Inpatient vs outpatient management of pregnancies with vasa previa: A historical cohort study

Abstract

Introduction

Vasa previa, a condition where unprotected fetal blood vessels lie in proximity to the internal cervical opening, is a potentially lethal obstetric complication. The precarious situation of these vessels increases the risk of fetal hemorrhage with spontaneous or artificial rupture of membranes, frequently causing fetal/neonatal demise or severe morbidity. As a result, in many centers, inpatient management forms the mainstay when vasa previa is diagnosed antenatally. This study aimed to determine whether a subpopulation of pregnancies diagnosed antenatally with vasa previa could be safely managed as outpatients.

Material and methods

We reviewed all cases of vasa previa in singleton pregnancies, with no fetal anomalies, diagnosed at Mount Sinai Hospital, Toronto, from January 2008 to December 2017. Cases were categorized into three arms for analysis: outpatients (OP), asymptomatic hospitalized (ASH) and symptomatic hospitalized (SH). The SH arm included patients admitted with any antepartum bleeding or suspicious fetal non‐stress test. Those that presented with symptomatic uterine activity/threatened preterm labor and delivered within 7 days of diagnosis were excluded from the study. Records were analyzed for details on hospitalization, antenatal corticosteroid administration, cervical length measurements, and fetal/neonatal mortality and morbidity.

Results

Of the 84 antenatally‐diagnosed cases of vasa previa, 47 fulfilled eligibility criteria. A total of 15 cases were managed as OP, 22 as ASH and 10 as SH. Unplanned cesareans were highest in the SH arm (40% vs. 0% ASH vs. 13.3% OP). Those in the SH arm delivered earliest (median 33.8 weeks, interquartile range (IQR) 33.2–34.3 weeks). Of the asymptomatic patients, those in the ASH arm delivered earlier than those in the OP arm (35.3 [34.6–36.2] weeks vs. 36.7 [35.6–37.2] weeks, p = 0.037). There were no cases of fetal/neonatal death, anemia or severe neonatal morbidity and no significant differences between groups based on cervical length or antenatal corticosteroid administration.

Conclusions

Our study suggests that asymptomatic women with an antenatal diagnosis of vasa previa, singleton pregnancies, and at low risk for preterm birth may safely managed as outpatients, as long as they are able to access hospital promptly in the event of antepartum bleeding or early labor.

Asymptomatic patients with an antenatal diagnosis of vasa previa, singleton pregnancies, and at low risk for preterm birth may safely be managed as outpatients, as long as they are able to access hospital promptly in the event of antepartum bleeding or early labor.

Abbreviations

ACS

antenatal corticosteroids

ASH

asymptomatic hospitalized

CL

cervical length

CPAP

continuous positive airway pressure

NST

fetal non‐stress test

IQR

interquartile range

OP

outpatients

RDS

respiratory distress syndrome

SOGC

Society of Obstetricians and Gynecologists of Canada

SD

standard deviation

SH

symptomatic hospitalized

Key message.

Asymptomatic patients with vasa previa at low risk of preterm labor may qualify for outpatient management.

1. INTRODUCTION

Vasa previa is a potentially lethal obstetric complication that affects approximately 0.46 (95% confidence intervals [CI]: 0.33–0.59) per 1000 births. ¹ In this condition, unprotected chorionic vessels lie in the vicinity of the internal cervical os, which increases the risk of fetal hemorrhage with rupture of membranes. ² , ³ , ⁴ , ⁵ If undetected, many of these cases will lead to severe neonatal compromise and rapid fetal exsanguination. ⁶ Prenatal diagnosis of this condition is therefore paramount and potentially life‐saving. While diagnostic screening protocols continue to be studied, ⁷ , ⁸ , ⁹ , ¹⁰ management of vasa previa has not been studied to the same extent and have thus far highlighted the need for planned cesarean births. ³ , ⁶ There is currently no consensus on the optimal management strategies between antenatal diagnosis and delivery. Prolonged antenatal admission (from approximately 30 weeks of gestation), prophylactic administration of antenatal corticosteroids (ACS) and planned late preterm cesarean births have been proposed by several national organizations, including the Society for Maternal‐Fetal Medicine. ¹¹ , ¹² However, some organizations, such as the Society of Obstetricians and Gynecologists of Canada (SOGC) have suggested a role for outpatient management in a select group of patients. ¹³

We conducted this historical cohort study to describe how patients with antenatally‐diagnosed vasa previa were managed in a tertiary care referral center, especially related to hospital admission, ACS administration and timing of delivery, with the aim of determining whether a subpopulation of these pregnancies could be safely managed as outpatients, as this would have considerable implications to families and healthcare systems.

2. MATERIAL AND METHODS

This was a historical cohort study conducted at Mount Sinai Hospital, a tertiary referral center in Toronto, Canada. The electronic medical record Powerchart and prenatal ultrasound software Astraia were searched for cases of vasa previa from January 2008 to December 2017. Prevalence was calculated based on cases diagnosed and delivered at Mount Sinai Hospital over the total number of deliveries during the study period (n = 67 426). There is currently no protocol for routine screening for vasa previa in Canada. All identified cases required confirmation of the vasa previa by transvaginal ultrasound with color Doppler. Only singleton, non‐anomalous gestations, with complete records regarding admission, management and delivery were included. Patients that presented in preterm labor and delivered within 7 days of diagnosis were excluded from categorical analysis as the diagnosis‐to‐delivery time was considered too short to exert significant differences in perinatal outcomes from the exposures studied. Cases that were not diagnosed antenatally were excluded from this study.

Exposure of interest was the type of antenatal and peripartum management received which was based on the status of hospital admission, as defined by days spent in inpatient care until the date of delivery. Cases were categorized into one of the following groups: outpatients (OP), asymptomatic hospitalized (ASH) and symptomatic hospitalized (SH). After discussing the pros and cons of inpatient vs outpatient management in the light of current recommendations, a group of patients opted for inpatient management, even in the absence of any symptoms that would otherwise warrant admission. They constituted the ASH group wherein the reason for admission was noted as maternal anxiety/request. The SH group included patients admitted for any degree of antepartum hemorrhage or suspicious fetal non‐stress test (NST) from any indication. Although vasa previa does not often result in fetal heart rate abnormalities in the antepartum period, theoretically, compression of fetal vessels especially in the presence of velamentous cord insertions, could result in suspicious findings on NST, which is therefore the mainstay of inpatient management. A suspicious NST in the context of vasa previa would warrant further surveillance, and these pregnancies were therefore included under the SH group. Those that presented with symptomatic uterine activity/threatened preterm labor and delivered within 7 days of diagnosis, were excluded from the study. Prior preterm birth, asymptomatic uterine activity picked up on routine monitoring, and cervical length at the index examination were not considered criteria for admission in the absence of bleeding or suspicious findings on NST. Outpatients underwent transvaginal ultrasound and biometry every 2–4 weeks. The hospitalized groups underwent daily NST, and weekly transvaginal ultrasound and biophysical profiles and 2‐weekly biometry in keeping with the unit protocol. Transvaginal ultrasound was performed to assess resolution/persistence of vasa previa and cervical shortening, if any. Cervical length (CL) shortening was calculated by subtracting the last known CL before delivery from the CL at diagnosis and then dividing by the number of weeks between these two measurements. Cesareans were planned between 35 and 37 weeks' gestation or at the discretion of the obstetrician and patient. An emergency (unplanned) cesarean was documented as an adverse outcome.

When analyzing ACS therapy for fetal lung maturation, cases were categorized by the administration of ACS, the timing of delivery, and whether delivery occurred within 48 h to 7 days of administration of ACS or >7 days post‐ACS administration. These data were also analyzed based on the admission status of the mother, as previously described.

Pathology reports were reviewed to assess cord insertion and placental lobes. Secondary outcomes included neonatal adverse events and mortality. Low neonatal hemoglobin at birth was defined as <125 g/L. ¹⁴ Cord gases at birth were considered abnormal if the umbilical cord pH was <7.1, or a base deficit >12 mmol/L. Neonatal morbidity data included the need for continuous positive airway pressure (CPAP) support or blood transfusion, and sequelae of prematurity (hypoxic ischemic encephalopathy, bronchopulmonary dysplasia, intraventricular hemorrhage and respiratory distress syndrome [RDS]). RDS was defined categorically as neonates requiring surfactant therapy in addition to oxygen and ventilatory support.

2.1. Statistical analyses

All results are expressed as either numbers and proportions, median with interquartile range (IQR) or mean with standard deviation (SD) as appropriate. Statistical analyses were performed using the GraphPad Prism 6 software. Freeman–Halton extension of the Fisher's exact test, unpaired t‐tests and one‐way ANOVAs were used when appropriate. A p‐value <0.05 was considered statistically significant.

2.2. Ethics statement

This study was approved by the Mount Sinai Hospital (MSH) Research Ethics Board [REB# 16‐0154‐C] on July 8, 2016.

3. RESULTS

As indicated in Figure 1, a total of 385 charts were reviewed over the study period, out of which 84 cases of vasa previa were confirmed. The calculated prevalence was 0.83 cases per 1000 births (95% CI: 0.61–1.05). The following were excluded: pregnancies where clinical symptoms required consideration of imminent delivery as opposed to expectant management at the time of initial diagnosis (n = 15), multiple gestations or congenital anomalies (n = 12), lack of information regarding delivery and management (n = 6 of 28 cases that were not delivered at MSH) or insufficient information on admission status (n = 4). Together, this left 47 cases available for descriptive (Table 1) and categorical (Table 2) analyses. Of these cases, 15 were managed as outpatients (OP arm), 22 cases were admitted based on maternal anxiety or request (ASH arm), and 10 cases were admitted for episodes of antepartum bleeding or a suspicious fetal NST (SH arm). Twelve cases had marginal cord insertion, but in all cases, umbilical vessels were unprotected by Wharton's jelly, overlying the internal cervical os, and therefore the criteria for antenatal diagnosis of vasa previa were fulfilled. As indicated in Table 2, the diagnosis of vasa previa was made between 25.1 weeks (IQR 22.4–27.0) in the SH arm to 27.9 weeks (IQR 25.1–31.3) in the ASH arm. Although patients in the SH group were admitted for a median of 27 days longer to the date of delivery compared to those in the ASH group, the wide confidence limits meant that there were no statistical differences between groups (15.5 days [IQR 10.3–39.0] compared to 42.5 days [IQR 22.3–57.8], p = 0.14). Similarly, the OP group lived at a median distance of 21.7 km (IQR 8.6–38.0) from the institution, while the ASH group lived at a median distance of 35.9 km (IQR 34.6–36.2) (p = 0.30). No relationship (R² = 0.0004) was found between the distance to the hospital and the number of days admitted between the three arms (Figure S1).

FIGURE 1.

Flow chart of eligibility of vasa previa cases for descriptive and categorical analysis.

TABLE 1.

Maternal/obstetric characteristics and risk factors (n = 47).

Characteristic/risk factor	Outpatient (n = 15)	Asymptomatic hospitalized (n = 22)	Symptomatic hospitalized (n = 10)	p‐value
Maternal age	34 (32.5–35.8) [21–46]	35 (30.8–37.3) [27–44]	35 (34.0–36.0) [28–48]	0.950
Gravidity	2 (2–3) [1 – 5]	2.5 (1–4) [1 – 9]	3 (2–3) [1 – 4]	0.510
Parity	1 (0–1) [0–2]	0 (0–1) [0–5]	1 (0–1) [0–2]	0.771
IVF conception (n = 29)	2 (13.3%)	4 (18.1%)	1 (10%)
Cord insertion (n = 30)				0.999
VCI	5 (33.3%)	9 (40.9%)	6 (60%)
MCI	2 (13.3%)	6 (27.2%)	4 (40%)
Placental lobes (n = 13)				1.000
Bilobed	2 (13.3%)	3 (13.6%)	3 (30%)
Succenturiate	2 (13.3%)	2 (9.0%)	1 (10%)
Low lying placenta	3 (20.0%)	1 (4.5%)	1 (10%)

Note: Age, gravidity and parity are presented as median (IQR) [range]. IVF (%) is the number of IVF pregnancies out of all methods of artificial reproductive technology per arm.

Abbreviations: IVF, in vitro fertilization; MCI, marginal cord insertion; VCI, velamentous cord insertion.

TABLE 2.

Maternal/obstetric and neonatal outcomes (n = 47).

Outcomes	Outpatient	Asymptomatic hospitalized	Symptomatic hospitalized	p‐value
Number of pregnancies	15	22	10
GA at diagnosis	26.3 (23.1–29.9)	27.9 (25.1–31.3)	25.1 (22.4–27.0)	0.221
Days admitted up to delivery	0	15.5 (10.3–39.0)	42.5 (22.3–57.8)	0.138
Distance to the hospital (km)	21.7 (8.6–38.0)	35.9 (10.3–73.1)	8.0 (7.1–46.5)	0.297
Cervical length at diagnosis (mm) (n = 44)	43.0 (35.0–46.0)	39.0 (32.8–48.3)	37.0 (36.0–41.0)	0.508
Last measured cervical length (mm) (n = 47)	37.0 (34.0–44.0)	37.5 (32.0–40.0)	38.5 (33.3–43.3)	0.869
Change in cervical length (mm) (n = 35)	0.13 (−0.48–1.09)	0.52 (−0.28–1.93)	−0.28 (−1.0–1.75)	0.411
Corticosteroid use				0.244
Admin and delivered 48 h–7 days	0	0	1 (10%)
Admin and delivered >7 days	5 (33.3%)	12 (54.5%)	7 (70%)
Withheld and delivered >34 weeks GA	10 (66.7%)	9 (41%)	1 (10%)
Withheld and delivered <33.9 weeks GA	0	1 (4.5%)	1 (10%)
GA at delivery	36.7 (35.6–37.2)	35.3 (34.6–36.2)	33.8 (33.2–34.3)	0.0002
Mode of delivery				0.003
Elective cesarean delivery	13 (86.7%)	22 (100%)	6 (60%)
Emergency cesarean delivery	2 (13.3%)	0	4 (40%)
Neonatal adverse outcomes				0.557
5 min Apgar score <7 (n = 47)	0	0	2 (20%)
CPAP	3 (20%)	7 (31%)	5 (50%)

Note: Distance to the hospital, gestational ages at diagnosis and delivery and cervical length at diagnosis are presented as median (IQR). There were no cases of low Hb (n = 20), abnormal cord cases (n = 28), need for blood transfusion, hypoxic ischemic encephalopathy (HIE), bronchopulmonary dysplasia (BPD), intraventricular hemorrhage (IVH) or mortality. One case (6.25%) of respiratory distress syndrome (RDS) occurred in a patient managed as asymptomatic hospitalized.

Abbreviations: CPAP, continuous positive airway pressure; GA, gestational age.

As indicated in Table 2, there were no significant differences between groups based on baseline CL at diagnosis, the last CL measurement before delivery, or the rate of CL shortening between these two time points.

As indicated in Table 3 of the 47 cases, ACS was administered routinely (upon diagnosis or admission) in 25 cases. Of these, only one patient in the SH arm delivered within the therapeutic window (48 h and 7 days) of ACS administration, that is, 6 days following ACS administration (Figure 2). The vast majority (n = 24) delivered >7 days after ACS administration (median 39 days [IQR 25–55 days]) after ACS administration. Of these, four (12.5%) delivered <34 weeks, three in the SH arm, and one in the OP arm, while the other 20 delivered after 34 weeks. In seven instances, ACS was administered before 28 weeks' gestation (Table S1), the time interval between ACS administration and delivery ranging from 7.9 to 9.3 weeks. Reasons for ACS administration before 28 weeks included an episode of antepartum bleeding, or the discretion of the primary physician based on the diagnosis of vasa previa. In 22 cases (46.8%), a shared decision was made not to routinely administer ACS, so as to optimize its effect by time administration when delivery was imminent and avoid unnecessary adverse effects of ACS. Of these 20, delivered after 34 weeks, in retrospect, not requiring ACS administration, while two delivered under 34 weeks, and may have benefited from ACS administration. As shown in Table 2, when stratified by the three groups, ACS was not administered in 10 (66.7%) cases managed as OP, all of whom delivered >34 weeks' gestation (after which ACS are not administered in Canada), compared to nine (41%) cases managed as ASH, only one of which delivered under 34 weeks of gestation. Patients in the ASH arm were more likely to receive ACS and deliver more than 7 days post‐therapy when compared to patients in the OP arm (54.5% vs. 33.3%).

TABLE 3.

Corticosteroid use in patients with vasa previa (n = 47).

	ACS withheld (n = 22)	ACS administered delivered 48 h–7 days (n = 1)	ACS administered delivered >7 days (n = 24)
GA at treatment in weeks (median, IQR)	—	29.1	29.9 (26.3–31.5)
GA at delivery in weeks (median, IQR)	36.5 (35.3–37.0)	29.7	34.6 (34.1–35.5)
>34 weeks GA	20 (90.9%)	0	20 (83.3%)
<33.9 weeks GA	2 (9.1%)	1 (100%)	4 (16.6%)
Difference	—	6 days	39 (25–55) days

Note: Difference, number of days between treatment and delivery.

Abbreviations: ACS, antenatal corticosteroids; GA, gestational age; IQR, interquartile range.

FIGURE 2.

Corticosteroid therapy timelines. Gestational age at the time of antenatal corticosteroid administration and delivery (n = 25). Data are presented as medians and interquartile ranges (IQR).

Women managed as SH delivered at a median of 33.8 weeks, compared to women managed as ASH at 35.3 weeks, and women managed as OP at 36.7 weeks (p = 0.0002, Table 2). Women in the SH arm had the highest frequency of unplanned cesareans (40%), compared to women in the ASH (0%) and OP (13.3%) arms (p = 0.003, Table 2), as the persistence of symptoms (antepartum bleeding and abnormal NSTs in the context of vasa previa) are most likely to be the indications for unplanned cesareans. Within the OP arm, the 2/15 (13.3%) required unplanned cesareans due to antepartum bleeding and early labor, respectively. In the SH arm, 4/10 (40%) required unplanned cesareans due to antepartum bleeding (n = 2), spontaneous rupture of membranes (n = 1), and regular uterine contractions (n = 1). Despite there being no unplanned cesareans in the ASH arm, patients delivered a median 1.4 weeks earlier compared to those managed as OP (p = 0.037, Figure 3). Neonates born to mothers in the ASH were more likely to require CPAP compared to neonates born to mothers managed as OP (31% vs. 20%, Table 2).

FIGURE 3.

Neonatal gestational age at delivery. Gestational age at delivery of neonates born to mothers with vasa previa managed as outpatients (OP), asymptomatic inpatients (ASH) or symptomatic inpatients (SH) n = 47. Data are presented as medians with interquartile ranges (IQR). An unpaired t‐test was performed, and p < 0.05 was considered statistically significant.

There were no cases of fetal or neonatal mortality, anemia or any severe neonatal morbidity. Two neonates born by emergency cesarean at 33 weeks' gestation to mothers being managed as inpatients (SH arm) had 5‐min Apgar scores <7—both attributable to prematurity and not to vasa previa. The first neonate was cyanotic and apneic at 4 min of life with a 5‐min Apgar of 6. The second had a 5‐min Apgar of 6 due to decreased airway entry and subcostal indrawing. Both neonates had normal hemoglobin, were placed on CPAP, and had 10‐min Apgar scores ≥8. A neonate in the ASH arm was treated with surfactant for RDS.

4. DISCUSSION

Vigilance is key in the management of pregnancies diagnosed with vasa previa. Clinical practice guidelines, even in asymptomatic women at low risk of preterm labor, recommend a cautious approach to maximize fetal survival and optimize pregnancy outcomes. Physicians and families must balance several decisions, including the need and timing of admission, potential prolonged hospital stays, empiric administration of ACS, and timing of cesarean to minimize the risk of fetal hemorrhage while reducing the risk of iatrogenic prematurity. ¹⁵ Our data demonstrate that in asymptomatic patients with a singleton fetus and at low risk for preterm birth, outpatient management may be an acceptable and safe option, and that inpatient management in the absence of antepartum bleeding or fetal heart rate abnormalities may not confer added advantage. This outpatient method of management for low risk, asymptomatic patients may enhance the optimal use of ACS and safely prolong pregnancies.

The Society for Maternal‐Fetal Medicine, as well as the SOGC, currently recommend considering inpatient surveillance from 30 weeks' gestation onwards. ¹¹ , ¹³ Swank et al. concur given that most of their emergency deliveries arose from 32 weeks onwards. ¹⁶ Since these recommendations are not based on strong clinical evidence, we counsel our patients based on the level of evidence of recommendations and alternative management strategies and encourage shared decision‐making based on patient preferences and individual circumstances. While symptomatic patients were managed as inpatients, asymptomatic patients were managed either as outpatients or inpatients based on shared decision‐making. Only two asymptomatic patients required unplanned cesareans—both of whom were managed as outpatients and delivered at 36 weeks' gestation and did not suffer major neonatal morbidity. There were also no fetal/neonatal deaths in our cohort, regardless of admission status. However, all asymptomatic mothers in our cohort who were managed as inpatients delivered without specific indications other than the diagnosis of vasa previa, by planned cesarean 1.4 weeks earlier compared to those managed as outpatients. It is conceivable that earlier cesareans in this group were due to perceived risk by the obstetrician. Likewise, Sullivan et al. reported a significant association between antepartum hospitalization for vasa previa and preterm birth, possibly due to the perceived increased risk of preterm labor, increased fetal surveillance, and pregnancies complicated by unrelated comorbidities. ¹⁷ Yet, similar to our use of a selective outpatient approach, Hasegawa et al. admitted patients solely on the basis of adverse events (antepartum hemorrhage, IUGR, abnormal fetal heart tracing, etc.). Although limited by sample size, no adverse outcomes were reported among asymptomatic patients managed as outpatients. ¹⁸

The pregnancy outcomes and ACS administration in our study highlight the possibility of a more selective approach to administration of ACS. When administered between 48 h and 7 days of delivery, ACS significantly reduced neonatal RDS and its associated complications. ¹⁹ The Society for Maternal‐Fetal Medicine and SOGC recommend that obstetricians provide ACS to patients with vasa previa at 28–32 weeks' gestation in order to prepare for potential emergency preterm deliveries, which occurs in 28% of their cited cases. ¹¹ , ¹³ In our study, 46.8% of patients did not receive ACS therapy, with more outpatients having ACS withheld than AHS cases (66.7% vs. 41%). None of the outpatients in whom ACS was withheld, delivered under 34 weeks of gestation (beyond which ACS are not administered in Canada), ²⁰ while only two inpatients in whom administration of ACS was withheld (one each in the SH and ASH arm) delivered under 34 weeks. This did not translate to adverse neonatal outcomes. It must be noted that despite our selective approach, when ACS was administered, in most cases (23/24), this was done >7 days prior to the cesarean birth and therefore outside the period of maximum effect. Similar results were found by Swank et al., where 76% of patients in their study who received ACS all had planned cesareans later than 33 weeks, ¹⁶ prompting the suggestion that ACS be deferred further than currently recommended. Golic et al. similarly reported no neonatal complications among their low‐risk pregnancies defined as no prior history of preterm birth or late abortion, and women who had not experienced antepartum bleeding, a shortened cervix, premature contractions or rupture of membranes in the current gestation suggesting that ACS should instead be reserved for high‐risk cases. ²¹ More recently, Bartal et al. published data on 109 antenatally‐diagnosed cases of vasa previa, 34 of whom were managed as outpatients. Compared with those managed as outpatients, those managed as inpatients were more likely to deliver earlier (36 vs. 36.4 weeks, p = 0.01), be admitted for longer (14 days vs. 1 day, p = 0.002) and receive ACS (57.3% vs. 26.4%, p = 0.002), but were less likely to require unplanned cesareans (34.6% vs. 58.8%, p < 0.001), with no difference in neonatal anemia or other complications. ²² This suggests that routine administration of ACS at 28–32 weeks should be revisited, and consideration should be given to administration when clinically indicated, to optimize its effect.

Cervical length has been used as a measure for predicting preterm birth, with conflicting results to its reliability, ²³ , ²⁴ , ²⁵ which is true of its application to vasa previa. ¹⁶ , ²⁶ A shortened cervix in the third trimester in cases of complete placenta previa and low‐lying placenta has been associated with a higher risk of emergency cesareans <34 weeks' gestation ²⁷ and antepartum hemorrhage. ²⁸ Administration of ACS in patients with vasa previa may benefit from CL measurements in anticipation of preterm labor due to the frequent incidences of antepartum hemorrhage and low‐lying placenta with vasa previa. ²⁹ However, our data did not show any significant differences in the rate of CL shortening between groups, nor were there an appreciable number of short cervices before delivery in our cohort. Based on this study, we are unable to comment on the use of serial CL measurements for monitoring patients with vasa previa. This would require larger prospective studies.

Our study had a number of limitations. First, our sample size was still too small to determine the role of predictors on adverse pregnancy outcomes, or to make firm recommendations with regard to management. Second, prevalence may have been inflated since a large portion of patients are referred to our tertiary care center. A significant number of patients included in the study did not deliver at this institution. Although management recommendations were provided to outside care centers, we are not able to comment on physician and/or patient compliance in external settings. Also, although electronic medical records are utilized in our system, some data were missing and/or insufficient. Finally, a key finding of the study is that a major determinant of whether a patient is likely to require an early unplanned cesarean birth is the presence of symptoms (antepartum bleeding or abnormal NST), and not whether the person was admitted (ASH arm) or not (OP arm). Since this was a retrospective study, groups were assigned following shared decision‐making based on current recommendations. In hindsight, it may be reasonable to conclude that admission of asymptomatic patients conferred no clinical benefit and that these patients could be safely managed as outpatients.

5. CONCLUSION

Based on the findings of this single‐center study, the decision to offer outpatient management to women with vasa previa and at low risk of preterm delivery was associated with an extended pregnancy duration and less administration of ACS, in comparison to asymptomatic women receiving inpatient care. Inpatient management of asymptomatic women did not confer any clinical benefit and resulted in more patients receiving inappropriately time ACS. Prolonged hospitalization may also increase the risks of venous thromboembolism, hospital‐acquired infections, loss of earning and psychosocial effects. It might therefore be prudent to offer patients with vasa previa that do not have antepartum bleeding or fetal heart rate concerns, the option of being managed as outpatients as long as they can have prompt access to the hospital in case of antepartum bleeding or signs of early labor. This approach could safely prolong pregnancies and avoid the unnecessary use of ACS. Establishing an international registry via Research Electronic Data Capture (REDCap) may go a long way in confirming these findings and addressing some knowledge gaps especially with regard to the role of cervical length measurement.

AUTHOR CONTRIBUTIONS

LAV extracted the data, completed the analysis and wrote the first draft of the manuscript. RAT helped with data extraction. PSS, RDS, JCK and JK helped with conceptualization of the study. RDS and JK revised all versions of the manuscript.

FUNDING INFORMATION

This study was partly funded by a grant from the International Vasa Previa Foundation to JK. The sponsor's role was limited to reviewing the manuscript prior to submission.

CONFLICT OF INTEREST STATEMENT

The authors report no conflict of interest.

Supporting information

Figure S1.

Table S1.

Villani LA, Al‐Torshi R, Shah PS, Kingdom JC, D’Souza R, Keunen J. Inpatient vs outpatient management of pregnancies with vasa previa: A historical cohort study. Acta Obstet Gynecol Scand. 2023;102:1558‐1565. doi: 10.1111/aogs.14595