The Feasibility, Accuracy, and Impact of Xpert MTB/RIF Testing in a Remote Aboriginal Community in Canada

Abstract

Xpert MTB/RIF testing for Mycobacterium tuberculosis and rifampin resistance is being used extensively in countries with a high burden of TB. However, recent evidence suggests that it may not have the same accuracy or impact in high-income, low-burden TB countries. A prospective, pragmatic study was done between March 2012 and March 2014 to determine the feasibility, accuracy, and impact on TB disease management provided by the Xpert test in a remote, medically underserved, predominantly Inuit population in Iqaluit, Nunavut, Canada. A total of 453 Xpert tests were run on sputum samples from 344 patients with suspected TB. Twenty-seven patients were identified as having active TB disease by culture. There were no cases of drug-resistant TB. Using culture as the gold standard, one Xpert test compared with one, two, or three sputum samples cultured per patient had a sensitivity of 85% (95% CI, 66%-95%) and a specificity of 99% (95% CI, 97%-100%) for detection of M tuberculosis. The indeterminate rate was 4.4% of all samples run. Treatment initiation was significantly shortened using Xpert vs the national standard of three smears (1.8 days vs 7.7 days, P < .007) and particularly shorter in smear-negative, culture-positive cases (1.8 days vs 37.1 days, P < .008). In a predominantly Inuit population in a remote region of Canada where the burden of TB is high and no TB testing facilities are available, onsite Xpert testing was feasible and accurate and shortened the time to TB treatment initiation.