Interventions for treating hyperemesis gravidarum: a network meta-analysis of randomized clinical trials

Abstract

Background: Several interventions were explored in clinical trials for treating hyperemesis gravidarum (HG). The present study is a network meta-analysis of such interventions.

Methods: Electronic databases were searched for appropriate randomized clinical trials comparing interventions for treatment of patients with HG. Control of HG symptoms was the primary outcome and emetic episodes, hospital stay, nausea scores, patients requiring rescue antiemetics, hospital readmission, adverse events, and adverse pregnancy outcomes were the secondary outcome measures. Random-effects model was used and odds ratio (OR) [95% confidence interval (CI)] was the effect estimate for categorical outcomes and weighted mean difference (WMD) [95% confidence interval] for numerical outcomes.

Results: Twenty studies were included in the systematic review and 18 in the meta-analysis. Acupuncture (OR: 18.9; 95% CI: 2.1, 168), acupressure (OR: 26.7; 95% CI: 2.5, 283.1) and methylprednisolone (OR: 6.7; 95% CI: 1.1, 38.8) were associated with better control of HG symptoms than standard of care. Acupressure decreases the requirement of rescue antiemetics (OR: 0.06; 95% CI: 0.01, 0.44); ondansetron with reduced hospital stay (WMD: −0.2; 95% CI: −0.31, −0.01) and diazepam with reduced risk of hospital admission (OR: 0.11; 95% CI: 0.01, 0.95). The quality of evidence is very low.

Conclusion: Acupuncture, acupressure, and methylprednisolone were observed with better therapeutic benefits than other interventions for treating HG. However, the pooled estimates may change with the advent of results from future head-to-head clinical trials.

Keywords:

• Acupressure
• acupuncture
• diazepam
• ginger
• methylprednisolone
• metoclopramide
• ondansetron
• prednisolone
• promethazine
• vitamin B6

Acknowledgments

We thank PROSPERO for registering the review protocol. We are also grateful to Cochrane for utilizing RevMan 5.3 software for generating risk of bias graph. We thank EpiGear for using MetaXL in generating mixed treatment comparison results for the outcome measures.

Disclosure statement

No potential conflict of interest was reported by the authors.