Objective: Tuberculosis (TB) is the leading cause of death in HIV-positive people. In Kenya, 140 000 new TB cases occurred in 2019, and 13 000 HIV-positive patients died due to TB. The objective of this study was to investigate the role of high-sensitivity C-reactive protein (HS-CRP) in TB diagnosis and the prediction of mortality in HIV-positive patients. Methods: The IDEA-TB Study enrolled HIV-positive adult patients attending three DREAM centres in Kenya who were suspected of having TB. A lateral flow urine lipoarabinomannan assay (LF-LAM), serum HS-CRP, and GeneXpert MTB/RIF assay (Xpert MTB/RIF) were performed. Six-month survival was evaluated. Results: A total of 574 patients were enrolled. The median (interquartile range) age, body mass index, and CD4 count were 45 years (37-54 years), 20.5 kg/m 2 (18.5-23.69 kg/m 2 ), and 477 cells/mL (290-700 cells/ mL), respectively. TB was confirmed in 87 (15.2%) patients. Concordance between the Xpert MTB/RIF and LF-LAM tests was 87.1%. HS-CRP was higher in TB patients (35.39 mg/l vs 9.21 mg/l). Malnutrition and elevated HS-CRP were associated with TB: odds ratio (OR) 2.5 (95% confidence interval (CI) 1.14-5.72) and OR 6.6 (95% CI 3.87-11.52), respectively. Nine (1.6%) patients died during follow-up. No single factor was associated with mortality. Only the combination of malnutrition and elevated HS-CRP was highly predictive of death (odds ratio (OR) 9.8, 95% CI 1.88-50.95); the association was stronger in TB patients (33.3% vs 1.0%; OR 47.6, 95% CI 7.03-322.23). Conclusion: TB diagnosis in HIV-positive patients remains challenging. HS-CRP could play a role in predicting early mortality in symptomatic patients.

No abstract