Medicaid expansions and differences in guideline-adherent cervical cancer screening between American Indian and White women

Abstract

Background: Although preventable through screening, cervical cancer incidence and mortality are higher among American Indian and Alaska Native women (AIAN) than White women. The Patient Protection and Affordable Care Act's (ACA) Medicaid expansions may uniquely impact access and use of cervical cancer screening among AIAN women and ultimately alleviate this disparity.

Methods: Using Medicaid eligible AIAN (N = 4681) and White (N = 57,661) women aged 18-64 years from the 2010-2020 Behavioral Risk Factor Surveillance System, we implemented difference-in-differences regression to estimate the association between the Medicaid expansions and guideline-adherent cervical cancer screening and health care coverage.

Results: The Medicaid expansions were not associated with guideline-adherent cervical cancer screening (AIAN: -1 percentage point [ppt] [95% confidence interval, CI: -4, 2 ppts]; White: 3 ppts [95% CI: -0, 6 ppts]), but were associated with a 2 ppt increase (95% CI: 0, 4 ppt) in having had a pap test in the last 5 years among White women. The Medicaid expansions were also associated with increases in having a health plan (AIAN: 5 ppts [95% CI: 1, 9]; White: 11 ppts [95% CI: 7, 15]) and decreases in avoiding medical care due to costs (AIAN: -8 ppts [95% CI: -13, -2]; White: -6 ppts [95% CI: -9, -4]).

Conclusions: While we observed improvements in health care coverage, we did not observe changes to guideline-adherent cervical cancer screening following the ACA's Medicaid expansions. Given the disproportionate burden of cervical cancer among AIAN women, identifying ways to improve cervical cancer screening uptake and delivery should be prioritized to reduce preventable deaths.

Keywords: American Indians or Alaska Natives; Medicaid; Papanicolaou test; early detection of cancer; healthcare disparities.

FIGURE 1

Trends in guideline‐adherent cervical cancer screening among Medicaid eligible AIAN and White women by year, BRFSS 2010–2020. The figure shows the prevalence of guideline‐adherent cervical cancer screening by year among US residents ages 18–64 that are likely Medicaid eligible. AIAN women are represented by the circles and solid line and White women by the triangles and dashed line. BRFSS provided survey weights were applied to calculate prevalence values. The x‐axis is survey year and the y‐axis guideline‐adherent cervical cancer screening prevalence. AIAN, American Indian and Alaska Natives; BRFSS, Behavioral Risk Factor Surveillance System.

FIGURE 2

Difference‐in‐differences estimators by study outcome and race/ethnicity, Behavioral Risk Factor Surveillance System, 2010–2020. All DID models included main effects for expansion status and a post expansion indicator (and their interaction), were adjusted for state and year‐specific unemployment rates (%) as well as individual level age group, income, education, dependent children, marital status, working status and included state and year fixed effects. Standard errors were clustered by state. Interaction term beta coefficients (interaction between expansion status and post expansion timeframe) were multiplied by 100 and are indicated by black circles (non‐significant) or red squares (p < 0.05) and their 95% confidence intervals by horizontal lines. Negative numbers indicate that expansion states experienced, on average, a decrease in the outcome compared to non‐expansion states, whereas positive numbers indicate that expansion states, on average, experienced an increase in the outcome compared to non‐expansion states. Due to missingness in outcome values, sample sizes for each outcome differ. cc, cervical cancer; DID, difference‐in‐differences; yr, year.