Vasa Previa: The Effect of Prenatal Diagnosis on Outcomes

Abstract & Commentary

Synopsis: Good outcomes with vasa previa depend primarily on prenatal diagnosis and cesarean delivery at 35 weeks of gestation or earlier should rupture of membranes, labor, or significant bleeding occur.

Source: Oyelese Y, et al. Obstet Gynecol. 2004;103: 937-942.

Oyelese and colleagues recently collected data to assess the clinical effect of diagnosing vasa previa prior to delivery. Their findings, though not surprising, were dramatic.

The group went to a repository of vasa previa information, the Vasa Previa Foundation, and to 6 large centers in order to collect data on 155 pregnancies complicated by vasa previa. The diagnosis was ultimately made in all cases by pathology. The idea was to compare outcomes in those who had the ultrasound diagnosis before delivery compared with those who did not.

The perinatal survival in the prenatally diagnosed cases was 96% (59/61) vs 44% (41/94) when not diagnosed by ultrasound. As expected, mean 1 and 5 minute Apgar scores were 1 and 4 in the undiagnosed group and 8 and 9 in the diagnosed group. Also, half of the undiagnosed surviving neonates needed transfusions.

Comment by John C. Hobbins, MD

The only way to avoid mortality and morbidity in vasa previa is to deliver by Cesarean prior to rupture of membranes; so it is no surprise that virtually all perinatal disasters could have been avoided by fore knowledge of this condition.1,2 However, now with large enough numbers, it confirms that looking for the cord insertion, especially in low-lying placentas, is worth the effort, even if the incidence of vasa previa in the overall population is said to be in 1 in 2500.

The American Institute of Ultrasound in Medicine/American College of Radiology (AIUM/ACR) guidelines for the performance of a basic ultrasound examination does not include locating the umbilical cord insertion on the placenta. However, it should, since in vasa previa the simple task can be lifesaving. Also, since marginal or velamentous insertions of the cord are associated with a higher incidence of intrauterine growth retardation (IUGR), this is information that would be useful to the alert clinician.

In placenta previa, vasa previa may actually evolve as pregnancy progresses. It is well known that low-lying placentas will seemingly move away from the cervix with lengthening of the lower uterine segment. However, years ago Bernischke postulated that another phenomenon, which he labeled "trophotropism," could contribute to this relative placental migration.3 The theory is that in some areas of the uterus, such as the lower uterine segment and cervix, the vascular environment is poorly suited to support placental development, so the placenta preferentially grows superiorly as pregnancy progresses while atrophying inferiorly. However, the umbilical cord may keep its relationship with the cervix, but lose its placental cushion, thereby finding itself in the membranes directly over the cervix.

Another condition that lends itself to vasa previa is an accessory lobe in which the connecting fetal vessels course over the cervix. This carries the same potential for disaster as an umbilical cord insertion in this area.

Although the reason for this is unclear, there is an inordinately high risk of vasa previa in (in vitro fertilization) IVF pregnancies. In one study, the rate was 1:293 IVF pregnancies vs 1:6000 in spontaneously conceived pregnancies. This is yet another example of the less than welcoming intrauterine environment inherent in some pregnancies resulting from assisted reproductive technology (ART).

It generally takes seconds to find the cord insertion site in the placenta (even without color Doppler) and I think that in every scan an attempt should be made to do this. If the cord appears to be inserting in the neighborhood of the cervix, then the diagnosis of vasa previa could be confirmed by transvaginal ultrasound. If the diagnosis is made early in pregnancy, then it should be confirmed prior to 35 weeks of gestation since, as pointed out above, the relationship of the placenta, cord, and cervix can change in later pregnancy.

In the Oyelese study, the average gestational age at delivery for the prenatally diagnosed vasa previa was 34.9 weeks, compared with 38.2 weeks for the undiagnosed vasa previa. They make a pitch for interruption of pregnancy around 35 weeks, and from their data I agree.


1. Oyelese KO, et al. Ultrasound Obstet Gynecol. 1998;12:434-438.

2. Sepulveda W, et al. Ultrasound Obstet Gynecol. 2003;21:564-569.

3. Bernischke K, et al. (eds). Pathology of the human placenta. 4th ed. New York (NY):Springer 2000;399-418.

John C. Hobbins, MD, Professor and Chief of Obstetrics, University of Colorado Health Sciences Center, Denver, and Associate Editor of OB/GYN Clinical Alert.