Quality of induced sputum using a human-powered nebuliser in a mobile human immunodeficiency virus testing service in South Africa.

OBJECTIVES: To investigate the quality of induced sputum samples using a human-powered (HPN) and an electric-powered nebuliser (EPN). METHODS: For each participant two sputum samples were induced using the HPN and the EPN. The sequence of the two nebulisers was allocated at random. The proportion of good quality sputum according to different assessment criteria was compared using an exact McNemar test. The difference in time to expectoration was compared using the Wilcoxon matched-pairs signed-rank test. RESULTS: A total of 123 individuals were eligible for the study. Nine individuals refused to participate and five were unable to produce a sputum sample. The proportion of good quality sputum was higher among sputum samples induced by the HPN compared to those obtained using the EPN. The median time to produce a sputum sample was 2.2 min (IQR 1.13-4.1) for the HPN and 2.5 min (IQR 1.4-4.1) for the EPN. CONCLUSION: The HPN induced good quality sputum within 3 min. The device operates without electricity and is suitable not only for remote clinics with unreliable electricity, but also for mobile services and community-based intensified tuberculosis (TB) case finding. Further research needs to investigate the yield of TB in sputum samples induced by the HPN.