Objective
The purpose of this study was to compare labor induction and cesarean delivery rates at term in community vs university hospitals.
Study Design
A population-based retrospective cohort study of births was performed. Primary outcomes were term gestation at <39 weeks, labor induction, and cesarean delivery. After we adjusted for comorbidities, malpresentation, and previous cesarean delivery, logistic regression assessed the association between hospital type and primary outcomes.
Results
Births occur less often in week 37 (n = 24390 [11%] vs 4006 [13%]; adjusted odds ratio [OR], 0.9; 95% confidence interval [CI], 0.8–0.9) and are similar in week 38 in community vs university hospitals. Inductions occur more commonly in community vs university settings at 37 weeks (n = 6440 [27%] vs 757 [19%]; adjusted OR, 1.7; 95% CI, 1.5–1.8) and at 38 weeks (n = 16586 [31%] vs 1530 [21%]; adjusted OR, 1.8; 95% CI, 1.7–1.9). Cesarean rates are no different between hospital types.
Conclusion
Induction is 70-80% more likely at community vs university hospitals before the optimal gestational age of ≥39 weeks, but cesarean delivery rates do not differ at term.
Induction of labor is a common practice among obstetricians in the United States, and rates have continued to rise since the early 1990s. In 2006, 22.5% of all births were induced, which is more than double the rate of 9.5% in 1990. In addition, 16% of preterm deliveries and 24% of term deliveries were the result of inductions. Secondary to potential maternal and fetal risks that are associated with induction of labor, the American College of Obstetricians and Gynecologists has set forth guidelines that outline the acceptable indications for and contraindications to this frequent procedure. In addition to maternal and fetal indications that may necessitate an expeditious delivery, the College also recognizes that psychosocial and/or logistic factors may be acceptable reasons for planning an elective induction of labor, but not at <39 weeks of gestation that has been confirmed by strict dating criteria.
Elective induction of labor historically has been considered a significant risk factor for cesarean delivery. Numerous reports approximate a 2- to 3-fold increased risk of cesarean delivery in nulliparous women with an unfavorable Bishop score. Cesarean delivery rates in 2007 were at a record high of 31.8% of all births. Because this procedure is associated with higher rates of maternal hemorrhage, infections, placenta previa/accreta, and maternal death when compared with vaginal delivery, the increased risk with induction of labor is alarming. Therefore, it is essential to identify the population of women that is more likely to undergo an induction of labor. To our knowledge, the timing of term births or induction and cesarean delivery rates in different hospital settings has not been studied before; thus, the objective of this study was to establish these rates in university vs community hospitals. In theory, university institutions are academic centers with a goal of education and patient care and may be more likely than community hospitals to adhere to strict gestational age guidelines for the induction of labor. Also, the constant availability of an on-site physician at university centers is often not available at community hospitals; this may lead to community physicians making earlier decisions to perform cesarean deliveries. We hypothesize that, despite an increase in medical comorbidities in patients delivered at university hospitals, patients who deliver at community hospitals are delivering more often at 37 and 38 weeks gestation and that induction and cesarean rates are higher at term in community vs university hospitals.
Materials and Methods
Using the Ohio Department of Health’s birth certificate database of births for 2006-2007, a population-based retrospective cohort study was performed. Primary outcomes were the frequency of delivery at 37 and 38 weeks gestation and the rates of induction and cesarean delivery in these groups. Each outcome was compared between university and community hospitals and was stratified as delivery at gestational ages 37 0/7 to 37 6/7, 38 0/7 to 38 6/7, and 39 to ≤42 weeks. The database included 308,380 total births. Births that occurred at gestational ages <20 weeks and >42 weeks and births that involved fetal death or major congenital anomalies were excluded. Of the remaining 291,756 births, there were 288,625 births (98.9%) that occurred in a hospital setting. Births that occurred outside of the hospital or in free-standing birthing centers were excluded. Of the hospital deliveries, 5255 (1.8%) births occurred in an undocumented hospital, for a total of 283,370 births included in this study.
There were 129 institutions considered: 5 university hospitals and 124 community hospitals. A facility was considered to be a university hospital if it was the primary site of a medical university training program. All others were considered community hospitals.
Recognizing that hospitals with obstetrics and gynecology residency programs are also teaching institutions and may treat patients similarly to university hospitals, we also chose to compare those hospitals with an accredited obstetric and gynecology residency programs (teaching hospitals) with those without a residency program (nonteaching hospitals) to further validate our conclusions. There were 18 teaching hospitals and 111 nonteaching hospitals.
Taking into consideration known limitations of accurate identification of indication for labor induction from Ohio birth certificate data, we chose to evaluate only the presence or absence of labor induction and not the indication of delivery. The presence or absence of labor induction was considered very likely to be recorded accurately in the birth certificate record based on labor induction rates in this study (approximately 22%), which is concordant with nationally reported rates and a previous study that compared labor induction recorded in the Ohio birth certificate database to documentation in the individual medical record.
Statistical analyses were performed with STATA Release 10 software (StataCorp, College Station, TX). Demographic characteristics of the 2 groups were analyzed with unpaired Student t tests for continuous variables and chi-squared tests for categoric variables. Logistic regression analyses were performed to assess the association between the hospital setting and birth timing, induction, and cesarean delivery; we adjusted for the medical comorbidities that would be expected to influence delivery. Risk estimates were also adjusted for the presence of gestational and pregestational diabetes mellitus, chronic hypertension, preeclampsia, previous cesarean delivery, and nonvertex presentation. Adjusted odds ratios (aORs) and 95% confidence intervals [CIs] were reported for each of the primary outcomes.
The protocol for this study was approved, and a deidentified data set was provided by the Child Policy Research Center of Cincinnati Children’s Hospital Medical Center. This study was exempt from review by the Institutional Review Board at the University of Cincinnati, Cincinnati, OH.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
