Introduction and Objectives Although pulmonary tuberculosis (pTB) remains a major health burden worldwide relatively little is known about the formation of droplets and aerosols by these patients by that are considered to be the primary source of transmission. We have developed an approach based on wearing face masks.

Our aim was to capture expectorated droplets and aerosols from patients with pTB in a convenient continuously wearable system and to quantify the retained Mycobacterium tuberculosis (Mtb).

Methods We designed a mask sampling system to be worn for one hour. A filter mounted in the mask was subjected to GeneXpert analysis (DNA based detection of Mtb and rifampicin resistance) and 16S rRNA analysis using in-house techniques. Where available, a concurrent sputum sample was also processed.

Results Table 1 shows results from 10 patients including two negative controls and one extra-pulmonary TB case.

Three of five pre-treatment pTB patients and one of two pTB patients at day 5 of treatment were filter GeneXpert positive. No Mtb was detected from our negative controls. In one case (patient 3) the mask was the only positive sample apart from a scanty smear. In background experiments we demonstrated a detection limit of 1200 CFU/filter. Preliminary studies on RNA analysis suggest that this may have advantages over DNA-based detection but these require confirmation.

Conclusions The mask sampling system detected expectorated Mtb in just over half of the patients diagnosed to have pTB and all four of those that were sputum smear positive. Compatibility with the GeneXpert system makes the approach widely applicable.

Further studies will be required before comparisons with the Cough Aerosol Sampling System and Guinea Pig sampling approach can be made. However, the mask sampling appears at least as sensitive as the former and the convenience and cost of the method recommend its potential for both diagnosis and research into TB transmission.