Skip to main content
. Author manuscript; available in PMC: 2016 Jan 25.
Published in final edited form as: Trop Med Int Health. 2014 Sep 17;19(12):1397–1410. doi: 10.1111/tmi.12386

Table 6.

Final multivariate model for predictors of attrition in multicountry retention study*

Adjusted hazard
ratio (95% CI)
Individual characteristics
  Age at start ART: <30 years (vs. ≥30 years) 1.30 (1.14, 1.47)
  WHO stage at start ART
    III vs. I & II 1.12 (0.95, 1.31)
    IV vs. I & II 1.56 (1.29, 1.88)
  Weight loss >10% 1.17 (1.00, 1.38)
  CD4 count (/log [tenfold] increase) 0.88 (0.78, 1.00)
  Functional status
    Ambulatory vs. working/active 1.29 (1.09, 1.54)
    Bedridden vs. working/active 1.54 (1.15, 2.07)
  Sex (men vs. women)
    In sites without community 1.33 (1.18, 1.50)
      ARV drug dispensing
    In sites with community 0.95 (0.67, 1.33)
      ARV drug dispensing
  Years of clinic operation by Type of health facility (/year)
    Government facility 1.17 (1.10, 1.23)
    Mission facility/non-religious NGO 1.03 (0.95, 1.11)
Programme characteristics
  ARV drug dispensing in community by Sex
    For women 0.55 (0.30, 1.01)
    For men 0.40 (0.21, 0.75)
  Type of health facility: Faith-based
facility or NGO vs. Government
(at start of the programme)
0.71 (0.42, 1.21)

ART, antiretroviral treatment; ARV, antiretroviral; CI, confidence interval; NGO: non-governmental organisation.

Working/active: able to perform usual work in or out of the house; ambulatory: able to perform activities of daily living but not able to work; bedridden: not able to perform activities of daily living.

*

The model was simplified using Akaike’s Information Criterion, retaining predictors and clinically plausible interaction terms that increased model fit, with a penalisation for increasing model complexity.