Vasa previa, estimated to occur in approximately one in 2,500 pregnancies, is an obstetric condition characterized by fetal vessels traveling within the amniotic membranes, unprotected by the placenta or umbilical cord, that overlie the cervix. With rupture of membranes or at labor onset, the umbilical vessels can lacerate with resultant fetal exsanguination and demise.1,2
When vasa previa has been diagnosed using ultrasonography, cesarean delivery can be scheduled prior to the onset of expected labor or rupture of membranes and thereby improve perinatal outcome.3–5 In fact, in one of the largest series published to date, most women with a prenatal diagnosis of vasa previa were delivered preterm.5 Yet, while the benefit of early delivery decreases the risk of fetal demise, it may expose a neonate to prematurity complications. Consequently, optimal care would strike a balance between limiting the risk of fetal demise and complications of prematurity. Unfortunately, there is little evidence to guide physicians in their choice of the optimal gestational age, which has led to consequent variability with regard to delivery timing in these individuals.3,6–8 In an effort to better quantify the tradeoffs in outcomes regarding timing of delivery, we have performed a decision analysis to determine at what gestational age individuals with ultrasonographic evidence of vasa previa should be delivered.
MATERIALS AND METHODS
This study was considered exempt by the Institutional Review Board at Northwestern University. A decision analytical model was designed using DATA 3.5 software that compared 11 different strategies for the timing of delivery in individuals with a singleton gestation and ultrasonographic evidence of vasa previa and no other complications of pregnancy (eg, preeclampsia, fetal growth restriction). All patients were presumed to have reached 32 weeks of gestation. The 11 strategies are displayed in Table 1. Strategies 1, 2, 3, and 4 include delivery at 32, 33, 34, and 35 weeks of gestation, respectively, after previous administration of antenatal corticosteroids for neonatal benefit. The remaining strategies do not involve routine administration of antenatal corticosteroids. Strategies 5, 7, 9, and 11 are expectant management until the target gestational age of 36, 37, 38, and 39 weeks of gestation, respectively. Strategies 6, 8, and 10 involve expectant management until an amniocentesis test for fetal lung maturity at the indicated gestational age has been performed. In these three strategies, in the event that testing shows lung immaturity, amniocentesis is repeated weekly and cesarean delivery is performed when either the fetal lung maturity is confirmed or at 39 weeks of gestation, whichever occurs first. In all strategies, if rupture of membranes, antepartum hemorrhage, or labor onset occurs, expectant management is abandoned and an emergent cesarean delivery performed.
The base-case estimates were chosen based on data in the published literature. Estimates of the weekly risk of premature rupture of membranes were derived from the National Vital Health Statistics data.9,10 The frequency of premature rupture of membranes increases with gestational age, and by term, 8% of women experience premature rupture of membranes before the onset of labor.11,12 This relationship was reflected in the model. Vasa previa was assumed not to alter the incidence of premature rupture of membranes. A positive predictive value of 83% was used for the ultrasonographic prediction of vasa previa.4
In one large series,5 28% of patients who were prenatally diagnosed with vasa previa were emergently delivered. This study, in combination with the gestational age distribution of all births in the United States,5,9,11 was used to estimate the probability that a woman with a vasa previa would require unscheduled delivery at a given gestational age.
Respiratory distress syndrome estimates were derived from a multinational observational study detailing neonatal morbidities of more than 20,000 neonates not routinely exposed to in utero corticosteroids.13 The frequency of respiratory distress syndrome in the model was decreased by approximately half for neonates with mothers receiving antepartum steroids.14 The probability that the amniocentesis revealed a mature value at each gestational age has been described previously.15 The risk of perinatal mortality in the setting of premature rupture of membranes and vasa previa was estimated based on prior studies detailing outcomes of women with vasa previa with particular emphasis given to those that documented timing of rupture of membranes and perinatal outcome.3–5,8,16–22 The risk of perinatal mortality in the setting of labor onset and any bleeding was chosen based on the risk of fetal and early neonatal demise in a large series examining outcomes related to vasa previa.5 If emergent delivery was required of individuals in groups 1 through 4 more than 1 week before their anticipated date of delivery, delivery was not delayed to administer steroids. Antenatal corticosteroids were assumed to decrease the gestational age-specific risk of mortality for preterm infants by approximately 30%.14 The assumption was made that amniocentesis would not alter the risk of gross membrane rupture, preterm labor, or antepartum bleeding.
Because vasa previa imposes potentially catastrophic risks to the fetus, and this model altered timing but not route of delivery, the effectiveness of each strategy was evaluated based on differences in perinatal but not maternal outcomes. Cesarean delivery was performed for all individuals. The life expectancy for the newborn was assumed to be 75 years; quality-adjusted life-years for each strategy were based on this anticipated life expectancy. Adverse perinatal outcomes considered in the model were perinatal mortality, respiratory distress syndrome, cerebral palsy, mental retardation, and infant mortality. Whenever a range of values for a given variable was found in the literature, sensitivity analysis was used to evaluate the range of values.
Newborn utilities for the various outcomes were determined on the basis of the literature (Table 2).23 Applied quality-adjusted life-years for both cerebral palsy and mental retardation were based on utilities for moderate cerebral palsy and moderate mental retardation. Utilities associated with respiratory distress syndrome were applied for shorter durations than utilities for chronic conditions such as mental retardation and cerebral palsy, and the decrement in quality of life associated with mental retardation and cerebral palsy was considered from 1 year of life onward. Infants dying within the first year were presumed to have died at 3.5 months of life.24 Baseline risks of stillbirth were assumed not to differ outside of the setting of rupture of membranes, antepartum hemorrhage, or labor onset.
When the baseline assumptions listed in Tables 2 through 4 are used, Strategy 3 –delivery at 34 weeks of gestation after administration of antenatal corticosteroids–results in the highest total quality adjusted life-years (Table 5). Additionally, delivery of individuals without amniocentesis is always favored over delivery only after confirmation of fetal lung maturity at a given gestational age.
One-way sensitivity analyses were performed for all variables. These analyses revealed that the base-case conclusion of the top two ranking strategies were affected only when one of three variables (the incidence of premature rupture of membranes, perinatal mortality in the setting of premature rupture of membranes, and the positive predictive value of ultrasonography for vasa previa) was altered. When the probability of preterm premature rupture of membranes was at the low end of the analyzed ranges, the preferred strategy switches from delivery at 34 weeks of gestation to delivery at 35 weeks of gestation. Figure 1 assesses the effect of varying the probability of perinatal mortality in the setting of premature rupture of membranes. There are two thresholds at which the most effective strategy is no longer delivery at 34 weeks of gestation. If the incidence of perinatal mortality in the setting of premature rupture of membranes falls to between 8.6% and 17.6%, the preferred strategy is to deliver at 35 weeks of gestation. Only when the incidence of perinatal mortality with premature rupture of membranes is less than 8.6% does expectant management until 37 weeks of gestation without amniocentesis become the most preferred strategy. When performing one-way sensitivity analysis on the positive predictive value of ultrasonography for vasa previa, the preferred strategy becomes delivery at 35 weeks of gestation at values of 55.8% or lower.
The results were also determined after incorporation of a more liberal betamethasone administration policy, in which mothers delivering in a nonemergent fashion between 36 and 38 weeks of gestation received betamethasone 2 days to 1 week before delivery. The only mothers who did not receive betamethasone were those for whom an amniocentesis was performed with mature results, those delivering emergently, or those delivered at 39 weeks of gestation. Thus, steroids were administered at 35, 36, and 37 weeks of gestation with patients delivered at 36, 37, and 38 weeks of gestation, respectively, for strategies 5, 7, and 9. For strategies 6 and 8, steroids were administered at 36 and 37 weeks of gestation, respectively, only if amniocentesis demonstrated lung immaturity, with all patients delivered 1 week later. Even when administered beyond 34 weeks, steroids were presumed to decrease the risks of respiratory distress syndrome by 50%. Incorporation of this more liberal antenatal corticosteroid policy did not result in any notable change in the quality-adjusted life-years associated with each strategy or ranking of the delivery strategies.
Two-way sensitivity analyses were performed for all combinations of variables. Analyses that led to a change in the preferred strategy were those that included one of the three variables identified in one-way sensitivity analysis as well as two other specific combinations in which varying the incidence of labor onset and bleeding, the incidence of fetal death in the setting of labor onset and bleeding, and the incidence of infant mortality affected ranking. These two specific combinations that contained variables distinct from those identified in the one-way analysis resulted in the preferred strategy switching to delivery at 35 weeks of gestation only under very select circumstances. Figure 2 illustrates a two-way sensitivity analysis examining the relationship between the probability of perinatal mortality in the setting of premature rupture of membranes and the positive predictive value of ultrasonography for vasa previa. The graph depicts that when the positive predictive value of ultrasonography for vasa previa is as low as 50%, the preferred strategy shifts to delivery at 35 weeks of gestation if the perinatal mortality risk with premature rupture of membranes is between approximately 23% and 38%. In the setting of less reliable ultrasonographic diagnoses for vasa previa and a perinatal mortality risk of less than 23%, the preferred strategy is to deliver at 37 weeks of gestation without amniocentesis. As the ultrasonographic reliability for vasa previa improves to estimates at the upper end of the range of published values, delivery at 34 weeks of gestation remains the preferred strategy except for when the incidence of perinatal mortality with premature rupture of membranes is lower than 17%.
Figure 3 illustrates a two-way sensitivity analysis examining the relationship between perinatal mortality with premature rupture of membranes and the frequency of premature rupture of membranes. When the frequency of premature rupture of membranes at 32 weeks of gestation is lowered to 0.13% (with a corresponding decrease in the frequency at other gestational ages as well), the preferred strategy is to deliver at 34 weeks of gestation if the perinatal mortality with premature rupture of membranes is 35% or higher. With this low incidence of premature rupture of membranes, when the risk of perinatal mortality with premature rupture of membranes falls between approximately 18% and 35%, the preferred strategy becomes delivery at 35 weeks of gestation. Only when the chance of perinatal mortality with premature rupture of membranes drops to less than 18% does expectant management until 37 weeks of gestation without amniocentesis become the preferred management strategy in the setting of a minimal risk of premature rupture of membranes. As both the frequency of premature rupture of membranes and perinatal mortality are increased to high estimates, delivery at 32 weeks of gestation becomes the preferred strategy.
Figure 4 is a two-way sensitivity analysis examining the relationship between the frequency of premature rupture of membranes and the positive predictive value of ultrasonography for vasa previa. If the chance of premature rupture of membranes at 32 weeks of gestation is higher than 0.3% (with a corresponding increase in the frequency at other gestational ages as well), the preferred strategy is to deliver at 34 weeks of gestation. In instances in which the positive predictive value of ultrasonography for vasa previa is less than 85%, delivery at 35 weeks of gestation becomes the preferred strategy in most situations in which the frequency of premature rupture of membranes is low, save in limited circumstances in which the positive predictive value of ultrasonography for vasa previa is low as well.
This decision analysis indicates that, in women with ultrasonographic diagnoses of vasa previa, the preferred strategy for timing of delivery under most but not all circumstances is scheduled delivery at either 34 or 35 weeks of gestation. Additionally, the model suggests that when delivery beyond 35 weeks is the preferred option, the advantages of amniocentesis for confirmation of fetal lung maturity fail to outweigh the disadvantages. It should be noted that this analysis is concerned with assessing the optimal approach when vasa previa has been diagnosed by ultrasonography and does not address what screening approach (ie, incidental compared with risk-based compared with universal) should be used in the quest to detect vasa previa.
The infrequency of patients with vasa previa likely precludes the undertaking of a prospective trial to address the proper timing of delivery. Decision analysis may be the only tool by which this question can be systematically explored. The values in the model were ascertained through a thorough search of the literature and include utilities dealing with the explicit pediatric health states of concern.23 Moreover, the elucidation of an optimal strategy for these patients is of particular relevance in the present day, given that in vitro fertilization, which is a major risk factor for vasa previa, is becoming increasingly common.5,25–26
This decision analysis demonstrates two particularly striking findings. The first is that in this population under no clinical circumstances is there demonstrable benefit to be gained by expectant management beyond 37 weeks of gestation, and in most cases, delivery at 34 or 35 weeks of gestation is preferred. The second point is that using amniocentesis to determine fetal lung maturity is never a more effective strategy than proceeding with delivery without the amniocentesis. This finding is likely the result of a combination of reasons: the relatively high frequency of falsely immature results at these gestational ages, the hazard of the rare but catastrophic event of perinatal mortality with expectant management outweighing the short-term consequences of respiratory distress syndrome, and the inability of fetal lung maturity testing to predict long-term consequences of preterm birth. Although the literature detailing the probability of events related to emergent delivery of women with vasa previa is not definitive enough to allow the unequivocal establishment of a single ideal gestational age at which to deliver these high-risk patients, the preferred gestational age is almost certainly at or beyond 34 weeks of gestation but no later than 37 weeks.
With respect to the differences in quality-adjusted life-years among the different strategies, the top two rankings were stable under most conditions (Table 5). Although controversy exists regarding the number of quality-adjusted life-years representing a clinically significant gain, Richardson et al has argued that a gain in quality-adjusted life-years of 2 or more months should be considered an important gain.27 Applying the 2-month standard (2 months of a 12-month year, or 0.166) to the differences in the ranking strategies suggests that the marginal benefit, in the base-case, of delivering at 34 weeks of gestation compared with 35 weeks of gestation achieves a gain in quality-adjusted life-years of borderline clinical significance (just over 1 month), whereas delivering at 34 weeks of gestation compared with 36 weeks of gestation or any strategy resulting in a delivery beyond 36 weeks of gestation achieves a gain of more than 3 months.
This analysis has several limitations. First, this is not a clinical trial or study. As a result of the relatively infrequent occurrence of vasa previa, it is unlikely that a randomized clinical trial or even properly powered observational study will ever be performed to address the decision of delivery timing in this population. Moreover, it is unlikely such a study would include long-term outcomes such as cerebral palsy included in the present analysis. In this setting, use of a decision analysis is a practical alternative to help inform our decision-making. Another limitation, inherent to all decision analyses, is the choice and precision of the estimates included in the model. Due to the relative paucity of data in the literature detailing the precise sequence of events and outcomes in women with prenatally diagnosed vasa previa, the frequency of preterm labor and premature rupture of membranes were derived from national statistics, with the data extrapolated to women with prenatal diagnoses of vasa previa. Nevertheless, because of the uncertainty of the base-case estimates, extensive sensitivity analyses were performed, and although some may disagree with the baseline values chosen, most would agree that the range of values over which the sensitivity analyses were conducted were likely to contain all reasonable estimates.
As demonstrated by the decision tree model, the preferred strategy for delivery timing in patients with ultrasonographically diagnosed vasa previa in the base-case as well as under a variety of circumstances is delivery at 34 weeks of gestation. This takes into account both long- and short-term outcomes for the child. This does not imply that all women with this prenatal diagnosis should be delivered at this gestational age. As is clear from the model, there are circumstances (eg, when the probability of premature rupture of membranes or the risk of perinatal mortality in the setting of premature rupture of membranes is low or when the positive predictive value of ultrasonography for vasa previa is low) that would lead one to advocate for awaiting planned delivery up to as late as 37 weeks of gestation. Unfortunately, no validated predictive models exist that allow physicians to confidently know which patients with vasa previa are most likely to experience premature rupture of membranes or have adverse outcomes related to premature rupture of membranes. Ultimately, clinical judgement and shared decision making are crucial. This decision analysis suggests that delivery at 34 to 35 weeks gestation may balance the risk of perinatal death with the risks of infant mortality, RDS, mental retardation, and cerebral palsy related to prematurity, and supports the concept that for any given gestational age at which delivery is planned for women with vasa previa, amniocentesis for FLM does not improve outcome.
1. Robert JA, Sepulveda W. Fetal exsanguination from ruptured vasa previa: still a catastrophic event in modern obstetrics. J Obstet Gynaecol 2003;23:574.
2. Oyelese KO, Turner M, Lees C, Campbell S. Vasa previa: an avoidable obstetric tragedy. Obstet Gynecol Surv 1999;54:138–45.
3. Catanzarite V, Maida C, Thomas W, Mendoza A, Stanco L, Piacquadio KM. Prenatal sonographic diagnosis of vasa previa: ultrasound findings and obstetric outcome in ten cases. Ultrasound Obstet Gynecol 2001;18:109–15.
4. Lee W, Lee VL, Kirk JS, Sloan CT, Smith RS, Comstock CH. Vasa previa: prenatal diagnosis, natural evolution, and clinical outcome. Obstet Gynecol 2000;95:572–6.
5. Oyelese Y, Catanzarite V, Prefumo F, Lashley S, Schachter M, Tovbin Y, et al. Vasa previa: the impact of prenatal diagnosis on outcomes. Obstet Gynecol 2004;103:937–42.
6. Gagnon R, Morin L, Bly S, Butt K, Cargill YM, Denis N, et al; Maternal Fetal Medicine Committee. Guidelines for the management of vasa previa. J Obstet Gynecol Can 2009;31:748–60.
7. Oyelese Y, Smulian JC. Placenta previa, placenta accreta, and vasa previa. Obstet Gynecol 2006;107:927–41.
8. Smorgick N, Tovbin Y, Ushakov F, Vaknin Z, Barzilay B, Herman A, et al. Is neonatal risk from vasa previa preventable? The 20-year experience from a single medical center. J Clin Ultrasound 2010;38:118–22.
9. Hamilton BE, Martin JA, Ventura SJ. Births: preliminary data for 2008. Natl Vital Stat Rep 2010;58(16):1–17.
10. Savitz DA, Blackmore CA, Thorp JM. Epidemiologic characteristics of preterm delivery: etiologic heterogeneity. Am J Obstet Gynecol 1991;164:467–71.
11. Davidoff MJ, Dias T, Damus K, Russell R, Bettegowda VR, Dolan S, et al. Changes in the gestational age distribution among U.S. singleton births: impact on rates of late preterm birth, 1992 to 2002. Semin Perinatol 2006;30:8–15.
12. Premature rupture of membranes. Clinical management guidelines for obstetrician-gynecologists. ACOG Practice Bulletin No. 80. American College of Obstetricians and Gynecologists. Obstet Gynecol 2007;109:1007–19.
13. Robertson PA, Sniderman SH, Laros RK Jr, Cowan R, Heilbron D, Goldenberg RL, et al. Neonatal morbidity according to gestational age and birth weight from five tertiary care centers in the United States, 1983 through 1986. Am J Obstet Gynecol 1992;166:1629–41; discussion 1641–5.
14. Roberts D, Dalziel S. Antenatal corticosteroids for accelerating fetal lung maturation for women at risk of preterm birth. The Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD004454. DOI: 10.1002/14651858.CD004454.pub2.
15. Robinson BK, Grobman WA. Effectiveness of timing strategies for delivery of patients with placenta previa and accreta. Obstet Gynecol 2010;116:835–42.
16. Carp HJ, Mashiach S, Serr DM. Vasa previa: a major complication and its management. Obstet Gynecol 1979;53:273–5.
17. Dougall A, Baird CH. Vasa previa—report of three cases and review of literature. Br J Obstet Gynecol 1987;94:712–5.
18. Jaovisidha A, Herabutya Y. Vasa previa in Ramathibodi Hospital: a 10 years review. J Med Assoc Thai 1998;81:998–1002.
19. Fung TY, Lau TK. Poor perinatal outcome associated with vasa previa: is it preventable? A report of three cases and review of the literature. Ultrasound Obstet Gynecol 1998;12:430–3.
20. Kouyoumdjian A. Velamentous insertion of the umbilical cord. Obstet Gynecol 1980;56:737–42.
21. Pent D. Vasa previa. Am J Obstet Gynecol 1979;134:151–5.
22. Jantarasaengaram S, Suthipintawong C, Kanchanawat S, Thanagumtorn K. Ruptured vasa previa in velamentous cord insertion placenta. J Perinatol 2007;27:457–9.
23. Carroll AE, Downs SM. Improving decision analyses: parent preferences (utility values) for pediatric health outcomes. J Pediatr 2009;155:21–5, 25.e1–5.
24. Tomashek KM, Shapiro-Mendoza CK, Davidoff MJ, Petrini JR. Differences in mortality between late-preterm and term singleton infants in the United States, 1995–2002. J Pediatr 2007;151:450–6, 456.e1.
25. Schachter M, Tovbin Y, Arieli S, Friedler S, Ron-El R, Sherman D. In vitro fertilization is a risk factor for vasa previa. Fertil Steril 2002;78:642–3.
26. Oyelese Y, Spong C, Fernandez MA, Mclaren RA. Second trimester low-lying placenta and in-vitro fertilization? Exclude vasa previa. J Matern Fetal Med 2000;9:370–2.
27. Richardson WS, Detsky AS. Users' guides to the medical literature. VII. How to use a clinical decision analysis. B. What are the results and will they help me in caring for my patients? Evidence Based Medicine Working Group. JAMA 1995;273:1610–3.
28. Bastek JA, Sammel MD, Pare E, Srinivas SK, Posencheg MA, Elovitz MA. Adverse neonatal outcomes: examining the risks between preterm, late preterm, and term infants. Am J Obstet Gynecol 2008;199:367.e1–8.
29. Gilbert WM, Nesbitt TS, Danielsen B. The cost of prematurity: quantification by gestational age and birth weight. Obstet Gynecol 2003;102:488–92.
30. Stotland NE, Lipschitz LS, Caughey A. Delivery strategies for women with a previous classic cesarean delivery: a decision analysis. Am J Obstet Gynecol 2002;187:1203–8.
31. Herbert WN, Chapman JF. Clinical and economic considerations associated with testing for fetal lung maturity. Am J Obstet Gynecol 1986;155:820–3.
32. Steinfeld JD, Samuels P, Bulley MA, et al. The utility of the TDx test in the assessment of fetal lung maturity. Obstet Gynecol 1992;79:460–4.
33. Parvin CA, Kaplan LA, Chapman JF, McManamon TG, Gronowski AM. Predicting respiratory distress syndrome using gestational age and fetal lung maturity by fluorescent polarization. Am J Obstet Gynecol 2005;192:199–207.
34. Bildirici I, Moga CN, Gronowski AM, Sadovsky Y. The mean weekly increment of amniotic fluid TDx-FLM II ratio is constant during the latter part of pregnancy. Am J Obstet Gynecol 2005;193:1685–90.
35. Nomiyama M, Toyota Y, Kawano H. Antenatal diagnosis of velamentous umbilical cord insertion and vasa previa with color Doppler imaging. Ultrasound Obstet Gynecol 1998;12:426–9.
36. Petrini JR, Dias T, McCormick MC, Massolo ML, Green NS, Escobar GJ. Increased risk of adverse neurological development for late preterm infants. J Pediatr 2009;154:169–76.
37. Donovan EF, Besl J, Paulson J, Rose B, Iams J; Ohio Perinatal Quality Collaborative. Infant death among Ohio resident infants born at 32 to 41 weeks of gestation. Am J Obstet Gynecol 2010;203:58.e1–5.