VALUE IN HEALTH ] (2016) ]]]–]]]
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Economic Evaluations Comparing a Trial of Labor with an Elective Repeat Cesarean Delivery: A Systematic Review Anna Joy Rogers, MA1,*, Nathaniel G. Rogers, MD2, Meredith L. Kilgore, RN, MSPH, PhD1, Akila Subramaniam, MD, MPH3, Lorie M. Harper, MD, MSCI3 1
Department of Health Care Organization and Policy, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA; Departments of Medicine and Pediatrics, University of Tennessee Health Science Center, Memphis, TN, USA; 3Maternal-Fetal Medicine Division, Department of Obstetrics and Gynecology, Center for Women’s Reproductive Health, School of Medicine, University of Alabama at Birmingham, Birmingham, AL, USA 2
AB STR A CT
Background: For women who have had a previous low transverse cesarean delivery, the decision to undergo a trial of labor after cesarean (TOLAC) or an elective repeat cesarean delivery (ERCD) has important clinical and economic ramifications. Objectives: To evaluate the cost-effectiveness of the alternative choices of a TOLAC and an ERCD for women with low-risk, singleton gestation pregnancies. Methods: We searched EMBASE, MEDLINE, CINAHL, Cochrane Library, EconLit, and the Cost-Effectiveness Analysis Registry with no language, publication, or date restrictions up until October 2015. Studies were included if they were primary research, compared a TOLAC with an ERCD, and provided information on the relative cost of the alternatives. Abstracts and partial economic evaluations were excluded. Results: Of 310 studies initially reviewed, 7 studies were included in the systematic review. In the base-case analyses, 4 studies concluded that TOLAC was dominant over ERCD, 1 study found ERCD
Introduction As the most frequently performed surgical procedure in the United States [1], cesarean deliveries accounted for 32.2% of nearly 4 million births in 2014 [2]—a substantial increase from 20% in 1996 and 5% in 1970 [3]. Although the optimum mode of delivery after a previous cesarean delivery is dependent on various individual characteristics and the indications for the primary cesarean birth [4,5], clinical consensus is that a trial of labor after cesarean (TOLAC) is generally a safe alternative to an elective repeat cesarean delivery (ERCD) for most women [6]. Despite these guidelines, more than 90% of women who have had a previous cesarean delivery have a repeat cesarean in subsequent pregnancies [7]. Although these women may avoid uterine rupture, they are increasingly susceptible to greater maternal morbidity than their successful TOLAC counterparts, especially across multiple cesareans [8]. In addition to clinical consequences, the decision to undergo a TOLAC or an ERCD has important economic ramifications. The cost incurred by hospitals, and thus passed on to payers in the
to be dominant, and 2 studies found that although TOLAC was more costly, it offered more benefits and was thus cost-effective from a population perspective when considering societal willingness to pay for better outcomes. In sensitivity analyses, cost-effectiveness was found to be dependent on a high likelihood of TOLAC success, low risk of uterine rupture, and low relative cost of TOLAC compared with ERCD. Conclusions: For women who are likely to have a successful vaginal delivery, routine ERCD may result in excess morbidity and cost from a population perspective. Keywords: cesarean delivery, cost, economics, effectiveness, elective surgical procedures, pregnancy complications, probability, qualityadjusted life-years, trial of labor, utility, vaginal birth after cesarean. Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc.
health system, is generally considered to be greater for an ERCD because of higher resource use and longer average hospital stays [9]. Nevertheless, because of the various types of costing methods, the estimated economic impact has varied by study. Although most studies have found cesarean births to be more costly than vaginal births when assessing payments made [10–12], a few studies have reported that vaginal deliveries offer only a modest reduction in cost [13,14]. Some, however, have found that cesarean deliveries may be equivalent to vaginal deliveries, or even cost saving, in terms of economic impact after accounting for variables such as oxytocin augmentation, epidural anesthesia, or failed trials of labor [15,16]. From a public health perspective, it is crucial to maximize health outcomes while stewarding constrained resources. Economic evaluations allow for the formal comparison of alternative interventions with simultaneous regard to their health effects, resource use, and costs [17]. The objective of this systematic review was to analyze the range and quality of studies that address the cost-effectiveness of the alternative choices of a
* Address correspondence to: Anna Joy Rogers, Department of Health Care Organization and Policy, School of Public Health, University of Alabama at Birmingham, SHEL 121, 1530 3rd Avenue South, Birmingham, AL 35294. E-mail:
[email protected]. 1098-3015$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jval.2016.08.738
2
VALUE IN HEALTH ] (2016) ]]]–]]]
TOLAC and an ERCD for women with a singleton gestation who have had a previous low transverse cesarean delivery.
Results Search Results and Quality Assessment
Methods Conduct of Systematic Review We developed a systematic review protocol using the Preferred Reporting Items for Systematic review and Meta-Analysis for Protocols (PRISMA-P) [18,19] and registered it in the PROSPERO database (CRD42015029177) [20]. In consultation with a librarian, we searched the databases EMBASE, MEDLINE, CINAHL, Cochrane Library, EconLit, and the Cost-Effectiveness Analysis Registry without language, publication, or date restrictions up until October 2015. Details of the search strategy are available in Appendix A. The reference lists of included articles were assessed for additional relevant studies.
Study Selection Studies were included if they met the following criteria: 1) had to be primary research; 2) compared TOLAC versus ERCD; and 3) provided information on the relative cost of each course of action. One investigator screened all the abstracts of retrieved articles according to the inclusion criteria, erring toward inclusion in the case of uncertainty. Two investigators then assessed full-text articles that passed the initial screen, with disagreements on which articles to include resolved by discussion. Conference abstracts for which full-text articles were unavailable and partial economic evaluations, which comprise “cost analyses, cost-description studies and cost-outcome descriptions,” were excluded [21].
Data Extraction and Quality Assessment Two investigators independently abstracted data from each article using the Guide to Community Preventive Services Economic Evaluation Abstraction Form [22,23] and the recommendations of the Panel on Cost-Effectiveness in Health and Medicine [24]. The principal summary measures of interest were incremental cost-utility ratios (i.e., the additional cost per utility—a measure of both length of life and subjective level of well-being), cost-effectiveness ratios (i.e., the additional cost per unit of effect —identical units of an outcome or event such as death), or costbenefit ratios (i.e., the additional cost per benefit—in which effects of alternative interventions are expressed in monetary units) [21]. A validated quality assessment tool was considered [25], but not used because there is a lack of consensus in the literature as to the ideal instrument for economic evaluations [26,27]. We used the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist because it offers more precise guidance as to which components are essential. We scored studies on whether they completely satisfy (1), partially satisfy (0.5), or do not satisfy (0) each component, with a maximum score of 24 (Appendix B) [28]. The level of agreement was assessed with kappa statistics, and differences in CHEERS scores were resolved by discussion.
Costs To make the studies comparable, all costs were adjusted to 2016 US dollars using purchasing power parities for the year of the data [29] and then adjusted for inflation using the medical care component of the Consumer Price Index [30].
Database searches yielded 310 records, 224 of which were unique citations (Fig. 1). Seven studies [31–37] met the inclusion and exclusion rubric for this systematic review (Table 1). Six studies presented a cost-utility analysis [31–36], and one study presented a cost-effectiveness analysis [37]. Six of the studies [31,32,34–37] were conducted in North America and one [33] in Europe. CHEERS scores ranged from 11 to 23, with most of the studies providing comprehensive information about their methods and results. The study by Chuang et al. [31] was an outlier in terms of its CHEERS score because of incomplete reporting on items such as study perspective, preference-based outcomes, and model assumptions; the other studies had scores ranging from 21 to 23. The inter-rater reliability for the raters was found to be κ ¼ 0.773 (95% confidence interval 0.649–0.897).
Study Characteristics Decision modeling and analytic horizon The included studies primarily defined their populations as being women who had one previous low transverse cesarean delivery and no contraindications to labor. All studies presented a decision tree comprising the decision node of a TOLAC versus an ERCD, with the TOLAC branch allowing for either a successful or failed trial of labor, and uterine rupture as a potential severe adverse event of TOLAC. Two studies [31,33] used a short-term analytic horizon, whereby costs and consequences accrued in the immediate time period surrounding the index pregnancy. The remaining studies used long-term analytic horizons and reported discount rates for both costs and utilities, thus evaluating future costs and consequences in terms of their present value. Two studies [34,37] allowed for the decision made in the index pregnancy to impact the outcomes and costs of future pregnancies by using a Markov model.
Study perspective The study perspective, the viewpoint from which the costs are calculated (hospital, payer, or society), varied among the studies. Three studies [33,36,37] reported the payer’s perspective—the insurance company or the government (in the case of Ireland)— as the entity responsible for paying for the costs. In these studies, cost components generally included short-term hospitalization costs (direct and indirect), professional fees, and long-term costs. Among the three studies [32,34,35] with a societal perspective— which considers the costs borne by the patient and is thus recommended by the Panel on Cost-Effectiveness in Health and Medicine—patient out-of-pocket costs and loss of employment productivity were added into calculations [24]. Chuang et al. [31] did not report the perspective taken, but it appears to be that of the hospital.
Utility inputs The summary measures that were reported differed among the studies. The study by Grobman et al. [37] calculated the incremental cost of an ERCD over a TOLAC for one major neonatal adverse outcome (death or cerebral palsy). The other six studies assigned a utility value to the various delivery outcomes and complications. Chung et al. [32] used the Quality of Well-Being classification system [38] to derive utilities and were subsequently quoted in articles by Fawsitt et al. [33] and Gilbert et al. [34,35]. Chuang et al. [31] used several studies that measured patient preference values with regard to labor and postpartum
VALUE IN HEALTH ] (2016) ]]]–]]]
3
Fig. 1 – Study flow chart. ERCD, elective repeat cesarean delivery; TOLAC, trial of labor after cesarean. complications. Wymer et al. [36] used utilities drawn from previously published cost-effectiveness analyses.
Outcome probability inputs Although all studies assessed maternal outcomes (such as operative injury, endometritis, hysterectomy, and mortality) and most assessed infant outcomes, Chuang et al. [31] and Fawsitt et al. [33] did not include infant outcomes (such as cerebral palsy, hypoxic ischemic encephalopathy, and mortality) in their analyses (see Tables 2 and 3).
Bias Ideally, to compare a TOLAC with an ERCD without confounding factors influencing the outcomes, women would be randomized in a trial to one of the delivery approaches. Nevertheless, challenges with obtaining consent to assign mode of delivery make such a trial unlikely, especially given the large sample size that would be required to power the study. Thus, outcome probabilities for the decision trees in five of the studies [31–33,36,37] were extracted from published observational studies, in which women had a choice of delivery mode. Although both the studies of Gilbert et al. [34,35] had the advantage of a large registry (the Cesarean Registry comprising 19 clinical centers by the Maternal-Fetal Medicine Units Network) from which to draw outcome probabilities, only one of the studies [35] was designed to address the lack of randomization by using propensity scores to balance observed and unobserved baseline covariates [39,40].
Synthesis of Results Base-case analyses When comparing an ERCD and a TOLAC on the two dimensions of effectiveness and cost, the ERCD could be dominant (less costly and more effective), dominated (more costly and less effective), less effective but also less costly, or provide additional effect but at an additional cost. In the latter two scenarios, society would
need to determine whether it is willing to pay a premium for the additional benefit, thus determining the cost-effectiveness of the alternatives. Four studies [33–36] (Table 4) found in their basecase analyses that ERCD was dominated by TOLAC. One study [31] found that ERCD dominated TOLAC, and two studies [32,37] concluded that TOLAC was more cost-effective when considering societal preferences. Among the latter three studies, Chuang et al. [31] reported in base-case calculations that ERCD provided additional utility at a lower cost when compared with TOLAC, but in their discussion concluded that the strategies were “more or less equivalent” when taking parameter uncertainties into account. Chung et al. [32] found that an ERCD would cost an additional $195,668 per quality-adjusted life-year (QALY) in 2016 US dollars, thus providing more utility than a TOLAC, but at a higher cost. Because the incremental cost-utility ratio of Chung et al. [32] was higher than a commonly used societal willingness-to-pay threshold of $50,000 per QALY, the authors concluded that a TOLAC was the cost-effective option when all model parameters were set at their most likely values. Even if Chung et al. [32] had used a higher threshold of $100,000 per QALY gained, TOLAC would have still been preferred by society. Grobman et al. [37] reported that averting one major adverse neonatal outcome (death or cerebral palsy) would require 1591 cesarean deliveries and come at a cost of $4.33 million in 2016 US dollars. Grobman et al. [37] ultimately concluded that routine ERCD would come at a high cost and result in an excess of maternal morbidity and mortality. Future pregnancies were considered in three studies. Wymer et al. [36] compared successive delivery outcomes and determined that the benefits of a TOLAC accrue over multiple cesareans, with a decreasing incremental cost and increasing probability of cost-effectiveness. Grobman et al. [37] and Gilbert et al. [34] did not compare successive models but rather built the events over a woman’s reproductive life into their base-case model and concluded qualitatively that the risks and costs associated with routine ERCD are potentially greater than societal willingness to pay.
4
Table 1 – Characteristics of included economic evaluation studies. Reference (year)
Country (perspective)
Chuang et al. [31] (1999)
United States (not reported)
Chung et al. [32] (2001)
Population definition and baseline risk
Analytic horizon and discount rates
Major cost analysis components
CHEERS score*
Pregnant women with one previous low transverse cesarean delivery No contraindications to labor
Decision tree Deterministic SA
Immediate NA
Hospitalization cost (mother only) Professional fee
11/22
United States (societal)
30-y-old pregnant woman with one previous low transverse cesarean delivery Baseline risk not reported
Decision tree Deterministic SA
Long-term Cost: 3% QALYs: 3%
Hospitalization cost (mother and infant) Professional fee Long-term medical costs
21.5/23
Fawsitt et al. [33] (2013)
Ireland (health system)
Pregnant women with an uncomplicated pregnancy No contraindications to labor
Decision tree Deterministic SA Probabilistic SA conducted by Monte-Carlo simulation with 10,000 iterations
6 weeks NA
Hospitalization cost (mother only) Professional fee
22/22
Gilbert et al. [34] (2013)
United States (societal)
28-y-old pregnant women (singleton, term, vertex gestation) with one previous low transverse cesarean delivery No contraindications to labor
Decision tree, extended through lifetime using Markov modeling Deterministic SA Probabilistic SA conducted by Monte-Carlo simulation with 10,000 iterations
Long-term (with lifecycle) Cost: 3% QALYs: 3%
Hospitalization cost (mother and infant) Professional fee Long-term medical costs Patient opportunity costs
23/23
Gilbert et al. [35] (2013)
United States (societal)
Pregnant women (singleton, term, vertex gestation) with one previous low transverse cesarean delivery No contraindications to labor
Decision tree Deterministic SA Probabilistic SA conducted by Monte-Carlo simulation with 10,000 iterations
Long-term Cost: 3% QALYs: 3%
Hospitalization cost (mother and infant) Professional fee Long-term medical costs Patient opportunity costs
22/23
Grobman et al. [37] (2000)
United States (payer)
Pregnant women with one previous low transverse cesarean delivery Baseline risk not reported
Decision tree, extended through reproductive life using Markov modeling Deterministic SA
Long-term (with lifecycle) Cost: 3% Effectiveness: NA
Hospitalization cost (mother and infant) Long-term medical costs
21/23
Wymer et al. [36] (2014)
United States (payer)
25-y-old pregnant women with one previous full-term, low transverse cesarean delivery No contraindications to labor
Reproductive life Cost: 3% QALYs: 3%
Hospitalization cost (mother and infant) Professional fee Long-term medical costs
23/23
Decision trees constructed for each subsequent delivery Deterministic SA Probabilistic SA conducted by Monte-Carlo simulation with 10,000 iterations
CHEERS, Consolidated Health Economic Evaluation Reporting Standards; NA, not applicable; QALYs, quality-adjusted life-years; SA, sensitivity analysis. * The CHEERS statement has a 24-item checklist. We assessed whether the studies completely satisfy (1), partially satisfy (0.5), or do not satisfy (0) each component, with a maximum score of 24. Nevertheless, because none of the studies conducted heterogeneity analyses and two studies conducted short-term analyses that did not require the inclusion of discount rates, the total possible score for any study was 22 or 23.
VALUE IN HEALTH ] (2016) ]]]–]]]
Modeling method and SA method(s)
Table 2 – Comparison of maternal outcomes assessed by each study. Reference (year)
Wound complication, infection
Hysterectomy
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Uterine rupture*
Operative injury
Blood transfusion
●
●
●
●
Endometritis
●
Thromboembolism
Urinary and/ or fecal incontinence
Death
●
● (both)
●
●
●
● (SUI only)
●
●
●
●
●
●
●
●
Placenta previa and accreta
●
●
●
VALUE IN HEALTH ] (2016) ]]]–]]]
Chuang et al. [31] (1999) Chung et al. [32] (2001) Fawsitt et al. [33] (2013) Gilbert et al. [34] (2013) Gilbert et al. [35] (2013) Grobman et al. [37] (2000) Wymer et al. [36] (2014)
Maternal outcomes
SUI, stress urinary incontinence; TOLAC, trial of labor after cesarean. * Uterine rupture could have been included in the model either as a final outcome or as a branch probability resulting from a TOLAC. The studies led by Chuang et al., Chunget al., and Grobman et al. defined uterine rupture as symptomatic uterine tears requiring operative intervention. The studies by Gilbert et al. defined uterine rupture as “a disruption or tear of the uterine muscle and visceral peritoneum or a uterine muscle separation with extension to adjacent structures.” The studies by Wymer et al. and Fawsitt et al. did not explicitly define uterine rupture.
5
6
VALUE IN HEALTH ] (2016) ]]]–]]]
● ●
HIE, hypoxic ischemic encephalopathy; NICU, neonatal intensive care unit; RDS, respiratory distress syndrome; TTN, transient tachypnea of the newborn.
● ● ● ● ●
●
● ● ●
●
Chuang et al. [31] (1999) Chung et al. [32] (2001) Fawsitt et al. [33] (2013) Gilbert et al. [34] (2013) Gilbert et al. [35] (2013) Grobman et al. [37] (2000) Wymer et al. [36] (2014)
●
●
● ● ●
● ● ●
● ● ●
●
●
Acidemia Infection/ sepsis RDS TTN NICU admission HIE Permanent neurological injury/ cerebral palsy Not assessed Reference (year)
Neonatal outcomes Table 3 – Comparison of neonatal outcomes assessed by each study.
●
●
●
●
Brachial plexus injury
●
Death
Sensitivity analyses To account for uncertainty in the model parameters, all studies conducted deterministic sensitivity analyses, varying the input values within their plausible ranges either individually or in conjunction with related variables. Threshold values were reported for sensitive variables that changed the preferred outcome when varied. There were several scenarios in which ERCD was the preferred strategy. First, if the probability of TOLAC success was low, the increased cost of an emergency cesarean delivery made the TOLAC attempt more costly. Although the heterogeneity in study methods make it impossible to summarize the cutoff value at which these scenario parameters changed the outcome conclusion, TOLAC was found to be cost-effective in univariate sensitivity analyses if the probability of TOLAC success was greater than 42% [35], 46% [34], 74% [32], or 95% [31], depending on the study. In the study reporting a 95% TOLAC success threshold, Chuang et al. [31] had assumed at baseline that a TOLAC was more expensive than an ERCD, making it an outlier among the studies. Second, if the probability of uterine rupture was high, the potential morbidity associated with TOLAC made it less effective. Threshold values reported in the various studies were 3.1% [34], 3.2% [32], and 4.2% [35]. Because several studies found both the probability of TOLAC success and the probability of uterine rupture to be sensitive variables, bivariate sensitivity analyses were conducted by Gilbert et al. [34] to determine the joint impact of the variables. They found that when the probability of uterine rupture was set at 0%, 0.8%, 1.5%, and 3.0%, TOLAC was preferred if the probability of success was 40.0%, 47.2%, 53.6%, and 67.2% or more, respectively. Third, if the cost of TOLAC was more than that of ERCD, the latter may be more cost-effective. Both the studies by Gilbert et al. [34,35] and the study by Wymer et al. [36] found the costeffectiveness of TOLAC to be sensitive to the costs of TOLAC and ERCD. Because TOLAC offered more benefits than ERCD in those studies, it may be inferred that its cost could be slightly higher than ERCD and still be preferred, all else being equal. Bivariate sensitivity analyses on the costs were not performed, and thus it is not possible to determine how much excess cost TOLAC could incur before ERCD becomes more cost-effective in different scenarios. Probabilistic sensitivity analyses were conducted in four studies [33–36] using Monte-Carlo simulations whereby parameter values were drawn from within a distribution of possible values during each iteration of the model. TOLAC was the more cost-effective option under most assumptions, a result that was robust as probabilistic sensitivity analyses found TOLAC to be preferred between 64.4% of the time and 100% of the time across the different studies. Nevertheless, when disability resultant from stress urinary incontinence (SUI) was included in probabilistic sensitivity analyses, Gilbert et al. [34] found it to substantially impact ERCD preference over TOLAC, causing TOLAC to be preferred 7.3% of the time instead of 64.6% of the time. Chung et al. [32] on the other hand did not find SUI or fecal incontinence to be sensitive variables and the remaining studies either explicitly excluded SUI because of inconclusive literature on the subject or did not mention its potential impact.
Discussion Main Findings We conducted a comprehensive systematic review to summarize economic evaluations that compare a TOLAC with an ERCD for women with a singleton gestation who had a previous low transverse cesarean delivery. Six studies [32–37] found TOLAC
Table 4 – Cost-utility and cost-effectiveness analysis of ERCD vs. TOLAC. Currency (base year)
Incremental effectiveness*
Incremental cost (2016 US $)*
Base-case ICER summary/ estimate
SA: Sensitive variable (threshold value), costs (2016 US $)†
Chuang et al. [31] (1999) Chung et al. [32] (2001)
US $‡ (1998)
1 utility
27,591
ERCD dominant
Univariate SA: Probability of successful TOLAC (495%); disutilities of ERCD (40.12), successful TOLAC (o0.06), and failed TOLAC (o0.10)
US $ (2000)
1 QALY
195,668
Univariate SA: Probabilities of successful TOLAC (476%, TOLAC dominant; o74%, ERCD cost-effective; o65%, ERCD dominant) and uterine rupture (43.2%, ERCD cost-effective)
Fawsitt et al. [33] (2013)
Euro (2010)
1 QALY
21,095
$112,023 per QALY ERCD costs more but has additional utility TOLAC dominant
Gilbert et al. [34] (2013)
US $ (2009)
1 QALY
99,378
TOLAC dominant
Gilbert et al. [35] (2013)
US $ (2009)
1 QALY
400,999
TOLAC dominant
Grobman et al. [37] (2000)
US $‡ (1999)
1 major adverse neonatal outcome (death or cerebral palsy)
4,332,930
NA
Wymer et al.§ [36] (2014)
US $‡ (2012)
1 QALY
295,967
TOLAC dominant
Univariate SA: Varying TOLAC success, which at base case was 67%, between 64% and 69% did not impact TOLAC cost-effectiveness Probabilistic SA: TOLAC was dominant 100% of the time Univariate SA: Probabilities of uterine rupture (43.1%), successful TOLAC (o46.4%), and excess SUI among TOLAC women (40.8%); costs of failed TOLAC (4$20,154), successful TOLAC (4$10,980), and ERCD (o$9,846) Bivariate SA: When probability of uterine rupture was set at 0%, 0.8%, 1.5%, and 3.0%, TOLAC was preferred if the probability of success was 40.0%, 47.2%, 53.6%, and 67.2% or more, respectively Probabilistic SA: TOLAC was cost-effective 64.6% (all variables without SUI) and 7.3% (all variables including SUI) of the time Univariate SA: Probabilities of uterine rupture (44.2%), successful TOLAC (o42.2%); costs of ERCD (o$9,466) and successful TOLAC (4$11,967); and QALY of failed TOLAC (427.00, which represented a disutility of 0.47 for 132 d or more) Probabilistic SA: TOLAC was cost-effective 91.9% of the time Univariate SA: At the upper bound of the frequency of uterine rupture (1.0%), the cost per averted neonatal adverse outcome decreases to approach $1.7 million Multivariate SA: With all risks and costs of TOLAC simultaneously increased to their upper limits, the ERCD policy continued to incur more than $3.4 million per neonatal adverse outcome, with a minimum of 806 cesarean deliveries Univariate SA: Maternal model was sensitive to the cost of the elective repeat and emergency cesarean deliveries (threshold values not reported) Probabilistic SA: TOLAC was cost-effective 74% (maternal model) and 95.5% (neonatal model) of the time; with each subsequent vaginal delivery taken into consideration, the model was increasingly robust toward TOLAC
VALUE IN HEALTH ] (2016) ]]]–]]]
Reference (year)
ERCD, elective repeat cesarean delivery; ICER, incremental cost-effectiveness ratio; NA, not applicable; QALY, quality-adjusted life-year; SA, sensitivity analysis; SUI, stress urinary incontinence; TOLAC, trial of labor after cesarean. * For comparability between studies, incremental effectiveness and incremental cost per unit of effectiveness was calculated from the reported ICERs. Negative incremental effectiveness indicates that ERCD had decreased effectiveness relative to TOLAC in the base-case scenario. All costs were adjusted to 2016 US $ using purchasing power parities for the year of the data and then adjusting for inflation using the medical care component of the Consumer Price Index. † All cost-effectiveness decisions were reported on the basis of a willingness-to-pay threshold of $50,000. ‡ If base year of currency was not explicitly reported, we assumed year before publication; if currency was not explicitly reported, we guessed on the basis of the context. § For the Wymer et al. study, the cost and utilities from the maternal and neonatal models for the second delivery were summed up.
7
8
VALUE IN HEALTH ] (2016) ]]]–]]]
to be either dominant or cost-effective under most assumptions. One study [31] found that ERCD was dominant in the base-case scenario, but concluded that on the basis of sensitivity analyses, TOLAC and ERCD were equivalent strategies. Thus, evidence suggests that a TOLAC is a cost-effective strategy for women with low-risk, singleton gestation pregnancies. There are, however, several scenarios in which ERCD may be the preferred strategy. These include low likelihood of TOLAC success, high likelihood of uterine rupture, high cost of TOLAC relative to ERCD, and high likelihood of disutility resultant from SUI. The best evidence suggests that in the United States, an average of 74% of women who attempt a TOLAC deliver vaginally, with higher rates of success being reported in studies outside of the United States [8]. Furthermore, literature on symptomatic uterine rupture among women who had a previous low transverse cesarean incision has found that the risk varies between 0.3% and 1%, making TOLAC the preferred option unless a population opts for an induction of labor or has a high incidence of other risk factors for uterine rupture such as unfavorable cervical status at the time of admission, obesity, or an interpregnancy interval of 18 months or less [6,8]. Thus, in most conditions, it is unlikely that the TOLAC success and uterine rupture rates would achieve the threshold values required to make ERCD more cost-effective. Only the study by Chuang et al. [31] considered TOLAC to have a higher cost at baseline relative to ERCD, and the literature is inconclusive as to the relative excess risk of SUI after a TOLAC. Thus, these factors did not feature greatly in the cost-effectiveness considerations of the included studies.
Progression of Evidence When assessing the contribution of successive studies to the state of evidence, we found a progression from simple but representative short-term decision-analysis models to complex models that accounted for the long-term consequences of the index decision. Chuang et al. [31] (1999) constructed a decision tree comparing a TOLAC with an ERCD that streamlined all possible outcomes into being major, minor, or no complications, and performed basic deterministic sensitivity analyses. This was the first known study to conduct a full economic evaluation that simultaneously compared both costs and health outcomes of TOLAC versus ERCD. Grobman et al. [37] (2000) expanded the analytic horizon by extending the decision tree with Markov modeling, allowing for downstream reproductive decisions and outcomes to be impacted by the choice of a TOLAC or an ERCD. Their results, presented as cost-effectiveness estimates for the avoidance of one major neonatal outcome, were the first to describe the cost and outcome trade-offs associated with the alternative delivery options. Chung et al. [32] (2001) conducted the first comprehensive cost-utility analysis, providing two-way sensitivity analyses interpreted in light of a willingness-to-pay threshold and a societal perspective. All in all, these early studies found neither strategy to be dominant, with ERCD seeming to offer reduced adverse events but at significant cost. In contrast, later studies all determined in their base-case scenarios that TOLAC was dominant compared with ERCD. Fawsitt et al. [33] (2013) conducted the first international economic evaluation, modeling considerations unique to the Irish health system, such as prevalent use of vacuum assistance during vaginal deliveries. In that setting, TOLAC was found to be dominant 100% of the time. Gilbert et al. conducted two studies [34,35], (2013) both of which incorporated patient opportunity costs. The studies differed in that one added to the literature by using a minimally biased cohort, whereas the other conducted a life-cycle Markov analysis; both conducted comprehensive sensitivity analyses confirming the robustness of their conclusion that TOLAC was cost-effective under a wide range of assumptions.
Wymer et al. [36] (2014) provided evidence that the morbidity and cost accruing from multiple cesarean deliveries further strengthened the argument for TOLAC in low-risk women who may choose to have future pregnancies. These results are particularly relevant considering that the risk of maternal complications such as placenta previa and accreta increases incrementally with an increased number of cesarean deliveries, resulting in elevated risk of poor maternal and neonatal outcomes [41,42]. Conversely, having had any previous vaginal delivery, including vaginal birth after cesarean, increases the likelihood of future TOLAC success [43,44].
Strengths and Limitations Although previous studies have reviewed economic evidence for a TOLAC versus an ERCD [45,46], we add to the literature by systematically summarizing and comparing the available evidence. This study has several notable limitations. Because the included studies primarily modeled women with low transverse uterine scars who had no contraindications to labor, the results are generalizable only to a profile of women with a higher TOLAC success likelihood and low risk of uterine rupture. Similarly, this study is limited by the bias inherent in the included studies because data on maternal and infant outcomes often have information only for the actual rather than the intended routes of delivery [8]. Only one of the reviewed studies addressed potential confounding resulting from the lack of randomization to a TOLAC or an ERCD by using propensity scores [35]. Another limitation is that although we tried to incorporate all evidence that met our inclusion and exclusion criteria, we are unable to assess the impact of reporting bias on our results. Finally, it should be noted that the studies by Chuang et al. [31] and Fawsitt et al. [33] assessed only short-term outcomes and did not include the impact of the TOLAC versus ERCD decision on neonatal outcomes. Although the usefulness of those studies may be limited by this omission, they still contributed valuable information for the purposes of this systematic review. In particular, the study by Fawsitt et al. [33] used a methodical bottom-up costing strategy that incorporated itemized resource use, labor duration, professional wages and benefits, and a range of costs including administrative, operational, and capital costs. It was the only study that explicitly modeled TOLAC as requiring more time spent in labor than ERCD. In spite of the fact that this led to higher staff costs for TOLAC, the study still found TOLAC to be dominant. One challenge to uniformity of cost-effectiveness analyses in general is the variability in costing methods [47]. The true cost incurred by hospitals is difficult to ascertain because it requires a microcosting approach whereby all resources used are identified, measured, and valued [48]. These costs are seldom published or generalizable to other facilities. As a result, studies included in this systematic review may have used different sources for their cost inputs. To begin to address the issue of nonstandardization, one important source of costing data in the United States is the Healthcare Cost and Utilization Project, which converts hospital charges minus physician fees into costs on the basis of standard cost-tocharge ratios [9]. It should also be noted that this systematic review was unable to provide a summary estimate of the cost-effectiveness of an ERCD versus a TOLAC because there is at present no agreedupon method for pooling combined estimates, because resource use and costs are sensitive to variability across settings, and because the included studies were presented from different perspectives [21,49].
Comparison with Existing Literature In 2003, the Agency for Healthcare Research and Quality published an evidence report on vaginal birth after cesarean deliveries that incorporated a review of the economic literature [45].
9
VALUE IN HEALTH ] (2016) ]]]–]]]
The report found that at the time, only the study by Chung et al. [32] provided adequate evidence on the relative value of TOLAC and ERCD. It suggested that future studies improve upon the state of evidence by performing bivariate and probabilistic sensitivity analyses, incorporating the impact of the index decision on future pregnancies, and including patient opportunity costs. Later studies by Gilbert et al. [34,35], Fawsitt et al. [33], and Wymer et al. [36] responded to all these calls for evidence. A host of factors have influenced the rate of a TOLAC, which ranges from 28% to 83% across delivery centers [8]. Factors that influence the rate and success of a TOLAC include hospital characteristics such as volume [53], policy [54], practice patterns [55], and rural/ urban location [56] as well as patient characteristics such as obesity [57], obstetric history [58], socioeconomic factors [59], and stage of labor dystocia causing the primary cesarean delivery [60]. In particular, the fear of regional, hospital, and individual liability may be a driving factor for cesarean rates [50,51], because the adverse consequences of a uterine rupture may be seen as more imminently damaging from a risk-assessment standpoint when compared with the complications of an ERCD [52]. Although these considerations may avoid an individual serious adverse event, they may not lead to an optimum patient outcomes and use of health care resources on a population level.
Conclusions and Implications This study lends support to the cost-effectiveness of a TOLAC when compared with an ERCD in most scenarios for low-risk women who have had one previous cesarean delivery. Thus, from a populationbased and economic perspective, decision makers and clinicians should seek to encourage a TOLAC for low-risk women. Nevertheless, patient preferences informed by education should continue to direct decisions on the planned mode of delivery. Future studies should assess the impact of heterogeneity in hospital and patient characteristics on cost-effectiveness estimates.
Acknowledgment We thank Carolyn M. Holmes, MLIS, for conducting the literature searches. Source of financial support: A.J. Rogers is supported by grants from the National Institute of General Medical Sciences (grant no. T32GM008361) and the National Center for Advancing Translational Sciences (grant no. 1TL1TR001418-01). The funders had no input in the systematic review.
Appendix A
Table A1 – Search strategy: Economic evaluations comparing a TOLAC with an ERCD: A systematic review * . Number
Searches in MEDLINE
Terms defining economic evaluations Economics [mh] Cost [tiab] Costs [tiab] Economic* [tiab] Economics [sh] OR/1–5 Terms defining a trial of labor 7 Trial of Labor [mh] 8 “Trial of labor” [tiab] 9 “Trials of labor” [tiab] 10 Labor-trial* [tiab] 11 “Trial of labour” [tiab] 12 “Trials of labour” [tiab] 13 Labour-trial* [tiab] 14 TOLAC [tiab] 15 OR/7–14 Terms defining vaginal birth after cesarean and cesarean delivery 16 Vaginal birth after cesarean [mh] 17 VBAC [tiab] 18 Cesarean section [mh] 19 Cesarean* [tiab] 20 “C section” [tiab] 21 “C sections” [tiab] 22 Abdominal-deliver* [tiab] 23 Caesarean* [tiab] 24 ERCD [tiab] 25 OR/16–24 Overall search string 6 AND 15 AND 25 1 2 3 4 5 6
MeSH, Medical Subject Headings; [mh], MeSH terms; [tiab],title/ abstract; [sh], MeSH subheadings. * Databases searched were Cochrane Library, EconLit, CostEffectiveness Analysis Registry, EMBASE, MEDLINE, and CINAHL without language, publication, or date restrictions up until October 2015.
10
VALUE IN HEALTH ] (2016) ]]]–]]]
Appendix B Table B1 – CHEERS statement scores. Item number
CHEERS checklist item
Chuang et al. [31]
Chung et al. [32]
Fawsitt et al. [33]
Gilbert et al. [34]
Gilbert et al. [35]
Grobman et al. [37]
Wymer et al. [36]
1 2 3
Title Abstract Background and objectives Target population and subgroups Setting and location Study perspective Comparators Time horizon Discount rate Choice of health outcomes Measurement of effectiveness Measurement and valuation of preference-based outcomes Estimating resources and costs Currency, price date, and conversion Choice of model Assumptions Analytical methods Study parameters Incremental costs and outcomes Characterizing uncertainty Characterizing heterogeneity Study findings, limitations, generalizability, and present knowledge Source of funding Conflicts of interest Total score
0.5 0.5 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
0.5
1
1
1
1
1
1
0 0 1 0 NA 0.5
1 1 1 0.5 1 1
1 1 1 1 NA 1
1 1 1 1 1 1
1 1 1 1 1 1
1 1 1 1 1 1
1 1 1 1 1 1
0.5
1
1
1
1
1
1
0
1
1
1
1
1
1
0.5
1
1
1
1
1
1
0
1
1
1
1
1
1
1 0.5 0.5 0.5 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
0.5
1
1
1
1
1
1
NA
NA
NA
NA
NA
NA
NA
1
1
1
1
1
1
1
1 0 11
1 0 21.5
1 1 22
1 1 23
1 0 22
0 0 21
1 1 23
4 5 6 7 8 9 10 11 12
13 14 15 16 17 18 19 20 21 22
23 24
CHEERS, Consolidated Health Economic Evaluation Reporting Standards; NA, not applicable.
R EF E R EN CE S
[1] Weiss AJ, Elixhauser A, Andrews RM. Characteristics of Operating Room Procedures in U.S. Hospitals, 2011. HCUP Statistical Brief #170. Rockville, MD: Agency for Healthcare Research and Quality, 2014. [2] Martin JA, Hamilton BE, Osterman MJ. Births in the United States, 2014. Atlanta, GA: National Center for Health Statistics, 2015. [3] Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2008. Natl Vital Stat Rep 2010;59:1, 3–71. [4] Wells CE, Cunningham FG, Lockwood CJ, et al. Choosing the route of delivery after cesarean birth. In: UpToDate, Post TW (Ed), UpToDate, Waltham, MA. [Accessed on November 25, 2015]. [5] Guise JM, Berlin M, McDonagh M, et al. Safety of vaginal birth after cesarean: a systematic review. Obstet Gynecol 2004;103:420–9.
[6] American College of Obstetricians and Gynecologists. ACOG practice bulletin no. 115: vaginal birth after previous cesarean delivery. Obstet Gynecol 2010;116:450–63. [7] Martin JA, Hamilton BE, Ventura SJ, et al. Births: final data for 2009. Natl Vital Stat Rep 2011;60:1–70. [8] Guise JM, Eden K, Emeis C, et al. Vaginal birth after cesarean: new insights. Evid Rep Technol Assess (Full Rep) 2010;1–397. [9] Podulka J, Stranges E, Steiner C. Hospitalizations Related to Childbirth, 2008. HCUP Statistical Brief #110. Rockville, MD: Agency for Healthcare Research and Quality, 2011. [10] Truven Health Analytics. The Cost of Having a Baby in the United States. Greenwood Village, CO: Truven Health Analytics, 2013. [11] Traynor JD, Peaceman AM. Maternal hospital charges associated with trial of labor versus elective repeat cesarean section. Birth 1998;25:81–4.
VALUE IN HEALTH ] (2016) ]]]–]]]
[12] Shorten A, Lewis DE, Shorten B. Trial of labour versus elective repeat caesarean section: a cost-effectiveness analysis. Aust Health Rev 1998;21:8–28. [13] Friedman AM, Ananth CV, Chen L, et al. An economic analysis of trial of labor after cesarean delivery. J Matern Fetal Neonatal Med 2016;29:1030–5. [14] Clark SL, Scott JR, Porter TF, et al. Is vaginal birth after cesarean less expensive than repeat cesarean delivery? Am J Obstet Gynecol 2000;182:599–602. [15] Bost BW. Cesarean delivery on demand: what will it cost? Am J Obstet Gynecol 2003;188:1418–21. [16] Kazandjian VA, Chaulk CP, Ogunbo S, et al. Does a cesarean section delivery always cost more than a vaginal delivery? J Eval Clin Pract 2007;13:16–20. [17] The Joanna Briggs Institute. The Joanna Briggs Institute Reviewers’ Manual 2014: The Systematic Review of Economic Evaluation Evidence. Adelaide, Australia: The Joanna Briggs Institute, 2014. [18] Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ 2015;349:g7647. [19] Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. [20] Rogers AJG, Rogers NG. A systematic review of economic evaluations comparing a trial of labor to an elective repeat cesarean delivery. PROSPERO 2015: CRD42015029177. Available from: http://www.crd.york. ac.uk/PROSPERO/display_record.asp?ID=CRD42015029177. [Accessed November 25, 2015]. [21] Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. [22] Carande-Kulis VG, Maciosek MV, Briss PA, et al. Methods for systematic reviews of economic evaluations for the Guide to Community Preventive Services. Task Force on Community Preventive Services. Am J Prev Med 2000;18:75–91. [23] Community Preventive Services Task Force. Guide to Community Preventive Services. Economic Evaluation Abstraction Form. Community Preventive Services Task Force, 2010: Available from: https://www.thecommunityguide.org/about/EconAbstraction_v5.pdf. [Accessed on November 25, 2015]. [24] Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the Panel on Cost-Effectiveness in Health and Medicine. JAMA 1996;276:1253–8. [25] Chiou CF, Hay JW, Wallace JF, et al. Development and validation of a grading system for the quality of cost-effectiveness studies. Med Care 2003;41:32–44. [26] Walker DG, Wilson RF, Sharma R, et al. Best Practices for Conducting Economic Evaluations in Health Care: A Systematic Review of Quality Assessment Tools. Rockville, MD: Agency for Healthcare Research and Quality, 2012. [27] Langer A. A framework for assessing Health Economic Evaluation (HEE) quality appraisal instruments. BMC Health Serv Res 2012;12:253. [28] Husereau D, Drummond M, Petrou S, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. Value Health 2013;16:e1–5. [29] Organisation for Economic Co-operation and Development (ODEC). StatExtracts, PPPs and exchange rates. 2015. Available from: http:// stats.oecd.org/Index.aspx?DataSetCode=MEI_PRICES. [Accessed on March 1, 2016]. [30] US Department of Labor. Bureau of Labor Statistics. Consumer Price Index. Available from: http://www.bls.gov/cpi/. [Accessed on March 1, 2016]. [31] Chuang JH, Jenders RA. Trial of labor versus elective repeat cesarean section for the women with a previous cesarean section: a decision analysis. Proc AMIA Symp 1999;226–30. [32] Chung A, Macario A, El-Sayed YY, et al. Cost-effectiveness of a trial of labor after previous cesarean. Obstet Gynecol 2001;97:932–41. [33] Fawsitt CG, Bourke J, Greene RA, et al. At what price? A costeffectiveness analysis comparing trial of labour after previous caesarean versus elective repeat caesarean delivery. PLoS One 2013;8: e58577. [34] Gilbert SA, Grobman WA, Landon MB, et al. Lifetime cost-effectiveness of trial of labor after cesarean in the United States. Value Health 2013;16:953–64. [35] Gilbert SA, Grobman WA, Landon MB, et al. Cost-effectiveness of trial of labor after previous cesarean in a minimally biased cohort. Am J Perinatol 2013;30:11–20.
11
[36] Wymer KM, Shih YC, Plunkett BA. The cost-effectiveness of a trial of labor accrues with multiple subsequent vaginal deliveries. Am J Obstet Gynecol 2014;211:56.e1–12. [37] Grobman WA, Peaceman AM, Socol ML. Cost-effectiveness of elective cesarean delivery after one prior low transverse cesarean. Obstet Gynecol 2000;95:745–51. [38] Kaplan RM, Anderson JP. A general health policy model: update and applications. Health Serv Res 1988;23:203–35. [39] Landon MB, Hauth JC, Leveno KJ, et al. Maternal and perinatal outcomes associated with a trial of labor after prior cesarean delivery. N Engl J Med 2004;351:2581–9. [40] Gilbert SA, Grobman WA, Landon MB, et al. Elective repeat cesarean delivery compared with spontaneous trial of labor after a prior cesarean delivery: a propensity score analysis. Am J Obstet Gynecol 2012;206:311.e1–9. [41] Silver RM, Landon MB, Rouse DJ, et al. Maternal morbidity associated with multiple repeat cesarean deliveries. Obstet Gynecol 2006;107:1226–32. [42] Getahun D, Oyelese Y, Salihu HM, Ananth CV. Previous cesarean delivery and risks of placenta previa and placental abruption. Obstet Gynecol 2006;107:771–8. [43] Landon MB, Leindecker S, Spong CY, et al. The MFMU Cesarean Registry: factors affecting the success of trial of labor after previous cesarean delivery. Am J Obstet Gynecol 2005;193:1016–23. [44] Cahill AG, Stamilio DM, Odibo AO, et al. Is vaginal birth after cesarean (VBAC) or elective repeat cesarean safer in women with a prior vaginal delivery? Am J Obstet Gynecol 2006;195:1143–7. [45] Guise J, McDonagh M, Hashima JN, et al. Vaginal Birth after Cesarean (VBAC) Evidence Report/Technological Assessment No. 71 (Prepared by the Oregon Health & Science University Evidence-Based Practice Center under Contract No. 290-97-0018). AHRQ Publication No. 03-E018. Rockville, MD: Agency for Healthcare Research and Quality, 2003. [46] Petrou S, Khan K. An overview of the health economic implications of elective caesarean section. Appl Health Econ Health Policy 2013;11:561–76. [47] Adam T, Koopmanschap MA, Evans DB. Cost-effectiveness analysis: can we reduce variability in costing methods? Int J Technol Assess Health Care 2003;19:407–20. [48] Morris S, Devlin N, Parkin D, et al. Economic Analysis in Healthcare (2nd ed). West Sussex, UK: Wiley, 2012. [49] Boehler CE, Lord J. Mind the gap! A multilevel analysis of factors related to variation in published cost-effectiveness estimates within and between countries. Med Decis Making 2016;36:31–47. [50] Murthy K, Grobman WA, Lee TA, et al. Association between rising professional liability insurance premiums and primary cesarean delivery rates. Obstet Gynecol 2007;110:1264–9. [51] Korst LM, Gregory KD, Fridman M, et al. Nonclinical factors affecting women’s access to trial of labor after cesarean delivery. Clin Perinatol 2011;38:193–216. [52] Riddell CA, Kaufman JS, Hutcheon JA, et al. Effect of uterine rupture on a hospital’s future rate of vaginal birth after cesarean delivery. Obstet Gynecol 2014;124:1175–81. [53] Chang JJ, Stamilio DM, Macones GA. Effect of hospital volume on maternal outcomes in women with prior cesarean delivery undergoing trial of labor. Am J Epidemiol 2008;167:711–8. [54] Cottrell EB, Snowden J, Guise JM, et al. Predictors of hospital VBAC policies. Am J Obstet Gynecol 2013;208:S332–3. [55] Coleman VH, Erickson K, Schulkin J, et al. Vaginal birth after cesarean delivery: practice patterns of obstetrician-gynecologists. J Reprod Med 2005;50:261–6. [56] Gochnour G, Ratcliffe S, Stone MB. The UTAH VBAC Study. Matern Child Health J 2005;9:181–8. [57] Carroll CS Sr, Magann EF, Chauhan SP, et al. Vaginal birth after cesarean section versus elective repeat cesarean delivery: weight-based outcomes. Am J Obstet Gynecol 2003;188: 1516–22. [58] Cieminski A. Mode of delivery in women with prior cesarean section: the effect of non medical factors and obstetric history. Ginekologia i Poloznictwo 2011;21:57–64. [59] Cecatti JG, Pires HM, Faundes A, et al. Factors associated with vaginal birth after previous cesarean section in Brazilian women. Pan Am J Public Health 2005;18:107–13. [60] Lewkowitz AK, Nakagawa S, Thiet MP, et al. Effect of stage of initial labor dystocia on vaginal birth after cesarean success. Am J Obstet Gynecol 2015;213:861.e1–5.