Remzi EKİCİ
Turkish Journal of Internal Medicine - 2026;8(1):174-175
This correspondence discusses the use of the Glasgow-Blatchford score (GBS) to assess the appropriateness of interhospital referral in suspected upper gastrointestinal bleeding, as reported by Ünal and Coşgun in this journal. Bleeding-related risk and the need for transfer are distinct constructs: referral decisions also depend on local endoscopic capacity, staffing, and access to blood products. Interpretation of the reported findings is limited by retrospective adjudication of referral appropriateness, inclusion of referred patients only, incomplete assessment of the outcomes the GBS was developed to predict, and the absence of confidence intervals for discrimination estimates. Clarification of mortality event coding and reporting inconsistencies would also improve interpretability. A prospective multicentre study incorporating all patients assessed at referring hospitals, prespecified independent adjudication, and centre-level resource data would better distinguish patient-level bleeding risk from system-level determinants of transfer.
