Table 3.
Peak CD4 cell count | |||||
Characteristics | >500 cells/μL | 351–500 cells/μL | 200–350 cells/μL | <200 cells/μL | P value for trend |
Patients | 468 | 163 | 102 | 60 | – |
Hepatic events | 13 | 7 | 10 | 13 | – |
Person-years | 3529 | 910 | 492 | 142 | – |
Incidence per 1000 PYs (95% CI) | 3.7 (1.7–5.7) | 7.7 (2.0–13.4) | 20.3 (7.9–32.8) | 91.5 (44.1–139.0) | |
Multivariable aOR (95% CI), model 1 | 1.00 (Ref) | 1.57 (0.62–4.01) | 3.80 (1.62–8.94) | 9.68 (4.24–22.10) | <0.001 |
P-value of model 1 | 0.345 | 0.002 | <0.001 | ||
Multivariable aOR (95% CI), Model 2 | 1.00 (Ref) | 1.54 (0.60–3.95) | 2.92 (1.19–7.13) | 7.36 (3.09–17.52) | <0.001 |
P-value of model 2 | 0.368 | 0.019 | <0.001 | ||
Multivariable aOR (95% CI), Model 3 | 1.00 (Ref) | 1.28 (0.49–3.33) | 2.54 (1.01–6.38) | 5.50 (2.23–13.60) | <0.001 |
P-value of model 3 | 0.619 | 0.048 | <0.001 | ||
Multivariable aOR (95% CI), Model 4 | 1.00 (Ref) | 0.98 (0.35–2.74) | 2.25 (0.87–5.86) | 3.96 (1.51–10.40) | 0.004 |
P-value of model 4 | 0.974 | 0.095 | 0.005 |
Multivariate logistics regression models, backward Wald method was used. Model 1, Adjusted for age at enrollment and gender (male and female); Model 2, Model 1 plus adjustment for the baseline CD4 cell count and initial antiretroviral regimen (NRTIs plus NNRTIs, NRTIs plus PIs, other regimens/unknown); Model 3, Model 2 plus adjustment for the liver cirrhosis (with and without), platelet, hemoglobin, ALT, albumin, total bilirubin; Model 4, Model 3 plus adjustment for the HBsAg status (positive and negative), HCV RNA, and anti-HCV treatment (with and without).
The CI is determined by assuming that the ratio conforms to the normal distribution.
ALT: Alanine aminotransferase; CI: Confidence interval; HBsAg: Hepatitis B surface antigen; HCV: Hepatitis C virus; HIV: Human immunodeficiency virus; NRTIs: Nucleoside reverse transcriptase inhibitors; NNRTIs: Non-nucleoside reverse transcriptase inhibitors; PIs: Protease inhibitors; PYs: Person years; RNA: Ribonucleic acid; –: Not applicable.