A Clinical Algorithm to Identify HIV Patients at High Risk for Incident Active Tuberculosis: A Prospective 5-Year Cohort Study

摘要: Background Predicting the risk of tuberculosis (TB) in people living with HIV (PLHIV) using a single test is currently not possible. We aimed to develop and validate clinical algorithm, baseline CD4 cell counts, viral load (pVL), interferon-gamma release assay (IGRA), identify PLHIV who are at high for incident active TB low-to-moderate burden settings where highly antiretroviral therapy (HAART) routinely provided. Materials Methods A prospective, 5-year, cohort study adult was conducted from 2006 2012 two hospitals Taiwan. HAART initiated based on contemporary guidelines (CD4 count < = 350/μL). Cox regression used predictors construct algorithm. The validation cohorts included 1455 HIV-infected individuals previous published studies. Area under receiver operating characteristic (ROC) curve calculated. Results Seventeen 772 participants developed during median follow-up period 5.21 years. Baseline 350/μL or pVL ≥ 100,000/mL predictor (adjusted HR 4.87, 95% CI 1.49–15.90, P 0.009). A positive IGRA predicted patients 6.09, 1.52–24.40, 0.01). Compared an IGRA-alone strategy, algorithm improved sensitivity 37.5% 76.5%, negative predictive value 98.5% 99.2%. untargeted spared 468 (60.6%) unnecessary preventive treatment. ROC 0.692 (95% CI: 0.587–0.798) 0.792 0.776–0.808) 0.766 2 cohorts. Conclusions validated incorporating count, load, status can be guide targeted treatment provided all PLHIV. implementation this will avoid exposure low-risk drug toxicity simultaneously, reduce universal healthcare system.