Vasa previa is a rare obstetric condition with an estimated incidence of 6–11 per 10,000 pregnancies.1–4 It can result in stillbirth or neonatal death and has been associated with a high perinatal mortality rate of more than 56% if not detected prenatally and appropriate care instituted.1–4 Early diagnosis and planned cesarean delivery have been found to mitigate the risk of perinatal death.1,5 However, most published research on vasa previa consists of small case series or multicenter studies, limiting generalizability of the findings of the studies.1,2,6,7
Vasa previa is a condition in which fetal vessels traverse the membranes over the cervix, under the fetal presenting part, unprotected by the placenta or the umbilical cord.8–11 Risk factors for vasa previa include velamentous cord insertion, placental accessory lobes, second-trimester low-lying placenta, multiple pregnancy, and in vitro fertilization.8,9,12 These risk factors are becoming common in high-income countries such as Australia, which are characterized by an aging reproductive population and high uptake of fertility treatment.13–15 The current professional college statements and guidelines on vasa previa are mainly based on consensus with the Royal Australian and New Zealand College of Obstetricians and Gynaecologists’ statement concluding that “perinatal outcomes improve significantly when prenatal diagnosis enables planned management that includes elective cesarean delivery by 35 weeks gestation before the onset of labor.”16,17
The aim of our study is to estimate the incidence, risk factors, timing of diagnosis, clinical practice, and perinatal outcomes of women with vasa previa in Australia.
MATERIALS AND METHODS
A national, population-based prospective cohort study was undertaken using the Australasian Maternity Outcomes Surveillance System. The Australasian Maternity Outcomes Surveillance System is a surveillance and research system that studies rare conditions during pregnancy and the puerperium.18 Almost all (97% [267/275]) eligible Australian maternity units (greater than 50 births per year) participated covering an estimated 96% of women giving birth in a hospital in Australia. Data collection occurred between May 1, 2013, and April 30, 2014, and included any woman who was diagnosed with vasa previa and gave birth in Australia during this period. Box 1 shows the study inclusion criteria. The main outcome measures included stillbirth, neonatal death, cesarean delivery, and preterm birth.
Web-based data collection forms were completed by Australasian Maternity Outcomes Surveillance System data collectors based in each hospital for all women who satisfied the study inclusion criteria. Before recruitment and data collection, the vasa previa study protocol was publicized through the Australasian Maternity Outcomes Surveillance System newsletter, some in-hospital presentations, and to Australasian Maternity Outcomes Surveillance System data collectors through monthly emails. There was no site-specific training because the data collection methods were unchanged from previous Australasian Maternity Outcomes Surveillance System studies. Australasian Maternity Outcomes Surveillance System telephone and email support was provided centrally for data collectors for any questions arising during the study with site visits as requested. Participants were identified through multiple sources: Australasian Maternity Outcomes Surveillance System surveillance, review of routine data collection within the hospital, audit committees, clinician notification, and request to clinicians of potential patients. Data were collected on baseline vasa previa diagnosis, previous obstetric history, current pregnancy factors, management of vasa previa, and obstetric, maternal, and perinatal outcomes.
Figure 1 details surveillance and confirmed cases of vasa previa with 99 cases notified to the Australasian Maternity Outcomes Surveillance System, of which 15 (15.2%) were outside the study period, 7 (7.1%) were duplicate notifications, and 14 (14.1%) did not meet the placental confirmation per the case definition. Of the latter 14 women, the methods of birth were vaginal birth for 1 and cesarean delivery for 13 women. Eleven of the 13 women who gave birth by cesarean delivery had vasa previa listed as the indication for cesarean delivery. The indications for cesarean delivery for the remaining two women were antepartum hemorrhage and placenta previa.
Incidence estimates with 95% CIs were calculated. The denominator of 294,045 women giving birth over the study period for incidence estimates was based on pro rata 2013–2014 birth data from Australia with an approximate coverage of 96% of all women giving birth in maternity units in Australia.14,19 χ2, Fisher exact test, Fisher-Freeman-Halton test, t tests or Mann–Whitney U test were used to investigate the difference in perinatal outcomes between women with vasa previa diagnosed prenatally and those not diagnosed prenatally. For women diagnosed prenatally, we also examined the effects of antepartum hospitalization and timings of cesarean delivery on maternal and perinatal outcomes. Data were analyzed using SPSS 22.0.
Ethics approval for this study (Ref: HREC/09/CIPHS/21) was granted on March 19, 2013, by the New South Wales Population and Health Services Research Ethics Committee and multiple human research ethics committees across Australia.20
Sixty-three women gave birth in Australia between May 1, 2013, and April 30, 2014, with a confirmed diagnosis of vasa previa. The estimated incidence of vasa previa was 2.1 per 10,000 women giving birth (95% CI 1.7–2.7/10,000), or 1 in 4,667 women giving birth. Of the 63 women with vasa previa, 58 (92%) were diagnosed prenatally and 5 (8%) were diagnosed intrapartum. Of the 58 women diagnosed prenatally, 51 (88%) were diagnosed by prenatal ultrasonogram and were confirmed on placental examination (Fig. 1). Overall, vasa previa diagnosis was confirmed by placental examination for 59 (94%) women with the placenta sent for pathologic examination for 33 women (Fig. 1).
Table 1 shows sociodemographic and pregnancy factors. The average age of women was 31.7 years and ranged from 20 to 41 years. There were one or more risk factors for vasa previa present in 55 (95%) of women diagnosed prenatally with velamentous cord insertion (62%) and low-lying placenta (60%) the most prevalent (Table 1). Use of assisted reproductive technology was reported for 11 women, all of whom had at least one placental abnormality(s) including low-lying placenta (n=6 [55%]), succenturiate placenta (n=3 [27%]), and velamentous cord insertion (n=7 [64%]). Almost half (n=29 [46%]) of women had a history of one or more prior uterine surgeries, including surgical termination of pregnancy (17%), dilatation and curettage (13%), cesarean delivery (13%), evacuation of retained products of conception (6%), manual removal of the placenta (5%), and dilation and evacuation (2%). Sixty (95%) women had at least one ultrasonic examination during pregnancy (Table 1); 53% of those 60 women had at least one transvaginal or color or power Doppler ultrasonograms documented, 40% had only transabdominal ultrasonogram documented, and four women had an unknown type of ultrasonogram. Four of five women diagnosed intrapartum had at least two transabdominal ultrasonograms but no transvaginal or color or power Doppler ultrasonograms documented. This gave a false-negative rate of 7% (4/60 patients).
Table 2 shows the maternal and perinatal outcomes for women with vasa previa stratified by timing of diagnosis. Spontaneous rupture of membranes occurred in four women, two diagnosed prenatally at 28 and 33 weeks of gestation and two not diagnosed prenatally at 35 and 37 weeks of gestation. Of the 10 women who went into labor, three were before 34 weeks of gestation, two between 34 and 36 weeks of gestation, and five started laboring at term with three of the latter not diagnosed prenatally. All women with vasa previa diagnosed prenatally had cesarean deliveries between 26 and 38 weeks of gestation compared with four women not diagnosed prenatally who had cesarean deliveries between 35 and 41 weeks of gestation. For the five women with vasa previa not diagnosed prenatally, one gave birth vaginally at 37 weeks of gestation to an antepartum stillbirth; the remaining four women had cesarean deliveries categorized as urgent threat to the life or the health of a woman or fetus (Royal Australian and New Zealand College of Obstetricians and Gynaecologists category 1).21 By contrast, only four (7%) women diagnosed prenatally had a similarly urgent cesarean delivery (P<.001).
There were no perinatal deaths for women diagnosed prenatally with vasa previa. There were two perinatal deaths (one antepartum stillbirth and one neonatal death) for the five women not diagnosed prenatally with a case fatality rate of 40% (P<.01) (Table 2). The overall perinatal case fatality rate was 3.1% (95% CI 0.8–10.5). Two thirds (n=44 [68%]) of the 65 neonates were preterm and 19 (29%) were low birth weight. Two (3%) neonates were diagnosed with anemia and both required blood transfusions with one a neonatal death complicated by renal failure and seizures resulting from hypoxic–ischemic encephalopathy.
Forty (69%) women diagnosed with vasa previa prenatally were hospitalized during pregnancy with gestational ages at hospitalization ranging from 23 to 38 weeks (Table 3). Antepartum hospitalization was associated with significantly higher rates of preterm birth for women with a prenatal diagnosis of vasa previa compared with women who were not hospitalized (Table 3). Forty-two (67%) women received betamethasone prenatally for fetal lung enhancement (median 34 weeks of gestation, range 23–37 weeks of gestation). Among cases diagnosed prenatally, neonates born to mothers admitted to the hospital prenatally did not differ significantly on perinatal outcomes from those born to mothers not admitted prenatally in terms of survival rate, birth weight, rates of admission to neonatal intensive care units or special care units, and 5-minute Apgar score less than 7 (Table 3).
Planned cesarean delivery was scheduled for all women diagnosed with vasa previa prenatally; 5 (9%) women had a cesarean delivery scheduled before 34 weeks of gestation, 33 (57%) between 34 and 36 weeks of gestation, and 18 (31%) at term (37 weeks of gestation or greater). Table 4 shows the actual timing of cesarean delivery with 12 (21%) women having a cesarean delivery before 34 weeks of gestation, 29 (50%) between 34 and 36 weeks of gestation, and 17 (29%) at term. The indications for cesarean delivery before 34 weeks of gestation were hypertension (n=1), placental abruption and nonreassuring fetal heart rate trace (n=1), vasa previa, premature rupture of the membranes and suspected chorioamnionitis (n=1), vasa previa and fetal growth restriction (n=1), and eight with either antepartum hemorrhage or placenta previa. Of the three women who gave birth at 34 weeks of gestation, one had hemorrhage and two had vasa previa and a nonreassuring fetal heart rate trace. Apart from a higher rate of neonatal intensive care unit or special care unit admission, there was no significant difference in maternal and perinatal outcomes between women who had cesarean delivery between 34 and 36 weeks of gestation compared with those who had a cesarean delivery at term (Table 4).
This prospective, national population-based study of vasa previa found no perinatal deaths in pregnancies diagnosed with vasa previa prenatally and delivered by cesarean. In contrast, the perinatal case fatality rate for pregnancies not diagnosed prenatally was 40% (two of five). These results were consistent with findings from previous studies and provide higher level evidence to support the recommendations from Royal College of Obstetricians and Gynaecologists and Royal Australian and New Zealand College of Obstetricians and Gynaecologists regarding prenatal diagnosis and birth by scheduled cesarean delivery.16,17 Oyelese and colleagues1 reported that the neonatal survival rate was 97% in cases diagnosed prenatally compared with 44% in cases not diagnosed prenatally. A retrospective review of 20 years of cases in a single medical center showed an increased prenatal detection rate from 25% to 60% and a decreased perinatal mortality rate from 25% to 0% between the two 10-year periods.2
Our study found better perinatal outcomes for the 58 women diagnosed prenatally with vasa previa than previously reported,1,2 notwithstanding the very high rate of planned preterm birth. This may be related to the very high rate of prenatal diagnosis, associated intention for all women to have planned cesarean delivery coupled with the low rate of unplanned, early preterm (less than 35 weeks of gestation) cesarean delivery.
Although this national study provides evidence of the benefit of prenatal diagnosis of vasa previa, it does not address whether there should be universal screening for the condition or just targeted screening using transvaginal ultrasonography with color Doppler in women with risk factors. Women in Australia routinely have an anatomy scan at 18–20 weeks of gestation. Vasa previa can be excluded in women who have risk factors at the anatomy scan by performing a transvaginal ultrasonogram with color Doppler examination of the area adjacent to the internal os of the cervix and pulsed Doppler examination to confirm fetal origin of identified vessels. However, women with low-lying placentas at the anatomy ultrasonogram should have a follow-up transvaginal color Doppler ultrasonogram in the third trimester to confirm resolution of the placenta previa and to exclude vasa previa. Universal screening for vasa previa is not offered in Australia; nevertheless, in our study, 92% of women with vasa previa were detected prenatally and 95% of women diagnosed prenatally had one or more risk factors. Our findings support further examination of the potential benefits of targeted screening for vasa previa in women with risk factors. The high rate of prenatal diagnosis may reflect increased awareness of vasa previa in Australia.
Hospitalization in the third trimester has been recommended for women with vasa previa because it allows access to urgent cesarean delivery in the case of unexpected rupture of membranes.1,16,17,22 Not surprisingly, our results showed a significant association between antepartum hospitalization for vasa previa and preterm birth and a proportionately lower rate of urgent cesarean delivery for fetal compromise compared with women who were not hospitalized prenatally. This may reflect an increased tendency for intervention for those admitted with an indication of vasa previa, which may or may not be related to severity of the condition or increased prenatal inpatient fetal surveillance. Alternatively, women who were hospitalized prenatally may have had more complex pregnancies complicated by comorbidities unrelated to vasa previa and these factors may have led or contributed to the disproportionately higher rate of earlier deliveries.
Our study is not able to definitively provide evidence of superior perinatal outcomes in support of the 2016 recommended practice of “admitting women with prenatally diagnosed vasa previa to a hospital with appropriate neonatal facilities from around 30 weeks gestation until delivery and administration of corticosteroids for lung maturity.”17 Furthermore, our study is not designed to establish an optimal gestational age for cesarean delivery. A decision analysis by Robinson and colleagues23 suggested that birth at 34–35 weeks of gestation provides an optimal balance between stillbirth risk and the risk of adverse perinatal outcomes. It has become commonplace to recommend delivery for vasa previa at 34 weeks of gestation. However, there is an imperative to reduce the rate of late preterm deliveries because recent data indicate that late preterm birth carries increased risks of both short- and long-term morbidity.24–26 There is now a drive to reduce the rate of late preterm deliveries. In our study, 46 of 58 deliveries took place at 34 weeks of gestation or greater with 17 deliveries at 36 weeks of gestation and 17 at 37 weeks of gestation or greater. It may be that previous recommendations are recommending preterm delivery without enough data and exposing the neonates to considerable risks of prematurity. For prenatally diagnosed women without complications, who are hospitalized and are clinically stable 36 weeks of gestation, this maybe considered as a more optimal gestational age for delivery. Nonetheless, extreme vigilance is warranted. As our results show, of the four women with spontaneous rupture of membranes, three occurred before 36 weeks of gestation reflecting the clinical dilemma in balancing the risks of prematurity compared with expectant management without a robust evidence base.
The major strengths of this study were the prospective population-based design with active case finding and the use of a robust case definition that required placental examination. The latter was critical to ensure the validity of the diagnosis but likely resulted in a lower reported incidence of vasa previa with 14 cases excluded as a result of a lack of examination of the placenta for velamentous cord insertion or fetal vessels in membranes. Placental examination for velamentous vessels is not currently routine in the clinical setting in Australia.
Our study has some limitations. Although the case ascertainment is believed to be high as a result of active surveillance, it is not possible to determine the level of ascertainment achieved. Only ultrasound reports were reviewed and not the actual ultrasound images. Furthermore, because there is no standard universal cervical assessment by transvaginal examination and color Doppler, nor universal placental assessment, and near misses may be underreported, there is likely underascertainment of vasa previa.
In Australia, the high rate of prenatal diagnosis of vasa previa is associated with good perinatal outcomes. Outcomes in pregnancies not diagnosed prenatally are poor with higher rates of perinatal morbidity and the two perinatal deaths in the study.
1. 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.
2. 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.
3. Rebarber A, Dolin C, Fox NS, Klauser CK, Saltzman DH, Roman AS. Natural history of vasa previa across gestation using a screening protocol. J Ultrasound Med 2014;33:141–7.
4. Ruiter L, Kok N, Limpens J, Derks JB, de Graaf IM, Mol BWJ, et al. Incidence of and risk indicators for vasa praevia: a systematic review. BJOG 2016;123:1278–87.
5. Baulies S, Maiz N, Munoz A, Torrents M, Echevarria M, Serra B. Prenatal ultrasound diagnosis of vasa praevia and analysis of risk factors. Prenat Diagn 2007;27:595–9.
6. Swank ML, Garite TJ, Maurel K, Das A, Perlow JH, Combs CA, et al. Vasa previa: diagnosis and management. Am J Obstet Gynecol 2016;215:223.e1–6.
7. Catanzarite V, Cousins L, Daneshmand S, Schwendemann W, Casele H, Adamczak J, et al. Prenatally diagnosed vasa previa: a single-institution series of 96 cases. Obstet Gynecol 2016;128:1153–1161.
8. D'Antonio F, Bhide A. Ultrasound in placental disorders. Best Pract Res Clin Obstet Gynaecol 2014;28:429–42.
9. Gagnon R, Morin L, Bly S, Butt K, Cargil YM, Denis N, et al. SOGC Clinical Practice Guideline: guidelines for the management of vasa previa. Int J Gynaecol Obstet 2010;108:85–9.
10. Nguyen D, Nguyen C, Yacobozzi M, Bsat F, Rakita D. Imaging of the placenta with pathologic correlation. Semin Ultrasound CT MR 2012;33:65–77.
11. Baumfeld Y, Gutvirtz G, Shoham I, Sheiner E. Fetal heart rate patterns of pregnancies with vasa previa and velamentous cord insertion. Arch Gynecol Obstet 2016;293:361–7.
12. Society of Maternal-Fetal (SMFM) Publications Committee, Sinkey RG, Odibo AO, Dashe JS. #37: Diagnosis and management of vasa previa. Am J Obstet Gynecol 2015;213:615–9.
14. Australian Institute of Health and Welfare. Australia's mothers and babies 2014—in brief. Perinatal statistics series no. 32. Cat no. Per 87. Canberra (Australia): Australian Institute of Health and Welfare; 2016.
15. Harris K, Fitzgerald O, Paul RC, Macaldowie A, Lee E, Chambers GM. Assisted reproductive technology in Australia and New Zealand 2014. Sydney (Australia): National Perinatal Epidemiology and Statistics Unit, the University of New South Wales; 2016.
16. Royal College of Obstetricians & Gynaecologists. Green-top guideline no. 27. Placenta praevia, placenta praevia accreta and vasa praevia: diagnosis and management. London (UK): Royal College of Obstetricians & Gynaecologists; 2011.
17. The Royal Australian and New Zealand College of Obstetricians and Gynaecologists. College statement c-obs 47: vasa praevia, review July 2016. Melbourne (Australia): The Royal Australian and New Zealand College of Obstetricians and Gynaecologists; 2016.
18. Australasian Maternity Outcomes Surveillance System (AMOSS). Annual report 2010–2011. Sydney (Australia): Australasian Maternity Outcomes Surveillance System (AMOSS); 2012.
19. Australian Institute of Health and Welfare. Australia's mothers and babies 2013—in brief. Perinatal statistics series no. 31. Cat no. Per 72. Canberra (Australia): Australian Institute of Health and Welfare; 2015.
20. Vaughan G, Pollock W, Peek MJ, Knight M, Ellwood D, Homer CS, et al. Ethical issues: the multi-centre low-risk ethics/governance review process and AMOSS. Aust N Z J Obstet Gynaecol 2012;52:195–203.
21. The Royal Australian and New Zealand College of Obstetricians and Gynaecologists. Categorisation of urgency for caesarean section. Sydney (Australia): The Royal Australian and New Zealand College of Obstetricians and Gynaecologists; 2015.
22. Johnston R, Shrivastava VK, Chmait RH. Term vaginal delivery following fetoscopic laser photocoagulation of type ii vasa previa. Fetal Diagn Ther 2014;35:62–4.
23. Robinson BK, Grobman WA. Effectiveness of timing strategies for delivery of individuals with vasa previa. Obstet Gynecol 2011;117:542–9.
24. Cheong JL, Doyle LW, Burnett AC, Lee KJ, Walsh JM, Potter CR, et al. Association between moderate and late preterm birth and neurodevelopment and social-emotional development at age 2 years. JAMA Pediatr 2017;171:e164805.
25. Rogers CE, Lenze SN, Luby JL. Late preterm birth, maternal depression, and risk of preschool psychiatric disorders. J Am Acad Child Adolesc Psychiatry 2013;52:309–18.
26. Reichman NE, Teitler JO, Moullin S, Ostfeld BM, Hegyi T. Late-preterm birth and neonatal morbidities: population-level and within-family estimates. Ann Epidemiol 2015;25:126–32.