Are there clinical scenarios in which you would start empiric treatment for pulmonary TB without microbiologic confirmation?