The Global Consortium for Drug-resistant Tuberculosis Diagnostics (GCDD): design of a multi-site, head-to-head study of three rapid tests to detect extensively drug-resistant tuberculosis

Background: Drug-resistant tuberculosis (DR-TB) remains a threat to global public health, owing to the complexity and delay of diagnosis and treatment. The Global Consortium for Drug-resistant Tuberculosis Diagnostics (GCDD) was formed to develop and evaluate assays designed to rapidly detect DR-TB, so that appropriate treatment might begin more quickly. This paper describes the methodology employed in a prospective cohort study for head-to-head assessment of three different rapid diagnostic tools. Methods: Subjects at risk of DR-TB were enrolled from three countries. Data were gathered from a combination of patient interviews, chart reviews, and laboratory testing from each site's reference laboratory. The primary outcome of interest was reduction in time from specimen arrival in the laboratory to results of rapid drug susceptibility tests, as compared with current standard mycobacterial growth indicator tube (MGIT) drug susceptibility tests. Results: Successful implementation of the trial in diverse multinational populations is explained, in addition to challenges encountered and recommendations for future studies with similar aims or populations. Conclusions: The GCDD study was a head-to-head study of multiple rapid diagnostic assays aimed at improving accuracy and precision of diagnostics and reducing overall time to detection of DR-TB. By conducting a large prospective study, which captured epidemiological, clinical, and biological data, we have produced a high-quality unique dataset, which will be beneficial for analyzing study aims as well as answering future DR-TB research questions. Reduction in detection time for XDR-TB would be a major public health success as it would allow for improved treatment and more successful patient outcomes. Executing successful trials is critical in assessment of these reductions in highly variable populations.