Comparative Performance of Modified Kenneth Jones Criteria Scoring, World Health Organization Criteria, and Antibodies in Lymphocyte Supernatant for Diagnosing Tuberculosis in Severely Malnourished Children Presenting With Pneumonia.

Background: The diagnosis of childhood tuberculosis (TB) can be difficult in severely malnourished children. This is mainly due to the fact of our perceived notion that clinical signs of TB are often subtle in severely malnourished children and we may rely on laboratory investigation for the diagnosis. However, comparative data on the performance of clinical and laboratory diagnostics of TB in such population are also very limited. Objectives: To compare the performance of composite clinical criteria and a technique that measures antibodies in lymphocyte supernatant (ALS) for the diagnosis of TB in severely malnourished children with pneumonia. Methods: Severely malnourished children under five with radiological pneumonia admitted to the Dhaka Hospital of International Centre for Diarrhoeal Disease Research, Bangladesh were enrolled consecutively following informed consent. We collected venous blood for ALS, gastric lavage fluid and induced sputum for microscopy, mycobacterial culture, and real-time PCR by XpertMTB/RIF. We compared the sensitivity, specificity, positive, and negative predictive values, and accuracy of modified Kenneth Jones criteria (MKJC) score, World Health Organization (WHO) criteria, and ALS in diagnosing TB in severely malnourished children with pneumonia for "Confirmed TB" and "All TB" ("Confirmed TB" plus "Probable TB") vs. "Not TB." Results: Compared to culture confirmed TB, the sensitivity, and specificity (95% CI) for MKJC were 60 (27-86) and 84 (79-87)% and for WHO criteria were 40 (14-73) and 84 (80-87)%, respectively. Compared to culture and/or Xpert MTB/RIF positive TB, the sensitivity and specificity (95% CI) for the criteria were 37 (20-58) and 84 (79-87)%; and 22 (9-43) and 83 (79-87)%, respectively. For both these comparisons, the sensitivity and specificity of ALS were 50 (14-86) and 60 (53-67)%, respectively. Conclusion: Our data suggest that WHO criteria and MKJC scoring mainly based on clinical criteria are more useful than ALS in diagnosing TB in young severely malnourished children with pneumonia. The results underscore the importance of using clinical criteria for the diagnosis of TB in severely malnourished children that may help to minimize the chance of over treatment with anti-TB in such population, especially in resource limited TB endemic settings.