Can CAP Proficiency Testing Samples be Adequately Processed with Laboratory Automation? A Proof of Concept Study



Microbiology laboratories are required to test proficiency samples to assess performance. The CAP provides multiple proficiency testing (PT) samples annually to laboratories to be tested per their routine procedures. For many bacterial challenges, PT samples consist of fiber swabs inoculated with organisms that are to be processed manually per CAP instructions. However, some laboratories are processing patient samples using laboratory automation (LA); thus, there is a need for these laboratories to have information on the appropriate automated method for processing PT samples. This study was undertaken to design a workflow which would allow CAP PT samples to be processed using LA.


The first step of the study consisted of defining a possible automated workflow that would equate to the manual procedure currently included in CAP PT instructions. For this, 15 blank CAP fiber swabs were inoculated with suspensions containing 100μl of 108, 106 and 104 CFU/mL of E. coli ATCC 25922 (5 swabs per concentration). One from each concentration was processed manually per CAP instructions (method 1), the remaining swabs were processed using 4 different automated methods (2A, 2B, 3A, 3B). The second step consisted of using the WASP®DT: Walk-Away Specimen Processor to process 3 retained CAP PT samples with the method selected as optimal from the first step. Plates were incubated using Full Laboratory Automation system. WASPLabTM and images of growth taken after 18, 24, 36 and 48 hours of incubation.


The swab inoculated with 104 CFU/mL of E. coli and processed with method 3A was determined to be optimal. Method 3A compared favorably to the CAP PT direct manual inoculation procedure (showing 63 CFU and 196 CFU, respectively) and it was the most efficient method studied. Briefly, method 3A consisted of vigorously mixing the PT fiber swab in 1ml of ESwabTM Amies medium, then pressing and rotating the swab against the side to the tube before discarding. For the 3 PT samples assayed using this method, the manual and automatic processing showed the same results: PT #1 grew gram negative rods, PT #2 grew gram positive cocci, and PT #3 also grew gram positive cocci, all in similar quantities.


Further testing with more challenging PT samples (including fastidious organisms, anaerobic organisms, and mixed samples) is needed, but this preliminary study suggests that the automatic processing of PT samples by LA is feasible. This will allow the laboratory to assess its performance using its routine automated workflow, as well as follow CLIA PT guidance in processing PT samples in a similar manner to that used for patient specimens.