1. A novel learning algorithm to predict individual survival after liver transplantation for primary sclerosing cholangitis.
- Author
-
Axel Andres, Aldo Montano-Loza, Russell Greiner, Max Uhlich, Ping Jin, Bret Hoehn, David Bigam, James Andrew Mark Shapiro, and Norman Mark Kneteman
- Subjects
Medicine ,Science - Abstract
Deciding who should receive a liver transplant (LT) depends on both urgency and utility. Most survival scores are validated through discriminative tests, which compare predicted outcomes between patients. Assessing post-transplant survival utility is not discriminate, but should be "calibrated" to be effective. There are currently no such calibrated models. We developed and validated a novel calibrated model to predict individual survival after LT for Primary Sclerosing Cholangitis (PSC). We applied a software tool, PSSP, to adult patients in the Scientific Registry of Transplant Recipients (n = 2769) who received a LT for PSC between 2002 and 2013; this produced a model for predicting individual survival distributions for novel patients. We also developed an appropriate evaluation measure, D-calibration, to validate this model. The learned PSSP model showed an excellent D-calibration (p = 1.0), and passed the single-time calibration test (Hosmer-Lemeshow p-value of over 0.05) at 0.25, 1, 5 and 10 years. In contrast, the model based on traditional Cox regression showed worse calibration on long-term survival and failed at 10 years (Hosmer-Lemeshow p value = 0.027). The calculator and visualizer are available at: http://pssp.srv.ualberta.ca/calculator/liver_transplant_2002. In conclusion we present a new tool that accurately estimates individual post liver transplantation survival.
- Published
- 2018
- Full Text
- View/download PDF