Unexplained stillbirth is the most frequent type of fetal death.1–5 Recognized risk factors for stillbirth include previous stillbirth, hypertension, diabetes, a small-for-gestational-age fetus, and advancing gestational age; women with these conditions routinely undergo antepartum testing.6–10 Approximately 65% of unexplained stillbirths occur after 35 completed weeks, with risk increasing as the gestation advances.2,3 The most significant risk factor for unexplained stillbirth is advanced maternal age, but other risk factors, such as prepregnancy obesity, and low socioeconomic status, also have been found to be important predictors.1–3 In otherwise healthy women without additional risk factors, evidence suggests that there are fetal benefits to delivery at 41 weeks without an increase in maternal morbidity and mortality when compared with expectant management with antepartum testing.11–17 Based on the estimated absolute reduction in stillbirth observed in a pooled analysis of randomized trials of induction versus expectant management, between 460 and 500 inductions would need to be performed to prevent one stillbirth.11–13 However, in a low-risk population, there is little information to assess whether antepartum testing before 41 weeks will improve fetal outcomes.13
Although advanced maternal age has been identified as a significant risk factor for stillbirth at term,2,3,18–21 older women do not routinely undergo antepartum testing on the basis of their age alone. Unexplained fetal death is one of obstetrics’ most distressing outcomes, and it is especially devastating in older women, who take longer to conceive and have fewer reproductive opportunities.22 Among women younger than age 35 in the general population, the risk of unexplained fetal death is relatively low, estimated at 1.1 per 1,000 births.2 For women aged 35 years and older, the estimated rate is 3.6 per 1,000 births, and for women 40 years of age or older, 4 per 1,000.2,3 Although unexplained fetal deaths are relatively rare even in older women, they occur 6 times more often than deaths because of sudden infant death syndrome, which was reported to occur at a rate of 0.6 per 1,000 births in 1998.23
A strategy of antepartum testing to reduce the risk of late fetal death in otherwise healthy women has not been subjected to a randomized trial; given the large sample size needed, such a trial seems unlikely for logistical reasons alone. In this setting, a decision analysis of alternative strategies offers the ability to compare the trade-offs of testing and induction, including the consequences of false-positive and false-negative tests, and to estimate the number of additional interventions required for each fetal death averted.
We developed a Markov state-transition model to compare 3 strategies for managing older pregnant women with no other risk factors for stillbirth. A Markov model simulates the transition of subjects at discrete time intervals among health states relevant to the clinical situation.24 Outcomes incurred during each time interval, or Markov cycle, are aggregated across the cohort when a specified number of cycles has elapsed. The strategies we evaluated were 1) usual care with no antepartum testing, 2) weekly antepartum testing starting at 37 weeks of gestation and induction if the patient has a positive test result, and 3) no antepartum testing, but induction of all women still undelivered at 41 weeks of gestation. We estimated the average number of fetal deaths, tests, labor inductions, and cesarean deliveries associated with each strategy.
Our antepartum testing model (Fig. 1) started with a cohort of women in their 37th week of pregnancy. During each 1-week Markov cycle, patients received an antepartum test or experienced spontaneous labor before testing. We assumed that a positive test result would lead to labor induction regardless of the Bishop score of the cervix. In both spontaneously laboring and induced women, delivery could be vaginal or cesarean. We assumed that all deliveries resulting from spontaneous labor before testing would result in a live birth. Labor inductions as the result of a positive test result also were assumed to result in a live birth, regardless of delivery method. Women with a negative test result remained undelivered or experienced an unexplained fetal loss.
The Markov process was repeated for a total of 5 cycles, until the end of 41 completed weeks of gestation (7 days after the estimated due date). We restricted the analysis to this time frame to separate antepartum testing in births at term from the issue of postdate pregnancy management. In the no-testing strategy, women faced a week-specific chance of spontaneous labor. Fetal death was expected to occur each week in a proportion of pregnancies. Those who remained undelivered at the end of 1 cycle repeated the Markov process until the end of 41 completed weeks of gestation. For the third strategy evaluated—induction at 41 weeks—we assumed that labor would be induced in all women who remained undelivered at 41 completed weeks of gestation. All computations were performed using a commercially available decision analysis program (DATA 4.0; TreeAge Software, Williamstown, MA) The base-case probabilities for the model are presented in Table 1.
The weekly risk of unexplained fetal death per 1,000 ongoing pregnancies was obtained from the Canadian hospital-based McGill Obstetrical Neonatal Database, which has previously been used to evaluate risk factors for stillbirth.2,4,5,21 The Royal Victoria Hospital is an academic medical center affiliated with McGill University. It serves a socioeconomically diverse population and has a computerized databank maintained by one researcher that currently contains information on over 100,000 births. Births of fetuses or infants weighing 500 g or more are included. Ninety-seven percent of stillbirths underwent fetopsy and placental histopathologic examination. All fetal deaths were reviewed collectively, and the primary cause of death was assigned. Unexplained fetal deaths were defined as deaths occurring before labor with no evident fetal, maternal, or placental abnormality. Fetal deaths were excluded if the mother had hypertension or diabetes or if the fetus had severe growth restriction (less than 2.4%) or a major congenital anomaly. Fetal deaths associated with placental pathology, such as abruption, placenta previa, or fetal-maternal hemorrhage, also were excluded, as were fetal deaths, because of intrapartum asphyxia or infection. Other fetal deaths were excluded if the mother had significant maternal disease such as lupus or severe renal disease. The incidence of unexplained fetal death at each gestational age per 1,000 fetuses at risk was determined from this data set; these estimates are similar to those reported by others.1,3,25
Using the McGill Obstetrical Neonatal Database data, we calculated weekly probabilities of unexplained fetal death, conditional upon remaining undelivered at the end of the prior week. These probabilities were estimated for the 5 weeks of our analysis (the 37th week to the 41st week or 36 weeks and 0 days to 40 weeks and 7 days). We similarly calculated weekly probabilities of spontaneous labor and delivery using these data. Weekly probabilities of unexplained fetal death and spontaneous labor were based on all pregnancies in the data set. The total fetal death rate in the cohort was 4.5 per 1,000, with the unexplained stillbirths representing 25.5% of all losses. The median maternal age in the McGill Obstetrical Neonatal Database data set was 29.8 years. The baseline weekly probabilities of unexplained fetal death used in our model (Table 1) reflect the risk for women under age 35, because this group was considered the reference population.
The effect of advanced maternal age on the risk of unexplained fetal death was also estimated from the McGill Obstetrical Neonatal Database data set. Using a logistic regression model with both maternal age and parity as predictors, we found that nulliparous women aged 35 years and older had a 3.3-fold increase in the risk of unexplained fetal death compared with women younger than 35 years of age. The odds ratio for maternal age 40 years and older was 3.7. We assumed that the increased risk of unexplained fetal death attributable to advanced maternal age did not vary by gestational age during the 5-week period of our analysis. In sensitivity analysis we varied the odds ratio for unexplained fetal death from a low value of 1.0 (same risk as women younger than 35 years of age) to a high value of 5.0 (high risk).
There are extremely limited data on the test characteristics of antepartum tests that explicitly use stillbirth as the “gold standard”; most studies use surrogate markers such as admission to the neonatal intensive care unit, fetal distress in labor, meconium staining, cesarean delivery for fetal distress, or another marker of fetal compromise.13,26,27 Because clinical decision makers were not blinded to the results of the antepartum tests in most of these studies, there is, if anything, a bias in favor of increased sensitivity for testing. Although formal meta-analysis is not possible because of the variability in methods and patient populations, a recent systematic review found that antepartum testing consistently had relatively low sensitivity and relatively high specificity for these surrogate markers.13 Based on this review, we assumed for the base case that the sensitivity, or true-positive rate, of antepartum testing was 70% and the specificity, or true-negative rate, was 90%. This would result in a 10% false-positive rate, leading to induction, and a 30% false-negative rate for fetal death in the interval before the next test. We did not assume a specific test (ie, nonstress testing versus biophysical testing) because there is no evidence that one type of testing is superior to another. We assumed that there was some benefit to antepartum testing, by identifying fetuses that were demonstrating impending fetal compromise (true-positive tests), but some benefit would also be derived from false-positive tests, by decreasing the number of women left undelivered and therefore at risk for a late fetal death. We assumed that there would be no significant long-term adverse fetal effects of elective delivery at the 37th week of gestation. Because there are no large studies specifically testing the utility of antepartum testing in low-risk older women, we performed a sensitivity analysis and varied the sensitivity and specificity from 60% to 90%,
Induction of labor has been associated with an increase in cesarean delivery in a variety of circumstances,28,29 although a large Canadian randomized trial of induction versus monitoring starting at 41 weeks of gestation did not show a significant increase in the risk of cesarean delivery.11 The possible increased risk of cesarean delivery may be justified when the benefits to mother or infant outweigh the maternal risks from surgery. However, among women who receive a false-positive antepartum test result, consequent labor induction and its attendant increase in risk of cesarean delivery are unnecessary procedures. To quantify these risks, we explicitly modeled the frequency of cesarean delivery with each management strategy.
We estimated the probability of cesarean delivery at each gestational age from a sample of 14,409 births among healthy women with uncomplicated pregnancies eligible for vaginal delivery in 1998 and 1999 at 2 large teaching institutions.28 This population was 62% non-Hispanic white; 14.6% received public assistance for medical care, and the average maternal age was 31 years. Parous and nulliparous women were analyzed separately, and the risk of cesarean delivery was estimated as a function of gestational age, maternal age, and induced labor.28 The probability of cesarean delivery after spontaneous labor was 14% among nulliparous women and 8% among parous women. The odds ratio for cesarean delivery after labor induction varied from 2.1 to 5.6 among nulliparas, depending on gestational age and maternal age. Among all women, the odds ratio varied from 1.4 to 3.4. These estimates are similar to those published by Seyb et al.29
Among nulliparous women aged 35 years and older, 5.2 unexplained fetal deaths per 1,000 pregnancies would be expected between the 37th and the 41st weeks of gestation in the absence of weekly antepartum testing (Table 2). Weekly antepartum testing initiated during the 37th week would avert 3.9 unexplained fetal deaths per 1,000 pregnancies and would require 863 antepartum tests, 71 inductions, and 14 additional cesarean deliveries to avert 1 fetal death. Each woman would receive an average of 3 tests with this strategy. If no antepartum testing was performed, induction of all women undelivered at 41 weeks would avert 0.9 unexplained fetal deaths per 1,000 pregnancies, with 469 inductions and 219 additional cesarean deliveries per fetal death averted (Table 2).
We varied the odds ratio for unexplained fetal death to that of a low-risk women younger than age 35 (odds ratio 1.0). A strategy of antepartum testing reduced 1.2 fetal deaths per 1,000 births and was associated with a higher ratio of tests (2,862), inductions (233), and cesarean deliveries (44) per fetal death averted (Table 3). As the risk of fetal death increased, the number of fetal deaths prevented by weekly antepartum testing increased. For women with a risk estimated to be increased 5-fold (which could occur if a women had more than one risk factor, such as advanced maternal age and black race or prepregnancy obesity), 5.9 unexplained fetal deaths would be averted with fewer antepartum tests (569), inductions (47), and additional cesarean deliveries9 per fetal death averted. The number of tests per pregnancy did not vary significantly depending on the risk estimate (Table 3).
Modeling an improved test sensitivity improved outcomes; however, improving test specificity actually worsened outcomes (Table 4). Holding specificity constant at its base-case value of 90%, a test with 90% sensitivity performed better than one with 70% sensitivity—more fetal deaths were averted with fewer inductions and cesarean deliveries per fetal death averted. Holding sensitivity constant at its base-case value of 70%, a test with 90% specificity averted fewer fetal deaths than one with 70% specificity. Figure 2 illustrates the probable mechanism for this finding: with lower test specificity, more women are delivered and, thus, are not pregnant and at risk of fetal death in subsequent weeks. However, the less specific test substantially increases the number of inductions and cesarean deliveries per fetal death averted
When we delayed the start of weekly antepartum testing, fewer fetal deaths were prevented compared with testing initiated during the 37th week (Table 5). With each week of delay in the initiation of testing, the number of inductions required to avert one fetal death dropped substantially. However, the number of cesarean deliveries per fetal death averted remained relatively stable, and the number of tests per fetal death averted remained high.
In making decisions about any strategy in pregnancy, clinicians and expectant mothers need to balance the risks and benefits to the mother and the fetus. This analysis, by quantifying the probabilities of risks and benefits of weekly antepartum testing for nulliparous women aged 35 years or older, does not provide the “right” answer but helps provide information that should be helpful to these women and the clinicians caring for them.
We focused on older women because these women have fewer reproductive opportunities, but this analysis could apply to other women who are at increased risk of stillbirth late in pregnancy (ie, overweight women or African-American women). Similar to nulliparous women, parous women also experience an increased risk of stillbirth late in pregnancy and would require the same estimated number of antepartum tests and inductions to avert a stillbirth but would have a lower estimated number of additional cesarean deliveries per fetal death averted.28 To determine whether a strategy of antepartum testing is reasonable for older women, we might consider other current practices in obstetrics. The magnitude of the risk of stillbirth in older women is similar to that of women with hypertension and diabetes, for whom antepartum testing late in pregnancy is routinely performed.7,8,10,20,21,30 Similarly, women aged 35 years and older routinely are offered invasive prenatal testing to detect a fetus with Down syndrome, with an estimated 365 amniocenteses (and one loss of a normal infant) required to detect one affected fetus.31 Induction at 41 weeks requires an estimated 460–500 inductions to prevent one perinatal death.12–14 Rouse and colleagues,32 also using a decision analysis model, estimated that in the diabetic patient, 443 cesarean deliveries for suspected fetal macrosomia (4,500 g) would be performed to avert a single permanent Erb's palsy.
When a strategy of antepartum testing and induction is compared with no testing but induction at 41 weeks, the latter strategy resulted in fewer fetal deaths averted and a higher ratio of inductions and cesarean deliveries performed per fetal death averted, largely because the majority of expected fetal deaths would have occurred before 41 weeks of gestation. We tried to quantify the maternal risks for a strategy of stillbirth prevention in an otherwise healthy woman of advanced maternal age. We did not model intrapartum deaths; the estimated risk of an intrapartum death in a nonanomalous fetus is less that 1 in 10,000 births.33 We are not aware of any data suggesting that intrapartum fetal deaths occur more or less often in a spontaneous versus an induced labor.
We assumed that once a fetus was identified as being at risk, induction would be performed, although the clinician or patient might elect a different strategy. For example, if a nonstress test was nonreassuring, additional tests, such as a complete biophysical profile or an oxytocin challenge test, might be performed. The characteristics of serial testing are not known, nor is the effect of fetal movement counting followed by formal antepartum fetal testing, although intuitively it should have the effect of decreasing the false-positive rate, thereby reducing the number of inductions. We also assumed that the Bishop score of the cervix would not influence the threshold for induction and that the threshold for induction after nonreassuring testing would be the same at 37 weeks as it would be at 41 weeks. However, clinician and patient thresholds for interventions, such as cesarean delivery, appear to be influenced by maternal age alone.34,35 It seems likely that the threshold for induction in the setting of abnormal fetal testing in women aged 35 years and older would be relatively low.
We assumed in our model that the patient had only one identifiable risk factor for unexplained fetal death, ie, advanced maternal age. We did not specifically model the scenario of a patient with one or more additional factors associated with late unexplained fetal death, such as prepregnancy obesity, black race, and lower socioeconomic status, because data on the interactions of these factors is limited. Instead, we performed a sensitivity analysis in which a clinician might estimate a patient's risk between odds ratios of 1.0 to 5.0. When we increased the underlying risk for stillbirth to 5 times that of the low-risk population, the testing profile changed, such that only 567 antepartum tests, 47 inductions, and 9 additional cesarean deliveries were required to avert one fetal death. Little research has been performed on women's attitudes toward antepartum testing. Roberts and Young36 showed that women's views of induction for a prolonged pregnancy changed, such that the idea of induction became more attractive as the gestation advanced.
One interesting finding of this analysis was that improved test specificity led to an increase in the number of expected fetal deaths. This is the result of the dynamic nature of the risk of stillbirth. Because antepartum testing has a relatively low sensitivity, the negative predictive value of testing is largely driven by the low risk of stillbirth rather than by test characteristics. As pregnancy progresses, the risk of stillbirth increases, resulting in declining negative predictive values with advancing gestational age (a finding consistent with randomized trial data on induction versus testing in postterm pregnancy).11,13,14 Thus, because lower specificity results in more inductions, fewer women are pregnant in subsequent weeks, and the absolute number of false-negative antepartum test results is decreased.
We did not perform a formal cost-effectiveness analysis. Although a cost-effectiveness analysis of antepartum testing in older women would provide useful information, there has been surprisingly little work focused on the cost-effectiveness of obstetric interventions. Unlike medical interventions in other populations and disease areas, obstetric interventions have consequences to 2 patients simultaneously—the mother and the fetus. This poses a number of methodological challenges to formal cost-effectiveness analysis, such as the estimation of utility weights for valuing quality of life and the aggregation of outcomes to capture all relevant impacts. Consequently, most cost-effectiveness analyses have used a disease-specific clinical measure of effectiveness, such as cost per case of neonatal human immunodeficiency virus prevented.37–39 The few studies that have reported effectiveness in the preferred metric of quality-adjusted life-years included both maternal and infant outcomes, but only infant survival was adjusted to reflect quality of life.40,41 Given the methodological challenges and wide variability in the cost of antepartum testing strategies, we elected to estimate the number of interventions required under different strategies, a clinically intuitive way of comparing relative resource utilization short of a formal cost-effectiveness analysis
Compared with no testing, a policy of weekly antepartum testing beginning at 37 weeks of gestation in otherwise low-risk women aged 35 years and older, with induction of labor for women with abnormal test results, would result in a reduction in unexplained fetal deaths. The efficiency of this strategy, in terms of number of inductions and cesarean deliveries, appears to be comparable to other accepted indications for antepartum testing. Ideally, this strategy should be evaluated in a randomized trial. However, a randomized trial to confirm our estimates of a decrease in unexplained fetal deaths from 5.2 per 1,000 to 1.3 per 1,000 would require approximately 7,000 subjects to have 80% power, which is beyond the capacity of any single institution. Before a policy of antepartum testing is implemented in clinical practice, either such a trial should be performed on a multi-institutional basis or the effectiveness of our strategy should be independently confirmed.
1. Yudkin PL, Wood L, Redman CW. Risk of unexplained stillbirth at different gestational ages. Lancet 1987;1:1192–4.
2. Huang DY, Usher RH, Kramer MS, Yang H, Morin L, Fretts R. Determinants of unexplained antepartum fetal deaths. Obstet Gynecol 2000;95:215–21.
3. Froen JF, Arnestad M, Frey K, Vege A, Saugstad OD, Stray-Pedersen B. Risk factors for sudden intrauterine unexplained death: epidemiological characteristics of singleton cases in Oslo, Norway, 1986–1995. Am J Obstet Gynecol 2001;184:694–702.
4. Fretts RC, Usher RH. Fetal death in women in the older reproductive age group. Contemp Rev Obstet Gynecol 1997;9:173–7.
5. Fretts RC, Usher RH. Causes of fetal death in women of advanced maternal age. Obstet Gynecol 1997;89:40–5.
6. Tenovuo A, Kero P, Piekkala P, Korvenranta H, Erkkola R. Fetal and neonatal mortality of small-for-gestational-age infants. Eur J Pediatr 1988;147:613–5.
7. Spinillo A, Capuzzo E, Piazzi G, Nicola S, Colonna L, Iasci A. Maternal high risk factors and severity of growth deficit in small for gestational age infants. Early Hum Dev 1994;38:35–43.
8. Harman C, Menticoglou S, Manning F, Albar H, Morrison I. Prenatal fetal monitoring abnormalities of fetal behavior. In: James DK, editor. High risk pregnancy management options. London: Saunders; 1994. p. 693–734.
9. Weeks JW, Asrat T, Morgan MA, Nageotte M, Thomas SJ, Freeman RK. Antepartum surveillance for a history of stillbirth: when to begin. Am J Obstet Gynecol 1995;172:486–92.
10. Smulian JC, Ananth CV, Vintzileos AM, Scorza WE, Knuppel RA. Fetal deaths in the United States: influence of high-risk conditions and implications for management. Obstet Gynecol 2002;100:1183–9.
11. Hannah ME, Hannah WJ, Hellmann J, Hewson S, Milner R, Willan A. Induction of labor as compared with serial antenatal monitoring in post-term pregnancy. The Canadian Multicenter Post-Term Pregnancy Trial Group. N Engl J Med 1992;326:1587–92.
12. Grant JM. Induction of labour confers benefits in prolonged pregnancy [review]. Br J Obstet Gynaecol 1994;101:99–102.
13. Myers ER, Blumrick R, Christian A, Gray R, Kolimaga J, Livingston E, et al. Management of prolonged pregnancy. Evidence Report/ Technology Assessment No. 3. Prepared by Duke University under contract No. 02-97-0014. AHRQ Publication No. 02-E012. Rockville (MD): Agency for Healthcare Research and Quality; 2002.
14. Crowley P. Interventions for preventing or improving the outcome of delivery at or beyond term (Cochrane Review). In: The Cochrane Library, Issue 1, 2003. Oxford: Update Software.
15. Hilder L, Costeloe K, Thilaganathan B. Prolonged pregnancy: evaluating gestation-specific risks of fetal and infant mortality. Br J Obstet Gynaecol 1998;105:169–73.
16. Goeree R, Hannah M, Hewson S. Cost-effectiveness of induction of labour versus serial antenatal monitoring in the Canadian Multicentre Postterm Pregnancy Trial. CMAJ 1995;152:1445–50.
17. Stallmach T, Hebisch G, Meier K, Dudenhausen JW, Vogel M. Rescue by birth: defective maturation and late fetal death. Obstet Gynecol 2001;97:505–9.
18. Prysak M, Lorenz RP, Kisly A. Pregnancy outcome in nulliparous women 35 years and older. Obstet Gynecol 1995;85:65–70.
19. Edge V, Laros RK Jr. Pregnancy outcome in nulliparous women aged 35 or older. Am J Obstet Gynecol 1993;168:1881–5.
20. Cnattingius S, Stephansson O. The epidemiology of stillbirth. Semin Perinatol 2002;26:25–30.
21. Fretts RC, Schmittdiel J, McLean FH, Usher RH, Goldman MB. Increased maternal age and the risk for fetal death. N Engl J Med 1995;333:953–7.
22. Menken J, Trussell J, Larsen U. Age and fertility [published erratum appears in Science 1986;234:413]. Science 1986;233:1389–94.
23. National Vital Statistics Reports. Hyattsville (MD): NCHS; 2000;48:93.
24. Sonnenberg FA, Beck JR. Markov models in medical decision-making: a practical guide. Med Decis Making 1993;13:322–8.
25. Cotzias CS, Paterson-Brown S, Fisk NM. Prospective risk of unexplained stillbirth in singleton pregnancies at term: population based analysis. BMJ 1999;319:287–8.
26. Thacker SB, Berkelman RL. Assessing the diagnostic accuracy and efficacy of selected antepartum fetal surveillance techniques [review]. Obstet Gynecol Surv 1986;41:121–41.
27. Nageotte M, Towers CV, Asrat A, Freeman R. Perinatal outcome with the modified biophysical profile. Am J Obstet Gynecol 1994;170:1672–6.
28. Heffner LJ, Elkin E, Fretts RC. Impact of labor induction, gestational age and maternal age on the cesarean delivery rates in nulliparous and multiparous women near term. Obstet Gynecol 2003;102:287–93.
29. Seyb S, Berka RJ, Socol ML, Dooley SL. Risk of cesarean delivery with elective induction of labor at term in nulliparous women. Obstet Gynecol 1999;94:600–7.
30. Rouse DJ, Owen J, Goldenberg RL, Cliver SP. Determinants of the optimal time in gestation to initiate antenatal testing: a decision-analytic approach. Am J Obstet Gynecol 1995;173:1357–63.
31. Hook EB, Lindsje A. Down syndrome in live births by single year maternal age interval in a Swedish study: comparison with results from a New York State study. Am J Hum Genet 1978;30:19–27.
32. Rouse DJ, Owen J, Goldenberg RL, Cliver SP. The effectiveness and costs of elective cesarean delivery for fetal macrosomia diagnosed by ultrasound. JAMA 1996;276:1480–6.
33. Fretts RC, Boyd ME, Usher RH, Usher HA. The changing pattern of fetal death, 1961–1988. Obstet Gynecol 1992;79:35–9.
34. Ecker JL, Chen KT, Cohen AP, Riley LE, Lieberman ES. Increased risk of cesarean delivery with advancing maternal age: indications and associations in nulliparous women. Am J Obstet Gynecol 2001;185:883–7.
35. Main DM, Main EK, Moore DH 2nd. The relationship between maternal age and uterine dysfunction: a continuous effect throughout reproductive life. Am J Obstet Gynecol 2000;182:1312–20.
36. Roberts LJ, Young KR. The management of prolonged pregnancy: an analysis of women’ s attitudes before and after term. Br J Obstet Gynaecol 1991;98:1102–6.
37. Ecker JL. The cost-effectiveness of human immunodeficiency virus screening in pregnancy. Am J Obstet Gynecol 1996;174:716–21.
38. Mauskopf JA, Paul JE, Wichman DS, White AD, Tilson HH. Economic impact of treatment of HIV-positive pregnant women and their newborns with zidovudine: implications for HIV screening. JAMA 1996;276:132–8.
39. Myers ER, Thompson JW Simpson K. Cost-effectiveness of mandatory compared with voluntary screening for human immunodeficiency virus in pregnancy. Obstet Gynecol 1998;91:174–81.
40. Binkin NJ, Koplan JP. The high cost and low efficacy of weekly viral cultures for pregnant women with recurrent genital herpes: a reappraisal. Med Decis Making 1989;9:225–30.
© 2004 The American College of Obstetricians and Gynecologists
41. Randolph AG, Washington AE, Prober CG. Cesarean delivery for women presenting with genital herpes lesions: efficacy, risks, costs. JAMA 1993;270:77–82.