Expectant management up to 72 h after term prelabour rupture of membranes: a retrospective cohort study in the Netherlands.

OBJECTIVE

To evaluate outcomes of the Dutch policy permitting expectant management up to 72 h after term prelabour rupture of membranes (PROM) and to assess whether maternal or neonatal infection risk increase approaching this threshold.

STUDY

DESIGN

Retrospective cohort study.

SETTING

Two-site regional teaching hospital providing secondary obstetric care, Northwest Clinics, the Netherlands (2019-2024).

POPULATION: Women with singleton term PROM (≥37 weeks) who have given birth >24 h after rupturing of membranes.

METHODS

Women were stratified by latency: 24-48, 48-72 h and >72 h. Primary outcomes were composite neonatal infection (suspected or culture-confirmed early-onset sepsis (EOS) requiring antibiotics treatment or high-care admission for suspected infection) and composite maternal infection (intrapartum or postpartum infection requiring antibiotic treatment). Primary outcomes were analysed using logistic regressions adjusted for Group B streptococcus (GBS) status.

MAIN OUTCOME MEASURES: Maternal and neonatal infectious morbidity.

RESULTS

A total of 914 women were included. Composite neonatal infection occurred in 13.9% in the 24-48 h group, 13.0% in the 48-72 h group and 17.7% in the >72 h group (p = 0.47). Composite maternal infection occurred in 9.7%, 8.9% and 12.4% (p = 0.58). No culture-confirmed EOS occurred. Suspected EOS increased with latency 5.2%, 9.7%, 10.6% (p = 0.02). Neonatal acidosis was significantly higher after >72 (17.6%) as compared to the 48-72 h group (8.9%) and the 24-48 h group (9.8%) (p = 0.03). Caesarean birth rates were 13.6%, 10.7% and 19.6% across the three latency groups; although rates appeared higher beyond 72 h, these differences did not reach statistical significance. No neonatal deaths occurred.

CONCLUSIONS

Women undergoing expectant management for term PROM should be counselled regarding the potential increase in adverse outcomes with prolonged latency, while recognizing that this observational study does not allow conclusions regarding the effectiveness of induction at specific time points. Further studies are needed to determine the optimal timing of intervention.