S2474 Quantifying the Tradeoff Between Medical Urgency and Psychosocial Fitness on Liver Transplant Listing Decisions

Introduction: While Model for End-Stage Liver Disease (MELD) prioritizes liver transplant (LT) candidates based on medical urgency; psychosocial risk factors are concurrently considered for listing decisions. Anecdotal evidence suggests medical urgency often outweighs psychosocial fitness for high acuity patients, but this remains unconfirmed. Methods: We identified 3,058 adults listed for LT at UCSF between 2012-2023. Using validated LLM-based NLP algorithm, we extracted 23 psychosocial risk factors from documentation and compared their prevalences across listing MELD score groups. We excluded Status 1A patients. Hepatocellular carcinoma patients were included based on their non-exception MELDs. We compared psychosocial risk factor prevalence by MELD score groups at listing with z-tests and calculated odds ratios using the Wald test. All P-values were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate. To determine whether differences persisted at LT, we also analyzed the subset of patients who underwent LT. Results: Of 3,058 patients listed for LT: 44.9% MELD ≤15, 22.3% MELD 16-20, 14.3% MELD 21-25, 9.1% MELD 26-30, 6.0% MELD 31-35, 3.5% MELD >35. Using MELD ≤15 as reference, there were consistent increases in psychosocial risks at higher MELDs. For recent alcohol use: prevalence was 11.5% in MELD ≤15 – increased to 18.7% MELD 26-30, 21.3% MELD 31-35, and 23.6% MELD >35 (all P < 0.01). For low transplant knowledge, prevalence was 3.3% in MELD ≤15 – increased to 11.5% MELD 31-35 and 12.3% MELD > 35 (all P < 0.01). For caregiver barriers: prevalence was 15.2% in MELD ≤15 – increased to 31.1% MELD >35 (P < 0.01). For history of nonadherence: prevalence was 7.8% in MELD ≤15 – increased to 17.5% MELD 31-35 and 20.8% MELD > 35 (all P < 0.01). For ongoing struggles with past trauma, prevalence was 5.5% in MELD ≤15 – increased to 12.0% MELD 31-35 (P < 0.01). Analyses of the subset of patients who underwent LT showed similar and persistent patterns, with higher prevalences of psychosocial risk in patients listed at higher MELDs. Conclusion: Psychosocial risk factors are more prevalent amongst listed patients with high MELDs, implying that medical urgency outweighs psychosocial risk in listing decisions. While this phenomenon was most apparent in MELD >35 where there is not enough time to “road test” or optimize psychosocial risk, the patterns also persisted at lower MELDs and at transplant. Future work will explore whether these pronounced psychosocial risk differences, especially in high MELD patients, impact post-LT outcomes.