Amsterdam institute for Immunology and Infectious diseases

Sepsis explained (3/3): Diagnosing sepsis

When a patient is admitted with suspected sepsis, doctors start antibiotics immediately, but for up to three days, they do not know exactly which microbe they are fighting. On Monday, we explained what sepsis is. On Tuesday, how it is treated. Today, in the final article of our series ahead of World Sepsis Day on 13 September, Dr. Iedan Verly, medical microbiologist at the Department of Medical Microbiology and Infection Prevention at Amsterdam UMC, Amsterdam institute for Immunology and Infectious Diseases,  explains how innovative diagnostics aim to shorten that delay, in theory to less than a single day, and how the research done within his department contributes to these new tools.

What are innovative diagnostics and why do we need them? 

For me, it all comes down to one goal: shortening the diagnostic delay. Bloodstream infections, which can lead to sepsis, are a medical emergency that requires prompt recognition and rapid treatment. To optimize that treatment, blood cultures are taken and incubated in the hope of identifying the cause. But this creates an inevitable diagnostic delay, and that delay consists of two parts. Part 1 is the time it takes for a blood culture to turn positive (to show sufficient growth) which usually happens within 15 to 24 hours. Part 2 is the time needed to identify the bacterium or fungus and to perform antimicrobial susceptibility testing (AST) which usually takes another one to two days. Together, that adds up to a delay of up to three days.

That delay is exactly why innovative sepsis diagnostics are so urgently needed. On the one hand, there is a severely ill patient who must be treated properly, on the other, we face a growing share of microorganisms that have become resistant to antibiotics. Concrete advice based on laboratory findings is, and probably always will be, essential to guide empirical therapy, the antibiotics doctors prescribe before the cause is known, and later the targeted therapy once it has been identified.

There are two ways to attack the delay. One is to focus on part 2: perform identification and susceptibility testing directly on the positive blood culture, skipping several laboratory steps, for example with the SepsiTyper® (Bruker) and VITEK REVEAL® (bioMérieux) systems. The other is more radical: bypass the blood culture altogether. Although commercial tests   that detect microbes directly from blood do exist,  they can identify only a limited panel of bacteria and usually give even more limited information regarding AST. A broader approach is to extract bacterial DNA directly from the blood and identify the cause using DNA PCR and/or sequencing and AI-algorithms. In theory, this could identify the cause of sepsis, including an estimate of its resistance pattern, in less than a single day. But the approach has drawbacks: it could be less sensitive, meaning it can miss the true cause (since the culture phase is skipped), and it is more prone to contamination, meaning it can detect a microbe that is not actually causing the disease. That is why it is not yet part of routine clinical practice.

"For me, it all comes down to one goal: shortening the diagnostic delay."

Dr. Iedan Verly

Medical microbiologist at the Department of Medical Microbiology and Infection Prevention of Amsterdam UMC and AI&I

How does your research at Amsterdam UMC contribute to the development, validation or implementation of innovative diagnostic tools for sepsis?

At our department, we are working on both sides of the delay. To improve the traditional blood culture, part 2 of the delay, we run several research projects with biotech spin-offs from TU Delft. Nostics: identifying microbes from a positive blood culture using SERS technology, a laser technique that recognizes microbes by their molecular fingerprint. BugSee: improving the preparation of positive blood culture samples for analysis. And SoundCell: performing susceptibility testing using nano-motion, thus detecting whether individual bacteria are still moving, and therefore alive, after antibiotic exposure.

To bypass the blood culture altogether, part 1 of the delay, we have created the PRISM consortium (Prevention and Rapid Identification of Sepsis via Molecular Pathways), which brings together the departments of pediatrics, hematology, intensive care and medical microbiology. Its studies compare traditional blood cultures with molecular culture: methods based on PCR, a technique that detects the genetic material of microbes, or on DNA sequencing, performed directly on blood. In this way, we hope to make sepsis diagnostics faster, while also ensuring diagnostic accuracy and logistic feasibility. The ultimate goal is to determine whether antibiotics could be safely stopped when the molecular culture turns out negative. 

Our department is also involved in projects using AI algorithms to decide whether blood cultures should be drawn from a patient at all and whether antibiotic therapy is indicated. In a way, this is the mirror image of the research we described on Monday: there, AI reads the patient's own data to recognize sepsis earlier, here, it helps decide which laboratory tests are truly needed.

What are the main challenges in translating novel sepsis diagnostics from research settings into routine clinical practice?

First, accuracy: tests performed directly on blood can be less sensitive than blood cultures and more prone to contamination. Second, logistics: a test that only works in a specialized laboratory does not help a patient at two in the morning, it has to fit into the daily routine of a hospital laboratory. Third, clinical utility: we must show that faster results actually change patient care, for instance, that a negative molecular result makes it safe to stop antibiotics. Without that evidence, new tests will not make it into routine practice.

This is the third and final article in our series. Read part 1, "What is sepsis?", and part 2, "Treating sepsis" on our website. 

Text: Iedan Verly and Esmée Vesseur