High burden of childhood tuberculosis in migrants: a retrospective cohort study from the Thailand–Myanmar border

Abstract

Background

Tuberculosis (TB) is a leading cause of morbidity and mortality in children but epidemiological data are scarce, particularly for hard-to-reach populations. We aimed to identify the risk factors for unsuccessful outcome and TB mortality in migrant children at a supportive residential TB programme on the Thailand–Myanmar border.

Methods

We conducted retrospective analysis of routine programmatic data for children (aged ≤ 15 years old) with TB diagnosed either clinically or bacteriologically between 2013 and 2018. Treatment outcomes were described and risk factors for unsuccessful outcome and death were identified using multivariable logistic regression.

Results

Childhood TB accounted for a high proportion of all TB diagnoses at this TB programme (398/2304; 17.3%). Bacteriological testing was done on a quarter (24.9%) of the cohort and most children were diagnosed on clinical grounds (94.0%). Among those enrolled on treatment (n = 367), 90.5% completed treatment successfully. Unsuccessful treatment outcomes occurred in 42/398 (10.6%) children, comprising 26 (6.5%) lost to follow-up, one (0.3%) treatment failure and 15 (3.8%) deaths. In multivariable analysis, extra-pulmonary TB [adjusted OR (aOR) 3.56 (95% CI 1.12–10.98)], bacteriologically confirmed TB [aOR 6.07 (1.68–21.92)] and unknown HIV status [aOR 42.29 (10.00–178.78)] were independent risk factors for unsuccessful outcome. HIV-positive status [aOR 5.95 (1.67–21.22)] and bacteriological confirmation [aOR 9.31 (1.97–44.03)] were risk factors for death in the secondary analysis.

Conclusions

Children bear a substantial burden of TB disease within this migrant population. Treatment success rate exceeded the WHO End TB target of 90%, suggesting that similar vulnerable populations could benefit from the enhanced social support offered by this TB programme, but better child-friendly diagnostics are needed to improve the quality of diagnoses.