Fast sepsis detection from a single drop of blood
Nature Biomedical Engineering
March 20, 2018
A simple test that can detect sepsis from a single drop of blood is reported in a paper published online this week in Nature Biomedical Engineering. The test is quick, cheap and accurate, and could be easily adapted to monitor patients at risk of sepsis.
Sepsis - a life-threatening condition in which the body’s extreme response to severe infection damages tissues and organs - is misdiagnosed in approximately 30% of patients. This is because current tests have poor specificity and are slow - taking days to provide results, which leads to unnecessary antibiotic prescriptions, thus promoting the spread of antibiotic-resistant strains of bacteria.
Daniel Irimia and colleagues designed a device in which a drop of blood fills a maze of microscopic channels. A machine-learning algorithm then correlates the movement of neutrophils - the immune system’s ‘first responders’ - in the maze with sepsis severity to calculate a ‘sepsis score’. The researchers show that the test runs in only a few hours and that in a double-blind observational study of 42 patients the sepsis score differentiated healthy individuals from those with sepsis with over 95% sensitivity and specificity.
Although the test needs to be validated using a larger and more diverse set of patients, it has the potential to increase the survival rates of patients at high risk of sepsis and to reduce antibiotic overuse.
doi: 10.1038/s41551-018-0208-z
Research highlights
-
May 12
Geoscience: Monitoring earthquakes at the speed of lightNature
-
May 4
Microbiology: Bacteriophage therapy helps treat multi-drug resistant infection in an immunocompromised patientNature Communications
-
Apr 27
Planetary science: Building blocks of DNA detected in meteoritesNature Communications
-
Apr 8
Health: Psilocybin use associated with lower risk of opioid addictionScientific Reports
-
Apr 5
Energy: Winterizing the Texan energy infrastructure pays off in the long termNature Energy
-
Mar 17
Neuroscience: Sample size matters in studies linking brain scans to behaviourNature