
Methods: We performed a retrospective cohort study of 297 HIV+ patients with first admission to MICU for > 24 hours between 7/2009 and 7/2014. This preliminary analysis includes 113 patients, randomly selected. Data on discharge diagnoses, comorbid conditions, disease severity scores, CD4 cell counts, HIV viral load (VL), HBV/HCV serology, ART use and short- and long-term mortality were collected. The relationship between hepatitis coinfection, ART use, qualitative VL and mortality was analyzed (χ2test).
Results: Of 113 patients, 65% (N=73) were male, 83% (N=94) were African American, and the mean age was 44±12 years. 43 patients (38%) had >= one current AIDS-DC and 8 patients had two. The VL was detectable in 82% (median: 59,700 copies/ml, IQR 235,653) and the median CD4 was 64 cells/mm3 (IQR 173). 53% (N= 60) were not receiving ART while in the MICU, of which 68% (N=41) were not on ART before admission. 52 patients (46%) had positive hepatitis serology (positive anti-HCV and/or anti-HBc and/or HBsAg). MICU mortality was higher in patients with hepatitis coinfection versus HIV monoinfection: 28.9% vs 6.1%, (OR 6.28, 95%CI: 1.33, 29.62, p=0.0105). Significant mortality differences persisted with follow up to 180 days. There was no difference in outcome by ART use or qualitative VL.
Conclusion: Our data reveal a large burden of uncontrolled HIV infection and AIDS in this inner city population with poor ART utilization. Hepatitis coinfection is associated with increased mortality in univariate analysis. The study emphasizes the importance of improved HIV and hepatitis treatment and control. ICU physicians can still expect to diagnose and treat a wide range of AIDS-DC.

E. Miller,
None
M. Doll, None
K. Saharia, None
U. Buchwald, None