Needle in a haystack: Looking for tuberculosis in a low-incidence setting

Abstract

Background: Canada is a low-incidence country for tuberculosis (TB). The BC Public Health Laboratory diagnostic algorithm for pulmonary TB includes acid fast bacilli (AFB) smear and mycobacterial culture of all submitted sputa. TB nucleic acid amplification testing (NAT) is routinely performed on AFB-smear-positive (AFB+) sputa only. We assessed the laboratory-associated costs of implementing the international recommendations for TB NAT on AFB-smear-negative (AFB–) sputa. Methods: Two data sets were obtained: ( 1 ) all AFB– samples for a 3-year period (October 1, 2014–September 30, 2017) and ( 2 ) all AFB–, TB-culture-positive samples for the same period. One AFB– sample/patient from each defined diagnostic set of sputa was deemed eligible for TB NAT. To stratify patients by ordering location, a 1-year subset of data (October 1, 2016–September 30, 2017) was examined. Results: In the 3-year period, 0.7% of all diagnostic sets were AFB− and culture-positive. In the 1-year period, the provincial TB Services clinics submitted 26% of all AFB– samples received, but these constituted 78% of AFB–, culture-positive samples. Conclusions: The annual cost of TB NAT on one AFB– sputum sample from each eligible diagnostic set would total approximately $247,000. Targeting only TB Services clinic patients would reduce this cost to approximately $64,000/year while capturing more than 75% of AFB–, culture-positive patients. On the basis of our provincial positivity rate, it would cost approximately $6,000 to provide an early TB diagnosis for an AFB–, culture-positive patient. The cost-effectiveness to public health of this approach in a TB low-incidence setting needs to be carefully evaluated.