BMC Public Health (Jul 2023)

Epidemiology of childhood tuberculosis and predictors of death among children on tuberculosis treatment in central Ethiopia: an extended Cox model challenged survival analysis

  • Abay Burusie,
  • Fikre Enquesilassie,
  • Nicole Salazar-Austin,
  • Adamu Addissie

DOI
https://doi.org/10.1186/s12889-023-16183-9
Journal volume & issue
Vol. 23, no. 1
pp. 1 – 12

Abstract

Read online

Abstract Background Childhood tuberculosis (TB) was poorly studied in Ethiopia. This study aimed to describe the epidemiology of childhood TB and identify predictors of death among children on TB treatment. Methods This is a retrospective cohort study of children aged 16 and younger who were treated for TB between 2014 and 2022. Data were extracted from TB registers of 32 healthcare facilities in central Ethiopia. Phone interview was also conducted to measure variables without a space and not recorded in the registers. Frequency tables and a graph were used to describe the epidemiology of childhood TB. To perform survival analysis, we used a Cox proportional hazards model, which was then challenged with an extended Cox model. Results We enrolled 640 children with TB, 80 (12.5%) of whom were under the age of two. Five hundred and fifty-seven (87.0%) of the enrolled children had not had known household TB contact. Thirty-six (5.6%) children died while being treated for TB. Nine (25%) of those who died were under the age of two. HIV infection (aHR = 4.2; 95% CI = 1.9–9.3), under nutrition (aHR = 4.2; 95% CI = 2.2-10.48), being under 10 years old (aHR = 4.1; 95% CI = 1.7–9.7), and relapsed TB (aHR = 3.7; 95% CI = 1.1–13.1) were all independent predictors of death. Children who were found to be still undernourished two months after starting TB treatment also had a higher risk of death (aHR = 5.64, 95% CI = 2.42–13.14) than normally nourished children. Conclusions The majority of children had no known pulmonary TB household contact implying that they contracted TB from the community. The death rate among children on TB treatment was unacceptably high, with children under the age of two being disproportionately impacted. HIV infection, baseline as well as persistent under nutrition, age < 10 years, and relapsed TB all increased the risk of death in children undergoing TB treatment.

Keywords