Table 3.
Subdistribution Hazard Ratios for Tracing Outcomes Death, Stop of Combination Antiretroviral Therapy (cART), Silent Transfer, and Retention on cART
| Characteristic | Death | Stop of cART | Silent Transfer | Retained on cART |
|---|---|---|---|---|
| SHR (95% CI) | SHR (95% CI) | SHR (95% CI) | SHR (95% CI) | |
| Sex | ||||
| Male | 1 | 1 | 1 | 1 |
| Female | 0.62 (.44–.88) | 0.97 (.87–1.07) | 1.32 (1.17–1.50) | 1.03 (.90–1.18) |
| CD4 counta (cells/µL) | ||||
| <50 | 1 | 1 | 1 | 1 |
| 50–99 | 0.75 (.64–.88) | 1.52 (1.21–1.90) | 1.13 (.86–1.49) | 0.88 (.66–1.19) |
| ≥100 | 0.40 (.35–.45) | 1.77 (1.47–2.13) | 1.17 (.93–1.47) | 1.47 (1.17–1.84) |
| WHO clinical stagea | ||||
| I–II | 1 | 1 | 1 | 1 |
| III | 1.23 (.93–1.63) | 0.77 (.64–.91) | 0.88 (.71–1.09) | 0.77 (.60–1.00) |
| IV | 1.57 (1.17–2.10) | 0.64 (.51–.80) | 0.71 (.56–.91) | 0.71 (.53–.95) |
| Last clinic visit | ||||
| Before 2009 | 1 | 1 | 1 | 1 |
| 2009 or later | 0.76 (.66–.87) | 1.04 (.92–1.18) | 1.56 (1.36–1.80) | 0.33 (.27–.40) |
| Ageb (y) | ||||
| <16 | 1 | 1 | 1 | 1 |
| 16–29 | 1.73 (.92–3.25) | 0.99 (.70–1.39) | 0.71 (.35–1.45) | 1.68 (1.00–2.82) |
| 30–39 | 2.21 (1.18–4.15) | 0.89 (.63–1.25) | 0.84 (.41–1.72) | 1.79 (1.04–3.08) |
| ≥40 | 2.84 (1.52–5.32) | 0.75 (.53–1.06) | 0.91 (.45–1.87) | 1.79 (1.04–3.09) |
| Time on cARTb (y) | ||||
| <1 | 1 | 1 | 1 | 1 |
| 1–2 | 0.51 (.44–.60) | 1.28 (1.13–1.45) | 1.02 (.87–1.20) | 1.89 (1.62–2.21) |
| ≥2 | 0.44 (.36–.53) | 1.14 (1.00–1.30) | 1.05 (.87–1.28) | 1.94 (1.60–2.37) |
| Interactiona | ||||
| Female clinical stage I–II | 1 | |||
| Female clinical stage III | 1.32 (.91–1.92) | |||
| Female clinical stage IV | 1.73 (1.17–2.56) | |||
Abbreviations: cART, combination antiretroviral therapy; CI, confidence interval; SHR, subdistribution hazard ratio; WHO, World Health Organization.
aAt cART initiation.
bAt last clinic visit. All models are stratified by cohort. Parameter estimates are pooled estimates from models fitted to 20 imputed datasets. Each model is fitted for the tracing outcome of interest, accounting for the alternative outcomes as competing risks.