylaxis is inconsistently associated with incidence of PONV across the guideline-defined risk spectrum. This phenomenon, and its attenuation with weighting, is consistent with 2 points: dichotomous risk-factor summation ignores differential effects of individual components, and prognostic information might exist beyond these risk factors. PONV risk at a given sum of risk factors is not homogeneous, but rather is determined by the unique composition of risk factors and other prognostic attributes. These differences appear to have been identified by clinicians, prompting use of more antiemetics. Even after accounting for these differences, however, addition of a third agent did not further reduce risk.