Intensive Care Admissions of HIV Infected Patients: The Effect of HAART on Outcomes in an Inner-City Hospital

Abstract

Purpose: Highly active antiretroviral therapy (HAART) improves survival and quality of life in HIV-infected patients. Intensive care unit (ICU) utilization in HAART era has shifted to admissions mainly due non-HIV related conditions. In the inner-city population, compliance with HAART is tempered by social, economic and cultural factors. The impact of HAART on ICU outcomes in the inner-city remains unclear, we compared HIV-infected patients on HAART with those not on HAART. Methods: Retrospective, observational study of HIV-infected patients admitted to ICU, from January to July 2012. Patients were divided into: HAART+therapy and HAARTtherapy. Demographics, comorbid conditions, indications for admission and laboratories were obtained. Primary outcomes were: mechanical ventilation (MV) and tracheostomy rates, ICU and hospital length of stay(LOS) and mortality. Secondary outcomes were discharge status, one-year mortality. Results: 293 HIV-infected patients were admitted to ICU during study period; 205(70%) HAART+ and 85(30%) HAART-. HAART+ therapy were older (55+/-9.2versus49.2+/-87;p=0.0001), had less alcohol abuse (18versus15;p=0.04), higher CD4 count (308+/-280 versus224+/-279;p=0.02) and undetectable viral loads (107versus16;p=0.0001). No other differences were found between the groups. Renal failure was more common as etiology for admission in HAARTtherapy (18versus15;p=0.04). Eighty-eight (30%) patients were found to have urine toxicology positive for opioids, cocaine or marijuana. There were no differences for use of MV (46versus23;p=0.6), tracheostomy rates (22versus 9;p=1.0), ICULOS(4.44+/-4.17versus 4.4+/-3.41;p=0.93),hospital LOS (11.04+/-10.5versus 10.3+/-7.0;p=0.55) and mortality(15versus12;p=0.11). Overall mortality for the cohort was 27/290(9.3%). Patients on HAART+therapy were more likely to be discharged home(100versus54;p=0.0278). One-year mortality was similar in both groups(37versus16;p=0.86). Subgroup analysis of patients with CD4 less than 200 revealed no differences for patient with or without HAART therapy. Conclusions: Mortality rates for HIV-infected patients admitted to ICU in HAART-era remains low. In our population, HAART therapy before admission had no effect in outcomes for patients admitted to ICU. We found a significant group of HIV-infected patients not on HAART in our population. Etiologies for ICU admission were mainly non-HIV specific. Clinical Implications: ICU triage for HIV-infected patients should follow general guidelines. We need to evaluate barrier for HAART treatment in our population.