Prospective Evaluation of ISTH-BAT As a Predictor of Bleeding Disorder in Adolescent Girls with Heavy Menstrual Bleeding

Background. Heavy menstrual bleeding (HMB) is a frequent complaint in adolescence and is often multifactorial. Of the possible causes, anovulation is likely to be the most common reason but there is mounting evidence that bleeding disorders (BDs) are often an unidentified cause of HMB. Wide ranges of reported prevalence, difficulty in discerning normal from excessive menstrual bleeding and the semi-empiric use of hormonal therapy makes identifying BDs in adolescents challenging. Bleeding assessment tools (BATs) have been developed- primarily in the adults - to improve diagnostic accuracy, predict bleeding phenotype and describe symptoms. The International Society of Haemostasis and Thrombosis (ISTH) BAT was specifically designed to quantify bleeding symptoms that are pediatric specific. An ISTH-BAT score of > 3 for children is considered abnormal. The ISTH BAT has not yet been specifically tested in adolescents presenting with HMB. The objective of this study was to examine the diagnostic utility of ISTH-BAT bleeding score (BS) of > 3 as a predictor of BDs in adolescents with HMB.

Methods. We prospectively analyzed clinical data on 70 adolescents without a known BD, referred for HMB, to the Adolescent, Gynecology and multidisciplinary Young Women's Blood Disorders Clinics at University of Texas Southwestern Medical Center from July 2014 to June 2016. This cohort is part of an ongoing prospective study investigating the incidence of BDs in adolescents with HMB (planned n=200). All subjects underwent a standardized comprehensive diagnostic approach, including the ISTH BAT to quantify bleeding symptoms accurately. The ISTH-BAT was applied by two trained investigators and any discrepancy in scores was settled by discussion. As per National Heart, Lung, and Blood Institute guidelines, VWF:Ag and/or VWF:RCo<30 IU/dL were labeled "definite von Willebrand disease (VWD)" while 30-50 IU/dL were labeled as "low VWF levels" which were grouped together to represent vWD. Receiver operating characteristics (ROC) of ISTH-BAT were determined to assess its value for predicting BD in our cohort.

Results. The mean age of study participants was 14.4+1.8 years (range: 11-18 years). Twenty-eight out of 70 patients (40%) were found to have a BD; 8 met criteria for VWD, 12 had low VWF levels, 2 were hemophilia A carriers, 5 were diagnosed with inherited platelet dysfunction and 1 had inherited thrombocytopenia. The mean BS was higher in subjects with VWD (N=20) as compared to those without a BD (N=42) (4.5 +1.6, vs. 3.6+ 1.0, p= 0.02). At least one other bleeding symptom was present in 8 (40%) of the 20 girls with vWD. The most commonly reported bleeding symptoms were epistaxis (35%), oral (15%), cutaneous (10%) and post-surgical (5%). There was no difference in patterns of menstrual bleeding (anovulatory vs. ovulatory) between girls with and without a BD (55% ovulatory vs. 41% ovulatory, p-=0.28). ROC analysis of the ISTH-BAT bleeding scores showed that at a BS of > 3, the sensitivity of the ISTH BAT was 100% but specificity was only marginal at 2.4% with an accuracy of 41%, whereas at a BS of > 5, sensitivity, specificity and accuracy were 30%, 88.10% and 69%, respectively. ROC analysis showed area under the curve of 0.66 (CI: 0.52-0.80) indicating poor discrimination for the ISTH-BAT score in determining BD in girls with HMB.

Conclusion. Our study is the first attempt to prospectively examine the applicability of using ISTH-BAT score as a screening tool to exclude the presence of BD in adolescent girls presenting with HMB. In this study cohort, ISTH-BAT bleeding score of > 3 demonstrated poor diagnostic accuracy in ruling out vWD. A score of > 5 had high specificity which can decrease false positive diagnosis and repetitive testing. Future data from the ongoing study will help understand how a combination of BATs and a laboratory testing algorithm can unravel hemostatic defects in adolescents with HMB.