Period analysis for more up-to-date graft and patient survival estimates in transplantation: an evaluation using united network for organ sharing data.

摘要

BACKGROUND: Traditional, cohort-based survival analysis approaches may provide outdated graft and patient survival estimates in times when clinical progress is rapid. Period analysis, a survival analysis method that uses left truncation and was shown to provide more up-to-date survival estimates than traditional, cohort-based methods in other medical fields, may improve the timeliness of survival monitoring in transplantation. METHODS: Using United Network for Organ Sharing/Organ Procurement and Transplantation Network data, we evaluated, through a series of comparisons, how well most up-to-date 5-year survival estimates potentially derivable by two commonly used cohort-based methods and the period method would have been able to predict the later observed survival of corresponding most recent transplants in the dataset between 1992 to 1994 and 2001 to 2003. RESULTS: In the analysis of overall survival, period analysis provided a best prediction for 93 of the 100 evaluated point estimates, whereas among 350 evaluated point estimates of age-specific survival, period analysis provided a best estimate on 254 occasions (72.6%), compared with 49 (14.0%) and 82 (23.4%) occasions for the cohort-based approaches. Mean average absolute differences between period estimates and the later observed survival were meaningfully lower than those obtained by traditional methods, indicating that period estimates may provide much better survival predictions for recently transplanted grafts and patients than estimates derivable at the same time by traditional survival analysis approaches. CONCLUSION: The timeliness of survival monitoring can be meaningfully improved by the application of period analysis. The use of period analysis for providing more up-to-date survival estimates in transplantation may be encouraged.