Clinicians Are Skeptical of Early Warning Systems for Sepsis

By Elaine Chen, MD

Assistant Professor, Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, Section of Palliative Medicine, Rush University Medical Center

Dr. Chen reports no financial relationships relevant to this field of study.

SYNOPSIS: While early warning systems for sepsis lead to clinical action, clinicians are skeptical and do not perceive them to be beneficial.

SOURCE: Guidi JL, et al. Clinician perception of the effectiveness of an automated early warning and response system for sepsis in an academic medical center. Ann Am Thorac Soc 2015;12:1514-1519.

Severe sepsis is very common, with high morbidity and mortality. Early recognition and intervention improves mortality. However, the diagnosis may often be missed in early sepsis. An academic health system developed an electronic early warning and response system (EWRS) for sepsis in 2012, monitoring real time vital signs and laboratory data for hospitalized, non-ICU, acute care patients and notifying clinicians when specific criteria were met. This EWRS accurately identified patients at increased risk for deterioration and death, resulting in more timely sepsis care and ICU transfer, and possibly reducing sepsis mortality.¹

All non-ICU medical and surgical inpatients were screened continuously for systemic inflammatory response syndrome (SIRS) criteria as well as criteria suggesting organ dysfunction. Whenever a patient fulfilled four or more criteria, a text page was sent to the covering provider (physician or advanced practice provider) and rapid response coordinator, and the bedside nurse received a pop-up notification in the electronic health record (EHR). Clinicians were instructed to meet at the bedside within 30 minutes to evaluate the patient and make any management changes. A patient could only trigger the alert once during hospitalization.

Successful implementation of new clinical systems includes clinician acceptance. The authors hypothesized that clinicians would perceive the EWRS as useful and effective. They tested their hypothesis by surveying clinicians immediately after receiving the EWRS alert to evaluate their perception of the value of the alert. A 16-item questionnaire examined the utility of implementation of the EWRS. The rapid response coordinator distributed paper surveys within 2 hours of an EWRS alert for 6 weeks. Subjects included providers and bedside nurses in a single academic medical center. Anonymously completed surveys were returned to a designated envelope.

The EWRS generated 247 alerts; 494 surveys were distributed, and 232 were returned (127 from providers and 105 from nurses) for an overall response rate of 47%. Both providers and nurses reported that patients were medically stable both before and after the alert in approximately 80% of cases and did not commonly perceive the presence of a new critical illness. Sepsis was the suspected trigger in one-third of cases and volume depletion in one-fifth. In one-third of cases, clinicians perceived the values to be erroneous at baseline or inconsequential.

Management changed in approximately half of patients, most commonly by way of closer monitoring, basic diagnostic testing, or therapies such as intravenous fluids and antibiotics. Less than half of providers or nurses found the alert helpful and less than one-third thought it improved patient care. Nurses thought more favorably about the EWRS than providers.

Overall, clinicians were lukewarm in their support of the EWRS. The authors noted that users received no formal education regarding the importance of early sepsis recognition and treatment. The authors suggested that the alert may have resulted in behavioral modifications that improved clinical outcomes, but that clinicians may not have appreciated that the tool was a catalyst to better patient care. Alternatively, the alert may have caused pressure to order tests or escalate care in patients who did not truly require it. That most patients were perceived to be stable both before and after the alert may be a sign of low signal-to-noise ratio, which has a risk of leading to alert fatigue. Further investigation can focus on acceptability, resource allocation, and system improvement.

COMMENTARY

Previously, the article by Umscheid et al described with great enthusiasm some potential benefits of the EWRS.¹ In contrast, this article tempers that enthusiasm, with only moderate clinician support. While clinicians are often initially skeptical of any changes that disrupt their usual work flow, the benefits may be borne out over a longer time period before clinicians recognize them. Thus, I would be hesitant to discount the results of lukewarm clinician support to a potentially beneficial early warning system.

However, as three clinicians were required to evaluate and communicate about each alert, this alert is highly personnel-intensive. EHRs have the potential to (and do) provide alerts at many points of contact. I have watched clinicians ignore myriad alerts in order to proceed with their work. Additionally, I would be hesitant to warmly welcome a system that identifies stable patients 80%of the time because it has the potential to overburden busy clinicians with low-yield clinical data. While I am optimistic about the EWRS and its potential to improve outcomes in sepsis, this system needs some improvements prior to widespread adoption.

REFERENCE

1. Umscheid JA, et al. Development, implementation, and impact of an automated early warning and response system for sepsis. J Hosp Med 2015;10:26-31.