Research

The 9-point caesarean gap nearly vanished after trial emulation

NICU360 Evidence Desk·7 min read·00

Among 74,423 very preterm infants, a crude 9.1-point gap in death or severe IVH narrowed to 0.3–1.8 points after intended-mode weighting. The breech signal remains observational.

THE HOT TAKE

Trial emulation turned a striking 9.1-point crude association into small, statistically inconclusive differences. That changes the inference, not the uncertainty.
— NICU360 Evidence Desk

The July 2026 PRECIOUS report used the National Neonatal Research Database to study 74,423 liveborn singleton or twin infants delivered at 22+0 to 31+6 weeks and admitted to NHS neonatal units in England (2012–2022) or Wales (2015–2022). Its prespecified primary outcome was death before neonatal-unit discharge or survival with grade 3–4 intraventricular haemorrhage on cranial ultrasound.

Weighted risk of death or severe IVH in the three emulations (%)

Whole population

CASSAVA+

PRECIOUS+

Intended vaginal birth

Was this worth carrying into your next shift?

Your reaction stays connected to your NICU360 profile.

Comments

Join the conversation

Loading conversation…

Emulation	Eligible frame	Intended vaginal	Intended caesarean	Risk difference
Whole population (n=18,856)	Whole cohort	13.3% (95% CI 12.5–14.0)	13.0% (95% CI 12.2–13.8)	−0.3 points (95% CI −1.4 to 0.8); P=0.609
CASSAVA+ (n=14,307)	22–31 weeks; cephalic or breech	12.9% (95% CI 12.1–13.6)	12.3% (95% CI 11.2–13.3)	−0.6 points (95% CI −1.8 to 0.6); P=0.351
PRECIOUS+ (n=5,553)	23–27 weeks; spontaneous labour	26.9% (95% CI 25.3–28.5)	25.1% (95% CI 23.3–27.0)	−1.8 points (95% CI −4.3 to 0.7); P=0.155

Emulation	Adjusted odds ratio	95% CI	Statistical reading
Whole population	0.73	0.64–0.83	Interval excludes 1
CASSAVA+	0.65	0.56–0.76	Interval excludes 1
PRECIOUS+	0.81	0.70–0.94	Interval excludes 1

The 9-point caesarean gap nearly vanished after trial emulation

THE HOT TAKE

Comments

HOW THE RAW GAP CHANGED

Observe actual mode

Infer intended mode

Identify likely equipoise

Weight measured confounders

Compare standardized risks

THE BREECH SIGNAL

WHAT THE STUDY CAN AND CANNOT SEE

WHY A DEFINITIVE TRIAL IS HARD

PRACTICAL INTERPRETATION