Detection of Mycobacterium tuberculosis in gastric aspirate and sputum collected from Ethiopian HIV-positive and HIV-negative children in a mixed in- and outpatient setting.

AIM: To investigate, through a prospective study, the detection rate of Mycobacterium tuberculosis in sputa and gastric aspirate from tuberculous children in a low-income country with high prevalence of tuberculosis and an increasing HIV epidemic.
METHODS: Gastric aspirates and/or sputum samples were collected from 355 children with pulmonary tuberculosis as follows: from 136 children under 5 y only gastric aspirate was taken, for 159 children aged 5 to 9 y both methods were used, and for 60 children over 10 y only sputum was analysed. The diagnosis of tuberculosis was based on clinical data, tuberculin test and chest radiography. All children were tested for HIV infection.
RESULTS: Direct microscopy for acid-fast bacilli was positive for 55 (15%) and mycobacterial culture for 183 (52%) children. The proportion of positive cultures was similar in all age groups. Among the 5 to 9 year-old children who could produce a sputum sample, sputum gave just as good culture yield of M. tuberculosis as gastric aspirate. Of the clinical or radiological findings only weight loss was associated with a higher yield. Repeat gastric aspirate increased the culture yield by 6%. Mycobacterial culture from HIV-positive children gave lower yield compared with HIV-negative children.
CONCLUSION: Our data suggest that one gastric aspirate for children less than 6 y and three sputum samples for the older children collected at an outpatient TB clinic, is enough to provide a close to 50% yield of M. tuberculosis available for culture and further analyses. However, with an increasing prevalence of HIV, this detection rate may be reduced.