(774) Hospital-level variation in cesarean birth after nulliparous term induction of labor

To estimate the variation in cesarean birth rates among nulliparous, term, singleton, vertex (NTSV) births that 1) underwent induction of labor (IOL) and 2) had an elective IOL.

Study Design:

This is a cross-sectional study across 68 birthing hospitals using clinically abstracted values from the Obstetrics Initiative, a quality improvement initiative supported by Blue Cross Blue Shield of Michigan. We included NTSV births undergoing IOL from 01/2020 to 12/2022. Risk and reliability adjusted cesarean rates were calculated using logistic linear mixed models with random intercepts for hospitals. The model included age, admission BMI, pre-gestational diabetes, gestational diabetes, hypertensive disorders of pregnancy, substance use disorder, social vulnerability index (SVI), hospital teaching status, and neonatal care level. 

Results:

Among 39,139 NTSV patients undergoing IOL, 13,364 (34.1%) had a cesarean birth. Most births occurred in a teaching hospital (87%). The median number of IOL per hospital was 352 (IQR: 192-844) with high variation in observed cesarean rates ranging from 18.8% to 46.3% (median 33.4%). In the model-based analysis, pre-pregnancy diabetes (aOR: 2.09), higher SVI (aOR: 1.57), older age (aOR: 1.06) and higher BMI (aOR: 1.06) were associated with cesarean birth (p< 0.05). Variation in cesarean was only slightly reduced after risk and reliability adjustment (median adjusted cesarean rate, 32.7%, range 21.9% to 44.4% [Figure]). In a sensitivity analysis among 10,971 elective IOL, we found similar results with risk and reliability adjusted cesarean rates ranging from 22.1% to 44.1% (median 30.2%). 

Conclusion:

High variability in risk-adjusted cesarean rates across hospitals after IOL highlights that lower rates of cesarean are achievable. Practice sharing regarding IOL management may lead to lower cesarean rates.