Effect of midwifery staffing on the occurrence of spontaneous vaginal birth and use of pain medication – a target trial emulation protocol

This is a target trial emulation protocol:

Abstract:

Studies have suggested that midwifery staffing is associated with the proportion of spontaneous vaginal births. Due to organisational and ethical considerations, however, randomised controlled trials are not possible to establish a causal link. The objective is to investigate the causal effect of midwifery staffing levels on the prevalence of physiological vaginal births and pain medication use.

The causal question of this target trial is to determine the effect of receiving continuous one-to-one midwifery care during labour and birth on the proportion of spontaneous vaginal births compared to receiving one-to-one midwifery care partially during labour and birth. The target trial is emulated with routine hospital data from one tertiary hospital from 01 January 2019 to 31 December 2022. The exposure, midwifery staffing, is defined as receiving one-to-one care during active labour and birth either for 100% or for less than 100% of the time. The eligibility criteria are assessed at admission to the labour ward. Women with a planned caesarean section, breech birth, multiple pregnancy and stillbirth will be excluded. All remaining women who enter active labour are eligible for inclusion. Women assigned to each staffing exposure are assumed to be comparable conditional on certain baseline covariates. We consider hypertensive disorders, diabetes, gestational age, parity, maternal age, country of birth, and labour induction as confounding variables. We use g-methods to adjust for confounding and derive the total effect. Specifically, we consider both marginal structural models fitted by means of inverse probability weighting and the g-formula so that we can assess robustness with respect to different misspecifications. Ethical approval to access the routine hospital data has been granted by the local ethics committee [2024-00300].