Modelling parity-specific timing of labour induction for prolonged pregnancy.

Background: More than 70,000UK women annually undergo induction of labour for prolonged pregnancy, as the risk of perinatal mortality increases with advancing gestation. Induction itself carries risks and is associated with a less positive birth experience. This study aims to investigate if timing of induction of labour for prolonged pregnancy can be optimised based on parity, to reduce the number of inductions without increasing adverse neonatal outcomes.

 

Methods: A mathematical model was developed to estimate the number of inductions, stillbirths and caesarean sections, and length of induced labour for different induction policy combinations for primigravidae and multigravidae. Model input was based on large population studies and records from a large UK maternity unit (n=35420). The model was validated by comparison with source and non-source historical data and sensitivity analyses with alternative inputs were performed.

 

Results: Model output agreed well with historical data. No policy combination led to a reduction in inductions and caesarean sections without increasing stillbirths, regardless of model input, and length of induced labour increased with gestational age for primigravidae. By combining the current policy of day 11 for primigravidae and any later gestation for multigravidae, the model suggests that inductions could potentially be reduced by more than 30 to 40% at a cost of a possible 1.5 to 2% increase in caesarean sections. The scope for delaying induction in multigravidae depended on the method used to calculate stillbirth risk, which varies in the literature.

 

Conclusion: Within the context of the model, there appears to be a potential to substantially reduce the number of inductions for prolonged pregnancy without an increase in stillbirths, by implementing different induction policies for primigravidae and multigravidae. Stillbirth risk calculations are currently subject to further testing with additional data.

Share this: