Globally, almost a quarter of us have latent tuberculosis infection – most are unharmed and remain oblivious. Unfortunately, because around 10% of latent infections reactivate, this represents an enormous potential source of infection that poses a serious obstacle to elimination.

For several years, TB rates have been dropping, which on the surface sounds like a very good thing. However, there has been increased reports of multi-drug resistant TB. MDR-TB is under-diagnosed meaning those with active disease are generally prescribed first-line drugs which are ineffective against resistant strains. This selection pressure is exacerbated by the fact that latent infection (with either susceptible or resistant strains) likely protects individuals from subsequent infection. Therefore, fewer individuals being infected with susceptible strains year-on-year may inadvertently provide more fuel for the resistant strains to burn through.

A new study led by LSHTM’s Gwen Knight used mathematical modelling to estimate the global burden of latent MDR-TB. Estimates of the proportion of new cases of tuberculosis that MDR-TB accounted for were provided by model fits to data from 138 countries. The study estimated that 19·1 million people were latently infected with MDR-TB in 2014, and that children were at significantly increased risk. Critically, and contrary to WHO’s current assessment that MDR-TB has stabilised, the study showed that rates of MDR-TB were still increasing – a trend likely to be continued over the next 10-20 years.