Von Willebrand disease (VWD) is the most common inherited bleeding disorder and is diagnosed via 3 cardinal features: 1) personal history of bleeding, 2) laboratory assays and 3) family history of VWD. The diagnosis of VWD in pediatric patients is complicated by von Willebrand factor (VWF) inter- and intra-assay variability, phlebotomy-induced physiologic stress increasing VWF levels from baseline, and a lack of prior hemostatic challenges. Given these challenges, NHLBI guidelines recommend repeated testing in patients with mildly low or normal levels and a high suspicion of VWD. However, no studies to date have evaluated the utility of repeat VWF testing in pediatric patients. Currently, our center's standard of care is to complete 3 separate sets of VWF testing to rule out VWD. The primary objective of this study was to determine the clinical variables associated with requiring more than 1 test to diagnose VWD and to establish a cutoff value for the first set of VWF assays above which further testing would not be informative.

This single center retrospective cohort study included patients ≤ 18 years of age who either had a diagnosis of or evaluation for VWD between January 2012 and July 2017. Patients were excluded if the VWD laboratory evaluation was completed at another institution or due to the presence of another bleeding disorder. Medical charts were abstracted for demographic information, medications, reason for testing, family history of VWD, results of VWF assays, and other illness at time of evaluation. All patients had a retrospective ISTH BAT score completed. Data were analyzed using SAS and are reported as median (IQR). Statistical analysis was done with non-parametric tests (Mann-Whitney or Wilcoxon sign-rank) for two groups comparisons. Odds ratios were calculated using Fisher's Exact test for clinical factors associated with a VWD diagnosis. Univariate logistic regression was performed, modeling the odds of requiring more than 1 diagnostic test to diagnose VWD.

One thousand unique patients were evaluated and 189 excluded, yielding a final cohort of 811 patients. Of these, 631 (77.8%) did not have VWD and 180 (22.2%) were diagnosed with VWD. Patients diagnosed with VWD were younger than those without (median age 5.8 vs 8.5 years, p=0.0019) and were more likely to have a family history of VWD (38% vs 22%, p < 0.0001) but there was no difference in race or sex between cohorts. As expected, patients in the VWD cohort had lower VWF activity (34 vs 78 IU/dl, p < 0.0001), VWF antigen (45 vs 89 IU/dL, p < 0.0001) and FVIII (57 vs 100 IU/dL, p < 0.0001) than those without VWD. ISTH BAT scores were higher in the VWD cohort (2.47 vs. 2.07, p = 0.027).

As shown in Table 1 and Figure 1, increased odds of VWD diagnosis were noted in those tested due to a family history of VWD (OR 1.75, 95% CI 1.21-2.51) or abnormal coagulation studies (OR 1.61, 95% CI 1.07-2.24). Subjects with VWD required fewer tests than those without VWD (median 1 vs 2, p < 0.001).

Univariate analysis failed to identify any significant associations with needing > 1 test for the diagnosis of VWD (Table 1), so a multivariable model was not performed. In 69.4% (125/180) of subjects with VWD, the first test was diagnostic.

In receiver-operating curve analysis, the first VWF antigen and activity have a high power for diagnosis of VWD with AUC of 0.88 and 0.92, respectively (Fig 1). A cutoff of 100 IU/dL for VWF antigen or activity on first test yielded a sensitivity of 95%, specificity of 38% and negative predictive value of 96.6% for VWF antigen compared to 98%, 38% and 98.6% for VWF activity, respectively.

Here we demonstrate that the majority of pediatric subjects had diagnostic VWF values on the first set of testing. Unfortunately, no clinical variables were identified for patients with VWD who required > 1 test for diagnosis. However increased odds of VWD diagnosis were noted in those with a family history of VWD and abnormal coagulation studies. A cutoff of 100 IU/dL for VWF activity or antigen on the first test resulted in > 95% negative predictive value to rule out the diagnosis. Pediatric patients without a family history of VWD and VWF levels > 100 IU/dL on initial test may not need further testing to rule out the diagnosis of VWD.


Doshi:Bayer Hemophilia Awards Program: Research Funding. Butler:Pfizer: Consultancy; Genentech: Consultancy; HemaBiologics: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.