Date of Award
3-31-2022
Document Type
Master's Thesis - Restricted (NYMC/Touro only) Access
Degree Name
Master of Science
Department
Pathology
First Advisor
Utsav Pandey
Second Advisor
Carol Carbonaro
Third Advisor
Neeru Chopra
Abstract
Bloodstream infections are a leading cause of mortality in hospitals, with sepsis falling under the top ten causes of death in the United States. Due to this, microbiology lab results are critical when it comes to patient care. Within the first six hours of sepsis onset, each hour of delay in initiating appropriate antimicrobial treatment results in significant reduction of survival. Physicians tend to treat patients with suspected bloodstream infections with broad spectrum antibiotics. However, it is not enough to treat the patient as the selected therapy may not be appropriate for the infecting organism. Additionally, it is known that overuse of broad-spectrum therapy and unnecessary treatment of contaminants can lead to adverse effects including treatment toxicity, unnecessary hospital costs and increasing antibiotic resistant levels. Timely diagnosis is incumbent as mortality rates increase each hour that improper antimicrobial agents are used to treat patients with sepsis. Thus, rapid, and efficient organism identification and antimicrobial susceptibility reports will aid to improve patient outcomes by providing faster care, and reducing hospitalization stays and costs.
Rapid and accurate organism identification and antimicrobial susceptibility reports are paramount for patient survival during sepsis. The Bruker MBT Sepsityper® Kit allows harvesting of bacteria directly from positive blood cultures and pairs it with the MALDI Biotyper System for rapid identification. In this study, the accuracy of the Sepsityper® kit for bacterial identification was compared to culture-based methods. Further, as an off-label use of the kit, rapid antimicrobial susceptibility results for gram-negative bacteria were setup directly using the harvested pellet and was compared to the established method in the laboratory.
A total of 130 blood cultures collected at Westchester Medical Center from February-September 2021 were included. Bacterial extraction was performed as per the Sepsityper® protocol and identification using MALDI Biotyper CA System. Log(score) values of ≥1.80 from MALDI-ToF MS are considered high confidence using the Sepsityper® kit. Antimicrobial susceptibility was setup for 50 blood cultures with gram-negative bacteria using the remnant bacterial pellet on the MicroScan® instrument. Antimicrobial susceptibility results were compared for essential and categorical agreement. Time from positive blood culture Gram stain results to final blood culture report was calculated for 20 cultures.
Out of the 130 blood cultures analyzed, 123 (94.6%) produced concordant identification. Among the 123 concordant results, 92 (74.8%) had a log(score) of ≥2.0, 17 (13.8%) between 1.80-2.0, and 14 between 1.60-1.79. While 100% of the identifications were accurate, 7 cultures, 3 with gram-positive and 4 with gram-negative bacteria, failed to identify using the Sepsityper® kit but were identified using established methods. Confidence scores for gram-negative bacteria were higher compared to gram-positive. Direct antimicrobial susceptibility results showed an overall 98.4% essential and 96.7% categorical agreement. The average minor, major and very major error rates were 2.5%, 0.6% and 1.0% respectively. The average time from Gram stain to bacterial identification and antimicrobial susceptibility results using the Sepsityper® kit was 21.83 hours versus 52.34 hours for established methods.
Sepsityper® kit can be implemented for rapid and highly accurate bacterial identification and antimicrobial susceptibility testing directly from positive blood cultures and can significantly decrease the time to final report.
Recommended Citation
Singer, Lindsey, "Rapid Bacterial Identification and Antimicrobial Susceptibility Testing from Positive Blood Cultures and its Impact on Patient Care" (2022). NYMC Student Theses and Dissertations. 33.
https://touroscholar.touro.edu/nymc_students_theses/33
