RESUMEN
INTRODUCTION: The Sequential Organ Failure Assessment (SOFA) score has been applied for the prediction of survival in critically ill patients. We analysed the value of the SOFA score for the prediction of short-term survival after liver transplantation in high-risk liver transplant recipients with a labMELD score ≥30. PATIENTS AND METHODS: We conducted a retrospective single-centre analysis including 88 consecutive liver transplants in adults between January 1, 2007 and December 31, 2010 with a pre-transplant labMELD score ≥30. The SOFA score was assessed preoperatively, directly after transplantation and on post-operative days (PODs) 1-10. Combined and living-related liver transplants were excluded. Receiver operating characteristic (ROC) curve analysis with the Hosmer-Lemeshow test and application of the Brier score were used to calculate sensitivity, specificity, overall model correctness and calibration. Cutoff values were selected with the best Youden index. RESULTS: ROC curve analysis showed areas under the curve (AUROCs) >0.8 for the SOFA score on PODs 1-10 for the prediction of hospital mortality, 30-day mortality and 3-month mortality with Hosmer-Lemeshow test results that confirmed good model calibration (p > 0.05). The Brier score demonstrated an accuracy of prediction (<0.25) of hospital mortality, 30-day mortality and 3-month mortality for the SOFA scores on PODs 4-9 indicating superior accuracy on PODs 7 and 8 with cutoff values for the SOFA score between 16.5 and 18.5. The pre-transplant SOFA score failed to reach AUROCs >0.7 (0.603-0.663) for the prediction of short-term survival. CONCLUSIONS: Our results confirm the usefulness of the SOFA score in high-risk liver recipients during the early post-operative course, especially on PODs 7-8 for the prediction of hospital mortality, 30-day mortality and 3-month mortality and may be useful to predict futile early acute retransplantation.
