Statistics from Altmetric.com
Editor,—We read with interest the study of Zaret al on the usefulness of sputum induction in infants and young children for the diagnosis of pulmonary tuberculosis.1 Bacteriological confirmation of pulmonary tuberculosis in infants and young children remains a problem because it is difficult to obtain sputum. Therefore, in young children, gastric lavage is the recommended method for the collection of respiratory secretions.2 Since the number of tubercle bacilli and the frequency of positive cultures in specimens recovered by gastric lavage are usually small, gastric washings are ideally performed on three consecutive mornings to maximise the yield.3
In this prospective study, children with acute pneumonia with a high risk of pulmonary tuberculosis were included. On 142 children both gastric lavage and sputum induction was performed. The yield ofM tuberculosis in sputum and gastric lavage was compared, as was the amount of positive cultures in sputum and gastric lavage. The influence of HIV status on the yield was also determined.
The authors found more positive cultures in the induced sputa compared to the gastric lavages. Therefore they conclude that sputum induction was a more sensitive method than gastric lavage for culture of M tuberculosis. However, in order to compare the sensitivity of two diagnostic tests, one should perform both tests in all patients. In this study, 39 patients underwent only one gastric lavage, 77 patients had lavages on two consecutive mornings, and only 26 patients underwent all three gastric lavages.
We therefore disagree with the authors on one of the conclusions, that induced sputum is better than gastric lavage for the isolation ofM tuberculosis in infants and children. In our opinion, in order to answer the question whether sputum induction is as good as or better than gastric lavage, only the results from the patients who underwent gastric lavages on three consecutive days should be used.
Drs Zar, Tannenbaum, Apolles, Hanslo, and Hussey comment:
Dr Wiersma and colleagues suggest that the yield from a single sputum induction should be compared only with the results of those children who had three consecutive gastric lavages. Only 26 of our patients had three gastric lavages; among this subset however, four children were culture positive on sputum while only three were positive on gastric lavage.
Although the yield from gastric lavage is improved with increasing number of specimens, it is frequently not feasible to perform this procedure on three consecutive days, particularly in developing countries with limited resources. Moreover, performing three repeated gastric lavages may be very unpleasant, both to the child and the health worker. In practice, even in tertiary institutions such as those in which our study was performed, obtaining three sequential gastric lavages is rarely feasible.
The yield from sputum induction may also be increased with increasing number of specimens.1-1 Therefore we would submit that the yield from consecutive gastric lavages should be compared with that of repeated induced sputa. Data from studies of adult patients using paired specimens of induced sputum and gastric aspirates have reported a higher yield from sputa specimens.1-2-1-4 In our study, the findings that a single induced sputum specimen yieldedM tuberculosis more frequently than repeated gastric lavages (in the majority of children) further strengthens our conclusion.