Objective: To compare the cost-effectiveness of three diagnostic algorithms incorporating a new rapid test, FASTPlaqueTB, with the current National TB Control Programme (NTCP) algorithm for diagnosis of pulmonary tuberculosis in new smear-negative suspects in South Africa.
Design: A model of the outcome of patients screened using each diagnostic algorithm was established using published costs and performance data for each component of the algorithm. Direct health care provider costs associated with provision of each diagnostic strategy were determined. Overall performance, total cost, speed and accuracy of diagnosis were estimated for screening 1000 new TB suspects using each algorithm.
Results: The use of FASTPlaqueTB and culture algorithms enabled 28% more smear-negative TB patients to be diagnosed overall and was cheaper to implement than the NTCP algorithm (18,312-18,581 US dollars compared with 20,079 dollars). Fewer clinic visits were required to establish a diagnosis, reducing clinic workload and patient costs. FASTPlaqueTB enabled rapid and specific diagnosis of at least 50% of smear-negative TB patients within 2 days.
Conclusions: The use of FASTPlaqueTB can assist in improving existing case detection strategies and may be cost-effectively integrated into the current diagnostic infrastructure, offering patients more rapid and reliable diagnosis whilst reducing the overall cost.