Bleeding Score as a Preoperative Predictor of Postoperative Bleeding in Type 1 Von Willebrand Disease.

Type 1 Von Willebrand Disease (VWD) is the most common congenital bleeding disorder, affecting 1% of the population, and caused by a quantitative deficiency of Von Willebrand Factor (VWF). In addition to mucosal bleeding, VWD patients often suffer postoperative bleeding, leading to significant morbidity. Thus, a preoperative diagnosis could potentially reduce postoperative bleeding. Because symptoms correlate poorly with VWD assays, subject to extragenic effects and lab variability, diagnosis is difficult. The bleeding score (BS) is a simple quantitative tool recently developed to rate bleeding symptom severity, with 99% specificity for VWD. To determine the potential utility of BS in predicting postoperative bleeding in VWD, we evaluated preoperative BS by retrospective review of type 1 VWD patients who suffered postoperative bleeding prior to diagnosis. Preoperative clinical bleeding symptoms and VWD assays, including VWF:RCo, VWF:Ag, and FVIII:C, were obtained. The severity of clinical bleeding symptoms present prior to surgery was rated by the 4-point BS scale: 0 = no/trivial; 1 = present; 2 = intervention required; 3 = replacement therapy. Statistical analysis was by chi square analysis and Fisher’s exact test for categorical data, and by student t test for continuous data. Of 260 registered type 1 VWD patients, 71 (27.3 %) experienced surgical bleeding prior to a diagnosis of VWD. Of these 56 (78.9%) were female, 48 (67.6%) were adults (≥ 18 yr), and 61 (85.9%) had a family bleeding history. The surgeries included general, gynecologic, genitourinary, and otolaryngologic procedures. The median preop BS, 3 in females and 4 in adults, was significantly higher than in males and children, each median 1, p<0.01, respectively. A BS ≥ 3 would have identified only 59.1% patients before surgery, but as many as 90.1%, if combined with one abnormal VWD test; 94.4%, with family bleeding history; or 97.2% with both family history and one abnormal VWD test. The proportion of children identified by BS was significantly lower than in adults, 26.1% vs 75.0 % with BS > 3, p = 0.001. Yet this significantly improved by combining BS with family history, 91.3% vs 95.8%, not different from adults, p = 0.591. We conclude that obtaining a preoperative BS and family bleeding history may reduce postoperative bleeding and promote timely diagnosis among individuals with type 1 VWD patients, particularly children.