A test of infection should not be a “standard” for guiding TB preventive therapy in at-risk populations