Table 3.
Factors associated with long-term survival among 543 patients who were living one year after establishing initial outpatient HIV care at the UAB 1917 HIV/AIDS Clinic, 2000–2005.
| Primary Analysisa | ||
|---|---|---|
| Univariate HR (95% CI) |
Multivariable HR (95% CI)a |
|
| Age (per 10 years) | 1.61 (1.21–2.17) | 1.58 (1.12–2.22) |
| Female | 1.11 (0.56–2.20) | 1.20(0.49–2.96) |
| African American | 1.05 (0.57–1.93) | 0.73 (0.33–1.64) |
| HIV risk factor | ||
| Men who have sex with men (MSM) | 1.0 | 1.0 |
| Heterosexual | 1.23 (0.62–2.43) | 0.91 (0.38–2.20) |
| Intravenous drug use (IVDU) | 2.06 (0.81–5.23) | 1.05 (0.33–3.31) |
| Health insurance | ||
| Private | 1.0 | 1.0 |
| Public | 2.11 (1.03–4.32) | 1.30 (0.54–3.14) |
| Uninsured | 0.79 (0.37–1.68) | 0.74 (0.32–1.71) |
| Affective mental health disorder | 0.80 (0.44–1.47) | 0.80 (0.39–1.62) |
| Substance abuse | 1.80 (0.96–3.35) | 2.08 (0.86–5.07) |
| Alcohol abuse | 1.12 (0.53–2.35) | 0.75 (0.31–1.84) |
| Baseline CD4 count | ||
| ≥ 350 cells/mm3 | 1.0 | 1.0 |
| 200 – 350 cells/mm3 | 1.32 (0.48–3.65) | 1.49 (0.47– 4.67) |
| < 200 cells/mm3 | 2.00 (0.95–4.20) | 2.70 (1.00–7.30) |
| Baseline Viral Load (log10) | 1.11 (0.84–1.48) | 1.02 (0.75–1.39) |
| Missed visit in first year | 2.34 (1.15–4.77) | 2.90 (1.28– 6.56) |
| Antiretroviral therapy started in first year | 1.00 (0.51–1.93) | 0.64 (0.25–1.62) |
| Propensity Score Sensitivity Analysisb | ||
|
Univariate HR (95% CI) |
Multivariable HR (95% CI)b |
|
| Missed visit in first year | 2.34 (1.15–4.77) | 2.52 (1.15–5.52) |
| Antiretroviral therapy started in first year | 1.00 (0.51–1.93) | 1.18 (0.58–2.41) |
| Propensity score for missed visit (per 10%) | 1.05 (0.87–1.26) | 0.98 (0.80–1.19) |
Primary analysis; Cox proportional hazards model (multivariable HR) includes variables listed in the table (age, sex, race, HIV risk factor, health insurance, affective mental health disorder, substance abuse, alcohol abuse, baseline CD4 count, baseline viral load, missed visit status in the first year, and antiretroviral therapy receipt in first year).
Sensitivity analysis; Cox proportional hazards model includes missed visit status in the first year, propensity score for missed visit in first year (derived from multivariable logistic regression model in table 2; includes age, sex, race, HIV risk factor, health insurance, affective mental health disorder, substance abuse, alcohol abuse, baseline CD4 count, baseline viral load), and antiretroviral therapy receipt in the first year.