Abstract
Bloodstream infections (BSIs) lead to sepsis, an acute inflammatory response associated with high mortality. Current clinical treatment pathways rely on culture-based diagnostics to identify the infectious agent and determine its susceptibility to antibiotics, delaying targeted therapy. Recent approaches to accelerate this focus on the isolation of bacteria directly from blood cultures for rapid downstream analysis using haemolytic agents to replace the gold-standard purity plate subculture. However, many of these methods impact bacterial viability during the early stages of growth. Rapid phenotypic antimicrobial susceptibility testing requires a pure inoculum of actively dividing bacteria to deliver early results (within 3 h), but most isolation chemistries are developed for mass spectrometry identification which does not require viable bacteria. This is particularly relevant for Gram-positive pathogens, as these are more susceptible to detergent agents used for haemolysis. This study investigated five different blood lysis protocols for the isolation of two different bacteria; Escherichia coli (Gram-negative) and Staphylococcus aureus (Gram-positive) from positive blood bottles. Bacterial viability was determined with microfluidic impedance cytometry and standard colony counting, which was also used to determine the extraction efficiency. For all five protocols, the numbers of extracted bacteria were generally high for the two organisms when determined with colony counting (following overnight incubation). However, when determined immediately after extraction with impedance cytometry, the counts were lower for S. aureus. The bacterial viability was also investigated with optical flow cytometry immediately after extraction showing inhibition of S. aureus growth with Sepsityper® but not with the milder detergent saponin.