论文部分内容阅读
Rapid antimicrobial therapy adapted to the susceptibility of bacteria responsible for bloodstream infections can reduce morbidity and mortality.In this study, self-constructed protein database combining with Surface enhanced laser desorption & ionization time of flight mass spectrometry (SELDI-TOF MS) was applied to identify the four most commonly encountered Gram-negative bacteria leading to bloodstream infection (E.coli, K.pneumoniae, A.baumanniiand P.aeruginosa).First, protein database was constructed with data of stably expressed protein peaks screened from ATCC reference strains and about 10 clinical isolates.The database was validated by 327 clinical isolates with identification rate of 92.44% (110/119) for E.coli, 91.34% (95/104) for K.pneumoniae, 98.41% (62/63) for P aeruginosa and 90.24% (37/41) for A.baumanniion agar plates, respectively.Then, protein fingerprints of 120 gram-negative bacteriafrom positive blood culture bottles were compared with the protein fingerprint database.The results showed that 93.75% of E.coli(45/48),86.96% of K.pneumoniae(20/23), 87.50% of P.aeruginosa(7/8)and A.baumannii (7/8) were correctly identified.Compared with the results of conventional approaches, the concordance was high for bacteria identification from blood culture bottles (Kappa=0.906).So, by using the protein fingerprint comparison, pathogenic bacteria causing bloodstream infection can be rapidly identified.