According to the UK Sepsis Trust, five people die with sepsis every hour in the UK and survivors can suffer long-term health problems. The challenge for health practitioners is that the initial symptoms of sepsis can be similar to symptoms of other conditions, meaning diagnosis of sepsis is often fatally delayed.
Positive news reported in the Scientific American journal last month is the progress of a sepsis detection algorithm developed in the US that reduced deaths from sepsis by nearly 20 per cent in the hospitals at which it was trialled.
The algorithm works by quickly combining a patient's observations, laboratory test results and factors known to increase sepsis risk to determine the overall likelihood that the patient is suffering from sepsis. The algorithm then alerts staff to patients who are at risk.
Early warning systems using patient observations to monitor for sepsis already exist but what this algorithm adds is an holistic picture to provide targeted warnings. As a result, patients with sepsis on average received treatment nearly two hours earlier.
In the UK, researchers at Imperial College London have used AI to review the health records of around 100,000 intensive care patients suffering from sepsis, in order to develop a tool for optimising treatment.
As technology improves, there is a push to develop the use of AI in healthcare. In 2019, the UK Government invested £250m in the creation of an NHS AI Lab. The Lab recognises that AI has the potential to analyse large quantities of complex data to make a significant difference in healthcare.
Meanwhile, the 10th anniversary of World Sepsis Day on 13 September 2022 focussed on improving understanding of this condition. Too many people still lose loved ones or sustain permanent disabilities as a result of failures to recognise and treat sepsis until it is too late. Initiatives to prevent this, both through traditional awareness raising and medical training and via the use of cutting edge AI technologies are encouraging.
Sign up to our email digest