Smart alerts for early sepsis detection
Bisepro/Bisepro for Covid - IIC
CUSTOMER: Balearic Islands Health Research Institute
TECHNOLOGY: Rule-based system, Artificial Intelligence, Big Data.
Addressing needs and bringing benefits
The Multidisciplinary Sepsis Unit at Hospital Son Llàtzer in Palma de Mallorca is the first in Europe where different specialists, internists, emergency medicine doctors, surgeons, pneumologists, microbiologists and pharmacologists work daily with serious community and hospital infections, such as sepsis.
With this innovative contribution from the Institute of Knowledge Engineering (IIC) based on the application of Big Data techniques to the health sector and the knowledge of its specialists, the hospital hopes to adapt the system to the characteristics of septic patients themselves and reduce the false positive rate by 14%. The first results show that the discrimination between false positives and real cases is undoubtedly better and can therefore shorten the response time for treating sepsis, thus optimising hospital management, cutting costs and directly benefitting patients, as it can mean either saving or not saving a life.
The IIC’s contribution to this unit of the Intensive Medicine Service consists in applying predictive analytics techniques to issue smart alerts that suggest the appearance of sepsis at an early stage. At present, in our country there are over 130,000 cases a year, with 17,000 patients losing their lives every year.
The hospital’s surveillance monitors offer real-time information of the patients’ condition and the results of the clinical tests performed. These data are contrasted with a computerised repository of sepsis data and are interpreted with the aim of detecting infection and foreseeing multiple organ dysfunction, thus improving patients’ prognosis. For this reason, algorithmic predictive models are developed in the IIC, which are capable of providing a detailed data analysis based on real sepsis cases, at a very early stage of the illness.
Additional information: Borges, M., Socias, A., Castillo, A., Aranda, M., Pruenza, C., Estrada, V., Mena, J. and Diaz, J. Detection of sepsis and sepsis shock in hospitalized adult patients using Artificial Intelligence (AI) and Machine Learning (ML) techniques. ESICM LIVES 2019. ICMx 7, 55 (2019).