The purpose of this study was to estimate costs and outcomes of subsequent trials of labor after cesarean delivery (TOLAC) compared with elective repeat cesarean deliveries (ERCD).
To compare TOLAC and ERCD, maternal and neonatal decision analytic models were built for each hypothetic subsequent delivery. We assumed that only women without previa would undergo TOLAC for their second delivery, that women with successful TOLAC would desire future TOLAC, and that women who chose ERCD would undergo subsequent ERCD. Main outcome measures were maternal and neonatal mortality and morbidity rates, direct costs, and quality-adjusted life years. Values were derived from the literature. One-way and Monte-Carlo sensitivity analyses were performed.
TOLAC was less costly and more effective for most models. A progression of decreasing incremental cost and increasing incremental effectiveness of TOLAC was found for maternal outcomes with increasing numbers of subsequent deliveries. This progression was also displayed among neonatal outcomes and was most prominent when neonatal and maternal outcomes were combined, with an incremental cost and effectiveness of –$4700.00 and .073, respectively, for the sixth delivery. Net-benefit analysis showed an increase in the benefit of TOLAC with successive deliveries for all outcomes. The maternal model of the second delivery was sensitive to cost of delivery and emergent cesarean delivery. Successive maternal models became more robust, with the models of the third-sixth deliveries sensitive only to cost of delivery. Neonatal models were not sensitive to any variables.
Although nearly equally effective relative to ERCD for the second delivery, TOLAC becomes less costly and more effective with subsequent deliveries.
Cesarean delivery has become increasingly common throughout the past 2 decades. Since 1996, cesarean delivery rates have risen >50% and reached an all-time high in 2009, comprising 32.3% of all births. In addition to increased procedural costs, cesarean delivery is associated with increased maternal morbidity when compared with vaginal delivery. With repeat cesarean deliveries comprising one-third of all cesarean deliveries in the United States, trial of labor after cesarean delivery (TOLAC) has been considered to be a potentially beneficial alternative. TOLAC has proved to be both effective and safe, with successful vaginal delivery occurring in approximately 75% of women who attempt TOLAC and with a significantly lower incidence of maternal and neonatal morbidity and death compared with elective repeat cesarean delivery (ERCD). However, TOLAC carries an increased risk of uterine rupture, which has been associated with increased neonatal complications. Such outcomes have influenced the high rate of ERCD and limited acceptance of TOLAC.
Previous cost-effectiveness analyses that examined TOLAC have found it to be cost-effective for women with 1 previous cesarean delivery. Although cost-effectiveness analyses have examined the effects of delivery method on a woman’s second pregnancy after a primary cesarean delivery, the literature regarding the implications of the decision between TOLAC and ERCD across multiple deliveries is limited. This question is particularly relevant because many maternal complications, which include placenta previa and accreta, increase incrementally with an increased number of cesarean deliveries. Both previa and accreta confer an elevated risk of maternal and neonatal morbidity and death, and successive repeat cesarean deliveries have been shown to result in higher cumulative maternal morbidity and deaths and a higher cumulative risk of neonatal morbidity. Previous comparison of the cost-effectiveness of TOLAC and ERCD across multiple deliveries used a population perspective and did not address the clinical value of a model that catered to specific future reproductive plans. A comparison of the cost-effectiveness of TOLAC and ERCD across a woman’s reproductive lifetime based on personalized reproductive plans has not yet been conducted. The objective of this study was to evaluate the costs and outcomes of TOLAC vs ERCD as a function of subsequent deliveries potentially to guide clinical decision-making and patient counseling after the initial cesarean delivery has occurred.
To compare the maternal and neonatal outcomes that result from successive TOLAC vs ERCD, separate decision analytic models for maternal and neonatal outcomes were developed for a hypothetic cohort of women with 1 previous full-term, low-transverse cesarean delivery. The analysis was conducted from a payer’s perspective, and effectiveness was measured as quality-adjusted life years (QALYs). Findings from the cost-effectiveness analysis were presented as incremental costs, incremental effectiveness, incremental cost-effectiveness ratio, and net monetary benefit. Incremental costs were calculated as the difference in the mean cost between the 2 strategies. Incremental effectiveness was calculated as the difference in the mean effectiveness (ie, QALYs). Incremental cost-effectiveness ratio was calculated as the incremental cost divided by the incremental effectiveness and was compared with a commonly accepted willingness to pay threshold of $50,000 per QALY to determine whether TOLAC was cost-effective compared with ERCD. Net monetary benefit (calculated as the willingness to pay × incremental effectiveness – incremental cost) was also used to assess the cost-effectiveness of TOLAC. If the net monetary benefit was greater than zero, TOLAC was cost-effective compared with ERCD. To evaluate the impact of route of delivery on maternal and neonatal deaths, the mortality rate per 1000 patients was also calculated for each model.
The models were developed with TreeAge Pro software (TreeAge Software, Inc, Williamstown, MA). Separate maternal and neonatal models were created for the initial delivery (ie, second overall delivery) after a primary cesarean delivery. In addition, separate maternal and neonatal models were created for each of the subsequent deliveries (third, fourth, fifth, and sixth overall deliveries; Figures 1 and 2 ). It was assumed that only women without contraindications that included, for example, placenta previa, malpresentation, and previous classic cesarean delivery underwent TOLAC for their second delivery. In addition, those women with a successful TOLAC during their second delivery were assumed to desire TOLAC for all subsequent deliveries, and those women who chose ERCD were assumed to undergo subsequent ERCD for future deliveries. All deliveries were assumed to be at term unless earlier delivery was indicated by abnormal placentation, and neonatal birthweight was assumed to be distributed normally. Last, it was assumed that the length of stay in the neonatal intensive care unit (NICU) was equal to the gestational age at delivery subtracted from 39 weeks’ gestation for births that were complicated by placenta previa or accreta. Women with fetal malpresentation were excluded from the model.
All modeling probabilities were obtained from the literature ( Table 1 ). If multiple estimates were available in the literature, a mean estimate was calculated by weighting the individual estimates by sample size and was used as the parameter value in the base case analysis.
|Variable||Baseline, %||Range, %||Reference|
|Previous cesarean deliveries|
|Accreta with previa|
|Previous cesarean deliveries|
|Accreta with no previa|
|Previous cesarean deliveries|
|Delivery with previa/accreta|
|No previous vaginal||26.48||19.00–36.50|
|≥1 previous vaginal delivery||10.40||9.65–11.20|
|Intensive care unit||3.12||2.50–4.27|
|Intensive care unit||33.15||16.50–66|
|Intensive care unit||1.41||1.38–1.45|
|Intensive care unit||0.23||0.22–0.25|
|Intensive care unit||1.41||1.38–1.45|
|Intensive care unit||18.46 a||9.92–30.03|
|Respiratory distress syndrome by gestational age|
|Mental retardation by gestational age|
|Cerebral palsy by gestational age|
|Death by gestational age|
|Severe or moderate cerebral palsy d||81.86||77.72–85.53|
|Respiratory distress syndrome by delivery method|
|Death by delivery method|
|Cerebral palsy by delivery method|
|Brachial plexus injury by delivery method|
|Permanent injury given brachial plexus injury||16.92||9.0–27.40|
|Neonatal intensive care unit|
The outcomes that were included in the maternal model were transfusion, hysterectomy, thromboembolism, intensive care unit admission, endometritis, and death. Neonatal outcomes included respiratory distress syndrome, brachial plexus injury, cerebral palsy, mental retardation, NICU admission, and death. The degree of severity of cerebral palsy was incorporated in the neonatal model and the occurrence of permanent brachial plexus injury.
The occurrence of placenta previa and placenta accreta was included in both the neonatal and maternal models. The incidences of previa and accreta were varied for each delivery based on the number of previous cesarean deliveries ( Table 1 ). Because both previa and accreta have been shown to be associated with an increased risk of preterm delivery, the incidences of neonatal complications that are associated with prematurity (cerebral palsy, mental retardation, respiratory distress syndrome, and neonatal death) were calculated based on gestational age for deliveries with previa or accreta. The distribution of the incidence of each complication by gestational age was matched with the corresponding distribution of gestational age for a delivery with a previa or accreta. To account for the heterogeneity of neonatal morbidity and death for deliveries at <34 weeks’ gestation, the distribution of these births in the setting of previa and accreta was further derived based on published literature, and a weighted average for the neonatal complications that are associated with prematurity was calculated. For women with normal placental presentation, delivery was assumed to occur at 39 weeks’ gestation, and neonatal complication rates were calculated accordingly for cerebral palsy and mental retardation. When uterine rupture occurred in the model, the rate of cerebral palsy was adjusted accordingly. For pregnancies without placenta previa or accreta, the probabilities of respiratory distress syndrome, neonatal death, and brachial plexus injury were based on route of delivery (ie, cesarean or vaginal). NICU incidence was based on route of delivery or presence of placenta previa or accreta.
Costs included hospital, obstetrician, pediatrician, and anesthesiologist ( Table 2 ). Obstetrician, pediatrician, anesthesiologist, intensive care unit admission, NICU admission, transfusion costs, and the long-term medical costs of cerebral palsy, permanent brachial plexus injury, and mental retardation were derived from the literature. All other hospital costs were obtained from the 2012 Hospital Cost and Utilization Project, which is a nationwide inpatient database that reports the national average costs for specific diagnoses and procedures. Total costs for short-term events were calculated by adding hospital and provider costs. Total NICU admission costs were calculated based on the length of stay as derived from gestational age or route of delivery. All costs were converted to 2012 dollars with the medical care component of the Consumer Price Index. All costs and all long-term outcomes and outcomes for future deliveries were discounted at a 3% annual rate to account for individuals’ time preference, valuation of present outcomes, and goods more than the same quantity of outcomes and goods in the future.
|Variable||Baseline, $||Range, $||Reference|
|Intensive care unit||11,184||5592–16,776|
|Respiratory distress syndrome||38,563||19,281.50–57,844.50|
|Brachial plexus injury||17,802||8901–26,703|
|Neonatal intensive care unit||3612||1806–5418|
|Brachial plexus injury||18,899.15||9449.58–28,348.73|