Presentation (CTI)

OBJECTIVES: Chronic kidney disease (CKD) is a progressive condition affecting an estimated 37 million Americans, yet estimates suggest roughly 87% of individuals with CKD are unaware of their condition. Given the large population of untreated patients and limited data on early-stage progression, we assess whether published lifetime cohort simulations in untreated CKD are replicable using a discrete event simulation model. The potential economic impact of expanded diagnosis in the Medicaid population was estimated.
METHODS: A six health-state model was constructed representing each CKD stage 1-5 and death. Parameters of the model included CKD-progression rates and mortality for untreated CKD. A targeted literature review coupled with backward searching was conducted for each parameter. The model was implemented as an individual-level, discrete-time, finite-state Markov simulation with annual cycles. Disease trajectories for 10,000 individuals beginning in CKD1 through death were simulated. Each annual cycle permits patients to remain in the current state, progress one stage, or die. To conservatively estimate the potential impact of expanded diagnosis, we evaluated Washington, the state with the highest per capita Medicaid expenditures. The incremental cost associated with diagnosing the additional 87% of the CKD population was calculated as the difference between the published cost of CKD and the per capita Medicaid expenditure.
RESULTS: This mirco-simulation yields an average survival time of 23.83 years with the majority of time spent in CKD1 (11.25 years) and CKD2 (6.24 years). This estimate is in the range of 28.2 years of survival reported in the literature of total survival. Additionally, we estimate a lower bound for incremental cost associated with diagnosing Medicaid patients in Washington to be approximately $275 million.
CONCLUSIONS: The DES-model reproduced survival estimates similar to what was reported in the literature, in untreated CKD progression, and the economic impact of diagnosing all Medicaid patients is significant.