icon-folder.gif   Conference Reports for NATAP  
  51th ICAAC
Chicago, IL
September 17-20, 2011
Back grey_arrow_rt.gif
Hospital Rate Lower With Atripla Than Other Combos in Medicaid Patients
  51st Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), September 17-20, 2011, Chicago

Mark Mascolini

HIV-positive Medicaid patients taking once-daily single-pill Atripla had a 25% to 30% lower risk of hospital admission than Medicaid patients taking regimens requiring two or more tablets daily, according to results of a retrospective analysis of almost 8000 patients [1]. Better adherence to Atripla than to other regimens may have contributed to the lower hospital admission rate.

Earlier research in US patients found that taking Atripla doubled chances of complete adherence compared with taking other regimens [2]. That study also found lower hospital admission and emergency room rates in people with better adherence.

To explore differences in adherence and hospitalization rates between Medicaid beneficiaries taking Atripla and more complex regimens, researchers retrospectively analyzed Medicaid claims of people diagnosed with HIV between January 2005 and December 2009 who took a triple combination including two nucleosides plus either a nonnucleoside, a protease inhibitor, a CCR5 antagonist, or an integrase inhibitor for at least 60 days [1]. Medicaid helps fund medical care for US citizens with limited income.

The investigators calculated adherence by refill timing, figuring the proportion of days between the start of the study regimen to stopping that regimen (or the end of follow-up) in which the person had a drug supply on hand for all components of the regimen (termed medication possession ratio). Medication possession ratios were grouped as 95% to 100%, 80% to 94%, 60% to 79%, and under 60%. The researchers measured hospitalization and adherence rates from when treatment began to (1) discontinuation, (2) switch from Atripla to another regimen or vice versa, or (3) the end of follow-up on March 31, 2009.

The analysis focused on 7783 people, 1838 (24%) taking Atripla and 5945 taking two or more pills daily. Overall, 49% of study participants in both study groups were women, 23% were younger than 35, 34% were 35 to 44, 32% were 45 to 54, and the rest were 55 or older. The groups did not differ significantly in rates of mental health disorders (23% overall) or illicit drug or alcohol use (16%). A substantially higher proportion of people taking Atripla than other regimens were antiretroviral naive when they started the study regimen (47% versus 27%). Average follow-up duration was much longer in the in the non-Atripla group (428 versus 347 days).

Patients taking Atripla versus other regimens were significantly more likely to include a higher proportion with a 95-100% medication possession ratio (25.3% versus 17.6%, P < 0.01) or an 80-94% ratio (44.8% versus 41.3%, P < 0.01). In contrast, people taking non-Atripla regimens were significantly more likely to have a 60-79% medication possession ratio (28.4% versus 23.2%, P < 0.01) or a ratio below 60% (11.6% versus 6.6%, P < 0.01).

Among people with no antiretroviral experience when follow-up began, those starting Atripla had significantly fewer hospital admissions per 100 people than those starting another regimen: 39.2 versus 53.3 (P < 0.001). Among people with antiretroviral experience when follow-up began, the Atripla group also had significantly fewer admissions per 100 people: 39.7 versus 53.9 (P < 0.001).

Among women of child-bearing age, those younger than 35 had a significantly lower number of admissions per 100 people if they took Atripla than if they took other regimens: 28.4 versus 47.3 (P < 0.001). Those rates for women from 35 to 44 years old were 30.8 in the Atripla group and 51.3 in the non-Atripla group (P < 0.001).

A Cox statistical model that accounted for daily tablet burden, demographics, comorbidities, and antiretroviral experience determined that taking Atripla lowered the risk of hospital admission 25% (hazard ratio [HR] 0.753, P < 0.001). A Poisson model figured almost a 30% lower hospital admission risk in the Atripla group (HR 0.705, P < 0.0001).

The multiple-event Cox model determined that women had almost an 18% higher risk of hospital admission than men (HR 1.178, P < 0.0001), while initially antiretroviral-naive people had almost a one third higher admission risk than people with antiretroviral experience (HR 1.329, P < 0.0001). People with a mental disorder had a 30% higher risk of going to the hospital (HR 1.301, P < 0.0001), and people with a drug or alcohol abuse diagnosis had a doubled risk (HR 2.052, P < 0.0001).

The investigators cautioned that this type of nonrandomized comparison cannot exclude the influence of unmeasured confounders. Notably, they could not factor CD4 count or viral load into their calculations because they did not have access to those numbers. With those limitations in mind, the researchers proposed that "higher observed adherence in the [Atripla] group may explain these associations" between Atripla and fewer hospital admissions.

The earlier study comparing adherence with Atripla and other regimens found that adherent patients had a significantly lower hospital admission rate (11% versus 28%, P < 0.0001) and used the emergency room less often (26% versus 34%, P < 0.09) than did poorly adherent patients [2]. Full adherence was twice more likely among people taking Atripla than among those taking other combinations (odds ratio 2.1, P < 0.05).


1. Cohen C, Davis KL, Meyers J. Association between daily antiretroviral pill burden and hospitalization risk in a Medicaid population with HIV. 51st Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). September 17-20, 2011. Chicago. Abstract H2-791.

2. Juday T, Gupta S, Grimm K, Wagner S, Kim E. Factors associated with complete adherence to HIV combination antiretroviral therapy. HIV Clin Trials. 2011;12:71-78.