The diagnostic accuracy of bleeding assessment tools for the identification of patients with mild bleeding disorders: A systematic review

INTRODUCTION: The bleeding assessment tool (BAT) has been developed to standardize and interpret bleeding history for mild bleeding disorders. However, a critical appraisal addressing the quality and results of validation studies is lacking. AIM: We performed a systematic review of diagnostic studies assessing the performance of the BAT in patients referred for evaluation of bleeding symptoms. METHODS: The electronic database PubMed was searched from inception through July 27, 2017. Eligible publications were original studies that assessed and validated the diagnostic accuracy of bleeding questionnaires for identification of adults with mild bleeding disorders. For each study, sensitivity, specificity and diagnostic odds ratio (DOR) were calculated. Quality was assessed using the Quality Assessment of Diagnostic studies-2 tool. To assess the influence of specific study characteristics on DOR, univariate meta-regression analyses were performed. RESULTS: Nine studies were included. Five studies investigating the ISTH-BAT or other bleeding questionnaires had a moderate to low DOR. Four studies investigating Vicenza-based BATs had a high DOR, with high specificity (>90%) and sensitivity of 59%-85%. Study characteristics such as case-control design, retrospective data collection and differences in reference standard were associated with optimistic estimates of diagnostic performance. Three of four studies with a high DOR had these study characteristics. Studies with good methodological quality mainly had a low DOR. CONCLUSION: The main advantage of the BAT is that it offers a complete and structured interview. However, the BAT is of limited diagnostic value to the workup of patients referred for bleeding evaluation in clinical practice