Repeat Induction Versus Expectant Management After Failed Primary Labor Induction at Term at Kenyatta National Hospital: a Randomised Clinical Trial.

Abstract

BACKGROUND: While the rates of labor induction have increased globally, occurring in about 20% of facilities, Africa reports of an average rate of 4.4% (Kenya 3.9%) with a high rate of unmet need (66.0 –80.2%). Despite it being a common obstetric procedure, criteria for successful and failed induction of labor (IOL) have not been standardized and no universal consensus has been reached. Caesarean delivery is thus often erroneously perceived to be the solution to failed labor induction. A failure rate of 26% is reported in Kenyatta National Hospital (KNH) with 24.1% in West Africa and 17.3% in Ethiopia. Primary caesarean delivery (CS) rate due to failed IOL in KNH in patients undergoing primary IOL is 51.5%, in contrast to 10% in the United States (U.S) and 3% in Australia. The need for uniformity in its definition and management in creating standards on both the duration and number of times the induction can be attempted creates the necessity to study the value of repeat induction after failed primary labor induction. OBJECTIVE: To determine the effect of repeat induction versus expectant management on rates of failed induction among women who have primary failure of induction. SETTING: Labor ward, antenatal wards and neonatal unit in Kenyatta National Hospital. STUDY DESIGN: A single center double arm block Randomized Clinical Trial non-blinded to both participants and researchers, blinded to the analyst. STUDY POPULATION: Eligible women with medical and elective indications for labor induction who have failed primary labor induction at KNH. SAMPLE SIZE: Based on the formulae for comparison of means between two groups. A total of 86 gravid women with failed primary induction were randomized to either expectant management (n = 43) or repeat induction (n = 43). OUTCOME MEASURES: Primary outcome was onset of active labor. Secondary outcomes included randomization-delivery interval, mode of delivery, adverse maternal and early neonatal outcomes. 14 DATA COLLECTION: Structured questionnaire in line with the study objectives was used. DATA ANALYSIS: Statistical analysis was performed using R Studio Version 3.5.1. Simple linear regression model was used to analyze variables measured on a continuous scale. Simple logistic regression model was used to assess association between outcomes measured on a binary scale and the treatment groups. Probability values <.05 were considered statistically significant. Intention to treat principle was utilized in data analysis. RESULTS: Demographics and clinical characteristics were similar between groups. There were significant differences in the primary outcome between the two treatment groups. The time from randomization to active phase was 8 hours shorter in the repeat induction group (18.20 vs 26.01 hours; mean difference, 7.81 h; 95% confidence interval 12.27 to 3.35, p=0.001). Time from randomization to vaginal delivery was >8 hours less in the repeat induction group (24.78 vs 33.37 hours, mean difference 8.59 h, 95 % CI 13.63 to 3.56, p=0.002). Failure rates was higher in the expectant management group (44.2 % vs 20.9%, OR = 2.99, 95% CI 1.18 to 8.02, p = 0.023). No significant differences in maternal outcomes, early neonatal outcomes and mode of delivery between the 2 groups was observed (p value>0.05). CONCLUSION: Repeat labor induction using vaginal Misoprostol was superior to expectant management in achieving active phase of labor in patients with a low Bishop score not responsive to primary labor induction using vaginal PGE1. RECOMMENDATION: Consider change of policy with regard to definition and management of failed primary labor induction at the hospital.