Sputum culture contamination and associated characteristics in a diagnostic clinical trial, Papua New Guinea

SETTING Papua New Guinea is a high-burden multidrug-resistant TB (MDR/RR-TB) country that reports low rates of bacteriological confirmation. Sputum drug susceptibility testing for second-line drugs is important. Access to culture is limited. OBJECTIVE To evaluate the prevalence of mycobacterial sputum sample culture contamination and determine factors associated with contamination. DESIGN A retrospective analysis of data from a TB diagnostic accuracy study that used culture as the diagnostic reference standard. Data included characteristics of people with presumptive pulmonary TB who provided sputum, sputum quality and culture results. RESULTS Sputum (1-3 samples) was collected from 174 adults. The median age was 33 years (IQR 24-47). Of 337 samples sent for culture, 28% were contaminated. Contamination was strongly associated with samples self-collected by participants outside clinic hours (aOR 5.69; 95% CI 2.62-12.38). Contamination risk increased with delays in shipping to the reference laboratory (aOR 1.19 per day, 95% CI 1.10-1.29) beyond the minimum 3 days. Contamination was less frequent among people aged 35-44 years compared to 18-24 years (aOR 0.27, 95% CI 0.10-0.73). Sputum quality was not associated with culture contamination. CONCLUSION Culture contamination could be reduced using spot sputum collection, expedited submission to laboratories and faster shipping when required.