icon-folder.gif   Conference Reports for NATAP  
48th Annual ICAAC / IDSA 46th Annual Meeting
October 25-28, 2008
Washington, DC
Back grey_arrow_rt.gif
Hospital Admission Rate Stays Higher for First 90 Days of HAART
  48th ICAAC, October 25-28, 2008, Washington, DC
Mark Mascolini
People who had a good virologic response to their first potent antiretroviral regimen (HAART) ran the same risk of landing in the hospital during their first 90 days of treatment as people who did not have a good virologic response to their first regimen. Blacks in this Baltimore, Maryland cohort were more likely than whites to respond poorly to their first regimen. Being a woman, injecting drugs, or starting HAART with fewer than 200 CD4s independently raised the risk of hospital admission.
Johns Hopkins University investigators studied 1327 HAART-naive people who began their first antiretroviral combination from 1997 through 2005. Defining virologic response as an undetectable viral load or at least a 10-fold drop in viral load during the first 6 months of therapy, the Hopkins team counted 965 responders (73%) and 362 nonresponders (27%). Among the 1036 blacks in the cohort, 30% ranked as nonresponders and 70% as responders, a much worse ratio than among 263 whites (16% nonresponders, 84% responders).
During the first 45 days of HAART, both responders and nonresponders had a hospitalization rate near 80 admissions per 100 person-years--the same as the pre-HAART rate in eventual responders. Over the next 45 days, the hospital admission rate dropped below 60 per 100 person-years in responders, a significant decline (P < 0.05), while the rate among nonresponders stayed flat. Up to 90 days after starting HAART, however, the hospitalization rate of responders did not differ significantly from that of nonresponders (P = 0.08). From day 91 through day 180 of HAART, the hospital check-in rate of responders fell to around 40 per 100 person-years, significantly lower than the 80-admission rate in nonresponders (P < 0.001). For day 181 to day 365 after starting HAART, the admission rate remained stable in responders and significantly lower than in nonresponders (P = 0.002).
Multivariate analysis that factored in calendar year (to account for improvements in antiretroviral regimens and care) uncovered four variables that independently raised the risk of hospital admission at the following incidence rate ratios (IRR) and 95% confidence intervals (CI):
⋅ Female gender: IRR 1.39, 95% CI 1.12 to 1.72, P = 0.002
⋅ Injection drug use: IRR 1.46, 95% CI 1.19 to 1.80, P < 0.001
⋅ Pre-HAART CD4 count below 50 (versus 200 or higher): IRR 2.49, 95% CI 1.81 to 3.42, P < 0.001
⋅ Pre-HAART CD4 count 50 to 199 (versus 200 or higher): IRR 1.44, 95% CI 1.07 to 1.92, P = 0.02
In other words, being a woman or injecting drugs, or starting HAART with 50 to 199 CD4s, independently made hospital admission about 40% more likely. Nonresponders were twice as likely as responders to be admitted to the hospital 91 to 180 days after starting HAART (IRR 2.25, 95% CI 1.48 to 3.43, P < 0.001) and almost 60% more likely to be admitted 181 to 365 days after starting therapy (IRR 1.58, 95% CI 1.11 to 2.24, P = 0.02).
This study did not appraise reasons for hospital admission in the first 90 days of HAART, which the investigators noted will require further work. The findings may not apply to other patient groups that differ substantially from the largely poor, urban Hopkins cohort in demographics or in the HIV care they receive. But the results suggest that even people who respond well to their first antiretrovirals need close monitoring for several months to anticipate and perhaps prevent serious illness.
1. Berry SA, Gebo KA, Moore RD, Manabe YC. A high risk of hospitalization immediately follows HAART initiation. 48th Annual International Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 25-28, 2008. Washington, DC. Abstract H-2292.