Anesthesia & Analgesia: October 2016 – Volume 123 – Issue 4 – p 977–988
AUTHORS: Heesen, Michael MD, PhD et al
BACKGROUND: Hypotension remains a frequent complication of spinal anesthesia, increasing the risk of nausea and vomiting, altered mental status, and aspiration. The aim of this systematic review and meta-analysis was to determine whether 5-hydroxytryptamine3 (5-HT3) receptor antagonists, administered before the initiation of spinal anesthesia, mitigate hypotension.
METHODS: After a systematic literature search in various databases, randomized placebo-controlled double-blind trials studying the preventive effect of 5-HT3 receptor antagonists were included. A random-effects model was applied, risk ratio (RR, binary variables) or weighted mean difference (continuous variables) with 95% confidence intervals (CIs) were calculated. The primary outcome was the incidence of hypotension.
RESULTS: Seventeen trials (8 obstetric, 9 non-obstetric) reporting on 1604 patients were identified. Ondansetron in doses from 2 to 12 mg was studied in 12 trials. Prophylactic 5-HT3 administration significantly reduced the risk of hypotension in the combined analysis of 17 trials, RR 0.54 (95% CI 0.36–0.81, I2 = 79%). In obstetric trials, the RR was 0.52, 95% CI 0.30–0.88, I2 = 87% (number needed to treat 4). In non-obstetric studies, the 95% CIs were wide and included a clinically relevant reduction in the risk of hypotension (RR 0.50, 95% CI 0.22–1.16; I2 = 66%). Contour-enhanced funnel plots confirmed publication bias. Meta-regression showed a significant ondansetron dose response in non-obstetric patients (β = −0.355, P = .04). In the combined and in the obstetric-only analysis, the risk of bradycardia was significantly reduced as was the use of phenylephrine equivalents.
CONCLUSIONS: 5-HT3 antagonists are effective in reducing the incidence of hypotension and bradycardia; the effects are moderate and are only significant in the subgroup of patients undergoing cesarean delivery. The effects in the non-obstetric population are not significant.