Medicaid Expansion and Mortality Among Persons Who Were Formerly Incarcerated

Abstract

Importance: Since 2014, Medicaid expansion has been implemented in many states across the US, increasing health care access among vulnerable populations, including formerly incarcerated people who experience higher mortality rates than the general population.

Objective: To examine population-level association of Medicaid expansion with postrelease mortality from all causes, unintentional drug overdoses, opioid overdoses, polydrug overdoses, suicides, and homicides among formerly incarcerated people in Rhode Island (RI), which expanded Medicaid, compared with North Carolina (NC), which did not expand Medicaid during the study period.

Design, setting, and participants: A cohort study was conducted using incarceration release data from January 1, 2009, to December 31, 2018, linked to death records from January 1, 2009, to December 31, 2019, on individuals released from incarceration in RI and NC. Data analysis was performed from August 20, 2022, to February 15, 2024. Participants included those aged 18 years or older who were released from incarceration. Individuals who were temporarily held during ongoing judicial proceedings, died during incarceration, or not released from incarceration during the study period were excluded.

Exposure: Full Medicaid expansion in RI effective January 1, 2014.

Main outcomes and measures: Mortality from all causes, unintentional drug overdoses, unintentional opioid and polydrug overdoses, suicides, and homicides.

Results: Between 2009 and 2018, 17 824 individuals were released from RI prisons (mean [SD] age, 38.39 [10.85] years; 31 512 [89.1%] male) and 160 861 were released from NC prisons (mean [SD] age, 38.28 [10.84] years; 209 021 [87.5%] male). Compared with NC, people who were formerly incarcerated in RI experienced a sustained decrease of 72 per 100 000 person-years (95% CI, -108 to -36 per 100 000 person-years) in all-cause mortality per quarter after Medicaid expansion. Similar decreases were observed in RI in drug overdose deaths (-172 per 100 000 person-years per 6 months; 95% CI, -226 to -117 per 100 000 person-years), including opioid and polydrug overdoses, and homicide deaths (-23 per 100 000 person-years per year; 95% CI, -50 to 4 per 100 000 person-years) after Medicaid expansion. Suicide mortality did not change after Medicaid expansion. After Medicaid expansion in RI, non-Hispanic White individuals experienced 3 times greater sustained decreases in all-cause mortality than all racially minoritized individuals combined, while non-Hispanic Black individuals did not experience any substantial benefits. There was no modification by sex. Individuals aged 30 years or older experienced greater all-cause mortality reduction after Medicaid expansion than those younger than 30 years.

Conclusions and relevance: Medicaid expansion in RI was associated with a decrease in all-cause, overdose, and homicide mortality among formerly incarcerated people. However, these decreases were most observed among White individuals, while racially minoritized individuals received little to no benefits in the studied outcomes.

Figure 1.. Association of Medicaid Expansion Implementation in 2014 in Rhode Island With All-Cause and Cause-Specific Mortality Compared With North Carolina, 2009-2019

Shaded areas indicate 95% CIs.

Figure 2.. Association of Medicaid Expansion Implementation in 2014 in Rhode Island With Unintentional Opioid and Polydrug Overdose Mortality Compared With North Carolina, 2009-2018

Shaded areas indicate 95% CIs.