Statistical Behavior of the Rank-order and Tiering Metric of the 2020 Centers for Medicare and Medicaid Services Organ Procurement Organization Rule.

Abstract

Background: The 2020 Centers for Medicare and Medicaid Services (CMS) Organ Procurement Organization (OPO) rule used an estimate of OPO organ donor potential that included contraindications to organ donation, termed cause, age, and location consistent with donation (CALC). CMS asserted that even though CALC was inaccurate, it could still be used to compare OPOs because it highly correlated with CALC-adjusted to exclude the contraindications. Quantitative bias analysis is required to determine whether CALC can, in fact, be reliably used to make certification decisions under the rule.

Methods: A Monte Carlo simulation was carried out with 10 000 simulations using statistical constraints defining the relationship between CALC and CALC-adjusted applied to the published data in the rule.

Results: In every simulation, at least 2 OPOs changed tier assignment, with a median of 10 OPOs changing tiers across the 10 000 simulations. Several validations were carried out to confirm the results. All the tier changes were in OPOs that were closely adjacent to one of the cutoffs that determined tier assignment.

Conclusions: CALC, as applied in the 2020 OPO rule, possesses a level of stochastic instability that exceeds the threshold for accurate regulatory decisions. Its use will result in misclassification of a significant number of OPOs.