Background:  It is widely held that improved outcomes are expected in clinical trials (efficacy) vs clinical practice (effectiveness) in part due to study inclusion and exclusion criteria, selection bias, and volunteer bias among patients, participating in research. The differential efficacy vs effectiveness of HAART among antiretroviral (ARV) -naïve patients initiating therapy is understudied.

Methods:  We conducted a retrospective study of the University of Alabama–Birmingham (UAB) 1917 HIV/AIDS Clinic Cohort from January 1, 2000 to December 31, 2006. We evaluated 6- and 12-month CD4 and viral load responses for ARV-naïve patients initiating HAART, stratified by treatment modality; clinical trial vs clinical practice. Student’s t test, χ² test, and univariate logistic regression were used for bivariate analyses. Multivariable logistic regression models evaluated factors associated with trial participation and with 6-month viral load >50 copies/mL.

Results:  Among 570 patients initiating HAART, 126 received ARV through a clinical trial (22%). Overall, mean age 37.5±9.5, 23% female, 54% African American, 51% men who have sex with men (MSM), 14% public insurance, 37% uninsured, 44% CD4 >200, mean log HIV RNA 4.7±1.0, 47% affective mental health disorder, 23% substance abuse, and 16% alcohol abuse. Receipt of HAART through clinical trial participation was less common among African Americans (OR = 0.41, 95%CI 0.26 to 0.67) and more common in patients with CD4 counts >350 cells/mm³ (OR = 2.88, 95%CI 1.58 to 5.22). Similar 6- and 12-month CD4 and viral load responses were observed in clinical trials and clinical practice (see the table). In multivariable logistic regression, treatment modality was not associated with 6-month viral load >50 copies/mL (OR = 1.04 for clinical practice, 95%CI 0.64 to 1.66). Viral load >50 copies/mL at 6 months was more common in patients with public insurance (OR = 2.37, 95%CI 1.26 to 4.44), and less common among patients with lower baseline CD4 counts (CD4 <50 OR = 0.22, 95%CI 0.13 to 0.38; CD4 50 to 199 OR = 0.31, 95%CI 0.19 to 0.57; CD4 200 to 349 OR = 0.23, 95%CI 0.11 to 0.37).

Conclusions:  In concordance with studies in the early HAART era, racial disparities in clinical trial participation persist in recent years. However, contrary to expectations and conventional wisdom, similar 6- and 12-month CD4 and viral load responses were observed among ARV-naïve patients initiating HAART through clinical trials vs clinical practice.