High Prevalence of Rifampicin Resistance Associated with Rural Residence and Very Low Bacillary Load among TB/HIV-Coinfected Patients at the National Tuberculosis Treatment Center in Uganda
Table 2
Multivariable logistic regression model for associations of RR among TB/HIV-coinfected patients.
Characteristic
Crude odds ratio 95% confidence interval
value
Adjusted odds ratio 95% confidence interval
value
Sex†
Male
1
1
Female
1.43 (0.64-3.18)
0.381
1.53 (0.45–5.26)
0.498
Residence
Urban
1
1
Rural
3.54 (1.52–8.21)
0.003
5.24 (1.51–18.21)
0.009
CTX and ART use
No
1
1
Yes
2.38 (1.05-5.36)
0.037
2.57 (0.78–8.47)
0.120
Night sweats
No
1
1
Yes
0.57 (0.25-1.33)
0.193
1.08 (0.24–4.92)
0.921
Weight loss
No
1
1
Yes
0.31 (0.13-0.74)
0.009
0.26 (0.07-1.01)
0.052
TB symptom burden
<4
1
1
≥4
0.53 (0.24-1.19)
0.124
1.20 (0.30-4.76)
0.797
Bacillary load
Low/medium/high
1
1
Very low
7.65 (2.81–20.84)
<0.001
13.52 (3.15–58.08)
<0.001
Temperature
Hypothermia
1
1
Normal
0.83 (0.35-1.96)
0.669
0.29 (0.07-1.17)
0.082
Hyperthermic
0.10 (0.01-0.85)
0.034
0.14 (0.01-2.19)
0.160
CD4/CD8 ratio
<0.52
1
1
0.52-4.1
2.17 (0.92-5.14)
0.078
1.27 (0.33–4.93)
0.732
White blood cell count
0.91 (0.80-1.04)
0.186
1.05 (0.86-1.29)
0.611
Mean corpuscular volume
1.05 (1.01-1.09)
0.011
1.06 (0.99-1.13)
0.085
TB: tuberculosis; CD: cluster of differentiation; CTX: cotrimoxazole; ART: antiretroviral therapy. †Included in the model due to effect on other risk factors for rifampicin-resistant TB [4, 22]. Note: Nagelkerke , that is, the covariates in the model explain 53% variation in RR among HIV/TB-coinfected patients. Cox Snell , that is, the covariates explain 37% of variation when using this criteria.
