Validation and digital implementation of the HADES score for hereditary angioedema

Background Hereditary angioedema due to C1 inhibitor deficiency (HAE-C1-INH) is frequently misdiagnosed, leading to delayed diagnosis and inappropriate management. We previously developed a predictive clinical score (HADES) to support early identification of patients at risk for HAE-C1-INH. Objective To evaluate the temporal performance of HADES in a new cohort of patients with suspected HAE-C1-INH and explore the feasibility of its digital implementation in a small prospective pilot study. Methods A total of 539 individuals with suspicion of HAE-C1-INH were evaluated. HADES scores were calculated using the previously predefined thresholds, and diagnostic performance and ROC analyses were compared with the derivation cohort. In parallel, a prospective pilot study using the HADES App was conducted in 18 patients to assess its implementation feasibility. In addition, 6 patients with HAE-FXII were analyzed separately for exploratory purposes. Results The validation cohort reproduced key performance characteristics of the original model. Sensitivity remained 100% at low thresholds, specificity increased across score categories, and no confirmed HAE-C1-INH cases occurred in score category 0. The AUC was 0.80 (95% CI, 0.73-0.86), matching the derivation dataset. The exploratory HADES-App cohort provided preliminary information on digital implementation feasibility. Patients with HAE- FXII showed score distributions overlapping those of HAE-C1-INH cases. Conclusion In a temporal validation cohort, HADES demonstrated discriminative performance consistent with the original study and reasonable overall category-level calibration, supporting its potential role as a clinical screening tool for HAE-C1-INH. Preliminary digital implementation findings warrant further prospective evaluation.