The algorithms diagnostic accuracy was high. Clinical and clinical-radiological scores triggered most of the decisions to initiate TB treatment and highlight the importance of these assessments as part of a multi-faceted approach to diagnose TB in children.

A high proportion of children with SAM were diagnosed with TB in nutritional centers using the new treatment decision algorithms. Xpert MTB/RIF positive rate was low, as expected, however, it remains an important adjunctive diagnostic tool and aids detection of TB resistance. Chest X-ray was a useful tool for diagnosis, where available. TB diagnosis was primarily based on clinical features or history of contact person with TB.

The systematic use of POCUS in addition to TDAs would lead to an increase of TB cases. Using POCUS signs highly suggestive of TB would diagnose a lower additional number of children with TB but with lower risk of overdiagnosis.