Alerts could lead to earlier interventions for patients with the condition.
A digital alert system could reduce deaths and hospital stays in patients with sepsis, according to the findings of a study published in Journal of the American Medical Informatics Association.
The investigators observed that patients who triggered the automatic alert had 24% lower odds of in-hospital death and a 35% increased chance of receiving timely antibiotics when compared to patients for whom the digital alert system was not implemented for.
“Often digital systems are implemented but research on their performance is not done” lead author Kate Honeyford, Ph.D., from the Global Digital Health Unit at Imperial College London said in a press release. “Our study shows for the first time that robust analysis of a digital alert system was associated with improvements in outcomes for patients and the system presents an opportunity to improve care for patients who may have sepsis.”
Sepsis is a serious and life-threatening condition but if diagnosed early, it can be treated effectively with antibiotics. The condition, however, is difficult to spot because symptoms are similar to that of other illnesses such as influenza.
To combat this, Cerner Corporation developed a digital sepsis alert system that was introduced and implemented in 2016 at Imperial College Health Care NHS Trust’s hospitals in the U.K.
The sepsis alert system works by monitoring changes in patients’ temperatures, heart rates and glucose levels. If these rates fall outside of the range of safety, the system automatically notifies clinicians to investigate further. The notification is delivered through a pop-up warning in the electronic health record (EHR) or dashboard, which highlights any patient with an active alert upon opening the record.
In addition to the automatic alerts, the Imperial College Healthcare team designed a multidisciplinary care plan which is launched in the EHR upon the clinician’s confirmation of a sepsis diagnosis.
From there, the care team is prompted to determine optimal treatment methods and ensure that the patient receives antibiotics within one hour, which is in-line with national targets.
As part of the evaluation, the authors of the report analyzed more than 27,000 hospital stays of patients who triggered the alert system between May 2018. These patients were admitted into emergency departments or acute or hematology wards at St. Mary’s Hospital, Charing Cross Hospital and Hammersmith Hospital, all of which are part of the Imperial College Healthcare NHS Trust.
The researchers determined that admitted patients had a 4% lower chance of stay in the hospital for more than seven days than patients with similar symptoms who received standard of care.
Through implementing this system, it has been easier to alert clinicians of deteriorating conditions in patients and as a result, investigations and treatment plans have been implemented more quickly, the researchers suggested.
In the future, the study team will conduct a larger study involving more NHS hospitals to determine if the results are consistent in a large study population.
Editor’s note: The original article appeared on Contagion Live, a sister publication of Inside Digital Health™.
Get the best insights in digital health. Register for our newsletter.
Geisinger and IBM Create Machine-Learning Model to Detect Sepsis
Machine Learning Tool Outperforms Clinicians in Identifying Sepsis in Infants