Cost-effectiveness analysis of prostaglandin E2 gel for the induction of labour at term.

OBJECTIVE To estimate the cost-effectiveness of prostaglandin E2 (dinoprostone) vaginal gel for the induction of labour at term from the perspective of the UK's National Health Service. DESIGN Economic evaluation conducted as part of a randomised controlled trial. SETTING Maternity department at a major teaching hospital in London, UK. POPULATION A cohort of 165 pregnant women presenting as cephalic between 36(+⁶) and 41(+⁶) weeks of gestation, for whom induction of labour was deemed necessary. METHODS Either 3-mg Prostin E2 vaginal tablets or 1- or 2-mg Prostin E2 vaginal gel were administered at 6-hourly intervals. MAIN OUTCOME MEASURES Incremental cost per hour prevented between induction and delivery. The nonparametric bootstrap method was used to construct cost-effectiveness acceptability curves and estimate net benefits at alternative cost-effectiveness thresholds. RESULTS Women receiving the gel accrued nonsignificantly higher costs (incremental cost £630; bootstrap 95% CI -£353, £2320; P = 0.43), and experienced a significantly reduced interval between induction and delivery (median of 1400 versus 1780 minutes; mean of 1711 versus 2765 minutes; P = 0.03). The incremental cost per hour prevented from induction of labour to delivery was estimated at £36. At a cost-effectiveness threshold of £100 per hour of care prevented, the probability that the gel is cost-effective was estimated at 0.83, and the mean net benefit to the health services was estimated at £1121 (bootstrap 95% CI -£1133, £3379). The results were sensitive to the inclusion of neonatal costs in the analysis and the value of the cost-effectiveness threshold. Notably, excluding neonatal costs increased the probability that the gel is cost-effective at a cost-effectiveness threshold of £100 per hour of care prevented to 0.99. CONCLUSIONS This study suggests that prostaglandin E2 gel is probably more cost-effective than prostaglandin E2 tablets for the induction of labour at term. Given that the results are applicable to the general obstetric population requiring induction of labour at term, decision-makers should consider the likely economic impacts of their implementation.