Development of a clinical scoring system for assessment of immunosuppression in patients with tuberculosis and HIV infection without access to CD4 cell testing--results from a cross-sectional study in Ethiopia.

Background: Currently, antiretroviral therapy (ART) is recommended for all HIV-positive patients with tuberculosis (TB). The timing of ART during the course of anti-TB treatment is based on CD4 cell counts. Access to CD4 cell testing is not universally available; this constitutes an obstacle for the provision of ART in low-income countries.

Objective: To determine clinical variables associated with HIV co-infection in TB patients and to identify correlations between clinical variables and CD4 cell strata in HIV/TB co-infected subjects, with the aim of developing a clinical scoring system for the assessment of immunosuppression.

Design: Cross-sectional study of adults with TB (with and without HIV co-infection) recruited in Ethiopian outpatient clinics. Clinical variables potentially associated with immunosuppression were recorded using a structured questionnaire, and they were correlated to CD4 cell strata used to determine timing of ART initiation. Variables found to be significant in multivariate analysis were used to construct a scoring system. Results : Among 1,116 participants, the following findings were significantly more frequent in 307 HIV-positive patients compared to 809 HIV-negative subjects: diarrhea, odynophagia, conjunctival pallor, herpes zoster, oral candidiasis, skin rash, and mid-upper arm circumference (MUAC) <20 cm. Among HIV-positive patients, conjunctival pallor, MUAC <20 cm, dyspnea, oral hairy leukoplakia (OHL), oral candidiasis, and gingivitis were significantly associated with <350 CD4 cells/mm(3). A scoring system based on these variables had a negative predictive value of 87% for excluding subjects with CD4 cell counts <100 cells/mm(3); however, the positive predictive value for identifying such individuals was low (47%).

Conclusions: Clinical variables correlate with CD4 cell strata in HIV-positive patients with TB. The clinical scoring system had adequate negative predictive value for excluding severe immunosuppression. Clinical scoring systems could be of use to categorize TB/HIV co-infected patients with regard to the timing of ART initiation in settings with limited access to laboratory facilities.