Table 2.

Univariable and multivariable regression analysis for associations between a baseline CD4 cell count ≥250 cells/μL and various ARV adherence measures in the first 3 months of therapy among a population of patients initiating anti-retroviral therapy in southwest Uganda (n=473).

	Univariable Model			Multivariable Model †
	Estimate¥	95% CI	p-value	Estimate¥	95% CI	p-value
Outcome 1: Any treatment interruptions > 72 hours
Baseline CD4 cell count ≥250 cells/μL	2.33	1.14 – 4.75	0.020	2.28	1.01 – 5.15	0.048
Outcome 2: Number of treatment interruptions > 72 hours
Baseline CD4 cell count ≥250 cells/μL	2.26	0.91 – 5.64	0.079	2.56	0.99 – 6.65	0.054
Outcome 3: Average Adherence > 90%
Baseline CD4 cell count ≥250 cells/μL	1.44	0.78 – 2.64	0.243	1.40	0.70 – 2.82	0.344
Outcome 4: Detectable HIV Viremia
Baseline CD4 cell count ≥250 cells/μL	2.24	0.95 – 5.29	0.065	2.83	1.14 – 7.00	0.024

^†Multivariable models were adjusted for the following: age, gender, marital status, educational attainment, employment status, Filmer-Pritchett asset index, self-reported distance to clinic, Hopkins Symptoms Checklist for Depression, positive screen for heavy drinking, ARV dosing frequency.

^¥Estimates of association for any treatment interruptions, average adherence, and detectable viremia represent odds ratios and adjusted odds ratios. Estimates of association for number of treatment interruptions represent incidence rate ratios and adjusted incident rate ratios.