Table 2. Predictors of Safety-Net Use among Non-Medicaid Insured Adults.
Odds Ratio (95% CI) | ||
---|---|---|
Unadjusted | Adjusted* | |
Age, years† | ||
18–64 | — | 2.44 (1.80–3.31)‡, § |
≥ 65 | — | [Reference] |
Female | 0.93 (0.80–1.07) | 0.89 (0.73–1.07) |
Race/ethnicity | ||
White, non-Hispanic | [Reference] | [Reference] |
Black, non-Hispanic | 2.39 (1.53–3.73)§ | 1.97 (1.06–3.66)‖ |
Hispanic | 3.46 (2.27–5.29)§ | 2.28 (1.32–3.93)‖ |
Other | 1.54 (0.88–2.69) | 1.46 (0.56–3.80) |
Dual eligible | 5.54 (3.18–9.68)§ | 2.13 (1.38–3.30)§ |
Rural area | 1.46 (0.68–3.11) | 1.09 (0.47–2.52) |
Geographic region | ||
Midwest | [Reference] | [Reference] |
Northeast | 3.03 (1.27–7.28)‖ | 5.87 (2.48–13.93)§ |
South | 1.58 (0.63–3.95) | 1.69 (0.70–4.10) |
West | 2.50 (1.13–5.54)‖ | 4.23 (1.83–9.76)§ |
Prevalence of poverty in zip code | ||
Lowest (<5%) | [Reference] | [Reference] |
Low (5–9.9%) | 3.39 (2.23–5.15)§ | 2.50 (1.34–4.68)‖ |
Moderate (10.0–19.9%) | 8.53 (5.19–14.02)§ | 5.34 (2.78–10.25)§ |
High (≥ 20%) | 23.93 (13.48–42.48)§ | 9.53 (4.65–19.53)§ |
Prevalence of bachelor’s degree in zip code | ||
Lowest (<12.8%) | 7.21 (4.18–12.44)§ | 1.90 (0.96–3.78) |
Low (12.8–19.7%) | 3.31 (1.97–5.55)§ | 1.12 (0.60–2.09) |
Moderate (19.8–31.7%) | 2.16 (1.56–2.98)§ | 0.79 (0.50–1.27) |
High (>31.7%) | [Reference] | [Reference] |
Frequent visits (≥5) in past 12 months | 2.15 (1.54–3.01)§ | 1.79 (1.14–2.81)‖ |
Polypharmacy (≥4 medications) | 1.85 (1.28–2.68) ‖ | 1.64 (1.08–2.49)‖ |
Multimorbidity (≥2 chronic conditions) | 1.69 (1.30–2.19)§ | 0.83 (0.59–1.16) |
Chronic conditions | ||
Diabetes | 1.74 (1.41–2.15)§ | 0.97 (0.76–1.23) |
Hypertension | 1.73 (1.34–2.24)§ | 1.40 (1.05–1.85)‖ |
Chronic obstructive pulmonary disease | 1.37 (1.05–1.78)‖ | 1.16 (0.85–1.58) |
* Adjusted for all characteristics listed in Table 1 and accounting for the complex survey sampling design.
† For the unadjusted analysis, we evaluated age stratified by primary payer (see Methods for further details). For individuals with Medicare, age 18–64 was associated with a 3.40 higher odds of safety-net use (95% CI 2.63–4.41) compared to age ≥ 65 years. For privately insured individuals, age 18–64 years was associated with a 2.12 higher odds of safety-net use (95% CI 1.53–2.93) compared to age ≥ 65 years.
‡ The association of age with safety-net use for all insured individuals. We omitted the interaction term for age and payer from the adjusted model, given it did not meaningfully change our findings (see Methods for further details).
§ p≤0.001.
‖ p≤0.05.