Regional and national variation in donor and recipient characteristics, transplant practice, and outcomes after liver transplantation (the IMPROVEMENT collaborative): a multicentre cohort study

Background Outcome variability in liver transplantation remains globally driven by differences in donor-recipient characteristics and in risk-mitigation strategies. Prediction models lack geographic balance and fail to account for current disease classification and advanced organ preservation methods. The IMPROVEMENT study (ClinicalTrials.gov, NCT05289609) aimed to characterise the donor-recipient risks and to develop and validate global prognostic algorithms (mortality and complication rate). Methods This non-competitive, international, cohort study involved 61 centres across Europe, Asia, North and South America, Oceania. Centres provided 83 balanced data series. 5218 patients were enroled in retrospective (2017–2019) and/or prospective (2022–2023) cohorts to avoid COVID-19 confounding. A four-step analytical hierarchy process (world maps, radar graphs, donor versus recipient risk plot, and mixed-effects generalised models) enables visualisation of risk propensity, namely, "aggressiveness", and risk-mitigation balance, namely "coverage adequacy". Outcome measures were 90-, 365-, and 730-day mortality and high complication rates. Findings Overall survival rates at 90 days, 1 year and 5 years were 93, 90 and 80, respectively. Predictors of mortality included recipient age, CAD, pre-operative complex portal thrombosis, preoperative ventilatory life support, MASLD, MELD 3.0 > 30, packed red blood transfusions (PRBC), and cold ischaemia time ≥360 min. The high complication rate resulted dependent from preoperative ventilatory life support, complex portal thrombosis, TIPS and PRBC. Notably, donor factors had a low impact on outcomes, reflecting successful donor-recipient matching and new graft perfusion modalities. Transplant at high-volume centres (>65 transplants/year) was a significant protective factor (mortality and complication rates). C-statistics of 90-day mortality and high complication rate models were 0.78 and 0.77, respectively. Interpretation The IMPROVEMENT study shows that, unlike before, recipient and system-level factors now appear to account for the largest proportion of outcome variability. Implementing risk-mitigation strategies can offset high-risk profiles. These findings provide a robust framework for equitable health resource allocation and clinical decision-making worldwide.