Diagnosis and Management of Tuberculosis in Candidates for Tumor Necrosis Factor Alpha Antagonists: An Experts Survey

Abstract

Background: There are some controversies regarding the management of latent tuberculosis infection and tuberculosis in patients with rheumatologic indications for biologic therapy. Objectives: To describe current expert opinions and preferences regarding the evaluation and management of latent tuberculosis infection and tuberculosis in candidates and recipients of tumor-necrosis factor-alpha blocking therapy. Methods: A questionnaire addressing preferences related to management and treatment of latent tuberculosis infection and active tuberculosis in tumor-necrosis factor-alpha blocking candidates was distributed to tuberculosis and rheumatology experts across the United States between August 18, 2009, and June 21, 2010. Survey responses were formulated as a 5-point Likert scale (strongly disagree to strongly agree), or as a priority rank order list (1 to 6 or 7), and data were analyzed for percent agreement and median rankings. Measurements and main results: The tuberculin skin test and interferon-gamma release assays for latent tuberculosis infection screening were highly accepted among tuberculosis and rheumatology experts. Most participants supported the use of daily isoniazid for 9 months for latent tuberculosis infection therapy, but responses were mixed regarding timing to initiation of tumor-necrosis factor-alpha blocking therapy. Most tuberculosis experts supported standard anti-tuberculosis therapy for treatment of tuberculosis, but preferences varied among rheumatologists. In contrast, most rheumatologists believed tumor-necrosis factor-alpha blocking therapy should be stopped in individuals with active tuberculosis, while opinions varied among tuberculosis experts. Conclusions: Agreement among experts was common regarding preferences for diagnosis and management of latent tuberculosis infection and tuberculosis under hypothetical but likely common clinical scenarios, but some differences exist.