Uterine Rupture Trends in Patients Pursuing Trial of Labor After Cesarean Delivery in the United States From 2010 to 2022

Abstract

National rates of trial of labor after cesarean delivery (TOLAC) have increased but the risk of uterine rupture remains. Clinical practice- and patient-level changes have also occurred in the past decade, and an evaluation of contemporary uterine rupture rates is needed. We conducted a repeated cross-sectional analysis aimed to evaluate U.S. trends in uterine rupture and perinatal outcomes among individuals undertaking TOLAC using natality files available in the National Vital Statistics System from 2010 to 2022. Temporal trends were characterized using Joinpoint regression with average annual percentage change (AAPC) and 95% CI reported. There were 2,888 uterine ruptures (0.28%) among 1,016,073 deliveries included. The rate of uterine rupture increased over time with a rate of 0.20% in 2010 and 0.37% in 2022 (AAPC 6.2, 95% CI 4.3–8.7). While rates of uterine rupture increased over time, it remains a rare event with a low absolute risk.

Précis:

Uterine rupture rates for those undertaking trial of labor after cesarean delivery have increased over time, but absolute rates remain low.

Introduction

Uterine rupture and perinatal morbidity are potential complications of trial of labor after cesarean delivery (TOLAC), but generally, TOLAC is considered a safe and reasonable alternative to repeat cesarean delivery (CD) among appropriately selected and counseled candidates.¹ A recent evaluation of U.S. trends in TOLAC using natality data found rising rates of attempted and successful TOLAC.¹ Simultaneously in the past decade, patient-level (e.g., advanced maternal age, obesity) and practice-level (e.g., rates of induction of labor) changes have also occurred.^2,3 Updated trends for rates of uterine rupture and perinatal complications among those undertaking TOLAC have not been reported at a national level during a contemporary time period. Therefore, we aimed to evaluate rates of uterine rupture and perinatal outcomes among individuals undergoing TOLAC using national vital statistics data.⁴

Methods:

This was a repeated cross-sectional analysis using natality files from the National Center for Health Statistics in the National Vital Statistics System (NVSS) from 2010 to 2022.^2,4 Individuals with singleton, cephalic, term pregnancies undertaking TOLAC with one or two prior CD were included. The primary outcome was uterine rupture defined as a ruptured uterus recorded for maternal morbidity on the birth certificate. Secondary outcomes included unplanned hysterectomy, maternal blood transfusion, maternal intensive care unit (ICU) admission (available for 2015 onward), neonatal ICU admission, and neonatal death.

Baseline characteristics were compared between those with and without uterine rupture using chi-square or two-sample t-tests as appropriate. Race and ethnicity were included in baseline characteristics of the study population but were not included in any modeling. Temporal trends were evaluated using annual rate and characterized using Joinpoint regression analysis with average annual percent change (AAPC) and 95% confidence interval (CI) reported. An adjusted log-binomial regression modeled rupture over time. Confounders were selected a priori based on available literature and clinical input and finalized in modeling via backwards selection. We modeled secondary maternal outcomes with the same parameterization as the adjusted model for the primary outcome.

Sensitivity analysis was performed for difference between year and rupture by history of CD, both by using an interaction term for history of CD (one or two prior) and year. This study was considered exempt from Institutional Review Board approval because data were deidentified. Analyses were performed using SAS and the Joinpoint Regression Program. A P value <0.05 was used for statistical significance.

Results:

Overall, there were 1,016,073 deliveries available for analysis and uterine rupture occurred in 2,888 (0.28%) deliveries. Those with uterine rupture were more likely to be older, have gestational diabetes, be diagnosed with a hypertensive disorder of pregnancy, undergo induction or augmentation of labor, and have two prior cesarean deliveries. Body mass index did not differ between those with and without uterine rupture (Appendix 1, available online at http://links.lww.com/xxx).

Uterine rupture rate increased over the study period from 0.20% in 2010–2014 to 0.37% in 2020–2022 with an annual percentage increase 6.2% (95% CI 4.3–8.7; Figure 1). Similarly, rates of transfusion and maternal ICU admission increased over time while the rate of unplanned hysterectomy did not substantially change (Table 1). In the log-binomial adjusted models for maternal outcomes, results were consistent and remained significant.

Figure 1.

Trends in uterine rupture and perinatal outcomes among those undertaking trial of labor after cesarean (TOLAC) from 2010 to 2022. Average Annual Percentage Change (AAPC) with 95% CI from Joinpoint analysis. Maternal outcomes with zero Joinpoints (i.e., linear), neonatal death estimated with two Joinpoints, and neonatal intensive care unit (NICU) admission with one Joinpoint. Maternal morbidity over time (A) and neonatal morbidity over time (B). Intensive care unit (ICU).

Table 1.

Unadjusted and adjusted estimate of uterine rupture and perinatal outcomes over time among individuals undertaking trial of labor after cesarean (TOLAC)

	Incidence			AAPC*
Outcomes	2010–14 (n=301,030)	2015–19 (n=438,407)	2020–22 (n=275,784)	Unadjusted	Adjusted†
Maternal
Uterine rupture	637 (0.21)	1,244 (0.28)	1,007 (0.37)	6.25 (5.14–7.38)	4.28 (3.13–5.44)
Unplanned hysterectomy	187 (0.06)	308 (0.07)	191 (0.07)	1.58 (−0.54–3.75)	0.14 (−2.08–2.41)
Blood transfusion	1162 (0.39)	2,126 (0.48)	1,859 (0.67)	7.17 (6.31–8.02)	5.99 (5.11–6.90)
ICU admission	400 (0.13)	690 (0.16)	485 (0.18)	3.32 (1.89–4.79)	2.09 (0.58–3.62)
Neonatal
Neonatal death	244 (0.1)	163 (0.0)	113 (0.0)	–	–
NICU admission	14,250 (4.7)	23,712 (5.4)	15,417 (5.6)	–	–

Data are n (%). Intensive care unit (ICU), neonatal intensive care unit (NICU), confidence interval (CI), Average Annual Percent Change (AAPC). Time is grouped into 5-year increments for tabular summary of incidence with the third group truncated due to availability of data; AAPC incorporates time in one-year increments for the 13 years included in the study period.

^*AAPC estimate from log binomial regression of outcome difference for 1-year difference, estimate transformed as 100*exp(B)-1. Year parameterized linearly for maternal endpoints is a validated assumption per Joinpoint analysis; data do not support assumption of linearity for neonatal outcomes and therefore estimates are not provided.

^†Adjusted model includes maternal age, delivery weight, tobacco use, interval since last delivery, induction of labor, augmentation of labor, year of delivery. – Modeling not completed.

For neonatal outcomes, neonatal death decreased over time with an annual percentage decrease of 6.4% (95% CI −8.5– −3.8) while admission to the NICU increased (AAPC 1.8%, 95% CI 1.3–2.2). The sensitivity analyses did not find a significant difference between the association between rupture by year or number of prior CDs.

Discussion:

While the uterine rupture rate increased over time, it remains a rare occurrence for those attempting TOLAC. The rate of uterine rupture (0.28%) is lower than that found in prior studies but is aligned with other recent data.^5,6,7 Similarly, rates of maternal ICU admission, NICU admission and transfusion increased, while neonatal death decreased. These findings extend our understanding of risks related to undertaking TOLAC in the modern era.

Limitations of this study include the potential for misclassification and under ascertainment from the use of birth certificate data, and the absence of granular clinical details around labor management. Natality data does not differentiate between intrapartum uterine rupture and incidental dehiscence identified at time of surgery. Strengths of the current study include use of nationally representative data. It is unknown whether our findings reflect changes in practice, approach to TOLAC management, or other factors.