Utility of Mycobacterium tuberculosis PCR in ruling out active disease and impact on isolation requirements in a low prevalence setting

Abstract Objective To analyze and interpret clinical microbiology data for specimens tested with the fluorochrome stain (AFB stain), mycobacterial culture and a laboratory-developed Mycobacterium tuberculosis (MTB) PCR in order to understand the performance of each test and to demonstrate the utility of MTB PCR to assist with decisions regarding discontinuation of airborne isolation. Methods Retrospective cohort analysis of 2798 respiratory specimens from 2006 patients in the period between November 1st, 2011 and January 1st, 2018. Results 53.7% were males, median age was 61 years, and 43 patients were HIV positive. Results demonstrated positive mycobacterial cultures for MTB in 52 specimens (1.9%) and for nontuberculous mycobacteria (NTM) or aerobic actinomycetes (eg., Nocardia spp.) in 435 specimens (16%). Using mycobacterial culture as the gold standard, AFB smear had a sensitivity of 48.1% while MTB PCR had a sensitivity of 96.0% in AFB smear positive specimens and an overall sensitivity of 57.7% with PPV of 94% and a NPV of 99%. Conclusions The combination of a positive AFB smear with a negative MTB PCR offers a rapid result to rule out active pulmonary MTB in a low prevalence setting. In this study, that combination reliably excluded active tuberculosis (NPV of 99.2%). The combination of a positive AFB smear with a negative MTB PCR indicated pulmonary NTM infection with the results available within 1 day. There was little benefit to pursuing collection and testing of more than 2 respiratory specimens in a low prevalence setting for both long term diagnostic or rapid isolation discontinuation purposes.