Medicaid Expansion and Optimal Starts for Incident Kidney Failure

Key Takeaways

• Optimal starts continued to rise in expansion states after 2014, while the adjusted pattern declined in nonexpansion states.
• Most of the widening gap was associated with stronger growth in home dialysis starts, whereas transplant and in-center hemodialysis with arteriovenous access did not show a similar trend shift.
• Expansion states also showed rising Medicaid coverage, falling uninsured rates, and improving pre-start nephrology care over time, although nephrology trends did not accelerate after 2014.

After 2014, optimal treatment initiation began to diverge between Medicaid-expansion and nonexpansion states, and by 2019 expansion states had a 3.9% higher adjusted percentage of patients with optimal starts. The analysis followed adults aged 26 to 64 years who were starting treatment for incident kidney failure across the United States. Optimal starts continued to rise in expansion states after Medicaid eligibility broadened, while the adjusted pattern declined in nonexpansion states. The widening gap reflected diverging treatment-initiation trends over time.

This observational cohort study used yearly US Renal Data System cohorts from 2008 through 2019 and included patients aged 26 to 64 years with incident treated kidney failure. Exposure was defined by residence in states that did or did not expand Medicaid on January 1, 2014. An optimal start meant preemptive kidney transplant, home dialysis initiation, or in-center hemodialysis begun with arteriovenous access, and interrupted time-series analysis estimated adjusted associations. The cohorts included 323,807 patients from expansion states and 271,725 from nonexpansion states; DC and 24 states expanded, 17 did not, and nine intermediate states were excluded. The design compared changes in pre-2014 and post-2014 trends rather than a single cross-sectional contrast.

Before expansion, adjusted optimal-start percentages were rising at similar rates in both state groups. In the postexpansion period, the adjusted percentage difference increased by 0.55% per year in expansion versus nonexpansion states, and the trend change was significant at P = 0.02. Most of the divergence was linked to home dialysis, which rose 0.29% per year more rapidly in expansion states after 2014. Preemptive transplant remained higher in expansion states throughout, and in-center hemodialysis with arteriovenous access also increased relatively, but neither showed a comparable post-2014 trend change. Home dialysis accounted for most of the growing overall separation between the two groups.

Coverage patterns moved in parallel, with Medicaid coverage increasing 1.97% more per year and uninsured rates falling 1.04% more per year in expansion states. By 2019, adjusted Medicaid coverage reached 46.7% in expansion states versus 23.2% in nonexpansion states. Pre-start nephrology care increased more in expansion states across the study period, although that advantage did not accelerate specifically after 2014. A separate incidence analysis did not support the idea that improved optimal starts simply reflected higher overall treatment incidence. These insurance and access shifts tracked with the broader pattern of treatment initiation without establishing a direct causal pathway.

The investigators noted that unmeasured confounding from state-level factors other than Medicaid expansion could have influenced the observed differences. They also could not exclude other state interventions, and they did not examine outcomes after treatment for kidney failure began. Investigators reported consistent results when 2014 alone or 2014 to 2015 were treated as transition periods. Follow-up ended in 2019 to avoid pandemic-era disruptions, so the association was observed only among adults aged 26 to 64 years starting treatment before COVID-19.