SESSION TYPE: AIDS/ Immunocompromised Patients Posters
PRESENTED ON: Wednesday, October 24, 2012 at 01:30 PM - 02:30 PM
PURPOSE: The CURB-65 pneumonia severity scoring system has been validated for predicting mortality in several studies. It is used as a clinical prediction rule for stratifying patients into outpatient, inpatient, and intensive care treatment strategies. Previous studies have excluded patients with human immunodeficiency virus (HIV) infection. As a result, the validity of CURB-65 in predicting mortality and ICU admission in HIV patients with community acquired pneumonia (CAP) is unknown.
METHODS: A retrospective chart review of 230 patient admissions, age 18-50 years, with CAP to the Saskatoon Health Region in 2010. 139 admissions fully satisfied all inclusion and exclusion criteria. Thirty-five HIV patients were compared to 104 non-HIV patients with respect to mortality, ICU admission, ICU and hospital length of stay, and CURB-65 score.
RESULTS: The mortality rate and ICU admission rate in the HIV population was 11.4% and 28.6% versus 4.8% and 27.9% in the non-HIV population. The ICU and hospital length of stay in the HIV population was 19.7 days (95% CI 10.2 to 29.2) and 18.7 days (95% CI 12.2 to 25.2) versus 7.3 days (95% CI 4.8 to 9.8) and 7.1 days (95% CI 5.8 to 8.4) in the non-HIV population. The CURB-65 score in the HIV population was 1.23 (95% CI 0.96 to 1.5) versus 1.41 (95% CI 1.21 to 1.61) in the non-HIV population.
CONCLUSIONS: The CURB-65 pneumonia severity scoring system may underestimate the mortality rate and need for ICU admission in CAP patients with HIV infection. Further prospective studies are required to determine the validity of CURB-65 in stratifying CAP patients with HIV infection into different treatment strategies.
CLINICAL IMPLICATIONS: At the present, no pneumonia severity scoring system has been validated in patients with CAP and HIV infection. We currently do not recommend the use of CURB-65 in patients with HIV infection.
DISCLOSURE: The following authors have nothing to disclose: Jordan Olfert, Stuart Skinner
No Product/Research Disclosure InformationUniversity of Saskatchewan, Saskatoon, SK, Canada