Skip to main content

Table 3 Unadjusted and adjusted Cox regression models for baseline factors associated with time to death

From: Mortality in an antiretroviral therapy programme in Jinja, south-east Uganda: a prospective cohort study

Factor Level Unadjusted HR (95% c.i.) Adjusted HR (95% c.i.) Adjusted P-value
Sex Male 1 (ref. level) 1 (ref. level) 0.002
  Female 0.73 (0.52; 1.02) 0.59 (0.44; 0.80)  
Study Arm Facility 1 (ref. level) 1 (ref. level)  
  Home-based 1.003 (0.77; 1.31) 0.92 (0.70; 1.20) 0.56
Age Per 10 year increase 1.002 (0.84; 1.20) 1.09 (0.91; 1.31) 0.30
WHO stage I 0.58 (0.08; 4.0) 0.72 (0.12; 4.42) 0.009
  II 1 (ref. level) 1 (ref. level)  
  III 1.99 (1.57; 2.53) 1.41 (1.09; 1.83)  
  IV 3.64 (2.53; 5.23) 2.23 (1.45; 3.43)  
CD4/25 Per transformed CD4 increase * 0.52 (0.44; 0.61) 0.61 (0.51; 0.73) < 0.001
Log10 plasma viral load Per 1 log increase 1.15 (1.004; 1.31) 1.04 (0.91; 1.18) 0.51
Weight Per 5 kg increase 0.74 (0.67; 0.82) 0.77 (0.70; 0.85) < 0.001
Haemoglobin Per 2 g/dl increase 0.82 (0.72; 0.93) 0.93 (0.81; 1.07) 0.35
Cotrimoxazole prophylaxis Yes
No
1 (ref. level)
2.03 (1.08; 3.81)
1 (ref. level)
2.18 (1.10; 4.34)
0.02
  1. * Because of the CD4 count transformation, per unit increases here means an increase in CD4 count from 0 to 25, or from 25 to 100, or from 100 to 225 etc;