'Revolutionary' blood test diagnoses childhood fever causes in less than an hour

Gene-based test could help doctors identify causes of fever, sepsis and other illnesses in children

A blood test that could help diagnose the cause of a child's fever in less than an hour is being called 'transformative' by researchers. PA

A ground-breaking blood test promises to identify the cause of a child's fever in less than an hour.

The innovation could revolutionise diagnostics, particularly in developing regions, by accelerating the treatment process, potentially minimising the overuse of antibiotics, and offering greater clarity on illness origin.

Led by a global team from Imperial College London, the blood test is differs from conventional diagnostic tests, which often hunt for pathogens and can take more than 72 hours to yield results.

Instead, the new method assesses how specific genes in a child's blood react to various infections and diseases.

After analysing data from 1,212 patients across 18 disease classes, including respiratory syncytial virus (RSV) and tuberculosis, the researchers identified specific genes that activate or deactivate in response to these illnesses.

With machine learning, they recognised patterns associated with particular conditions, narrowing their focus to 161 pivotal genes.

To validate its efficacy, this approach was tested on 411 children admitted to hospital due to sepsis or severe infections from 13 of the identified 18 diseases.

Published in the Cell Press Med journal, the results showed the method's potential, emphasising its ability to enhance childhood disease diagnosis, curtail delayed or missed diagnoses, and notably affect global health care.

Study shows gene-based test can identify cause of fever in children quickly and accurately, with the potential to revolutionise diagnosis and treatment. AFP

Prof Michael Levin from Imperial College London described the test as "transformative for health care". He encapsulated the urgency of the situation: “When a child is hospitalised with a fever, doctors often treat based on a 'gut feel' of the probable illness causes.

"Our primary information comes from the child's symptoms, details from the parents, and our own medical training.

"However, confirming if the fever results from bacterial, viral, or other sources might take hours to days after admission".

According to Prof Levin these delays hinder prompt, accurate treatment. If implemented into point-of-care devices, this diagnostic method could be revolutionary.

Dr Myrsini Kaforou, a key contributor to the research, elaborated on the study's implications.

She said: “Out of the myriad genes in the human genome, we've pinpointed the molecular signature of several diseases based on just 161 genes. Our multi-disease, machine-learning diagnostic method is a testimony to the power of interdisciplinary collaboration and large research consortia, bridging infectious disease know-how, molecular science, and bioinformatics.”

While this promising development heralds a new era in quick diagnostics, Dr Kaforou emphasised that more work lies ahead to bring the test to clinical environments.

Yet the ultimate goal is clear: administer the right treatment, to the correct patient, at the optimal time, optimising antibiotic use, and slashing the protracted diagnosis duration for inflammatory diseases.

The researchers caution that before the RNA transcript panel enters clinical practice, further adaptations and rigorous testing are needed.

One potential benefit is its ability to quickly diagnose respiratory conditions such as bacterial pneumonia or tuberculosis.

Even if the causative pathogen resides in an organ like the lung, this innovative test could still detect its molecular signature in the bloodstream.

The next phase entails testing the method on thousands of patients across Europe, Africa and Asia.

As Prof Levin concluded: “The overarching drive behind this research is the well-being of our patients. Transitioning this platform into clinical use could truly revolutionise health services.”