Big data used in treatment of sepsis

2 September 2016
Combining up to now fragmented patient data with information on sepsis treatment the moment a patient is received, should help clinicians in deciding what the right dosage of antibiotics should be. According to project leader Paul Elbers each week 300 sepsis patients are admitted to Dutch IC’s, 14,000 each year. Based upon US studies Elbers states that 28 percent of patients with sepsis dies almost immediately or shortly after admission. After a year the mortality rate goes up to 40 percent. 

80 percent of people going into septic shock die. One cause of this high death rate is the lack of data about a patients physical condition. Most specialists administer antibiotics based upon their experience, apply standard dosages or take in data of limited use, such as weight or length of the patient.

Giving the right treatment quickly, based upon personal patient data such as liver-, heart- and kidney functions, might help more people survive sepsis, Elbers believes. This data is collected realtime by medical devices monitoring the admitted patient. Elbers’ teamed developed analytic software that combines this data with anonymous sepsis patient data collected by IC’s (mostly dosages of antibiotics prescribed and the amount of antibiotics ending up in the bloodstream). This way a much more personalized treatment can be given.

Elbers says the adoption of EHR’s has been very helpful in collecting data on a large scale. If the project is successful, the software will be shared with other hospitals.