Skip to main content

Table 4 The crude and adjusted estimate of cox regression model fitting IPT and ART treatment status as time dependent covariates for TB or death in HIV cohort of Dilla University hospital from January 2007 to August 2013

From: Effect of isoniazid preventive therapy on tuberculosis or death in persons with HIV: a retrospective cohort study

Covariates Crude HR 95 % CI  Adjusted HRa 95% CI
Combination therapy On ART, no IPT 1 1
No ART, no IPT 1.08 (0.79, 1.47) 1.36 (0.97, 1.91)
No ART, treated with IPT 1.22 (0.45, 3.28) 1.86 (0.68, 5.10)
On ART, treated with IPT 0.35 (0.16, 0.77) 0.40 (0.18, 0.87)
Covariates: Crude HR 95 % CI Adjusted HRa 95% CI
Gender (male) 1.44 (1.12, 1.83) 1.56 (1.18, 2.06)
Age (years)b 1.01 (1.00, 1.03) 1.01 (0.99, 1.02)
Baseline body weight (Kg)c 0.98 (0.97, 0.99) 0.99 (0.97, 1.00)
Baseline CD4 count (Cells/mm3)d 1.00 (1.00, 1.01) 1.00 (0.999, 1.001)
Cotrimoxazole (ever use) 0.56 (0.43, 0.72) 0.72 (0.51, 1.02)
Baseline Function functional status Ambulatory 1 1
Bedridden 2.53 (1.72, 3.72) 2.47 (1.62, 3.75)
Working 0.56 (0.43, 0.73) 0.67 (0.50, 0.89)
Baseline WHO HIV disease stage Stage I 1 1
Stage II 1.12 (0.54, 2.31) 1.11 (0.53, 2.30)
Stage III 2.46 (1.33, 4.58) 2.14 (1.13, 4.06)
Stage IV 3.71 (1.87, 7.37) 2.25 (1.07, 4.74)
Patients transferred from other hospital 0.33 (0.17, 0.64) 0.39 (0.18, 0.82)
Entry from From the same facility 1 1
Self-referred 0.61 (0.20, 1.85) 0.71 (0.23, 2.15)
Referred from other facilities 0.61 (0.41, 0.87) 0.73 (0.48, 1.10)
  1. aIPT & ART were adjusted to the baseline covariates: gender, age, baseline body weight, baseline CD4 count, cotrimoxazole prophylaxis (ever use), baseline WHO functional status, baseline WHO stages, transfer history, and the entry point of care; HR Hazard Ratio, CI Confidence Interval, IPT Isoniazid Preventive Therapy, ART Antiretroviral Therapy; bAge interval is in every five years; cBaseline body weight is in every 10kgs; dBaseline CD4 count is in every 50 cells/mm3