Abstract
BACKGROUND
TB contributes to increased morbidity and mortality in children with vulnerable immune systems such as the severely malnourished, HIV infected or the infants. Clinical features might be poor predictors of active TB infection especially in children with severe malnutrition. This study determined the prevalence; estimated additional yield of TB cases on routine screening compared to targeted screening approach and studied the clinical predictors of tuberculosis disease in children between 2 months and 5 years with severe acute malnutrition.
METHODS
A cross sectional study of newly admitted children with severe malnutrition aged between 2 months and 5 years was conducted between March and September 2014 at Mbarara Regional Referral Hospital. Written informed consent was obtained from guardians. A detailed history, general physical examination, and investigations which included specimen collection by gastric, nasopharyngeal and or lymph node aspirate methods, as well as TST and CXR were done. Children were classified according to level of certainty of TB diagnosis as “confirmed”, “probable”, “possible”or “TB unlikely”. The proportional yield by routine screening and subgroup of targeted screening was determined. Logistic regression was done to determine independent predictors of TB.
FINDINGS
A total of 172 children had complete TB evaluation. The prevalence of confirmed/probable TB (TB cases) was 6.4% (11/172); Of the 11 TB cases, 4 were confirmed; 3 of whom had a positive smear, Xpert/MTB/RIF and culture results while 1 had a smear positive result only. Although, there was no statistical difference in TB yield between targeted and routine screening of TB in this population (p-value>0.05), there were 4 more TB cases identified through routine screening. Severe wasting, cervical lymphadenopathy and age group below 1 year had a statistically significant association with tuberculosis (p¼0.0002).
INTERPRETATION
We found a high prevalence of TB cases using NIH criteria, but low rates of Xpert/culture-confirmed TB among severely malnourished hospitalized children. Due to very unspecific presentation of TB in this population, evidenced by lack of statistical associations with documented predictors of TB, routine screening of all severely malnourished children for TB may offer clinical benefits
