Multiple sampling in one day to optimize smear microscopy in children with tuberculosis in Yemen

The diagnosis of pulmonary Tuberculosis (TB) in children is difficult and often requires hospitalization. We explored whether the yield of specimens collected for smear microscopy from different anatomical sites in one visit is comparable to the yield of specimens collected from a single anatomical site over several days.

Children with signs/symptoms of pulmonary TB attending a reference hospital in Sana'a Yemen underwent one nasopharyngeal aspirate (NPA) the first day of consultation and three gastric aspirates (GA) plus three expectorated/induced sputa over 3 consecutive days. Specimens were examined using smear microscopy (Ziehl-Neelsen) and cultured in solid media (Ogawa). Two hundred and thirteen children (aged 2 months-15 years) were enrolled. One hundred and ninety seven (93%) underwent nasopharyngeal aspirates, 196 (92%) GA, 122 (57%) expectorated sputum and 88 induced sputum. A total 1309 specimens were collected requiring 237 hospitalization days. In total, 29 (13.6%) children were confirmed by culture and 18 (8.5%) by smear microscopy. The NPA identified 10 of the 18 smear-positives; three consecutive GA identified 10 and induced/expectorated sputa identified 13 (6 by induced, 8 by expectorated sputum and one positive by both). In comparison, 22 (3.7%) of 602 specimens obtained the first day were smear-positive and identified 14 (6.6%) smear-positive children.

The examination of multiple tests the first day of consultation identified a similar proportion of smear-positive children than specimens collected over several days; would require half the number of tests and significantly less hospitalization. Optimized smear microscopy approaches for children should be explored further.