External Transportability Validation, Implementation Readiness, and Outcome Association of the TransHepMet-Mort Liver Metabolic Mortality Risk Model

Study data were derived from existing electronic health records at our hospital's health examination center between January 1, 2018, and December 31, 2024. These records encompassed fatty liver-related checkups, liver ultrasound/elastography, laboratory tests, anthropometric measurements, and medical histories. Following ethical approval, these datasets were linked to mortality registry, hospitalization records, and liver-related events. This study does not alter participants' past or current clinical pathways, introduces no new examination items, and requires no additional follow-up visits.Concurrently, this study serves as an external implementation evaluation of a clinical prediction model. The original model was derived from prior analyses of NHANES, CHARLS, and VCTE datasets. Rather than re-developing the primary model within the hospital data, this study focuses on evaluating the real-world implementation readiness of the existing model, conducting phenotypic validation, and performing conditional outcome verification. The application of the model within the hospital cohort is strictly limited to research purposes and will not serve as the sole baseline for clinical diagnostic or therapeutic decision-making.