|
|
Application of infection biomarkers in early diagnosis of multiple resistant
bacteria bloodstream infection
XIA Fei, HU Guang-xu, LIAO Ya-ling, XU Xin, CHEN Min, DU Ming
2020, 48 (1):
50-54.
doi: 10.11958/20190785
Objective To evaluate the role of bacterial infection related inflammatory biomarkers in identifying multi-drug resistant (MDR) bloodstream infections (BSIs) confirmed by blood culture (BC) and predicting pathogen types.
Methods A retrospective analysis was conducted on 852 MDR infection patients in the Third People’s Hospital of Hubei
Province, while 526 patients with concurrent laboratory test results including blood culture, routine blood tests, C reaction
protein (CRP), procalcitonin (PCT) and serum pancreatic amyloid (SAA). The differences of the above indicators were
analyzed in negative blood culture and positive blood culture groups, and the diagnostic value for the bloodstream infections
was also evaluated. Results In 526 patients, there were 108 BSIs cases, including 42 cases (38.9%) were infected with
Gram-positive bacteria and 66 cases (61.1%) were infected with Gran-negative bacteria. MRSA (23 cases), ESBL-positive
Escherichia coli (39 cases) and ESBL-positive klebsiella pneumoniae (20 cases) were the major pathogens in BSIs. In BC
positive cases, there were significant differences in CRP, PCT, SAA and NEU between Gram-negative bacteria group and
Gram-positive bacteria group (Z=2.448, 5.647, 3.368, 4.905, all P＜0.05). Within cases of Gram-positive bacteria, there
was statistical significance in CRP (H=19.021, P＜0.001), with the highest level of MRSA (114.35 mg/L). Within cases of
Gram-negative bacteria, both CRP and PCT showed statistical significance (H=19.369, 15.013, P＜0.01), with the highest
level in ESBL-positive escherichia coli (164.60 mg/L, 97.42-217.50; 7.361 µg/L, 3.51-9.95). The area under the receiver
operating characteristic curves (ROC-AUCs) of CRP, PCT, SAA, WBC and NEU for discriminating positive MDR-BC from
MDR-BC negative cases were 0.778, 0.728, 0.658, 0.578, and 0.645, with statistical significance in CRP and PCT (P＜
0.01). For discriminating Gram-negative bacteria BC, AUCs of CRP and PCT were 0.692 and 0.883 with statistical
significance (P＜0.01). Conclusion CRP is more applicable for the early diagnosis of BSI compared with PCT. After
confirming BSI, PCT is suitable for discriminating Gram-negative bacteria infection. Therefore, the combination of PCT and
CRP may help medical staff to judge the BSI and type the possible pathogens, and to rationally use drugs
Related Articles |
Metrics
|
