Utility of Von Willebrand Testing during Acute Bleeding Episodes

Abstract

Background: The diagnosis of mild bleeding disorders such as von Willebrand disease (VWD) poses many diagnostic challenges. Several patient-specific factors including physiologic, inflammation, anemia and medications can affect VWD testing and lead to falsely elevated values. Given that the physiologic stress may cause falsely normal levels of VW factor and associated tests, determining the appropriate timing for testing remains an area of debate. Testing is often performed during inpatient or emergency department (ED) visits when patients present for bleeding complaints, although these results may be skewed by stress or illness. Initial normal results may falsely reassure a patient and their health care provider, preventing them obtaining repeat testing or pursuing outpatient hematological evaluation. Further study is needed to determine the reliability of VWD evaluation when obtained during episodes of acute bleeding. Methods: This single-center retrospective cohort study included subjects who were evaluated for acute bleeding complaints in the UPMC Children's Hospital of Pittsburgh ED or inpatient units who had VWD testing sent during that visit over a five-year period. Bleeding complaints included were epistaxis, heavy menstrual bleeding, spontaneous ecchymosis, post-procedural hemorrhage or intracranial hemorrhage. VWD test results including von Willebrand antigen (VWF:Ag), factor VIII (FVIII) and ristocetin co-factor (VWF:RCo) levels obtained at the time of ED or inpatient care were compared with those obtained in outpatient follow-up visits. T-tests and Wilcoxon rank sum tests were used to compare VWD test levels across the two time points. Results: Over a 5-year period, 144 patients had VWD testing completed during ED or inpatient evaluations for acute bleeding complaints. The population was primarily Caucasian (71%) with a median age of 9 years old. Fifty-two percent of the population was male. Epistaxis and post-tonsillectomy/adenoidectomy bleeding were the most common presenting complaints, making up 47 and 22% of the population, respectively. Other presenting bleeding complaints included spontaneous bruising (18%), menorrhagia (6%), post-procedural hemorrhage of a digestive organ (4%), and nontraumatic intracranial hemorrhage (3%). Six subjects (4%) met laboratory criteria for a diagnosis of VWD during an acute bleeding episode. Twenty-four percent of the total population was evaluated subsequently by outpatient hematology, and 23% of patients had repeat VWD testing performed. Median levels of VWD tests obtained at initial presentation were 116% for VWF:Ag, 112% for FVIII, and 85% for VWF:RCo, and were not statistically different from those obtained on repeat testing in outpatient follow up. An additional two patients that were not initially diagnosed with VWD met laboratory criteria for VWD on repeat outpatient testing. Conclusions: Among children presenting for emergency or inpatient care for acute bleeding episodes, VWD testing was unlikely to be diagnostic. Interestingly, our study did not find that median VWF:Ag and FVIII levels were higher during acute bleeding as has been suggested in other studies. A minority of patients who were evaluated in the hospital for acute bleeding were subsequently evaluated by hematology, which may have limited these findings. It is possible that obtaining VWD testing at the time of an acute bleeding episode may provide false reassurance to a patient, thus decreasing hematology referrals and follow up. Based on these results, we recommend that for patients evaluated in the ED with mild mucocutaneous bleeding who warrant a bleeding disorder evaluation, VWD testing be postponed until the patient is seen in steady-state by hematology.