Deceased Donor Kidney Transplantation for Older Transplant Candidates: A New Microsimulation Model for Determining Risks and Benefits

Abstract

Background: Under the current US kidney allocation system, older candidates receive a disproportionately small share of deceased donor kidneys despite a reserve of potentially usable kidneys that could shorten their wait times. To consider potential health gains from increasing access to kidneys for these candidates, we developed and calibrated a microsimulation model of the transplantation process and long-term outcomes for older deceased donor kidney transplant candidates.

Methods: We estimated risk equations for transplant outcomes using the Scientific Registry of Transplant Recipients (SRTR), which contains data on all US transplants (2010-2019). A microsimulation model combined these equations to account for competing events. We calibrated the model to key transplant outcomes and used acceptance sampling, retaining the best-fitting 100 parameter sets. We then examined life expectancy gains from allocating kidneys even of lower quality across patient subgroups defined by age and designated race/ethnicity.

Results: The best-fitting 100 parameter sets (among 4,000,000 sampled) enabled our model to closely match key transplant outcomes. The model demonstrated clear survival benefits for those who receive a deceased donor kidney, even a lower quality one, compared with remaining on the waitlist where there is a risk of removal. The expected gain in survival from receiving a lower quality donor kidney was consistent gains across age and race/ethnic subgroups.

Limitations: Limited available data on socioeconomic factors.

Conclusions: Our microsimulation model accurately replicates a range of key kidney transplant outcomes among older candidates and demonstrates that older candidates may derive substantial benefits from transplantation with lower quality kidneys. This model can be used to evaluate policies that have been proposed to address concerns that the current system disincentivizes deceased donor transplants for older patients.

Highlights: The microsimulation model was consistent with the data after calibration and accurately simulated the transplantation process for older deceased donor kidney transplant candidates.There are clear survival benefits for older transplant candidates who receive deceased donor kidneys, even lower quality ones, compared with remaining on the waitlist.This model can be used to evaluate policies aimed at increasing transplantation among older candidates.

Keywords: calibration; kidney transplantation; microsimulation; older candidates; posttransplant; waitlist.

Figure 1/

Cohort Construction Cohort Inclusion/Exclusion Criteria Flow Chart. 18,135 recipients and 71,522 candidates were used to construct risk prediction equations. 4,534 recipients and 17,881 candidates were used to calibrate the model.

Figure 2/

Model Diagram Model diagram where individuals start in the Waitlist Markov Model. If they receive a deceased-donor transplant, they have one of the 30-day outcomes before entering the Post-Transplant Markov Model.

Figure 3/

Simulated Outcomes Compared to Observed Targets Results of calibration comparing observed to simulated targets, showing strong fit to our targets.

Figure 4/

Kaplan-Meier Curves - Simulated vs. Observed Outcomes Kaplan-Meier curves comparing simulated to observed survival. The gray curves each represent one of the simulated best-fitting parameter sets, while the gray curve represents the observed data.

Figure 5/

Expected Survival by Subgroup Life expectancy by subgroup. The top row is for all candidates, the middle row is by race/ethnicity, and the bottom row is by age at listing. We see consistent benefits of transplantation across all subgroups.