The SOFT Cluster Score as a Multifaceted Predictive Model for Postoperative Outcomes

Consideration of both donor and recipient factors at the time of an organ offer is crucial in achieving the most optimal outcome given the shortage of available donor organs and waitlist mortality risk. Our aim was to achieve a comprehensive risk assessment for liver transplant patients by creating a series of independent risk index equations for 6 distinct posttransplant endpoints in the transplant continuum of care.

The Survival Outcomes Following Liver Transplantation (SOFT) Cluster score includes risk indices for patient mortality at 90 d, 1 y, 3 y, and 5 y posttransplant, graft failure at 1 y after transplantation, and prolonged length of stay (>30 d). Using a total sample of 61 231 patients, significant risk and protective factors were identified through univariable and multivariable analysis from a training set of 40 821 patients from the Organ Procurement and Transplantation Network database.

When applied to a validation set of 20 410 patients, the SOFT Cluster score has a C-statistic of 0.66 for 90-d mortality, 0.64 for 1-y mortality, 0.63 for 3-y mortality, 0.63 for 5-y mortality, 0.76 for prolonged length of stay, and 0.60 for 1-y graft failure.

The SOFT Cluster score represents a novel proof-of-concept model for posttransplant risk assessment and provides physicians with a more holistic understanding of patient risk by examining a spectrum of outcomes after liver transplantation up to 5 y posttransplant using a modern cohort of recipients.