Skip to main content
. Author manuscript; available in PMC: 2013 Oct 29.
Published in final edited form as: AIDS. 2012 Jan 28;26(3):10.1097/QAD.0b013e32834ed814. doi: 10.1097/QAD.0b013e32834ed814

Table 2.

Hazard ratios for mortality, loss to follow up and retention on ART.

Mortality Loss to follow up Retention on ART
Adjusted sHR (95% CI) p-value Adjusted sHR (95% CI) p-value Adjusted HR (95% CI) p-value
Age group (years)
 0–1 1.37 (1.17–1.60) 1.37 (1.18–1.59) 0.71 (0.64–0.79)
 2–5 0.81 (0.73–0.90) 0.84 (0.75–0.94) 1.25 (1.16–1.35)
 6–14 0.63 (0.57–0.70) 0.63 (0.56–0.70) 1.66 (1.54–1.79)
 15–24 1.17 (1.10–1.25) 1.27 (1.19–1.35) 0.81 (0.77–0.85)
 25–34 1 1 1
 35–44 0.94 (0.91–0.98) 0.81 (0.77–0.85) 1.15 (1.11–1.18)
 45– 0.99 (0.94–1.04) <0.001 0.79 (0.75–0.83) <0.001 1.13 (1.09–1.17) <0.001
Gender
 Female 1 1 1
 Male 1.39 (1.34–1.44) <0.001 1.21 (1.17–1.26) <0.001 0.75 (0.73–0.77) <0.001
WHO stage at start
 I or II 0.48 (0.45–0.52) 0.59 (0.55–0.63) 1.93 (1.84–2.02)
 III or IV 1 <0.001 1 <0.001 1 <0.001
Year of starting ART
 2004 1.65 (1.54–1.76) 0.56 (0.51–0.62) 0.96 (0.91–1.02)
 2005 1.57 (1.50–1.64) 0.88 (0.84–0.93) 0.82 (0.79–0.85)
 2006 1.34 (1.28–1.39) 0.90 (0.86–0.94) 0.90 (0.88–0.93)
 2007 1 <0.001 1 <0.001 1 <0.001
Region
 Northern 1.26 (1.20–1.32) 0.85 (0.80–0.90) 0.94 (0.90–0.97)
 Central 1.05 (1.01–1.10) 1.11 (1.07–1.16) 0.93 (0.90–0.95)
 Southern 1 <0.001 1 <0.001 1 <0.001
Degree of urbanization
 Urban 1 1 1
 Semiurban 1.03 (0.98–1.08) 1.01 (0.96–1.07) 0.99 (0.96–1.03)
 Rural 1.02 (0.98–1.07) 0.41 0.82 (0.78–0.86) <0.001 1.08 (1.04–1.11) <0.001
Health care facility
 Central hospital 0.69 (0.65–0.73) 1.30 (1.23–1.38) 1.07 (1.03–1.12)
 District hospital 1 1 1
 Rural hospital 0.86 (0.79–0.94) 1.11 (1.01–1.21) 1.05 (0.99–1.12)
 Health centre 0.87 (0.82–0.92) 0.71 (0.66–0.76) 1.28 (1.23–1.34)
 Others* 0.80 (0.76–0.85) <0.001 1.09 (1.03–1.15) <0.001 1.10 (1.05–1.14) <0.001

Mortality and loss to follow were analyzed according to the competing risk method. The hazard ratios (HR) for retention on ART are calculated as inverses of hazard ratios obtained by fitting a Cox regression model with failure defined as meeting any of the following endpoints: death, loss to follow up or stopping treatment.

All analyses were adjusted for all the variables shown in the table. sHR=subhazard ratio; HR=hazard ratio; CI=confidence interval

*

NGOs, armed forces, occupational clinics