ER-Based Protocol Improves the Accuracy of von Willebrand Disease Testing in Young Patients With Heavy Menstrual Bleeding

Watch our interview with Megan C. Brown, MD.


Researchers at Children’s Healthcare of Atlanta (CHOA) have developed an emergency department (ED)–based protocol that helps standardize both the evaluation and treatment of von Willebrand disease (VWD) in girls with heavy menstrual bleeding (HMB), according to research presented at the 2019 ASH Annual Meeting. The protocol also includes values for von Willebrand factor antigen (VWF:Ag) or von Willebrand factor ristocetin cofactor (VWF:RCo) that identify patients who are unlikely to have VWD and may not require repeat VWD testing, lead investigator Megan C. Brown, MD, from CHOA, reported.

HMB is a common reason for ED visits among adolescent females; in addition to the physiological effects, it has a significant impact on quality of life and leads to school absenteeism, Dr. Brown said during her presentation. While most patients with HMB do not have a coagulation disorder, HMB is a common presentation of inherited bleeding disorders.

The timing and contents of a hemostatic work-up for younger patients with HMB is challenging, as data to guide testing are extrapolated from adult data, despite age-related differences in the causes of acute HMB. “Adolescents are often stressed, both physiologically, from iron deficiency, and from a psychological standpoint, due to being in the clinic or the ED,” Dr. Brown told ASH Clinical News. “We know that VWF and factor VIII (FVIII) can be affected by this,” she added, with previous studies showing higher serum levels of each factor in the setting of physiologic stress and supplemental estrogen use.

Researchers at CHOA therefore developed a standardized protocol to improve the reliability of these test results and find the optimal time to perform testing – in the acute setting in the ED or at a follow-up visit with a hematologist, when bleeding and anemia have resolved.

The protocol was tested in all CHOA EDs in the metropolitan Atlanta area. Patients with a previously diagnosed bleeding disorder, immune thrombocytopenia, active rheumatologic disease, cancer, or anticoagulant use were excluded.

Eligible patients presented with acute HMB, per an adapted Philip Menorrhagia Screening Tool. If girls met the criteria for HMB, the provider performed a uniform bleeding inventory and a standardized set of laboratory tests based on the American College of Obstetricians and Gynecologists consensus statement for adults. These measures helped to standardize the care delivered to girls with HMB, which Dr. Brown noted was “quite variable in the past,” and differed according to the ED provider and the consulting hematologist.

During a 2-year period, 221 adolescent girls (median age at presentation = 13.5 years; range = 12.5-15.2) presented to the ED with acute HMB. Of the entire population, 126 had VWD studies at presentation, and 40 had studies obtained at follow-up.

As the investigators expected, FVIII levels at presentation were “quite high,” Dr. Brown reported, and significantly decreased at follow-up (p=0.0017). This pattern was also observed with VWF:Ag and VWF:RCo levels. In addition, while girls with normal FVIII levels at presentation had a negligible change in VWF:Ag or VWF:RCo levels between presentation and follow-up, there was an approximately 35% difference in patients with elevated FVIII at presentation (p=0.03 and 0.02).

In a secondary analysis, the investigators looked at whether a cutoff value of >100 IU/dL for baseline VWF:Ag or VWF:RCo could rule out patients who do not require repeat testing for VWD, which U.S. and U.K. guidelines note may be necessary to ensure an accurate diagnosis. They found that both measures were associated with a high negative predictive against VWD:

  • VWF:Ag >100 IU/dL: 92.3%
  • VWF:RCo >100 IU/dL: 95%
  • together: 92.6%

The diagnostic accuracy was even higher in girls with normal FVIII levels at presentation. The >100 IU/dL cutoff had a 100% negative predictive value in these patients, suggesting that repeat VWD testing is not necessary in this population.

Overall, 16 of the 126 girls with initial VWD labs were diagnosed with type 1 VWD or low VWF. Most of these diagnoses (69%) were made with the initial labs, Dr. Brown noted, “indicating that despite the fact that these levels are elevated at the time of presentation, we are picking up a significant number of these girls by screening at the time of acute HMB presentation.”

Based on these results, the CHOA electronic medical records system now includes an automatic referral to hematology for any adolescent girl who presents to the ED with HMB, Dr. Brown added.

Still, these findings are limited by the low rates of adherence to follow-up visits with a hematologist or gynecologist, which Dr. Brown said could give false reassurance against a diagnosis of VWD.

“In our population of more than 200 girls, only 17% had the follow-up necessary to go through the process of ensuring whether or not they had a bleeding disorder,” she noted. “In an ideal world, we would test these girls at follow-up, which would eliminate the stress factor that can lead to false diagnoses, [but] we feel like it’s best to continue some upfront testing with the knowledge that if we can improve the system of follow-up to get [patients] in to the hematology clinic as we had hoped, we would move that testing into a time where they are not acutely stressed.”

The study authors report no relevant conflicts of interest.

Reference

Brown MC, White MH, Sidonio RF. Obtaining a von Willebrand evaluation at time of acute heavy menstrual bleeding presentation leads to overestimation of von Willebrand levels. Abstract #627. Presented at the 2019 American Society of Hematology Annual Meeting, December 9, 2019; Orlando, FL.