Rapid diagnosis of sepsis pathogens

Abstract

Sepsis is a growing global healthcare concern and is related to high morbidity and mortality. Studies have repeatedly demonstrated that the rapid administration of appropriate antimicrobial treatment is crucial for patient survival. Thus, receiving timely and actionable information from the laboratory on identification of the microorganism causing sepsis is crucial for patient management.

In today’s routine diagnostics, blood culture is the standard method for diagnosing sepsis and identification of microorganisms is based on sub-culturing the positive blood culture bottles.

The overall aim of this thesis was to evaluate and improve the use of rapid identification methods in identification of microorganisms directly from blood culture bottles.

The thesis focused on two methods, FilmArray and MALDI-TOF MS. The present studies showed that the identification of microorganisms from a blood culture bottle by FilmArray and MALDI-TOF MS were ready in 65 min and 30 minutes respectively.

The most common form of sepsis is caused by a single microorganism. In paper 1 we studied the performance of FilmArray in identification of microorganism’s form blood culture bottles. The FilmArray could identify the microorganism in 91.6% of the blood culture bottles.

The anaerobic bacteria are not covered by current rapid identification methods including FilmArray. In study 3, we analyzed the performance of MALDI-TOF MS in identification of anaerobic bacteria from four different blood culture bottle types. MALDI-TOF MS could identify anaerobic bacteria in between 75-79% of the different blood culture bottle types.

The incidence of detection of polymicrobial growth in blood culture bottles is increasing. This is an obvious challenge both for conventional and rapid identification methods. In study 1 and 4 we evaluated the performance of rapid methods in identification of polymicrobial growth directly from blood culture (BC) bottles after positivity. FilmArray correctly identified all microorganisms in 17/24 (71%) and 99/115 (86.1%) of the BC bottles in study 1 and 4 respectively. In contrast, the present MALDI-TOF MS method showed poor performance and could identify both microorganisms in only 2/115 (1.7%) blood culture bottles.

The high analytical performance of the current rapid methods stimulated us to ask the question if we can identify microorganisms from bottles before the blood cultures signals positive. We called this unique approach as semi-culture based identification since the full- term culture is not needed.

In study 2, we analyzed the semi-culture based identification by FilmArray and MALDI-TOF MS. We analyzed both simulated and clinical blood cultures with this approach. MALDI- TOF MS failed to identify the microorganisms prior to positivity even in the simulated blood culture bottles. Interestingly, FilmArray could identify microorganisms from bottles before the blood cultures signals positive both in simulated and clinical blood culture bottles. In simulated samples, the median time to detection (TTD) of growth for the bottles in the blood culture system was 11.1 h, whereas FilmArray could identify microorganisms after 5 h incubation in the system.

In conclusion, the present thesis shows that the FilmArray is a reliable method for identification of microorganisms from positive blood culture bottles with mono- as well as polymicrobial growth. The data from the studies showed also that it is possible to improve the use of rapid identification methods as in the case of semi-culture based identification.