1176: EFFECT OF HIV INFECTION ON SHORT-TERM MORTALITY IN SEPSIS: A POPULATION-BASED COHORT STUDY

Abstract

Introduction: The reported short-term mortality among septic patients infected with the human immunodeficiency virus (HIV) is markedly higher compared to the general population. However, the evidence base on the prognostic impact of HIV in sepsis is generally dated, based mostly on small cohorts, and the temporal trajectories of sepsis-related short-term mortality in HIV-infected patients are unknown. Methods: We used a statewide dataset identify hospitalizations aged ≥18 years with sepsis in Texas during 2014-2017. Sepsis was defined by “explicit” ICD-9 and ICD-10 codes for severe sepsis (995.92, R65.20) and septic shock (785.52, R65.21). HIV was identified using ICD-9 and ICD-10 codes 042 and B20. Hierarchical logistic models were used to estimate the association of HIV with short-term mortality (defined as in-hospital death or discharge to hospice) among sepsis hospitalizations and for sensitivity analyses restricted to ICU admissions and those with septic shock. Results: Among 283,025 sepsis hospitalizations, 2,079 (0.7%) had HIV. Compared to those without HIV, sepsis hospitalizations with HIV were younger (aged ≥65 years, 7.6% vs 57.0%), more commonly racial minority (76.0 % vs 47.8%), uninsured (20.9% vs 8.5%), and had higher mean [SD] Deyo comorbidity index (6.0 [2.1] vs 2.6 [2.4]); [p < 0.0001 for all comparisons]. The rates of septic shock were similar among sepsis hospitalizations with and without HIV (61.5% vs 59.6%; p = 0.0786). Short-term mortality among sepsis hospitalizations with and without HIV was 29.1% vs 31.2%, respectively. On adjusted analyses, HIV was not associated with short-term mortality (adjusted odds ratio [aOR] 1.04 [95% CI 0.93-1.16]), with similar findings on sensitivity analyses. Short-term mortality among sepsis hospitalizations did not change statistically over time among those with HIV (aOR 1.01/year [95% CI 0.85-1.20]), but decreased among those without HIV (aOR 0.98/year [95% CI 0.97-0.99]). Conclusions: HIV infection was not associated with adverse impact on short-term mortality in sepsis. However, short-term mortality remained statistically unchanged over time among sepsis hospitalizations with HIV, contrasting the downtrend among those without HIV. Further studies are needed to determine the factors underlying the diverging mortality trends.