Are a past history of tuberculosis and WHO clinical stage associated with incident tuberculosis in adults receiving antiretroviral therapy?, reply to Lawn et al.