Table 3.
Adjusted Odds Ratios of Non-Retention in HIV Care Among Miami-Dade County Ryan White Program Clients, 2017, Models Stratified for Women and Men.
| Women (N = 1606) | Men (N = 5311) | |||
|---|---|---|---|---|
| Adjusted odds ratio | 95% confidence interval | Adjusted odds ratio | 95% confidence interval | |
| Demographic characteristics | ||||
| Age group | ||||
| 18–34 | 1.98 | (1.27–3.09) | 1.86 | (1.50–2.32) |
| 35–49 | 1.35 | (0.95–1.93) | 1.42 | (1.18–1.72) |
| ≥50 | Ref | Ref | ||
| Race/Ethnicity | ||||
| Non-Hispanic Black | 0.81 | (0.39–1.67) | 1.38 | (0.99–1.91) |
| Hispanic | 1.17 | (0.53–2.60) | 0.99 | (0.72–1.38) |
| Haitian | 1.01 | (0.43–2.38) | 1.22 | (0.78–1.92) |
| Non-Hispanic White/Other | Ref | Ref | ||
| US-borna | ||||
| Yes | 2.11 | (1.30–3.43) | 1.34 | (1.06–1.70) |
| No | Ref | Ref | ||
| Household income, percent of Federal Poverty Level | ||||
| 200%- >300% | Ref | Ref | ||
| 100%–199% | 0.70 | (0.43–1.13) | 0.93 | (0.75–1.16) |
| <100% | 0.71 | (0.43–1.16) | 1.04 | (0.82–1.31) |
| Number of minors in household | ||||
| None | Ref | Ref | ||
| 1 | 1.13 | (0.73–1.73) | 0.96 | (0.62–1.49) |
| 2 | 1.34 | (0.80–2.26) | 0.88 | (0.43–1.81) |
| 3 or more | 2.03 | (1.10–3.76) | 0.56 | (0.16–2.00) |
| Need characteristics | ||||
| Diagnosis of AIDS at any time | ||||
| Yes | 1.20 | (0.89–1.61) | 0.98 | (0.83–1.16) |
| No | Ref | Ref | ||
| Has HIV-related symptoms at time of assessment | ||||
| Yes | Ref | Ref | ||
| No | 0.67 | (0.30–1.47) | 0.54 | (0.34–0.86) |
| Vulnerable/enabling variables | ||||
| Drug use in the last 12 months | ||||
| Yes | 1.18 | (0.42–3.33) | 1.22 | (0.78–1.90) |
| No | Ref | Ref | ||
| Drug use resulted in problems with daily activities or legal issue or hazardous situation | ||||
| Yes | 1.39 | (0.48–4.05) | 1.27 | (0.75–2.13) |
| No | Ref | Ref | ||
| Drug use affects adherence | ||||
| Yes | 0.77 | (0.28–2.13) | 0.96 | (0.59–1.54) |
| No | Ref | Ref | ||
| Would like substance use treatment now | ||||
| Yes | 2.25 | (0.91–5.54) | 1.57 | (0.89–2.76) |
| No | Ref | Ref | ||
| Feeling depressed or anxious | ||||
| Yes | 0.90 | (0.58–1.39) | 1.18 | (0.94–1.49) |
| No | Ref | Ref | ||
| Receives or needs mental health services | ||||
| Yes | 0.92 | (0.60–1.41) | 0.93 | (0.74–1.17) |
| No | Ref | Ref | ||
| Ever experienced domestic violence | ||||
| Yes | 0.71 | (0.40–1.24) | 0.97 | (0.64–1.46) |
| No | Ref | Ref | ||
| Disclosure of HIV status to partner or an adult in household | ||||
| Does not have a partner or adult in household | 1.34 | (0.97–1.84) | 1.13 | (0.95–1.34) |
| At least 1 adult in household or partner, but partner and adult do not know status | 1.03 | (0.63–1.67) | 0.98 | (0.73–1.33) |
| Adult in household or a partner knows status | Ref | Ref | ||
| Has a social support system to depend on | ||||
| No | 1.13 | (0.76–1.68) | 1.13 | (0.93–1.38) |
| Yes | Ref | Ref | ||
| Work related barriers to attending care appointments | ||||
| Not working | 1.72 | (1.21–2.43) | 1.80 | (1.48–2.20) |
| Yes | 1.13 | (0.45–2.85) | 1.35 | (0.77–2.36) |
| No | Ref | Ref | ||
| Client has access to transportation to appointments | ||||
| No | 1.19 | (0.71–1.99) | 1.24 | (0.96–1.58) |
| Yes | Ref | Ref | ||
| Client getting food he/she needs | ||||
| No | 1.53 | (0.51–4.64) | 1.98 | (1.24–3.14) |
| Yes | Ref | Ref | ||
| Homeless | ||||
| Yes | 1.30 | (0.67–2.53) | 1.32 | (0.98–1.78) |
| No | Ref | Ref | ||
| Infected perinatally with HIV | ||||
| Yes | 3.04 | (1.16–7.93) | 0.89 | (0.35–2.26) |
| No | Ref | Ref | ||
| Health care environment | ||||
| Number of Ryan White clients that client’s clinician cares for | ||||
| 1–9 | 1.91 | (1.03–3.56) | 2.89 | (1.93–4.30) |
| 10–29 | 0.63 | (0.27–1.46) | 1.57 | (1.06–2.33) |
| 30–99 | 1.10 | (0.74–1.65) | 1.02 | (0.80–1.29) |
| 100–199 | 0.66 | (0.44–1.00) | 1.24 | (0.99–1.54) |
| ≥200 | Ref | Ref | ||
| Unknownb | 1.41 | (0.79–2.51) | 2.08 | (1.52–2.85) |
| Neighborhood environment | ||||
| Neighborhood low socioeconomic status (SES) Index (higher score indicates lower SES) c | 1.00 | (0.82–1.23) | 0.94 | (0.85–1.04) |
| Neighborhood residential instability/homicide indexd | 1.12 | (0.92–1.36) | 0.98 | (0.90–1.07) |
The intercluster correlation coefficients were 0.047 for the same ZIP code and medical case management site, 0.016 for the same ZIP code and different medical case management site and 0.032 for a different ZIP code and same medical case management site.
Bold indicates statistically significant P < 0.05.
Results exclude19 men and 3 women with missing values from one or more of the following variables: feeling depressed or anxious, receives or needs mental health services, experienced domestic violence social support needs, and neighborhood indices.
a Those born in Puerto Rico or other US territories are classified as non-US born.
b Clients who could not name HIV care provider during health assessment or patient intake.
c Higher score indicates lower SES.
d Higher score indicates more instability and homicides.