Differences in Insurance Coverage for Individuals With Schizophrenia After Implementation of the Patient Protection and Affordable Care Act

Abstract

This cross-sectional study uses Medical Expenditure Panel Survey data to assess differences in insurance coverage for individuals with schizophrenia before vs after implementation of the Patient Protection and Affordable Care Act (ACA).

Schizophrenia impacts 0.25% to 0.64% of people in the US and is associated with comorbid diseases, reduced life expectancy, and substantial health care costs.¹ Health insurance is critically important to ensure adequate health care access for ongoing treatment of psychiatric and medical needs.

Research before implementation of the Patient Protection and Affordable Care Act (ACA) showed that although people with schizophrenia had lower uninsurance rates than the general population, 12% annually were uninsured, with most covered by Medicaid and/or Medicare.² Studies conducted after the ACA have shown high rates of transitions between insurance types for young adults with schizophrenia.³ ACA provisions that may impact coverage for those with schizophrenia include Medicaid expansion, creation of health insurance Marketplaces, individual insurance mandate, access to parental insurance for young adults, behavioral health services as an essential health benefit, and eliminating exclusions for preexisting conditions.⁴

The national uninsurance rate for individuals younger than 65 years decreased after ACA implementation from 16.6% in 2010 to 11.0% in 2021,⁵ but it is not known whether this has changed for people with schizophrenia. Using nationally representative data, this cross-sectional study assessed insurance coverage and sources for people with schizophrenia, including differences between the pre-ACA (2008-2013) and post-ACA (2014-2020) periods.

Methods

Using the Medical Expenditure Panel Survey (MEPS) from 2008 to 2020,⁶ we identified people with schizophrenia using ICD-9 code 295, ICD-10 code F20, and clinical classification (revised) codes, limited to those with an associated health care event or medication in each 2-year panel and complete data on insurance coverage and covariates. We calculated descriptive statistics for the sample using χ² tests for differences in pre-ACA vs post-ACA periods. We then calculated unadjusted and adjusted probabilities of insurance coverage using logistic regression with the primary independent variable of pre-ACA vs post-ACA, accounting for respondent-reported sex, age category, marital status, ethnicity, race, education level, and region. Two-sided P = .05 was considered significant. Population weights were used to make the sample representative of the US civilian, noninstitutionalized population and SEs corrected for sampling design. The study was determined to be not human participant research by the University of Massachusetts Amherst Human Research Protection Office. The study followed the STROBE reporting guideline.

Results

The weighted analytic sample (2008-2020) included 9 173 644 individuals (unweighted n = 1035) with schizophrenia (original sample n = 1149 before exclusion criteria). Of these, most were male (57.4%), 41 to 55 years of age (39.9%; mean [SE] age, 44.1 [0.7] years), and White (68.2%) (Table). Comparing the pre-ACA and post-ACA periods, there were significant differences in marital status but not for other characteristics. Regression-adjusted analyses showed a significant reduction in the percentage who were uninsured after the ACA vs before the ACA (8.4% vs 4.0%; difference, −4.5 percentage points [pp]; 95% CI, −8.3 to −0.6 pp) (Figure). There was a corresponding increase in coverage by Medicaid (difference, 8.7 pp; 95% CI, 0.3-17.1 pp) and Medicare (difference, 5.0 pp; 95% CI, −4.3 to 14.3 pp; not significant). Unadjusted analyses were similar, but differences were not statistically significant.

Table. Descriptive Statistics of Individuals With Schizophrenia From 2008 to 2020^a.

Characteristic	Individuals, %			P valueb
	Overall (N = 9 173 644)	Period
		Pre-ACA (n = 4 721 714)	Post-ACA (n = 4 451 929)
Sex
Female	42.6	37.5	47.9	.07
Male	57.4	62.5	52.1
Age category, y
18-25	10.5	10.5	10.5	.16
26-40	27.7	31.8	23.4
41-55	39.9	39.9	39.8
56-64	22.0	17.8	26.3
Hispanic ethnicityc	14.2	14.1	14.2	.97
Racec
Black/African American	23.7	21.9	25.6	.17
White	68.2	72.0	64.2
Otherd	8.2	6.2	10.3
Married	15.8	10.8	21.1	.03
Education level
Less than high school degree	24.3	23.7	24.9	.72
High school diploma or equivalent	36.1	38.2	33.9
Some college including AA degree	23.8	24.3	23.3
Bachelor’s degree or higher	15.8	13.8	17.9
Geographic region
Northeast	20.2	20.7	19.6	.47
Midwest	20.2	19.1	21.4
South	37.0	34.4	39.7
West	22.6	25.8	19.3

^a The pre-ACA period was from 2008 to 2013, and the post-ACA period was from 2014 to 2020. Population weights were used to make the sample representative of the US civilian, noninstitutionalized population. N values shown are weighted. Unweighted values were 1035 for the overall sample, 528 for the pre-ACA period, and 507 for the post-ACA period.

^b Calculated using χ² tests.

^c Ascertained by respondent self-report.

^d Includes American Indian, Asian, multiple races, and other.

Figure. Regression-Adjusted Probability of Insurance Coverage Types for Individuals With Schizophrenia by Period From 2008 to 2020.

Differences were compared using average marginal effects calculated after multivariable logistic regression accounting for sex, age category, marital status, ethnicity, race, education level, and geographic region. Insurance coverage categories are not mutually exclusive. The pre-ACA period was from 2008 to 2013, and the post-ACA period was from 2014 to 2020. CHAMPVA indicates Civilian Health and Medical Program of the Department of Veterans Affairs.

Discussion

Following implementation of ACA provisions, the percentage of individuals with schizophrenia who were uninsured declined, accompanied by a rise in Medicaid coverage. Of people with schizophrenia, 4.0% remained uninsured after the ACA.

This study was limited as we could not measure whether expanded coverage changed the clinical characteristics of those with Medicaid coverage or specific contributors to continued uninsurance. MEPS may underrepresent those with serious mental illness, although schizophrenia prevalence estimates derived from MEPS are consistent with other estimates.¹ This study found a reduction in uninsurance rates among those with schizophrenia after the ACA, but continued attention is warranted to ensure adequate health care access for those with and without insurance.

Supplement.

Data Sharing Statement