Comparison of bacterial identification by MALDI-TOF mass spectrometry and conventional diagnostic microbiology methods: agreement, speed and cost implications.

Identification of microbial pathogens still relies primarily on culture and phenotypic methods, which is labour-intensive and time-consuming. In this study, identification of bacteria with valid standard identification using BD Phoenix, API panels and other recommended procedures is compared to identification with matrix-assisted laser desorption/ionisation-time of flight (MALDI-TOF) mass spectrometry using the MALDI Biotyper (Bruker Daltonics) in the setting of a routine NHS diagnostic microbiology laboratory. In total, 928 bacterial isolates obtained from blood (n=463), wounds and pus (n=208), respiratory tract (n=100), faeces (n=86) and urines (n=71) were analysed. There were 721 (77.7%) isolates with a MALDI Biotyper score > or =2.0, indicating secure genus and probable species identification; and 149 (16.1%) isolates with a score > or =1.7 and <2.0 indicating probable genus identification. The isolates with scores of > or =2.0 and > or =1.7 comprised 31 and 33 genera and 65 and 67 species, respectively. Overall, 99.4% and 99.1% of organism identifications were in agreement between the MALDI Biotyper and conventional identification at the genus level, and 89.3% and 87.8% at species level when analysing organisms with MALDI Biotyper scores > or =2.0 and > or =1.7, respectively. With many but not all organisms, identification at the genus level is sufficient; however, MALDI Biotyper separation of 208 staphylococci into Staphylococcus aureus and coagulase-negative staphylococci was always correct when scores were > or =1.7. First results were obtained after 5-10 min and analysis of a full 96-well target plate was completed in approximately 90 min. Substantial savings of between pounds 1.79 and pounds 2.56 per isolate, depending on the cost model of acquisition of the MALDI Biotyper system and number of isolates tested, would be realised when all 928 isolates were identified using the MALDI Biotyper and disk-susceptibility testing when compared to the cost for 618 Phoenix ID panels and 158 API panels and disk-susceptibility tests only (i.e., not taking into account costs incurred for identification of the remaining 152 mixed isolates). Microbial identification by MALDI Biotyper offers a rare opportunity for significant cost-neutral or even cost-saving quality improvements in medical diagnostics.