RESUMO
BACKGROUND: Continuous labor support is widely acknowledged for potentially enhancing maternal and neonatal outcomes, the physiological labor process, and maternal satisfaction with the labor experience. However, the existing literature lacks a comprehensive analysis of the optimal characteristics of labor companions, particularly in comparing the effects of trained versus untrained and familiar versus unfamiliar labor companions across diverse geographical regions, both pre-and post-millennial. This meta-analysis addresses these research gaps by providing insights into the most influential aspects of continuous labor support. METHODOLOGY: A thorough search of PubMed, Google Scholar, Science Direct, International Clinical Trials Registry Platform (ICTRP), ClinicalTrials.gov, Research4Life, and Cochrane Library was conducted from 25/06/2023 to 04/07/2023. Study selection utilized the semi-automated tool Rayyan. The original version of the Cochrane Risk of Bias tool was used to assess the quality of Randomized Controlled Trials (RCTs) while funnel plots gauged the publication bias. Statistical analysis employed RevMan 5.4, using Mantel-Haenszel statistics and random effects models to calculate risk ratios with 95% confidence intervals. Subgroup analyses were performed for different characteristics, including familiarity, training, temporal associations, and geographical locations. The study was registered in INPLASY (Registration number: INPLASY202410003). RESULTS: Thirty-five RCTs were identified from 5,346 studies. The meta-analysis highlighted significant positive effects of continuous labor support across various outcomes. The highest overall effect without subgroup divisions was the improvement reported in the 5-minute Apgar score < 7, with an effect size of 1.52 (95% Confidence Interval (CI) 1.05, 2.20). Familiar labor companions were better at reducing tocophobia, with an effect size of 1.73 (95% CI 1.49, 2.42), compared to unfamiliar companions, with an effect size of 1.34 (95% CI 1.14, 1.58). Untrained labor companions were the better choice in reducing tocophobia and the cesarean section rate compared to trained companions. For the analysis of tocophobia, the pooled effect sizes were 1.34 (95% CI 1.14, 1.57) and 1.84(95% CI 1.60, 2.12) in trained versus untrained subgroup comparisons. For the cesarean rate, they were represented as 1.22 (95% CI 1.05, 1.42) and 2.16 (95% CI 1.37, 3.40), respectively. The pooled effect size for the duration of labor was 0.16 (95% CI 0.06, 0.26) for the subgroup of RCTs conducted before 2000 and 0.53 (95% CI 0.30, 0.77) for the subgroup of RCTs conducted after 2000. A significant subgroup difference (<0.1) was found in relation to the duration of labor, cesarean section rate, oxytocin for labor induction, analgesic usage, and tocophobia in the subgroup analysis of geographical regions. DISCUSSION AND CONCLUSION: The beneficial effects of a labor companion are well-established in the literature. However, studies systematically assessing the characteristics of labor companions for optimal beneficial effects are lacking. The current study provides insights into the familiarity, training, temporal association, and geographical settings of labor companions, highlighting the differing impact of these characteristics on measured outcomes by evaluating the current randomized controlled trials on the topic. There is insufficient evidence to define the 'best labor companion' owing to the heterogeneity of labor companions and outcome assessment across different studies. We encourage well-designed further research to fill the research gap.