Severe maternal and neonatal morbidity after attempted operative vaginal delivery.

ABSTRACT BACKGROUND Operative vaginal delivery (OVD) is a critical tool in reducing primary cesarean birth, but declining OVD rates and concerns about provider skill necessitate a clear understanding of risks. These risks are ambiguous because most studies compare outcomes with OVD to spontaneous vaginal delivery, rather than to second stage cesarean which is usually the realistic alternative. OBJECTIVE Our objective was to compare severe maternal and neonatal morbidity by mode of delivery among patients with a prolonged second stage of labor who had a successful OVD, a cesarean birth after failed OVD, or a cesarean birth without an OVD attempt. STUDY DESIGN We used a population-based database to evaluate nulliparous, term, singleton, vertex livebirths in California between 2007 and 2012 among patients with a prolonged second stage of labor. Birth certificate and ICD-9-CM coded diagnoses and procedures were used for ascertainment of exposure, outcome, and demographics. Exposure was mode of delivery among patients who had any OVD attempt versus cesarean without OVD attempt. The outcomes were severe maternal morbidity (SMM) and severe unexpected newborn morbidity (UNM), defined using established indices. Anticipating that the code for prolonged second stage of labor would represent only a fraction of true OVD candidates, a secondary analysis was conducted removing this restriction in order to explore granular outcomes in a larger cohort with unsuccessful labor. Multivariable logistic regression was used to compare outcomes by mode of delivery adjusted for measured confounders. Sensitivity analyses were done excluding patients with combined vacuum-forceps and birthweight >4000g. RESULTS 9,239 prolonged second stage births were included; 6,851 (74.1%) were successful OVDs, 301 (3.3%) were failed OVDs, and 2,087 (22.6%) were cesareans without OVD attempts. Of successful OVDs, 6,195 (90.4%) were vacuums and 656 (10.6%) were forceps. Of failed OVDs where OVD type was specified, 83 (47.4%) were vacuums, 38 (21.7%) were forceps, and 54 (30.9%) were combined vacuum-forceps. Of note, all 54 combined vacuum-forceps OVD attempts that we identified failed. Patients with failed OVD differed from those with successful OVD, with higher rates of comorbidities, use of combined OVD, and birthweight >4000 g. Successful OVD was associated with reduced SMM (aOR 0.55, 95% CI 0.39-0.78) without a difference in severe UNM (aOR 0.99, 95% CI 0.78-1.26). In contrast, failed OVD was associated with increased SMM (aOR 2.14, 95% CI 1.20-3.82) and severe UNM (aOR 1.78, 95% CI 1.09-2.86). Findings were similar in secondary analysis of 260,585 patients with unsuccessful labor. CONCLUSION In this large cohort of nulliparous, term, singleton, vertex births, successful OVD was associated with a 45% reduction in SMM without differences in severe UNM when compared to cesarean birth after prolonged second stage of labor. OVD failed infrequently, but when it did it was associated with a 214% increase in SMM and a 78% increase in severe UNM; combined OVDs were major contributors to this, since all combined OVDs failed. Optimization of OVD success rates through means such as improved patient selection, enhanced provider skill, and dissuasion against combined OVD could reduce maternal and neonatal complications.