Our literature search identified 369 trials (Figure 1); 17 trials reporting on 1604 patients were included in the meta-analysis.7,8,14–28 Nine studies were performed in a non-obstetric population14–19,21,22,28 and 8 studies included parturients undergoing cesarean delivery.7,8,20,23–26,28 The methodological quality of the included studies is shown in Tables 1 and 2; Table 3 shows details of the studies. The study by Rashad and Farmawy29 was excluded because it was not double-blinded.
Prevention of Hypotension: All Studies Combined
When combining obstetric and non-obstetric trials, prophylactic administration of a 5-HT3 antagonist significantly reduced the risk for hypotension, RR 0.54, 95% CI 0.36–0.81, I2 = 79% (Figure 2). The NNT was 6.3. In the obstetric studies, RR was 0.52, 95% CI 0.30–0.88, I2 = 87% (NNT = 4); in non-obstetric trials, the RR was 0.56, 95% CI 0.28–1.12, I2 = 59%.
Ondansetron was the most frequently used drug. RR of hypotension for the 12 obstetric and non-obstetric trials studying ondansetron 2–12 mg was 0.67, 95% CI 0.48–0.92, I2 = 65% (Supplemental Digital Content 1, http://links.lww.com/AA/B471). The NNT was 9.6.
Prevention of Hypotension: Obstetric Studies
Of the 8 obstetric trials, 7 studied ondansetron7,8,20,23–25,28 and 1 study used granisetron.26 The doses of ondansetron used in these trials varied from 2 to 12 mg. When analyzing the 7 ondansetron studies,7,8,20,23–25,28 the RR was 0.70, 95% CI 0.49–0.99, I2 = 71%. In addition, we performed an analysis with the obstetric studies that used higher doses of ondansetron (>4 mg),8,23,25,27 The RR was 0.77, 95% CI 0.41–1.41, I2 = 86%.
Prevention of Hypotension: Non-obstetric Studies
Of the non-obstetric trials, 7 trials14,15,17,19,21,22,28 studied ondansetron in doses ranging from 4 to 12 mg. One study used granisetron17 and another trial18 studied ramosetron. Subgroup analysis of the non-obstetric trials using ondansetron,14,15,17,19,21,22,28 did not find a statistically significant effect of 5-HT3 antagonists on the incidence of hypotension: RR 0.50, 95% CI 0.22–1.16, P = .10, I2 = 66%; however, the CI was wide and did not exclude a clinically significant reduction in the risk of hypotension. We performed a meta-analysis excluding the 2 studies14,19 using low-dose ondansetron (4 mg). Analysis of the trials of high-dose ondansetron15,17,21,22,28 only (6–12 mg) revealed a significant reduction in the risk of hypotension, RR 0.23, 95% CI 0.06–0.86, I2 = 70%.
We performed sensitivity analyses including only those studies that had hypotension as the primary outcome variable. In a combined analysis of obstetric and non-obstetric studies,7,8,17,20,21,23–28 the RR was 0.47, 95% CI 0.28–0.77, P = .003, I2 = 86%. Analysis of obstetric studies alone7,8,20,23–27 gave a RR 0.52, 95% CI 0.30–0.88, P = .01, I2 = 87%. Analysis of non-obstetric trials17,21,28 gave a RR 0.16, 95% CI 0.01–2.17, I2 = 84%.
Analysis of all studies combined, only the obstetric studies found that the incidence of bradycardia was significantly lower in the 5-HT3 antagonist group (Figure 3). In these studies, only ondansetron was used. Subgroup analysis of the 3 non-obstetric studies15,21,22 showed no significant difference, but the CI was wide.
Need for Vasopressors
Five studies15,17,22,27,28 reported on the number of patients needing treatment with a vasopressor; all 5 studies used ephedrine. The use of 5-HT3 antagonist did not reduce the number of subjects who required ephedrine (RR 0.6, 95% CI 0.35–1.04; P = .07, I2 19%) (Supplemental Digital Content 2, http://links.lww.com/AA/B472). The total dose of vasopressor was analyzed in 5 trials7,20,23,24,26 of parturients undergoing cesarean delivery. One study used ephedrine7 and 420,23,24,26 used phenylephrine. We computed the phenylephrine equivalents based on the assumption that 5 mg ephedrine are equivalent to 62 µg phenylephrine.30 5-HT3 antagonists significantly reduced the amount of vasopressor needed for the treatment of hypotension (weighted mean difference −48 µg, 95% CI −86 to −10, I2 = 77%) (Supplemental Digital Content 3, http://links.lww.com/AA/B473). We performed a sensitivity analysis without the study by Sahoo et al,20 because the phenylephrine dose was an order of magnitude lower than in the other trials. The weighted mean difference was −64 µg, 95% CI −99 to −30, P = .0003, I2 = 36%.
Investigation of Publication Bias
Funnel plots of the analyses of the incidence of hypotension combining obstetric and non-obstetric studies, the subset of those trials that only used ondansetron, the obstetric and the non-obstetric trials were asymmetric and suggestive of bias (data not shown). Contour-enhanced funnel plots with trim-and-fill method were done for obstetric and non-obstetric trials combined (Figure 4) and for the subgroup of obstetric and non-obstetric trials that used ondansetron (Supplemental Digital Content 4, http://links.lww.com/AA/B474). The plots confirmed publication bias.
Meta-regression for Ondansetron Dosing
Meta-regression analysis found no dose-response relationship of ondansetron in both the combined analysis and in the obstetric trials alone (Supplemental Digital Content 5, http://links.lww.com/AA/B475). There was a significant dose-response relationship for prevention of hypotension in the non-obstetric trials, β = −0.355, P = .04. There was no evidence of a dose-responsive relationship for bradycardia (Supplemental Digital Content 6, http://links.lww.com/AA/B476).
Other Side Effects
Other side effects included tachycardia, dysrhythmia, headache, drowsiness, discomfort, rigor, and skin flushing. No significant difference between the 5-HT3 antagonist and the placebo group was observed (data not shown).
Our study found that prophylactic administration of 5-HT3 receptor antagonists significantly reduced the incidence of spinal anesthesia–induced hypotension. Most studies used ondansetron, and limiting the analysis to only these studies did not change the results. A similar decrease in risk of hypotension was observed when limiting the analysis to only obstetric trials, but was not observed for non-obstetric trials. However, the 95% CIs for RR in non-obstetric studies were wide, suggesting that the data were inadequate to draw any definitive conclusions. 5-HT3 antagonist use did not lower the number of patients who required vasopressor for the treatment of hypotension; however, the dose of vasopressor was significantly lower in 5-HT3 antagonist–treated patients. The incidence of bradycardia was also significantly reduced by the use of a 5-HT3 antagonist. Maternal side effects were unaffected by 5-HT3 antagonists.
No dose responsiveness of ondansetron was observed for the prevention of hypotension in the combined analysis of obstetric and non-obstetric trials or in obstetric trials alone. However, a statistically significant dose-response relationship was observed for the prevention of hypotension in the non-obstetric population. A possible explanation for these findings is that the overall analysis included trials using low doses (eg, 4 mg) as well as high doses of ondansetron. Indeed, an analysis without these low-dose non-obstetric studies revealed a significantly reduced incidence of hypotension.
Numerous studies have dealt with strategies to reduce the incidence of spinal-related hypotension in parturients. Preload with crystalloids has been found to be ineffective. Volume load with crystalloids starting immediately after spinal injection (coload) has been suggested as well as colloid preload and coload.31,32 Among vasopressors, the pure α-adrenergic receptor agonist phenylephrine has become the first-line treatment,33 because of a more favorable effect on neonatal pH compared with ephedrine.34 Proactive administration of phenylephrine is currently advocated31 because it is more effective than an administration only after blood pressure has fallen.35 After continuous infusion of 25–50 µg/min phenylephrine, hypotension was still noted in 15%–30% of the patients.36
Bradycardia is a known side-effect of phenylephrine, affecting approximately 10% of the women receiving this vasopressor.34,37 In a trial of continuous phenylephrine infusion, as many as 32% of the patients experienced bradycardia with infusion rates of 75 and 100 µg/min.36 Heart rate was suggested as a surrogate parameter for cardiac output by Dyer et al38 as well as Langesaeter et al.4 It is cardiac output that reflects more accurately uteroplacental perfusion than maternal blood pressure,39 and the well-being of the fetus, unable to store oxygen, finally depends on maternal cardiac output.40 Often one phenylephrine bolus or the chosen infusion rate of phenylephrine is not sufficient to reverse low blood pressure. It would be of interest to study the use of 5-HT3 antagonists in this situation. Recently, Ngan Kee et al41 compared phenylephrine with the nonspecific adrenergic agonist, norepinephrine, highlighting the ongoing search for alternatives to the negative chronotropic effects of phenylephrine. Indeed, norepinephrine use was associated with less bradycardia and consequently higher cardiac output, with no difference in blood pressure between the 2 treatment arms.41
The Bezold-Jarisch reflex has been proposed to explain perioperative hypotension associated with bradycardia.42 A rationale for the use of 5-HT3 antagonists to prevent Bezold-Jarisch reflex is based on chemoreceptors sensitive to 5-HT3; 5-HT3 agonists activate the reflex.5 Animal studies have demonstrated the effectiveness of 5-HT3 antagonists in blocking Bezold-Jarisch reflex.6
The effects of 5-HT3 antagonists in obstetric patients were moderate, as reflected by NNTs of 5.3 and 7.6 for the prevention of hypotension and bradycardia, respectively. Interestingly, George et al12 estimated that the NNT for prevention of nausea and vomiting in obstetric patients under spinal anesthesia are 8.6 and 11, respectively. Bonnet and colleagues11 found a NNT of 6 for the combined outcome postoperative nausea and vomiting.
We acknowledge several limitations of our study: First, heterogeneity in all our analyses was high with the exception of bradycardia. Funnel plots were indicative of publication bias. Differences in the spinal anesthesia technique, the fluid management, and the vasopressor used are other factors contributing to heterogeneity. Second, the definitions of hypotension varied across the studies included in our meta-analysis. The definition of hypotension affects the incidence3 and this makes it difficult to compare results. Third, definition of bradycardia was also different or was not even defined in some of the included studies. Fourth, there is a paucity of neonatal outcome data. Moreover, the data were obtained from healthy parturients with uncomplicated term pregnancies undergoing elective cesarean delivery. Our results, therefore, cannot be extrapolated to emergency delivery and compromised fetuses.
While our manuscript was under review, a similar meta-analysis43 was published. Gao et al43 included 10 studies (6 obstetric, 4 non-obstetric) reporting on 863 patients receiving prophylactic ondansetron. However, there are several major differences between the report by Gao et al43 and our study. The meta-analysis by Gao et al43 did not include 5 studies14,16,18,19,25 that we included. A study29 that had methodological flaws (no patient blinding) was included by Gao et al43 but excluded by us. Our meta-analysis included sophisticated tools to evaluate publication bias and a meta-regression to study dose-responsiveness. However, our results largely confirm those of Gao et al.43
The heterogeneity was high in many analyses and funnel plots were performed in the study by Gao et al43 and in our study. However, funnel plot asymmetry can be caused by reasons other than publication bias. We, therefore, added contour-enhanced funnel plots that are more specific for the evaluation of publication bias. We also performed a sensitivity analysis of trials that hypotension had as a primary outcome.
Taken together, our results suggest that 5-HT3 antagonists have an effect in preventing hypotension and bradycardia when analyzing obstetric trials alone or when combining obstetric and non-obstetric surgery trials. Ondansetron was the most frequently used substance in the trials; it remains unclear whether the effects seen in our study are specific for this drug only or for all 5-HT3 antagonists. Only when trials of 4 mg ondansetron were excluded from the analysis was significant effect observed. In the trials entered in our meta-analysis, 5-HT3 antagonists were studied prophylactically. We suggest that these drugs should be studied for the treatment of hypotension.
Name: Michael Heesen, MD, PhD.
Contribution: This author was responsible for study design, literature search, data extraction, quality analysis of the retrieved trials, meta-analysis, interpretation of results, writing of the manuscript.
Name: Markus Klimek, MD, PhD, DEAA, EDIC.
Contribution: This author was responsible for literature search, data extraction, writing of the manuscript.
Name: Sanne E. Hoeks, MD, PhD.
Contribution: This author was responsible for meta-analysis (Contour-enhanced funnel plots), interpretation of results, writing of manuscript (revised version).
Name: Rolf Rossaint, MD, PhD.
Contribution: This author was responsible for study design, interpretation of results, writing of the manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
1. Liu SS, McDonald SB. Current issues in spinal anesthesia. Anesthesiology. 2001;94:888–906.
2. Carpenter RL, Caplan RA, Brown DL, Stephenson C, Wu R. Incidence and risk factors for side effects of spinal anesthesia. Anesthesiology. 1992;76:906–916.
3. Klöhr S, Roth R, Hofmann T, Rossaint R, Heesen M. Definitions of hypotension after spinal anaesthesia for caesarean section: literature search and application to parturients. Acta Anaesthesiol Scand. 2010;54:909–921.
4. Langesaeter E, Rosseland LA, Stubhaug A. Continuous invasive blood pressure and cardiac output monitoring during cesarean delivery: a randomized, double-blind comparison of low-dose versus high-dose spinal anesthesia with intravenous phenylephrine or placebo infusion. Anesthesiology. 2008;109:856–863.
5. Campagna JA, Carter C. Clinical relevance of the Bezold-Jarisch reflex. Anesthesiology. 2003;98:1250–1260.
6. Yamano M, Ito H, Kamato T, Miyata K. Characteristics of inhibitory effects of serotonin (5-HT)3-receptor antagonists, YM060 and YM114 (KAE-393), on the von Bezold-Jarisch reflex induced by 2-Methyl-5-HT, veratridine and electrical stimulation of vagus nerves in anesthetized rats. Jpn J Pharmacol. 1995;69:351–356.
7. Trabelsi W, Romdhani C, Elaskri H, et al. Effect of ondansetron on the occurrence of hypotension and on neonatal parameters during spinal anesthesia for elective caesarean section: a prospective, randomized, controlled, double-blind study. Anesthesiol Res Pract. 2015;2015:158061.
8. Ortiz-Gómez JR, Palacio-Abizanda FJ, Morillas-Ramirez F, Fornet-Ruiz I, Lorenzo-Jiménez A, Bermejo-Albares ML. The effect of intravenous ondansetron on maternal haemodynamics during elective caesarean delivery under spinal anaesthesia: a double-blind, randomised, placebo-controlled trial. Int J Obstet Anesth. 2014;23:138–143.
9. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269, W64.
10. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011] Editors: Julian PT Higgins and Sally Green.
11. Bonnet MP, Marret E, Josserand J, Mercier FJ. Effect of prophylactic 5-HT3 receptor antagonists on pruritus induced by neuraxial opioids: a quantitative systematic review. Br J Anaesth. 2008;101:311–319.
12. George RB, Allen TK, Habib AS. Serotonin receptor antagonists for the prevention and treatment of pruritus, nausea, and vomiting in women undergoing cesarean delivery with intrathecal morphine: a systematic review and meta-analysis. Anesth Analg. 2009;109:174–182.
13. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–1558.
14. Pirat A, Tuncay SF, Torgay A, Candan S, Arslan G. Ondansetron, orally disintegrating tablets versus intravenous injection for prevention of intrathecal morphine-induced nausea, vomiting, and pruritus in young males. Anesth Analg. 2005;101:1330–1336.
15. Kelsaka E, Baris S, Karakaya D, Sarihasan B. Comparison of ondansetron and meperidine for prevention of shivering in patients undergoing spinal anesthesia. Reg Anesth Pain Med. 2006;31:40–45.
16. Sagir O, Gulhas N, Toprak H, Yucel A, Begec Z, Ersoy O. Control of shivering during regional anaesthesia: prophylactic ketamine and granisetron. Acta Anaesthesiol Scand. 2007;51:44–49.
17. Owczuk R, Wenski W, Polak-Krzeminska A, et al. Ondansetron given intravenously attenuates arterial blood pressure drop due to spinal anesthesia: a double-blind, placebo-controlled study. Reg Anesth Pain Med. 2008;33:332–339.
18. Kim MS, Kim DW, Woo SH, Yon JH, Lee S. Effect of ramosetron on shivering during spinal anesthesia. Korean J Anesthesiol. 2010;58:256–259.
19. Shakya S, Chaturvedi A, Sah BP. Prophylactic low dose ketamine and ondansetron for prevention of shivering during spinal anaesthesia. J Anaesthesiol Clin Pharmacol. 2010;26:465–469.
20. Sahoo T, SenDasgupta C, Goswami A, Hazra A. Reduction in spinal-induced hypotension with ondansetron in parturients undergoing caesarean section: a double-blind randomised, placebo-controlled study. Int J Obstet Anesth. 2012;21:24–28.
21. Marashi SM, Soltani-Omid S, Soltani Mohammadi S, Aghajani Y, Movafegh A. Comparing two different doses of intravenous ondansetron with placebo on attenuation of spinal-induced hypotension and shivering. Anesth Pain Med. 2014;4:e12055.
22. Safavi M, Honarmand A, Negahban M, Attari M. Prophylactic effects of intrathecal Meperidine and intravenous Ondansetron on shivering in patients undergoing lower extremity orthopedic surgery under spinal anesthesia. J Res Pharm Pract. 2014;3:94–99.
23. Wang M, Zhuo L, Wang Q, et al. Efficacy of prophylactic intravenous ondansetron on the prevention of hypotension during cesarean delivery: a dose-dependent study. Int J Clin Exp Med 2014;7:5210–5216.
24. Wang Q, Zhuo L, Shen MK, Yu YY, Yu JJ, Wang M. Ondansetron preloading with crystalloid infusion reduces maternal hypotension during cesarean delivery. Am J Perinatol. 2014;31:913–922.
25. Terkawi AS, Tiouririne M, Mehta SH, Hackworth JM, Tsang S, Durieux ME. Ondansetron does not attenuate hemodynamic changes in patients undergoing elective cesarean delivery using subarachnoid anesthesia: a double-blind, placebo-controlled, randomized trial. Reg Anesth Pain Med. 2015;40:344–348.
26. Eldaba AA, Amr YM. Intravenous granisetron attenuates hypotension during spinal anesthesia in cesarean delivery: a double-blind, prospective randomized controlled study. J Anaesthesiol Clin Pharmacol. 2015;31:329–332.
27. Marciniak A, Owczuk R, Wujtewicz M, Preis K, Majdyło K. The influence of intravenous ondansetron on maternal blood haemodynamics after spinal anaesthesia for caesarean section: a double-blind, placebo-controlled study. Ginekol Pol. 2015;86:461–467.
28. Owczuk R, Wenski W, Twardowski P, et al. Ondansetron attenuates the decrease in blood pressure due to spinal anesthesia in the elderly: a double blind, placebo-controlled study. Minerva Anestesiol. 2015;81:598–607.
29. Rashad MM, Farmawy MS. Effects of intravenous ondansetron and granisetron on hemodynamic changes and motor and sensory blockade induced by spinal anesthesia in parturients undergoing cesarean section. Egypt J Anaesth. 2013;29:369–374.
30. Saravanan S, Kocarev M, Wilson RC, Watkins E, Columb MO, Lyons G. Equivalent dose of ephedrine and phenylephrine in the prevention of post-spinal hypotension in Caesarean section. Br J Anaesth. 2006;96:95–99.
31. Ngan Kee WD. Prevention of maternal hypotension after regional anaesthesia for caesarean section. Curr Opin Anaesthesiol. 2010;23:304–309.
32. Cooper DW. Caesarean delivery vasopressor management. Curr Opin Anaesthesiol. 2012;25:300–308.
33. Heesen M, Stewart A, Fernando R. Vasopressors for the treatment of maternal hypotension following spinal anaesthesia for elective caesarean section: past, present and future. Anaesthesia 2015;70:252–257.
34. Veeser M, Hofmann T, Roth R, Klöhr S, Rossaint R, Heesen M. Vasopressors for the management of hypotension after spinal anesthesia for elective caesarean section. Systematic review and cumulative meta-analysis. Acta Anaesthesiol Scand. 2012;56:810–816.
35. Heesen M, Kölhr S, Rossaint R, Straube S. Prophylactic phenylephrine for caesarean section under spinal anaesthesia: systematic review and meta-analysis. Anaesthesia. 2014;69:143–165.
36. Allen TK, George RB, White WD, Muir HA, Habib AS. A double-blind, placebo-controlled trial of four fixed rate infusion regimens of phenylephrine for hemodynamic support during spinal anesthesia for cesarean delivery. Anesth Analg. 2010;111:1221–1229.
37. Shen CL, Ho YY, Hung YC, Chen PL. Arrhythmias during spinal anesthesia for Cesarean section. Can J Anaesth. 2000;47:393–397.
38. Dyer RA, Reed AR, van Dyk D, et al. Hemodynamic effects of ephedrine, phenylephrine, and the coadministration of phenylephrine with oxytocin during spinal anesthesia for elective cesarean delivery. Anesthesiology. 2009;111:753–765.
39. Robson SC, Boys RJ, Rodeck C, Morgan B. Maternal and fetal haemodynamic effects of spinal and extradural anaesthesia for elective caesarean section. Br J Anaesth. 1992;68:54–59.
40. Ngan Kee WD. Chestnut DH, Polley LS, Tsen LC, Wong CA. Uteroplacental blood flow. Obstetric Anesthesia: Principles and Practice
. 2009:4th ed. Philadelphia: Mosby Elsevier; 37–53.
41. Ngan Kee WD, Lee SW, Ng FF, Tan PE, Khaw KS. Randomized double-blinded comparison of norepinephrine and phenylephrine for maintenance of blood pressure during spinal anesthesia for cesarean delivery. Anesthesiology. 2015;122:736–745.
42. Kinsella SM, Tuckey JP. Perioperative bradycardia and asystole: relationship to vasovagal syncope and the Bezold-Jarisch reflex. Br J Anaesth. 2001;86:859–868.
43. Gao L, Zheng G, Han J, Wang Y, Zheng J. Effects of prophylactic ondansetron on spinal anesthesia-induced hypotension: a meta-analysis. Int J Obstet Anesth. 2015;24:335–343.
Supplemental Digital Content
© 2016 International Anesthesia Research Society