Bleeding events and other complications during pregnancy and childbirth in women with von Willebrand disease

Introduction

There are multiple reports of women with von Willebrand disease (VWD) who have experienced bleeding complications during pregnancy [1–5]. Bleeding is not an uncommon problem during pregnancy, but 102 women with VWD surveyed by the United States (US) Centers for Disease Control and Prevention (CDC) had more pregnancies that ended with heavy bleeding compared with controls [6]. Despite the primary role of uterine contractions in controlling bleeding at the time of delivery, women with VWD are at an increased risk of postpartum hemorrhage. There are multiple case series [7] and two case–control studies [6,8] documenting the increased incidence of postpartum hemorrhage among women with VWD, but there are no large epidemiological studies of VWD in pregnancy. There are no data on the frequency of the diagnosis of VWD among women giving birth and only limited data on the incidence of bleeding events and other complications that women with VWD experience during pregnancy and childbirth.

VWD is a relatively rare condition. No one medical center delivers a sufficiently large enough number of women with VWD to obtain meaningful estimates of the frequency of the diagnosis and the incidence of bleeding events and complications. The purpose of this study was to estimate the incidence of bleeding events and other complications in women with VWD during pregnancy and childbirth using a large US data base.

Methods

Results

During the period from 2000 to 2003, there were 16 824 897 deliveries among 18 345 538 pregnancy-related discharges. Among the deliveries, there were 4067 with a diagnosis of von VWD. Women with VWD were slightly older than women without VWD, 27.9 years vs. 27.4 years old (P < 0.01). The age distribution of the women with VWD is presented in Table 1. Women aged 40 and over were significantly more likely to carry the diagnosis than women under the age of 20. One out of 2700 white women, one out of 12 000 black women and one out of 9000 Hispanic women carried the diagnosis.

At the time of delivery, women with VWD were more likely to be diagnosed with hypertension, cardiomyopathy, anemia and thrombocytopenia than women without VWD. Women with VWD were no more likely to be diagnosed with diabetes or obesity (Table 2). Although women with VWD were no less likely to experience a pulmonary embolism (based on an extremely limited number of cases) than women without VWD, there were no cases of deep vein thrombosis, stroke or myocardial infarction among the women with VWD (Table 3).

Women with VWD were no more likely to develop the pregnancy complications of pre-eclampsia, eclampsia, abruption of the placenta or intrauterine (fetal) growth restriction (Table 4). They were less likely to develop gestational diabetes. They were slightly more likely to undergo Caesarean delivery. Even though they were no more likely to experience abruption of the placenta, they were 10 times more likely to experience other antepartum bleeding (OR, 10.2; 95% CI: 7.1, 14.6). While antepartum bleeding is a risk factor for premature delivery and stillbirth, women with VWD were no more likely to experience preterm labor or suffer an intrauterine fetal demise.

Women with VWD were more likely to experience a postpartum hemorrhage (OR, 1.5; 95% CI: 1.1, 2.0). Six percent (6%) of women with VWD experienced postpartum hemorrhage compared with only 4% of other women. While the numbers were small and the odds ratio did not reach statistical significance, it also appeared that women with VWD were more likely to experience a perineal hematoma (OR, 3.3; 95% CI: 0.8, 13.4). Women with VWD were also more likely to require transfusion. Transfusions were reported for 152 or 4% of the 4067 women sometime during pregnancy or childbirth. The odds of receiving a transfusion were 4.7 (95% CI: 3.2, 7.0) for women with VWD compared with women without VWD. Hospitalization for childbirth was longer for women with VWD. The length of stay was 2.83 ± 0.031 days vs. 2.57 ± 0.001 days (P < 0.01) for women without VWD.

There were five deaths among the 4067 women with VWD. All of the records with a death also had a diagnosis of transfusion, but none had a diagnosis of hypertension, cardiomyopathy, pulmonary embolism or postpartum hemorrhage. The maternal morality rate for women with VWD was 123 per 100 000 deliveries, 10 times higher than the rate for all other women, which was 12.7 per 100 000 deliveries.

Discussion

The reported incidence of VWD depends on the population studied and the definition of disease used. We found the frequency of a diagnosis of VWD among women giving birth in the United States to be 0.024% or one in 4000 women. This figure is lower than the expected incidence based on identification of persons with bleeding symptoms, low von Willebrand factor and a positive family history, which has been estimated to be between 0.6% and 1.3% [11–13], but is two to ten times higher than the expected incidence based on the number of symptomatic patients seen at hemostasis centers, which is 0.002% to 0.01% [14]. One explanation for the higher incidence found in this study compared with the estimate based on patients seen at hemostasis centers is that many patients with VWD are not seen in hemostasis centers, but are managed by hematologists who practise outside of a center.

We found the incidence of VWD to be higher among white women (one in 2700 women) compared with African American women (one in 12 000 women). One possible explanation for this finding is that white women may be more likely to be evaluated for the condition than African American women. An alternative explanation is that African American women have higher levels of von Willebrand factor antigen [15] and are less likely to have the disease. A lower incidence of VWD in African American women has previously been reported in another population, women with menorrhagia [16–18]. The incidence of VWD among Hispanic women was one in 9000 women, intermediate between white and African American women.

Women with VWD were more likely to have certain medical conditions at the time of delivery than women without VWD. Women with VWD were more likely to have hypertension and cardiomyopathy. Other than the fact that women with VWD were slightly older than women without VWD, there is no obvious explanation for this finding. One possible explanation is that women with cardiomyopathy could have had the condition as a result of aortic valve disease and, therefore, could have had acquired VWD. Women with VWD were more likely to have anemia and thrombocytopenia. The increased incidence of anemia among women with VWD is probably explained by their bleeding tendency, which would have predisposed them to pre-existing anemia [19] and anemia as a consequence of postpartum hemorrhage. Although women with VWD were no less likely to experience a pulmonary embolism than women without VWD, there were no cases of deep vein thrombosis. From the same data base, the rate of pregnancy-related deep vein thrombosis was 1.36 per 1000 deliveries [20]. Among 4067 pregnant women, six cases would have been expected, yet none was reported among the women with VWD. There is no obvious explanation why women with VWD would be less likely to experience deep vein thrombosis, yet no less likely to experience pulmonary embolism.

Women with VWD were no more likely to develop the pregnancy complications of pre-eclampsia, eclampsia, abruption of the placenta or intrauterine (fetal) growth restriction. Even though they were no more likely to experience abruption of the placenta, they were 10 times more likely to experience other antepartum bleeding. While antepartum bleeding is a risk factor for premature delivery and stillbirth, women with VWD were no more likely to experience preterm labor or suffer an intrauterine fetal demise.

Not surprisingly, as women with VWD would be expected to have lower von Willebrand and factor VIII levels, women with VWD were more likely to experience a postpartum hemorrhage, but the odds, 1.5 (95% CI: 1.1, 2.0), were not as high as expected. As the women in this study had diagnosed cases of VWD, perhaps providers, in at least some cases, were able to anticipate the possibility of hemorrhage and provide prophylaxis. The rates of postpartum hemorrhage were lower for both cases and controls than in the three other published studies that compared women with VWD with controls. In a study by Silwer [21], 23.5% of women with VWD reported postpartum hemorrhage compared with 19.5% of controls. In a survey of 102 women with VWD published by the CDC, 59% reported a history of postpartum hemorrhage compared with 21% of controls [6], and in another survey, 31% of women with VWD reported a history of postpartum hemorrhage compared with 10% of controls [8]. Case series have reported an incidence of postpartum hemorrhage in women with VWD of between 18% and 80% [7,22]. Compared with the incidence reported in these studies, the incidence estimated by two-thirds of the members of the Royal College of Obstetricians and Gynaecologists who responded to a recently reported survey, was 5–10% [22]. In the case series and case control studies, the incidence of postpartum hemorrhage was based on retrospective chart review or patient recall, which may explain why the rates were higher than the 6% among women with VWD and the 4% among other women found in this study. Nonetheless, women with VWD appear to be at an increased risk of postpartum hemorrhage, and given the fact that in the world as a whole, postpartum hemorrhage is the leading cause of maternal mortality [23], identifying women at increased risk would be expected to be an important strategy in reducing maternal mortality from postpartum hemorrhage.

While the numbers were small and the odds ratio did not reach statistical significance, it also appeared that women with VWD were more likely to experience a perineal hematoma. Perineal hematoma, a rare complication of vaginal birth, has previously been reported to occur with some frequency in women with VWD. Greer et al. [1] reported one hematoma in 13 vaginal deliveries and Kadir et al. [24] reported three in 49 vaginal deliveries. These relatively high frequencies are compared with a rate of only 2.2/1000 in a cohort of 26 187 spontaneous or operative vaginal deliveries [25].

Technically, the number of deaths was statistically significant (P = 0.03), but practically speaking, given that the number of deaths was only five, and five was based on sampling, no meaningful conclusions can be drawn about the risk factors for maternal mortality. While the rate of maternal death was 10 times higher among women with VWD, the rate was still low, 123 per 100 000 deliveries, or approximately one in 1000 women. This should be reassuring to most women with VWD who are contemplating a pregnancy.

There are several limitations to this study. Data were limited to the data in the NIS, which are derived from discharge record abstractions, allowing for the possibility of errors in coding. Nonetheless, the NIS is the most reliable source of US data on hospital admissions and discharges [9,10] and there is no reason to believe that in this case–control study, the errors in coding would introduce bias by appearing more often in the records of women with VWD than in the records of women without VWD. On the other hand, because the women who had a diagnosis of VWD had sufficiently symptomatic disease for the condition to be mentioned in their medical records, the results of this study are likely to be biased toward type 3, type 2 and more symptomatic cases of type 1 VWD. Because data were derived from discharge record abstractions, detailed and precise information was not available to validate the diagnosis of VWD and other comorbidities. Also, the nature of the data base does not allow for the verification of missing data, nor does the nature of the data base allow for the identification of individual women. Therefore, during the 4-year study period, some women will have delivered more than once and their comorbidities will be overrepresented. An alternative explanation for the increased incidence of anemia and thrombocytopenia and the increased incidence of bleeding events among the women with VWD was that these conditions were more likely to be reported among women known to have a bleeding disorder. Additionally, the discharge record abstractions do not allow for attribution of the timing of events. Therefore, it is possible that transfusions were administered for prevention rather than treatment of bleeding. There were insufficient data on the incidence of miscarriage. The data in the NIS are from hospital admissions. Women who miscarry, in most cases, are managed as outpatients without admission to the hospital.

Despite its limitations, the NIS, with its large numbers of cases, allowed for the estimation of the incidence of VWD in pregnancy and childbirth and for the estimation of the incidence of bleeding events and complications in this study. Data were collected prospectively and, therefore, not subject to patient recall, a limitation of other studies. We were able to ascertain that women with VWD do not appear to be at an increased risk of poor fetal outcomes, but are at an increased risk of bleeding events and possibly death during pregnancy and childbirth.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.