Infectious disease/brief research reportProspective Trial of Real-Time Electronic Surveillance to Expedite Early Care of Severe Sepsis
Introduction
Emergency departments (EDs) initially care for most community-acquired sepsis in the United States, doing so more than 500,000 times annually.1 This task is challenging in that patients typically present with undifferentiated symptoms rather than a diagnosis, and data distinguishing sepsis from other causes of serious illness may lag behind the need to initiate therapy. Furthermore, potentially septic patients are not cared for in a clinical vacuum, but in a setting characterized by limited resources and other patients who compete for caregiver time and attention.
For research purposes, numeric definitions of sepsis are commonplace and vary from relatively complex, such as the Sequential Organ Failure Assessment (SOFA) score, to simple, as in defining sepsis as 2 or more systemic inflammatory response syndrome (SIRS) criteria and a suspected source of infection.2, 3 The need for expedient detection and the use of numeric definitions make sepsis an excellent target for computer decision support. Our group has previously shown that a simple detection rule embedded in an ED clinical information system reliably identifies a cohort that is at significant risk of having a life-threatening infection; when infection is confirmed, this cohort faces high 90-day mortality.3, 4 In the current work, we prospectively studied an automated electronic medical record query and caregiver notification system to improve the care of ED patients with severe sepsis. We sought to increase the frequency and timeliness of 4 key interventions: measurement of blood lactate, collection of blood cultures, performance of a chest radiograph, and initiation of antibiotic therapy.
Section snippets
Study Design
This was a single-site, before-and-after study conducted between April and October 2009. From April through June, the surveillance algorithm operated in the background, was monitored for performance issues, and recorded all patients meeting activation criteria. Caregivers were not notified. From July through October, the physicians, physician assistants, nurses, and technicians responsible for detected patients were notified with recommendations by alphanumeric paging and a text entry into the
Results
During the study, 33,460 patients were screened. Among these, 398 (1.2%) patients manifested 2 or more SIRS criteria and 2 or more systolic blood pressures less than or equal to 90 mm Hg. Of all patients activating the system, 184 (46%) went on to receive an admission diagnosis of infection. To better understand the strategy's ability to detect severe sepsis, we reviewed in detail all patients treated in 1 week during which the system was operational, encompassing 1,386 visits. These results
Limitations
The work was performed at a single academic site with a relatively ill patient population. Typically, 1 or more studies related to sepsis are ongoing at any time in our department, such that general awareness of the condition may be higher than at some sites. These features would seem to risk diluting, rather than exaggerating, the effectiveness of our strategy. Furthermore, our ED clinical information system was developed at our institution and therefore conveniently could be modified to
Discussion
We found that an ED clinical information system notification system triggered on SIRS criteria and hypotension was modestly effective in increasing the performance of key tasks in the early resuscitation of sepsis in the ED. In approximately 50% of patients in whom the 4 key interventions were going to take place, the interventions had already occurred at detection. Thus, the primary benefit appeared to be increasing the awareness of the need for these interventions because time-to-event
References (7)
- et al.
Validation study of an automated electronic acute lung injury screening tool
J Am Med Inform Assoc
(2009) - et al.
National estimates of severe sepsis in United States emergency departments
Crit Care Med
(2007) - et al.
The SOFA (Sepsis-Related Organ Failure Assessment) score to describe organ dysfunction/failure
Intensive Care Med
(1996)
Cited by (92)
Implementation and evaluation of sepsis surveillance and decision support in medical ICU and emergency department
2022, American Journal of Emergency MedicineCitation Excerpt :A few studies have shown the potential for predicting severe sepsis and septic shock hours before the onset [18-21]. Earlier identification of patients at risk of severe sepsis without any changes in management is unlikely to affect outcomes [10,12,22,23]. A few studies have investigated the feasibility and efficacy of decision support algorithms [13,24,25].
Interventions for rapid recognition and treatment of sepsis in the emergency department: a narrative review
2021, Clinical Microbiology and InfectionThe development an artificial intelligence algorithm for early sepsis diagnosis in the intensive care unit
2020, International Journal of Medical InformaticsCitation Excerpt :Artificial intelligence (AI) is an emerging science capable of extensive data analysis that is ideal for EMR analysis. EMR also facilitates the evolution of AI medicine on a complex issue such as sepsis through expanding and analysis of EMR content [13,14]. One of the significant difficulties encountered when introducing AI medicine is data cleansing.
Emergency department sepsis huddles: Achieving excellence for sepsis benchmarks in New York State
2020, American Journal of Emergency MedicineCitation Excerpt :It has become obvious that one major hurdle is the early identification of sepsis. Many hospital systems have tried to implement ways to recognize septic patients by implementing triggering systems [5], or devising various sepsis scoring systems such as the MEDS score [6]. As of 2003, in the United States alone, there were approximately 751,000 sepsis related deaths annually [7].
Scope and Influence of Electronic Health Record–Integrated Clinical Decision Support in the Emergency Department: A Systematic Review
2019, Annals of Emergency MedicineCitation Excerpt :Another focus was clinical decision support interventions that applied decision support at a radiology order entry interface, with the goal of reducing unnecessary or inappropriate imaging for diagnostic evaluation of conditions such as pulmonary embolus26,27 or intracranial hemorrhage28; these studies accounted for 17 (40%) of the total articles included in the review. More comprehensive technologies supported several decisions along the diagnostic and treatment pathway to improve both detection and care for specific conditions, including syncope,29 sepsis,30 asthma management,31 pneumonia,32 and appendicitis.33,34 Thirty-five studies (83%) reported positive effects on measured outcomes.
Prediction of sepsis patients using machine learning approach: A meta-analysis
2019, Computer Methods and Programs in Biomedicine
Supervising editor: Alan E. Jones, MD
Author contributions: JLN, BLS, and JGY conceived the study and designed the trial. JLN obtained research funding. JGY supervised the conduct of the trial and data collection. JLN and BLS organized and collected data. The algorithm for the electronic surveillance system used in this study was written and maintained by JDJ. JLN and JGY analyzed the data and drafted the article. All authors contributed to its revision. JLN takes responsibility for the paper as a whole.
Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article that might create any potential conflict of interest. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. Supported by grant UL1RR024986 from the National Center for Research Resources of the National Institutes of Health (Ms. Nelson).
Publication date: Available online January 12, 2011.
Reprints not available from the authors.
A podcast for this article is available at www.annemergmed.com.
Please see page 501 for the Editor's Capsule Summary of this article.