News: EHR sepsis alert volume increased during pandemic even though census declined

A study published by the Journal of the American Medical Association (JAMA) found that though the COVID-19 pandemic caused reduced censuses for most hospitals, the amount of sepsis EHR alerts increased 43% during the same timeframe.

The study collected the number of alerts generated by the Epic Sepsis Model (ESM) in 24 hospitals for the months before and during the pandemic with the goal of evaluating the extent to which nursing reports of sepsis over-alerting were linked to the alert volume. Additionally, the study aimed to examine the variation in alert volume across United States-based hospitals.

In the three weeks before and after the first reported case of COVID-19 in the United States health system in the study, the proportion of sepsis alerts per day more than doubled from 9% of patients to 21%. During that time period, however, the total hospital census declined by 35%. The total number of alerts per day increased by 43%, despite the lower hospital census.

According to the study, “although the increase in the proportion of patients generating sepsis alerts in this study can be explained by the cancellation of elective surgeries and a higher average acuity among the remaining hospitalized patients, the 43% increase in total alerts illustrates the increased alerting burden imposed by COVID-19 on a sepsis model.”

Researchers note that the study was limited as they did not evaluate the model’s accuracy, however even if the alerts were accurate, many existing sepsis workflows are built around bacterial sepsis and thus may not be entirely appropriate in the context of COVID-19.

The study concludes that “being able to rapidly assess and disable AI-based alerts is a responsibility faced by health systems using AI to support clinical care. Given the susceptibility of AI-based systems to changes in alerting patterns, clinical AI governance within health systems may play a role in monitoring and supporting deployed AI systems.”

Editor’s note: The full JAMA study can be found here. On Tuesday, December 7, ACDIS and AHIMA will be hosting a free webinar focused on best practices in CDI technology use and functionality, including AI systems. Click here to register.