The primary measure of effectiveness was health-related quality of life calculated as QALY, which include a length of time component (eg, 1 year) and a quality of life component (ie, utility). Health utility is the numerical valuation of one's quality of life on a linear scale from 0.00 (death) to 1.00 (perfect health). For example, one QALY for an individual in perfect health (with a utility of 1.0) for 1 year is considered equivalent to 2 years in a health state with a utility of 0.5 (as might occur with a disability). Maternal health outcomes and fetal health outcomes were weighed equally.
Table 5 has the estimated utilities for maternal and neonatal health states. We considered three long-term maternal health states: perfect health (utility assigned a value of unity), well health after emergent hysterectomy, and death (utility assigned a value of zero). We assumed that quality of life after menopause was equivalent whether or not the patient had undergone a hysterectomy.
For maternal complications that prolonged hospitalization or resulted in short-term (less than 1 year) morbidity, a reduction in quality of life or disutility was included. We calculated disutilities (ie, validated deductions from the utility value), using the Quality of Well-Being classification system (Table 6).31 This system has established lists of scaled reductions in a patients quality of life or disutilities in each of four categories: mobility, physical activity, social activity, and symptom complexes. For example, we assumed that a woman with a vaginal delivery would have a reduction in quality of life equal to 0.35 lasting 7 days of recovery, whereas a woman having an elective cesarean delivery would have a reduction of quality of life postoperatively of 0.45 per day lasting on average 21 days.
Quality-adjusted life-years were calculated by combining these estimates with 1997 life table data from the National Center for Health Statistics (http://www.cdc.gov/nchs).
The cost of a medical service from society's point of view is the total net cost to all the different components of society. Direct hospital costs were included in the numerator of the cost-effectiveness ratio whereas indirect costs (or productivity costs) caused by morbidity or mortality were incorporated by the effectiveness measure, QALY, in the denominator. We used data from actual obstetric patients cared for at Stanford University Medical Center in 1999. Financial data were extracted using the hospital's financial management software (Eclipsys Corporation, Delray Beach, FL). The system used actual supply costs, wage rates, and labor effort as measured by hospital department managers to compute variable and fixed costs of patient care.32 Costs were calibrated to actual cash expenses on a quarterly basis.
We computed the mean hospital cost at our institution for 550 patients with Diagnosis Related Group 371 (cesarean section without complication and/or comorbidity) and actual mean hospital cost for 2,878 patients with Diagnosis Related Group 373 (vaginal delivery without complicating diagnosis) in 1999 (Table 7). For the failed trial of labor-cesarean delivery group, we assumed the hospital cost would be that of elective cesarean delivery plus the cost of resources consumed in the labor and delivery unit before the operative delivery. From billing records, we determined that patients had an average length of stay of 10 hours (expressed as a fraction of the cost of 1 day equaled [10/24 × $2150]) in the labor and delivery unit before the cesarean delivery.
Physician professional costs were included. The costs of the obstetrician's professional service were obtained using Current Procedural Terminology codes, reimbursed at the 1999 Medicare rate without geographic adjustment.33 Using Medicare rates to estimate costs of physician services is the most common method used in health services research. The costs of the obstetrician's professional service were obtained for CPT 59610 (routine obstetric care including antepartum care, vaginal delivery, and postpartum care, after previous cesarean section), CPT 59618 (routine obstetric care including antepartum care, cesarean delivery, and postpartum care, following attempted vaginal delivery after previous cesarean delivery), and CPT 59590 (routine obstetric care including antepartum care, cesarean delivery, and postpartum care).
The cost of an obstetrician's time from society's point of view is the result of society having given up the opportunity to use the physician's time and expertise for some other purpose. The ACOG's revised version of the practice parameter on VBAC states that physicians be “immediately available to provide emergency care.”3 We did not include this increased societal cost of providing physician availability in the model.
The cost of an anesthesiologist's service for patients with trial of labor was based on Medicare reimbursement rate of $19/unit using the American Society of Anesthesiologists Relative Value System (with 5 base units startup + 4 units/first hour + 1 unit/hour, thereafter × 4 hours). The duration of epidural analgesia for a vaginal delivery use was determined from billing records for 222 patients during December 1999. For cesarean delivery, the cost of the anesthesiologist's service constituted 7 base units startup + 1 additional unit/15 minutes × 75 minutes.
For each maternal and neonatal outcome, we assigned an additional cost based on society's point of view (Table 8). We used a “bottom up” cost methodology in which costs of supplies, drugs, and labor were added as required for each complication or outcome. Costs of incremental nursing care (because of complications) were included because changes in the intensity of nursing care reflect the sacrifice of resources that could have been used elsewhere. This is standard in health economics research. Costs associated with maternal morbidities assume successful treatment without subsequent complications.
Details of cost calculations are as follows:
Hemorrhage = $267 (3 units of blood) + $61 (additional 1.5 hours nursing cost only for patients with successful trial of labor as operative deliveries already incur operating room cost).
Hemorrhage requiring hysterectomy | successful trial of labor = $1285 (75 minutes of operating room time) + $2310 (2 additional ward days) + $919 (additional obstetrician service) + $228 (anesthesiologist).
Hemorrhage requiring hysterectomy | failed trial of labor or elective cesarean = $771 (45 minutes additional operating room time) + $457 (obstetrician) + $57 (anesthesiologist).
Infection = $51 (antibiotics) + $30 (disposable supplies) + $61 (1 hour nursing/day × 1.5 days) + $1155 (1 additional ward day only for patients with successful trial of labor).
Thrombosis = $8 (heparin) + $150 (supplies for infusion pump) + $122 (1 hour nursing/day × 3 days) + $45 (warfarin × 3 months) + $326 (coagulation testing × 14) + $3464 (3 additional ward days only for patients with successful trial of labor) or + $1154 (additional ward day for patients with failed trial of labor or elective cesarean).
Uterine rupture requiring hysterectomy | successful trial of labor = $1285 (75 minutes of operating room time) + $2310 (2 additional ward days) + $919 (obstetrician) + $228 (anesthesiologist).
Uterine rupture requiring hysterectomy | failed trial of labor or elective cesarean = $771 (45 minutes additional operating room time) + $457 (obstetrician) + $57 (anesthesiologist).
Uterine rupture requiring uterine repair | successful trial of labor = $1285 (75 minutes of operating room time) + $2310 (2 additional ward days) + $611 (obstetrician) + $228 (anesthesiologist).
Uterine rupture requiring uterine repair | failed trial of labor or elective cesarean = $515 (30 minutes additional operating room time) + $38 (anesthesiologist).
Operative injury = $771 (45 minutes additional operating room time) + $57 (anesthesiologist).
Hospital cost of maternal death assuming resuscitative measures equivalent to 1 intensive care unit day = $2150.
None or mild perinatal morbidity = $179 (2 hours of observation in neonatal intensive care unit) + $82 (perinatologist).
Moderate perinatal morbidity = $50,075 (mean hospital cost for Diagnosis Related Group 389 [full-term neonate] and Principal Diagnoses 770.1 [meconium aspiration], 771.8 [perinatal infection], 770.8 [postbirth respiratory problem] from 1999 fiscal year) + $2117 (perinatologist for average length of stay of 7 days).
Severe perinatal morbidity = $57,978 (mean hospital cost for Diagnosis Related Group 389 [full-term neonate] and Principal Diagnosis 768.5 [severe birth asphyxia] from 1999) + $5675 (perinatologist for average length of stay of 13 days) + $17,440 (cost of health expenditures for average of 1 year of life as a multiple of normal [40 × $436]).
We calculated the hospital cost of neonatal death to equal $35,553 (mean hospital cost for Diagnosis Related Group 385 [neonates, died] from 1999) + $4861 (perinatologist for 11 hospital days). We attributed the mean hospital costs of severe neonatal morbidity from birth asphyxia to be $81,093 (Principal Diagnosis 768.5, severe birth asphyxia, range $50,000–$500,00024) added to the cost of a neonatologist's time and expertise during the average length of stay of 13 days.
We also estimated cost of health care expenditures for these infants after discharge from the hospital for a lifetime survival of 1 year. Although one published study reported a life expectancy of 17 years for patients with cerebral palsy,34 we decided to use a conservative estimate of 1 year to take into account those infants with extremely severe outcomes who died early in life outside of the newborn period.
Future health care costs were also estimated for those maternal and neonatal outcomes with short-term morbidities, based on age-specific average annual health care expenditures obtained from the Bureau of Labor Statistics' 1997 Consumer Expenditure Survey (http://188.8.131.52/cgi-bin/dsrv?cx). These figures include costs for health insurance, medical services, medications (prescription and nonprescription), and medical supplies. All costs are reported in 2000 United States dollars. We discounted all future costs and QALY at 3% per annum in the base case scenario.
We used sensitivity analysis to test a range (eg, from the low probability to the high probability of a particular event occurring) of values for the specified variables. Sensitive factors change the decision about whether an intervention is considered cost-effective. Two of the key a priori estimates were the probability of successful vaginal delivery and the probability of uterine rupture. We also ran sensitivity analyses on the probabilities for different neonatal outcomes given each possible type of delivery. Sensitivity analysis was also performed on cost estimates as well as the discount rate applied to all future costs and QALY, using a range between 2% and 5%.
DATA software, version 3.5 for Windows (Tree-Age Software, Inc., Boston, MA) was used to build the decision tree and to perform model calculations.
Elective repeat cesarean delivery in our base case, which assumes a 75% likelihood that a trial of labor will result in a vaginal birth, would result in a cost-effectiveness ratio of $112,023 per QALY. This is greater than the commonly accepted threshold of less than $50,000 per QALY for an intervention to be considered cost-effective. Therefore, a trial of labor was the cost-effective method of delivery in our base case.
We explored the effect on cost-effectiveness produced by varying probabilities and costs of complications and outcomes and also by changing basic model assumptions (Figure 2). The incremental cost-effectiveness was extremely sensitive to the probability of successful vaginal birth after a trial of labor. If the probability of successful VBAC was less than 0.65, elective repeat cesarean delivery was the dominant choice, ie, elective repeat cesarean delivery was both less costly and more effective. Between 0.65 and 0.74, elective repeat cesarean delivery was the cost-effective delivery method, because, although it cost more than VBAC, it was offset by improved outcomes (ie, the cost-effectiveness ratio was less than the $50,000 per QALY threshold). Between 0.74 and 0.76, elective repeat cesarean delivery still had better outcomes than trial of labor; however, the cost was prohibitive with a cost-effectiveness ratio greater than $50,000 per QALY. Therefore, trial of labor was cost-effective in this range. If the probability of successful VBAC was greater than 0.76, however, trial of labor was dominant because it was both less costly and now more effective.
Costs associated with a moderately morbid neonatal outcome, as well as the probabilities of infant morbidity occurring, heavily impacted our results. However, the probability (range 0.3–1.5%) of uterine rupture was not a sensitive factor. Even at the highest incidence of 1.5%, the cost-effectiveness ratio did not go below the $50,000 per QALY threshold in our base case. Only when the incidence of uterine rupture exceeded 3.2% in the trial of labor group did it change the overall cost-effectiveness results. The probability of urinary incontinence, the probability of hemorrhage, infection, thrombosis, and operative injury, as well as the choice of discount rates were also not sensitive factors.
We found that the cost-effective delivery method in patients with one previous, low transverse cesarean depends on the a priori estimate of the likelihood of successful trial of labor. Taking a societal perspective and assuming a threshold cost/effectiveness ratio of $50,000 per QALY, our modeling suggests that a trial of labor is cost-effective if the probability of a successful vaginal delivery is greater than 0.74. However, if the probability of a successful vaginal delivery is below 0.65, elective repeat cesarean delivery costs less and has better outcomes than a trial of labor. For the range between 0.65 and 0.74, elective repeat cesarean delivery costs more than trial of labor but remains the cost-effective method of delivery because of improved outcomes. The usefulness of this information depends on clinicians being able to accurately estimate the probability of a successful vaginal delivery. The estimated probabilities of a moderately morbid neonatal outcome, as well as the concomitant costs, also heavily impacted our results.
Maternal preferences for one delivery method over another (eg, women who prefer to experience labor) also need to be considered in the delivery decision.35 Because many patients actively participate in this decision,36 patient education as to the choices is important.
For the results of this study to be applied, improved algorithms for computing the likelihood of a particular patient having a successful vaginal delivery after a cesarean are necessary. Predicting the likelihood of success of a trial of labor for a given patient is difficult. This is especially true with enough precision (eg, whether a particular patient has a 65% chance or a 75% chance of having a successful vaginal delivery) to benefit from our modeling.
Factors positively associated with successful vaginal birth include: prior vaginal delivery, prior cesarean because of a nonrecurring condition (eg, breech presentation), and favorable Bishop score (based on exam of cervical dilation, effacement, consistency, and position plus the station of the fetus).6,15,17,37–39
Flamm et al examined nonrecurrent indications, such as breech presentation, fetal distress, or the presence of herpetic lesions as the indication for the primary cesarean, and, based on these factors, the predicted success rate of trial of labor ranged from 78% to 87%.15 These success rates are point estimates with ranges associated with them. Recommendations about cost-effectiveness based on success rates need to take into account the variability of the success rates among studies. For patients whose a priori probability is greater than 74%, our results suggest that trial of labor is cost-effective. In fact, routine elective repeat cesarean delivery was found to result in an excess of maternal morbidity and mortality at a high cost to society.40
On the other hand, factors that may reduce the likelihood of successful vaginal birth include a history of more than one previous cesarean, induction of labor, or failure to progress (cephalopelvic disproportion) during the previous pregnancy.5,15 If the primary indication for the prior cesarean was cephalopelvic disproportion or failure to progress, the rate of successful vaginal birth ranged from 62% to 68%. Our results suggest that for these patients, elective repeat cesarean delivery may be a reasonable option from a societal economic viewpoint. Further elucidation of the relative impact of each of the risk factors may help to stratify the population of candidates for VBAC.
The cost-effectiveness of elective repeat cesarean delivery was also sensitive to how the morbidity of an infant injured by uterine rupture was considered. Reduced infant mortality from elective repeat cesarean delivery was a significant contributor to the cost-effectiveness of elective repeat cesarean delivery.
The accuracy of our results depends on the accuracy of our estimates for variables in the model. Quantifying certain variables is difficult. By performing sensitivity analyses, we found that the probabilities of a moderately morbid neonatal outcome (infants with meconium aspiration, infection/sepsis screening, and respiratory distress/failure), as well as the concomitant costs, heavily impacted our results. Although this is a broad category that includes a range of conditions, studies often do not quantify adverse neonatal outcomes when evaluating the costs and morbidity associated with trial of labor. This is in part because there is no clear standard of practice in cost-effectiveness analysis with respect to the treatment of infant morbidity. Studies further defining the probability for neonatal outcomes and associated utilities after trial of labor versus elective repeat cesarean are necessary.
We addressed a range of perspectives for probabilities and costs of a morbid neonatal outcome by performing sensitivity analysis with a wide spectrum of estimates. As the weighing of the injured infant's QALY is increased or as costs of health care, litigation, or settlements for poor infant outcomes are increased, elective repeat cesarean delivery has an improved economic profile. For example, as the cost of a moderate morbid neonatal outcome was increased to the upper range of $ 150,000, elective repeat cesarean delivery was found to be increasingly cost-effective. On the other hand, if the costs or incidence of adverse neonatal outcomes are decreased, trial of labor becomes more cost-effective.4–6
The probability of uterine rupture, within the range from 0.3% to 1.5%, was not a sensitive factor. In our modeling, only when the incidence of uterine rupture reached 3.2% in the trial of labor group, did elective repeat cesarean delivery become cost-effective assuming that the a priori probability of successful vaginal delivery was 0.75. Interestingly, Grobman et al confirmed that at a uterine rupture rate of 3.2%, the increased infant morbidity and mortality from attempted trial of labor exceeded the benefit of reduced cost of trial of labor.41
Clinicians are not currently able to estimate who has a 65–75% chance of successful delivery. Further research to assist physicians in assessing an individual patient's likelihood of successful trial of labor is needed. This risk factor stratification then can be used to educate parturients, and in conjunction with the patient's preference for one delivery method, be used to make a delivery decision.
1. Muchinski M. Average charges for uncomplicated vaginal, cesarean and VBAC deliveries: Regional variations, United States, 1996. Stat Bull Metrop Insur Co 1998;75:17–28.
3. American College of Obstetricians and Gynecologists. Vaginal birth after previous cesarean delivery. ACOG Practice Bulletin no. 5. Washington DC: American College of Obstetricians and Gynecologists, July, 1999.
4. Miller D, Diaz F, Paul R. Vaginal birth after cesarean: A 10-year experience. Obstet Gynecol 1994;84:255–8.
5. Martin J, Harris B, Huddleston J, Morrison J, Propst M, Wiser W, et al. Vaginal delivery following previous cesarean birth. Am J Obstet Gynecol 1983;146:255–63.
6. McMahon M, Luther E, Bowes W, Olshan A. Comparison of a trial of labor with an elective second cesarean section. N Engl J Med 1996;335:689–95.
7. Leung A, Leung E, Paul R. Uterine rupture after previous cesarean delivery: Maternal and fetal consequences. Am J Obstet Gynecol 1993;169:945–50.
8. Eisenberg J. Clinical economics. JAMA 1989;262:2879–86.
9. Russell L, Gold M, Siegel J, Daniels N, Winstein M. The role of cost-effectiveness analysis in health and medicine: Panel on cost-effectiveness in health and medicine. JAMA 1996;276:1172–7.
10. Drummond MF, O'Brien B, Stoddart GL, Torrance GW. Methods for the economic evaluation of health care programmes. 2nd ed. New York: Oxford University Press, 1997.
11. MacArthur C, Bick DE, Keighley MR. Faecal incontinence after childbirth. Br J Obstet Gynaecol 1997;104:46–50.
12. Groutz A, Fait G, Lessing JB, David MP, Wolman I, Jaffa A, et al. Incidence and obstetric risk factors of postpartum anal incontinence. Scand J Gastroenterol 1999;34:315–8.
13. Rageth J, Juzi C, Grossenbacher H. Delivery after previous cesarean: A risk evaluation. Obstet Gynecol 1999;93:332–7.
14. Obara H, Minakami H, Koike T, Takamizawa S, Matsubara S, Sato I. Vaginal birth after cesarean delivery: Results in 310 pregnancies. J Obstet Gynaecol Res 1998;24:129–34.
15. Flamm B, Goings J, Liu Y, Wolde-Tsadik G. Elective repeat cesarean delivery versus trial of labor: A prospective multicenter study. Obstet Gynecol 1994;83:927–32.
16. Cowan R, Kinch R, Ellis B, Anderson R. Trial of labor following cesarean delivery. Obstet Gynecol 1994;83:933–6.
17. Rosen M, Dickinson J, Westhoff C. Vaginal birth after cesarean: A meta-analysis of morbidity and mortality. Obstet Gynecol 1991;77:465–70.
18. Macklon N. Venous thromboembolic disease in obstetrics and gynaecology: The Scottish experience. Scott Med J 1996;41:83–6.
19. Wilson PD, Herbison RM, Herbison GP. Obstetric practice and the prevalence of urinary incontinence three months after delivery. Br J Obstet Gynaecol 1996;103:154–61.
20. Iosif S. Stress incontinence during pregnancy and in puerperium. Int J Gynaecol Obstet 1981;19:13–20.
21. Thorp J, Norton P, Wall L, Kuller J, Eucker B, Wells E. Urinary incontinence in pregnancy and the puerperium: A prospective study. Am J Obstet Gynecol 1999;181;266–73.
22. Viktrup L, Lose G, Rolff M, Barfoed K. The symptoms of stress incontinence caused by pregnancy or delivery in primiparas. Obstet Gynecol 1992;79:945–9.
23. Nielsen T, Hokegard K. Cesarean section and intraoperative surgical complications. Acta Obstet Gynecol Scand 1984;63:103–8.
24. Waitzman NJ, Romano PS, Scheffler RM, Harris JA. Economic costs of birth defects and cerebral palsy — United States, 1992. MMWR 1995;44:694–9.
25. Hook B, Kiwi R, Amini SB, Fanaroff A, Hack M. Neonatal morbidity after elective repeat cesarean section and trial of labor. Pediatrics 1997;100:348–53.
26. Meehan FP, Burke G, Casey C, Sheil JG. Delivery following cesarean section and perinatal mortality. Am J Perinatol 1989;6:90–4.
27. Farmer RM, Kirschbaum T, Potter D, Strong TH, Medearis AL. Uterine rupture during trial of labor after previous cesarean section. Am J Obstet Gynecol 1991;165:996–1001.
28. Jones RO, Nagashima AW, Hartnett-Goodman MM, Goodlin RC. Rupture of low transverse cesarean scars during trial of labor. Obstet Gynecol 1991;77:815–7.
29. Scott JR. Mandatory trial of labor after cesarean delivery: An alternative viewpoint. Obstet Gynecol 1991;77:811–4.
30. Eden RD, Parker RT, Gall SA. Rupture of the pregnant uterus: A 53-year review. Obstet Gynecol 1986;68:671–4.
31. Kaplan RM, Anderson JP. The general health policy model: Update and applications. Health Serv Res 1988;23:203–35.
32. Macario A, Vitez T, Dunn B, McDonald T. What does perioperative care really cost? Analysis of hospital costs and charges for inpatient surgical care. Anesthesiology 1995;83:1138–44.
33. Gordon B, ed. American Medical Association. Current procedural terminology. Chicago: American Medical Association, 1999.
34. Waitzman NJ, Romano PS, Scheffler RM. Estimates of the economic costs of birth defects. Inquiry 1994;31:188–205.
35. Chuang J, Jenders R. Trial of labor versus elective repeat cesarean section for the women with a previous cesarean section: A decisions analysis. Proc AMIA Symp 1999;226–30.
36. Deber RB, Kraetschmer N, Irvine J. What role do patients wish to play in treatment decision-making? Arch Intern Med 1996;156:1414–20.
37. Weinstein D, Benshushan A, Tanos V, Zilberstein R, Rojansky N. Predictive score for vaginal birth after cesarean section. Am J Obstet Gynecol 1996;174:192–8.
38. Troyer L, Valerie MP. Obstetric parameters affecting success in a trial of labor: Designation of a scoring system. Am J Obstet Gynecol 1992;167:1099–104.
39. Pickhardt M, Martin J, Meydrech EF, Blake PG, Martin RW, Perry KG, et al. Vaginal birth after cesarean delivery: Are there useful and valid predictors of success or failure? Am J Obstet Gynecol 1992;166:1811–9.
40. Clark SL, Scott JR, Porter TF, Schlappy DA, McClellan V, Burton DA. Is vaginal birth after cesarean less expensive than repeat cesarean delivery? Am J Obstet Gynecol 2000;182:599–602.
© 2001 The American College of Obstetricians and Gynecologists
41. Grobman WA, Peaceman AM, Socol ML. Cost-effectiveness of elective cesarean delivery after one prior low transverse cesarean. Obstet Gynecol 2000;95:745–51.