A cloud-based screening system has shown its effectiveness at reducing deaths or other severe outcomes from patients with sepsis, according to a study released by a group of Los Angeles researchers. The program, called Clinical Decision Support, shows strong precision and an ability to protect patients from a highly dangerous complication that often strikes indiscriminately.
Sepsis is a life-threatening complication that happens when an infection triggers an inflammation response across the entire body. This can result in a cascade of further damage. In worst-case scenarios, the patient will progress to a state of septic shock that can lead to organ failure and either death or a severely compromised quality of life. Sepsis can occur in almost anyone but is most frequently triggered by bacterial lung or kidney infections in the elderly or those with weak immune systems. Detecting the symptoms and acting quickly is often key to preventing fatal or long-term damage.
The Clinical Decision Support (CDS) is a cloud-based program that embeds itself in a hospital’s electronic health records. It assesses patient medications, vital signs, lab results, nursing notes, and other data that is inputted regularly to monitor for signs of a septic reaction. If the CDS believes such a reaction is occurring, it alerts a nurse who then contacts the provider within five minutes of receiving the warning. Assessment of the patient’s condition and treatment then ensues.
The sample size was 6,200 hospitalizations that occurred within the participating healthcare facilities over a year-long period. Of these cases, the CDS issued alerts for 817 patients. Out of this number, three general groups were identified by researchers. The first, and perhaps most important, was the quarter of patients where the CDS correctly identified and issued a sepsis alert before the patient’s healthcare provider had ordered diagnoses or treatment for sepsis, but who did so after receiving the alert. The second group made up about half of the issued alerts and was undergoing diagnoses or treatment for sepsis at the time the CDS alert was delivered. The third and final group was where an alert was issued but no orders for diagnoses or treatment were issued before or after, presumably a false positive. Out of the remaining 5,383 patients that the CDS did not flag, 80 were later determined to have been false negatives.
These numbers, along with other forms of statistical data, were used to determine that the CDS method had an 83% sensitivity or ability to detect the appropriate condition. CDS was also found to flag sepsis with a 92% specificity, referring to how often it would react only to sepsis and not another condition. Overall, the use of CDS was associated with a 30% drop in hospice discharge or patient death during the time period observed during the study.
CDS is just the latest in a series of sepsis-detecting programs integrated into hospital electronic health records. A similar alert system, dubbed POC Advisor, showed similar results when it was assessed at Huntsville Hospital in Alabama in 2014.
Sources for Today’s Article:
Carey, B., “New Approach Advised to Treat Schizophrenia,” New York Times web site, October 20, 2015; http://www.eurekalert.org/pub_releases/2015-10/sp-ama101915.php.
Kane, J.M., “Comprehensive Versus Usual Community Care for First-Episode Psychosis: 2-Year Outcomes From the NIMH RAISE Early Treatment Program,” The American Journal of Psychiatry; http://www.medscape.com/viewarticle/845051.