Complications of operative vaginal delivery and provider volume and experience

Abstract

Objective To evaluate associations between operative vaginal delivery complications and provider experience (operative vaginal delivery volume and time since residency). Methods We included all operative vaginal deliveries between 2008 and 2014 at a tertiary care teaching hospital, stratified into forceps-assisted and vacuum-assisted deliveries. Complications included severe perineal lacerations (3rd and 4th degree) and neonatal injuries (subgaleal/subdural/cerebral hemorrhage, facial nerve injury, and scalp injury), which were identified by International Classification Diagnosis-9 codes. Providers were categorized by operative vaginal delivery volume (mean annual forceps- or vacuum-assisted deliveries over the study interval) and time since residency. Regression analyses were used to compare complication rates by provider volume and time since residency, adjusting for potential confounders, using 0–1 deliveries per year and <5 years since residency as reference groups. Results Nine hundred and thirty-four forceps and 1074 vacuums occurred. For forceps-assisted deliveries, severe perineal injury was decreased among providers with >10 forceps per year (aOR 0.50 [95%CI 0.30–0.81]) and at 15–19 years (aOR 0.45 [95% CI 0.22–0.94], and ≥25 years (aOR 0.45 [0.27–0.73]) since residency. There were no associations with neonatal injuries. Among vacuum-assisted deliveries, severe perineal injury decreased at ≥25 years since residency (aOR 0.35 [95%CI 0.17–0.74], with no association with provider volume. Neonatal injury decreased at 5–9 years (aOR 0.53 [95%CI 0.30–0.93]), and 15–19 years since residency (aOR 0.53 [95%CI 0.29–0.97]), due to differences in scalp injuries. Neonatal injuries other than scalp injury were rare. Conclusion Severe perineal lacerations decreased with increasing operative vaginal delivery experience, primarily among forceps-assisted vaginal delivery. Providers >5 years since residency may have lower scalp injury with vacuums, but this cohort was largely underpowered for neonatal injury.