Abstract:

Objective To investigate the contamination rate and contaminating bacterial species in a MGIT 960 system inoculated with pretreated sediment from sputum samples with different properties. Methods Sputum specimens from outpatients with suspected pulmonary tuberculosis were collected from January to February 2018 at the Beijing Chest Hospital. One hundred and sixty-two randomly-selected sputum samples, including 20 caseous sputum samples, 122 mucus sputum samples and 20 bloody sputum samples were included. Each sputum sample was divided into two parts and each part was then treated for 15 min with a sputum treatment solution, recording the weight of the sediment resulting after pretreatment for one part, and using the other part to inoculate two liquid culture tubes for culture (one with 0.5 ml supernatant (supernatant inoculation group) and the other with 0.5 ml sediment (sediment inoculation group). Bacterial species in contaminated samples were identified by sequencing the 16S rDNA segment. Results The average weight of pretreatment liquid sediment for ‘caseous sputum’ samples was (0.17±0.14)g and was significantly higher than that for ‘mucus sputum’ samples ((0.09±0.07)g)(F=14.56, P=0.00) and ‘bloody sputum’ samples ((0.10±0.07)g)(F=4.29, P=0.01). The quality of the pretreatment liquid sediment from ‘bloody sputum’ samples was slightly higher than that from ‘mucus sputum’ samples, but the difference was not statistically significant (F=0.09, P=0.80). The contamination rate for the sediment inoculation group was 16.05% (26/162) and was higher than that for the supernatant inoculation group (0.00% (0/162)). This difference was tested using Fisher’s exact test and was statistically significant (P=0.00). The contamination rate for the sediment inoculation group was highest for ‘bloody sputum’ (50.00%, 10/20), followed by ‘caseous sputum’ (15.00%, 3/20) and ‘mucus sputum’ (10.66%, 13/122), and differences were statistically significant (χ ²=15.61, P=0.00). Ten bacterial species were found among 21 cases of contamination. DNA could not be amplified from the remaining 5 cases of contamination. Bacillus sp. and Streptococcus sp. were dominant, accounting for 23.08% (6/26) and 15.38% (4/26) cases, respectively. Conclusion Pretreatment liquid sediment in sputum specimens is one of the main reasons for the high contamination rate. Avoiding sediment inoculation can effectively reduce contamination rates.

Key words: Mycobacterium tuberculosis, Sputum, Specimen handling, Culture media, Microbial contamination, Factor analysis, statistics