It can anticipate the risk of sepsis by 96%.
It has arisen from a research project at Son Llàtzer University Hospital in collaboration with the Institute of Knowledge Engineering and the Balearic Islands Health Research Institute Foundation.
This morning BiAlert was presented, the first Spanish system that uses artificial intelligence (AI) to anticipate the risk of sepsis with a 96% predictive capacity. This system has already obtained CE level IIa certification as a healthcare device.
TDB keeps you informed. Follow us on Facebook, Twitter and Instagram
Son Llàtzer University Hospital presents BiAlert
The presentation ceremony was attended by the Regional Minister of Health, Manuela García; the director general of the Health Service, Javier Ureña; the manager of Son Llàtzer University Hospital, Soledad Gallardo; the head of the Sepsis Unit of Son Llàtzer University Hospital, Dr. Marcio Borges; the director general of the Institute of Knowledge Engineering (IIC), Alberto Barrientos, and the manager of the Balearic Islands Health Research Institute Foundation (IdISBa), Carlos Enrique Herrero.
BIAlert Sepsis is the result of a research project at Son Llàtzer University Hospital in collaboration with the IIC, a pioneering R+D+i centre in AI and an expert in big data technologies for 35 years, and IdISBa.
Sepsis is a severe clinical syndrome caused by bacterial, viral or fungal infection, with an estimated mortality of 32-50%. It is the leading cause of hospital death worldwide and exceeds other adverse conditions such as ischaemic heart disease, stroke and most tumours. The incidence of sepsis is increasing by approximately 3% per year, partly due to the ageing of the population and the increase in patients with multiple pathologies, immunosuppressed patients and those undergoing surgery and aggressive treatments.
Son Llàtzer University Hospital has played a key role in the validation and testing of the BIAlert Sepsis system. After two and a half years of rigorous testing, the hospital has demonstrated the efficacy of the system in the early prediction of sepsis.
BIAlert Sepsis collects information every 30 minutes from the moment the patient is admitted and issues alerts to the medical team indicating the risk of sepsis 24 hours in advance.
The success of Hospital Universitario Son Llàtzer in implementing BIAlert Sepsis has led to its incorporation in two other hospitals. This tool not only improves the response capacity of medical staff, but also adapts to the specific characteristics of each hospital centre and allows for crucial personalisation to be effective.
During the prospective testing period, which ran from January 2019 to April 2022, the system issued 64,524 alerts, of which approximately 30 % were reviewed. In these cases, it demonstrated tremendous accuracy, with less than 1% false negatives. Compared to traditional methods of sepsis diagnosis, which offer 65% reliability and usually detect the condition when organ dysfunction is already present, BIAlert Sepsis has a predictive ability of 96% up to 24 hours earlier compared to other methods, and significantly reduces false positives, from more than 60% to less than 9%, and negatives.
Hospital Universitario Son Llàtzer has not only pioneered the implementation of BIAlert Sepsis, but has also actively contributed to its development and validation. Since 2018, the hospital has worked closely with the Institute of Knowledge Engineering using retrospective data from more than 200,000 patients to generate the tool and train the system.
This project has been possible thanks to the involvement of the different services of the Hospital Universitario Son Llàtzer, the Multidisciplinary Sepsis Unit, the Computer Service of the Hospital Universitario Son Llàtzer, the IdISBa Sepsis Group and the IIC.