Existing strategies to control the spread of tuberculosis (TB) miss important opportunities to interrupt transmission. This study compares the microscopic observation drug susceptibility (MODS) assay with conventional reference methods of automated mycobacterium culture or culture on Lowenstein–Jenson medium.

This study, conducted in Peru in 2003-2004, targeted three patient groups: unselected patients with suspected TB, pre-screened patients at high risk of TB or multidrug-resistant (MDR) TB and unselected hospitalised patients infected with HIV. From 10 government clinics and two hospitals 1980 patients were recruited; 3760 sputum samples were collected, out of which 401 (10.7%) yielded culture positive results for MDR compared with 220 (5.9%) sputum smear positive results. The sensitivity of detection using MODS assay, automated mycobacterium culture and Lowenstein–Jenson medium culture was 97.8%, 89% and 84%, respectively. The median time to culture positivity was 7 days for MODS assay, 13 days for automated mycobacterium culture and 26 days for Lowenstein–Jenson medium. The median time to results of susceptibility tests was 7 days for MODS assay compared with 22 days for automated mycobacterium culture and 68 days for Lowenstein–Jenson medium. Agreement between MODS assay and reference standard for susceptibility was 100% for rifampicin, 96.7% for isoniazid, 98.8% for combination of rifampicin and isoniazid, 95.4% for ethambutol and 91.7% for streptomycin.

The study was designed to include broader groups of patients with or without disease and in pertinent patient groups. The authors concluded that MODS assay detected mycobacterium TB in sputum with greater sensitivity and speed and reliably identified MDR TB in less time than conventional reference methods. However, the study did not comment on application of this assay in clinical scenarios and the role of this assay in the future.

This intriguing study hints to future developments in the diagnosis of TB and application of the MODS assay in normal day practice.