TABLE 3.
Adjusted probability of underinsurance by sociodemographic characteristics, insured non‐elderly cancer survivors, 2011–2015
| Probability of underinsurance (%)a | p b | |
|---|---|---|
| Age (years) | ||
| 18–49 [ref.] | 6.21 | |
| 50–59 | 8.17 | 0.114 |
| 60–64 | 13.02 | <0.001 |
| Sex | ||
| Male [ref.] | 9.09 | |
| Female | 8.68 | 0.696 |
| Race/ethnicity | ||
| Non‐Hispanic White [ref.] | 9.8 | |
| Black | 4.57 | <0.001 |
| Hispanic | 7.79 | 0.117 |
| Asian/others | 6.21 | 0.047 |
| Education | ||
| HS education/diploma [ref.] | 7.46 | |
| Some college | 9.37 | 0.142 |
| College degree or above | 11.07 | 0.021 |
| Income levelc | ||
| Low income [ref.] | 45.36 | |
| Middle income | 5.77 | <0.001 |
| High income | 0.7 | <0.001 |
| Insurance status | ||
| Private MC [ref.] | 11.99 | |
| Private non‐MC | 11.86 | 0.941 |
| Medicaid | 3.65 | <0.001 |
| Medicare/dual‐eligible | 6.54 | 0.008 |
Abbreviations: AAP, average adjusted prediction; AME, average marginal effect; FI, family income; FPL, federal poverty level; HS, high school; MC, managed care; MEPS, Medical Expenditure Panel Survey.
AAP from a logistic model. Estimation model was adjusted for age, sex, race/ethnicity, marital status, income level, education, census region, insurance status, number of MEPS priority conditions, and self‐reported health status.
p‐values represent statistical significance of AME contrasting the AAP of each category to the AAP of the reference category (the first row) for each variable.
Low income represents FI < 200% of FPL, middle income represents FI 200% to <400% of FPL, and high income represents FI ≥ 400% of FPL.