Evaluating the Impact of a Molecular Diagnostic Algorithm on Tuberculosis and Nontuberculous Mycobacterial Infections in Newfoundland and Labrador, Canada

<b>Background/Objectives</b>: The diagnosis of <i>Mycobacterium tuberculosis</i> complex (MTBC) and nontuberculous mycobacterial (NTM) infections is accomplished by three main diagnostics methods: smear microscopy, culture, and molecular testing. Diagnostic algorithms used by laboratories can significantly impact clinical and infection control management. Current Canadian Tuberculosis Standards recommend the use of nucleic acid amplification testing (NAAT) for smear-positive patients and smear-negative patients upon request. An alternative algorithm is to utilize NAAT in the Panel approach on all samples, pulmonary and extrapulmonary, to potentially reduce time to diagnosis and treatment. This alternative approach was implemented in November 2019 at the Newfoundland and Labrador Public Health and Microbiology Laboratory (NL PHML) using a laboratory-developed multiplex real-time PCR (LDT m-qPCR) assay targeting <i>Mycobacterium</i> spp. (<i>Myco</i> spp.) and MTBC, performed in parallel with smear and culture. <b>Methods:</b> To investigate the impact of this alternate testing approach, we conducted an observational retrospective analysis of laboratory diagnostic and treatment data, recognizing that temporal changes in epidemiology, clinical practice, and laboratory workflow may also have influenced outcomes. To complete this, study data from three years before and four years after implementation were gathered. <b>Results</b>: The sensitivity/specificity of the smear, m-LDT qPCR-MTBC, m-LDT qPCR-<i>Myco</i> spp., and culture assays in this study were 18.1%/100%, 96.7%/99.8%, 47.6%/99.0%, and 96.8%/100%, respectively. The gold standard utilized for these calculations was clinical diagnosis for active MTBC disease and culture for NTM infections, recognizing that the use of clinical diagnosis may introduce subjectivity. The Panel approach reduced the time to diagnosis of tuberculosis MTBC by 29 days (<i>p</i> < 0.0001) for NL PHML, and when modelled for a laboratory with rapid culture identification, diagnosis was reduced by 14 days (<i>p</i> = 0.003). Among non-empirically treated tuberculosis patients, the time to treatment was decreased by 25.5 days (<i>p</i> < 0.001). For NTM infections, rapid diagnostics only affected one patient's treatment. This finding agrees with clinical management guidelines, which do not routinely utilize rapid diagnostics for the diagnosis of disease or treatment decisions. The cost implications of additional NAAT testing were calculated to be an increase of CAD 23.62 per sample. <b>Conclusions</b>: Our findings support the adoption of a molecular assay for MTBC as an initial diagnostic tool to decrease time to diagnosis and time to treatment, depending on local epidemiology and irrespective of smear status. Utilizing a molecular assay for genus level identification of NTM had minimal impact on clinical management suggesting its limited diagnostic utility in a broad population setting.