Pediatric Bleeding Questionnaire (PBQ)

Overall utility:

• Most useful in the pediatric clinical setting to assess which children should be further evaluated for VWD
• May also be useful in platelet function disorders
• May be used as a self-administered tool once validated appropriately
• Quick and easy to use

Limitations:

Very few studies documented a need for the pediatric specific bleeding symptoms in the “other” category, therefore the need for them is unclear.

1. Time to complete: 20 minutes
2. Equipment/space required: None
3. Training required: Physician, research nurse or research associate (“trained expert”) administered. One study utilized a self-administered version of the PBQ.⁵
4. Cost: None
5. Scoring/scaling/interpretation of results:

• Scale –1 to 4 for 2 categories of the 13 symptoms tested: dental extraction and surgery. Maximum score for cutaneous bleeding is 2; maximum score for gastrointestinal tract or post-partum bleeding is 3; score of –1 is assigned for lack of bleeding on 2 or more challenges for each of post-surgery bleeding, post-partum hemorrhage, and bleeding post-tooth extraction.
• Possible range: –3 to 48
• Differential features include modification to include pediatric-specific bleeding symptoms in the “other category” (post-circumcision bleeding, umbilical stump bleeding, cephalohematoma, macroscopic hematuria, post-venipuncture bleeding, conjunctival hemorrhage)
• Abnormal bleeding score (BS) ≥2

Psychometric properties:

1.  Construct validity:

Convergent validity

• 32% of boys with VWD scored a clinically significant bleeding score (BS) with post-circumcision bleeding.²
• 96% of children with a platelet function disorder scored a positive BS.⁴
• BS was inversely related to VWF:Ag and VWF:RCo when using a self-administered version of the PBQ.⁵
• BS was significantly higher in affected vs. unaffected family members with VWD.⁶

Group differences

• Abnormal BS was derived from assessment of questionnaire in a general control population of children attending school association meetings. The normal range of BS was between –1.5 to 2.5 and only 13% had a score ≥2.¹
• Median BS was significantly different between children with VWD > positive BS without VWD > negative BS without VWD.¹
• Median BS was significantly higher in all forms of VWD vs. controls as well as all forms of definite VWD vs. possible VWD.²

2.  Criterion validity:

• For the laboratory diagnosis of VWD, the BS was able to accurately differentiate between those affected and unaffected with VWD.¹

3.  Reliability:

• Repeat testing was performed in one study but results were not recorded.²

4.  Responsiveness/sensitivity:

• Responsiveness by means of ROC analysis had an AUC of 0.88.¹
• Responsiveness of tool to increased BS with increasing age was shown (Spearman’s correlation coefficient was 0.35, p=0.0004).²
• The sensitivity of an abnormal BS for diagnosis of VWD was 83%, with a specificity of 79%, positive predictive value (PPV) of 0.14, and negative predictive value (NPV) of 0.99.¹
• When using a screening question of “Does your child have a problem with bleeding or bruising?” the sensitivity and specificity for a positive BS and diagnostically proven VWD was similar (sensitivity of 94% vs. 91% and specificity of 94% vs. 95%, respectively).
• Internal consistency of questions within one version of the self-PBQ was good, with a Crohnbach’s alpha of 0.88.⁵
• PBQ ROC-AUC was 0.768, similar to the ISTH BAT ROC-AUC of 0.764.⁸
• However, in a tertiary care setting when evaluating for a mild bleeding disorder, the PBQ was found to have an ROC-AUC<0.6 for qualitative BS cut off of 2:
  – Using a cut off of >2, hemorrhagic symptoms sensitivity was 60.5%, and specificity was 50.8%
  – Using a cut off of BS >3 in males and >5 in females, sensitivity was 41.9%, and specificity was 82%.⁷

Languages studied: English, German ⁸

Groups tested with this measure:

• Controls^1,2
• Children being evaluated for VWD because of personal or family history of bleeding^1,6
• Children with known diagnosis of VWD (type 1–3)^2,3,6,7
• Children with a known diagnosis of a platelet function disorder⁴
• Hemophilia A⁵

Age: Children and some adults