Résumé
Objective To determine the value of differential time to positivity ( DTTP ) of blood culture for the diagnosis of catheter-related bloodstream infection ( CRBSI ) in patients with solid tumors in intensive care unit ( ICU ). Methods A retrospective study was conducted. 615 pairs of peripheral vein blood cultures and instantaneous catheter tip blood culture of 615 patients admitted to ICU of Tianjin Medical University Cancer Institute and Hospital were collected from August 2011 to March 2014. The DTTP method and ( or ) semi quantitative culture of catheter tip were compared. CRBSI was diagnosed when both cultures were positive for the same microorganism and DTTP ≥2 hours ( 120 minutes ). The result of this procedure was compared with that of organism obtained using the semi quantitative culture of blood at catheter tip with≥15 cfu. Based on the clinical diagnosis, the reliability of two kinds of laboratory examination was compared for the diagnosis of CRBSI by plotting receiver operator characteristic curve ( ROC curve ). Results The result of 615 cases suspected of having CRBSI were analyzed during the study period. Of these, 440 episodes were excluded because cultures were negative for blood obtained through peripheral vein and central vein. Eight episodes were excluded because only peripheral vein blood culture was positive and 57 episodes were excluded because of only central vein blood culture was positive, 68 pairs of blood cultures were excluded due to the presence of multiple catheters and repeated blood withdrawals. Two cases of polymicrobial cultures were excluded from the final analysis due to the difficulty in determining the time of positive result for each individual microorganism. Ten cases in 42 cases of suspected cases of CRBSI were excluded from analysis because catheter was not removed, therefore culture from catheter tip could not be obtained. Using the DTTP method, 14 out of 17 CRBSI cases were diagnosed with DTTP≥120 minutes, while 3 cases were missed;the semi quantitative catheter tip culture was positive in 13 cases, and in 4 cases it was neglected. In 2 cases of CRBSI it was missed by both methods. The area under the ROC curve ( AUC ) of DTTP, catheter tip culture and the combination method was 0.912, 0.882 and 0.941 for diagnosis of CRBSI, respectively. Validity values for the diagnosis of CRBSI for DTTP were:sensitivity 82.35%, specificity 92.31%, positive predictive value 93.33%and negative predictive value 80.00%, and they were higher than those of the catheter tip culture method only ( 76.47%, 84.62%, 86.67% and 73.33%). The specificity and positive predictive CRBSI combination of the two methods in the diagnosis value were up to 100%, the sensitivity ( 88.24%) and negative predictive value ( 86.67%) was also increased, but no significant differences were found with DTTP method (χ2=0.00, P=1.00;χ2=0.00, P=0.98;χ2=0.00, P=0.98;χ2=0.00, P=0.98 ). Conclusions DTTP can be a valid method recommended for CRBSI diagnosis in critically ill patients with acceptable sensitivity, good specificity as well as positive predictive value. DTTP combined with other clinical symptoms can not only avoid unnecessary catheter withdrawal, but it also can help obtain the optimal treatment time and strategy.
Résumé
BACKGROUND: The diagnosis of catheter-related bloodstream infection (CRBSI) should demonstrate catheter colonization of the same organism as the isolate from peripheral blood cultures, by catheter tip culture or by differential time to positivity (DTP) of catheter-drawn blood cultures versus peripheral blood cultures. The purpose of this study was to compare the sonication and the roll-plate methods of catheter tip culture. METHODS: One hundred and sixty-one catheter tips from 122 patients were submitted for catheter tip culture. Distal segments of the catheter were first inoculated using a roll-plate, and then inoculated by sonication. Sonication was performed using a BactoSonic device (Bandelin GmbH, Germany). A total of 1,018 sets of blood cultures from 7 days before to 1 day after catheter removal were analyzed for isolated organisms and DTP. Cutoffs of catheter colonization were > or =15 CFU for the roll-plate method, > or =100 CFU for sonication, and > or =2 h for DTP. RESULTS: Twenty-four catheter tips (14.9%) showed colonization with at least one of the two methods: 21 (13.0%) with the roll-plate method and 22 (13.7%) with sonication. The positivity rates for the two methods showed no significant difference, and the concordance rate for the two methods was 96.9% (k=0.866, P<0.001). Blood culture was positive in 56 episodes in 44 patients, and 14 episodes of CRBSI were diagnosed in 12 patients: 10 by tip culture (two by sonication only) and 8 by DTP. Of the 122 specimens that were negative according to both methods, 4 were from the episodes of CRBSI diagnosed by DTP. CONCLUSION: Roll-plate and sonication methods are comparable in diagnostic sensitivity for catheter colonization. The roll-plate and sonication catheter tip culture methods and DTP are complementary for diagnosis of CRBSI.