The Power of a Standardized Bleeding Score in Diagnosing Pediatric Type 1 Von Willebrand Disease.

Type I von Willebrand disease (VWD) is the most common inherited bleeding disorder. Repetitive testing of von Willebrand factor (VWF) levels is necessary before the diagnosis can be safely ruled out, as VWF levels fluctuate in response to genetic and environmental factors. A predictive bleeding score (BS) could reveal individuals that may benefit from repetitive testing and those for whom repetitive testing is unlikely to be of benefit. While a standardized questionnaire (the Vicenza score) was developed to evaluate hemorrhagic symptoms, it was never prospectively validated for a pediatric population in a tertiary care setting.

The study targeted children, ages 0 to 17 years, referred to the Hemostasis and Thrombosis Center (HTC) of Nationwide Children's Hospital for a coagulation evaluation as a result of bleeding symptoms, family history of a bleeding disorder and/or abnormal coagulation labs found during pre-operative screening. Children were excluded if they had a previously diagnosed bleeding disorder, if their caregiver did not speak English or if the child did not undergo VWF:Ag and VWF:RCo testing.

Prior to the diagnosis or exclusion of a bleeding disorder in the child, caregivers consented to answer the questionnaire over the telephone. Descriptions of the Vicenza score are available online (http://www.euvwd.group.shef.ac.uk/bleed_score.htm).

A single VWF:Ag or VWF:RCo <30 IU/dL was classified as “Definite Type 1 VWD” while levels from 30-50 IU/dL were classified as “Low VWF” (http://www.nhlbi.nih.gov/guidelines/vwd). Platelet function analysis (PFA-100) screened for platelet function defects, with some patients undergoing follow-up platelet aggregation studies and/or platelet electron microscopy. Laboratory studies from other institutions were excluded from analysis. Patients' medical records were reviewed after hematologic evaluation, and the resultant data was analyzed with STATA 10.1 (Stata Corp., College Station, TX).

A total of 104 children (52 females and 52 males) with a mean age of 7.53 years (range 1 month to 17 years) were included. At least one hemorrhagic symptom was present in 99 of the 104 children (95%) with the mean number of symptoms being 2.87 (range 0 to 7). The mean Vicenza score was 3.24 (range -1 to 13). Of the 104 children, 8 met criteria for “Definite Type 1 VWD,” 23 met criteria for “Low VWF,” 14 were diagnosed with a “Platelet Function Defect,” and 2 children had bleeding secondary to Ehlers Danlos syndrome. Children with non-bleeding disorders (e.g. Factor XII deficiency) or no laboratory evidence of a bleeding disorder were classified as “No Bleeding Disorder.” In general, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and positive diagnostic likelihood ratio of the bleeding questionnaire demonstrated poor predictive value in our patient population with the exception of high specificity in ruling out “Definite Type 1 VWD” (Table). The NPV was comparably high with both qualitative (>2 bleeding symptoms) and quantitative (BS ≥2) criteria.

The Vicenza score, previously validated in adults and in a pediatric primary care setting, appears to have limited predictive value in a pediatric tertiary care setting when evaluating patients with platelet function defects or low VWF levels. While the Vicenza score has a high NPV to exclude “Definite Type 1 VWD,” the use of simpler qualitative criteria is similarly predictive.

No relevant conflicts of interest to declare.