Identifying organisms directly from blood culture bottles has expected benefits to clinical care. Many techniques have been published, typically requiring about 20 minutes of hands-on time.
Serum Separator Tubes (SSTs) can be used to separate bacterial cells from human red blood cells in blood culture broth. It was noted that the bacterial layer forming atop the gel in these tubes after centrifugation can be easily collected and plated directly to MALDI-TOF MS acquisition slides using a 1microL loop, thus producing a direct identification technique involving 1 centrifugation and less than 2 minutes of hands-on time per isolate. This technique was analysed prospectively to assess its accuracy and utility. On-spot protein extraction with 70% formic acid was used when gram-positive organisms were seen on gram stain. Identifications were accepted with a confidence score of >94.9%, only if concordant with gram staining. 363 isolates were analysed between August and December 2020, simultaneously with normal laboratory workflow, with identifications from the spin-and-stick method compared to identification made by MALDI-TOF MS from colonies after subculture to solid media.
The correct identification rate was highest for gram negative bacilli (93.5%) and Enterococcus faecalis (93.8%), and Streptococcus pneumoniae (100%), but low for coagulase-negative Staphylococcus spp. (25%). A marked difference in correct identification rate was noted for S. aureus between the aerobic and anaerobic bottles (21.4% versus 88.9%). This difference was not noted for other species. By applying probability theory to the data generated, the expected rate of identifying new Staph. aureus bacteraemia episodes was calculated as 80.4%. False identifications belonged to a recurring list of 8 species, the most common being Gemella sanguinis. False identifications met the pre-specifcied accepting criteria for 9/363 isolates, 7 of which were of G. sanguinis.
The spin-and-stick method was shown to correctly identify most organisms detected in blood culture, requiring negligible extra hands-on time for laboratory staff.