Table 3.
Predictors of loss to follow-up, mortality and viral load rebound by down-referral status and amongst HIV treatment patients in Johannesburg, South Africa (n=2,772)
| No. Events (%) |
Person Time (years) |
Rate (per 100 person yrs.) |
Adjusted HR† (95% CI) |
No. Events (%) |
Person Time (years) |
Rate (per 100 person yrs.) |
Adjusted HR† (95% CI) |
No. Events (%) |
Adjusted RR§≠ (95% CI) |
|
|---|---|---|---|---|---|---|---|---|---|---|
|
| ||||||||||
| Mortality£ | Loss to follow-up¥ | Viral load rebound€ | ||||||||
| Down referral status | ||||||||||
| Not down-referred | 32 (1.5) | 2025.6 | 1.6 | Reference | 87 (4.2) | 2025.6 | 4.3 | Reference | 100 (5.6) | Reference |
| Down-referred | 2 (0.3) | 687.7 | 0.3 | 0.2 (0.04-0.8) | 10 (1.4) | 687.7 | 1.5 | 0.3 (0.2-0.6) | 22 (3.3) | 0.6 (0.4-0.9) |
| Gender | ||||||||||
| Female | 21 (1.2) | 1787.6 | 1.2 | Reference | 69 (3.8) | 1787.6 | 3.9 | Reference | 74 (4.6) | Reference |
| Male | 13 (1.4) | 931.8 | 1.4 | 1.1 (0.5-2.3) | 28 (2.9) | 931.8 | 3.0 | 0.8 (0.5-1.3) | 48 (5.7) | 1.3 (0.9-1.9) |
| Age (years) | ||||||||||
| ≥50 | 6 (2.8) | 209.6 | 2.9 | Reference | 5 (2.3) | 209.6 | 2.4 | Reference | 5 (2.6) | Reference |
| 40-49.9 | 10 (1.6) | 613.9 | 1.6 | 0.6 (0.2-1.6) | 14 (2.2) | 613.9 | 2.3 | 0.9 (0.3-2.3) | 25 (4.6) | 1.9 (0.7-4.9) |
| 30-39.9 | 11 (0.8) | 1306.6 | 0.8 | 0.3 (0.1-0.9) | 43 (3.2) | 1306.6 | 3.3 | 1.3 (0.5-3.0) | 57 (4.7) | 1.9 (0.8-4.7) |
| 25-29.9 | 6 (1.4) | 429.5 | 1.4 | 0.6 (0.2-1.8) | 22 (5.0) | 429.5 | 5.1 | 1.9 (0.8-4.7) | 27 (7.2) | 3.1 (1.2-8.0) |
| 18-24.9 | 1 (0.7) | 148.2 | 0.7 | 0.3 (0.04-3.0) | 12 (7.8) | 148.2 | 8.1 | 3.1 (1.1-8.6) | 8 (6.0) | 2.6 (0.8-8.2) |
| CD4 count | ||||||||||
| ≥500 cells/mm3 | 8 (1.1) | 708.4 | 1.1 | Reference | 19 (2.6) | 708.4 | 2.7 | Reference | 24 (3.7) | Reference |
| 400-499 cells/mm3 | 4 (0.7) | 557.4 | 0.7 | 0.7 (0.2-2.2) | 21 (3.7) | 557.4 | 3.8 | 1.4 (0.7-2.5) | 25 (5.0) | 1.3 (0.7-2.2) |
| 300-399 cells/mm3 | 12 (1.3) | 875.3 | 1.4 | 1.1 (0.5-2.8) | 35 (3.9) | 875.3 | 4.0 | 1.6 (0.9-2.8) | 38 (4.8) | 1.3 (0.8-2.1) |
| 200-299 cells/mm3 | 10 (1.7) | 572.2 | 1.7 | 1.5 (0.6-3.8) | 22 (3.8) | 572.2 | 3.8 | 1.5 (0.8-2.9) | 35 (6.8) | 1.7 (1.0-2.9) |
| AZT- vs. d4T-based regimen at down-referral eligibility | ||||||||||
| AZT | 11 (1.4) | 801.9 | 1.4 | Reference | 21 (2.6) | 801.9 | 2.6 | Reference | 26 (3.6) | Reference |
| d4T | 23 (1.2) | 1911.5 | 1.2 | 0.8 (0.4-1.8) | 76 (3.9) | 1911.5 | 4.0 | 1.4 (0.8-2.2) | 96 (5.6) | 1.4 (0.9-2.1) |
| NVP- vs. EFV-based regimen at down-referral eligibility | ||||||||||
| EFV | 32 (1.3) | 2485.5 | 1.3 | Reference | 89 (3.5) | 2485.5 | 3.6 | Reference | 110 (4.9) | Reference |
| NVP | 2 (0.9) | 219.9 | 1.0 | 0.9 (0.2-3.9) | 8 (3.6) | 219.9 | 3.6 | 0.8 (0.4-1.7) | 12 (6.0) | 1.2 (0.7-2.2) |
| Valid South African national identification number | ||||||||||
| No | 6 (0.8) | 720.0 | 0.8 | Reference | 34 (4.6) | 720.0 | 4.7 | Reference | 36 (5.6) | Reference |
| Yes | 28 (1.4) | 1993.3 | 1.4 | 1.6 (0.7-3.9) | 63 (3.1) | 1993.3 | 3.2 | 0.7 (0.5-1.1) | 86 (4.8) | 0.9 (0.6-1.3) |
Hazard ratios (HR) are from a Cox proportional regression models
Relative risks (RR) are from log-linear regression models
Loss to follow-up was defined as >4 months since last visit
Deaths are reported to the clinic by the hospital or the patients family notifying the clinic staff, the counselors identify the death through active tracing or through linkage with the South African National Vital Registration Infrastructure Initiative for those loss to follow-up
Viral load rebound was defined as having an unsuppressed viral load at 12-months after being suppressed at down-referral eligibility
All variables adjusted for in the models are listed in the table
CI, confidence interval; AZT, zidovudine; d4T, stavudine; NVP, nevirapine; EFV, lamivudine