Title
Optimizing Sepsis Guideline-Directed Treatment by Embracing Diagnostic Uncertainty in an Electronic Health Record Workflow
Authors
Isabella Slaby, DO1, Sharath Kharidi, MD1, Edward Ewen, MD1 ; 1Christiana Care, Newark, DE
Introduction
For many inpatient providers, the electronic health record (EHR) sepsis alert is often dreaded not only for what it implies of the patient's status, but also for the amount of required clicking and documentation that must inevitably follow. Sepsis diagnosis can be challenging due to its varied clinical manifestations that can appear at different times in the clinical course, often mimicking other pathologies. Furthermore, prompt treatment of sepsis is critical. There is a considerable amount of research on early screening and diagnosis in sepsis, as well as the benefits of order sets, and standardized clinical protocols in sepsis1-3. However, most EHR alerts and workflows for sepsis clinical decision support (CDS) ask the provider for a definitive decision on whether sepsis is present, right at the very earliest detection of sepsis criteria. This may lead a provider to disengage with the CDS workflow if they are unsure of the clinical picture, which may lead to delayed initiation of treatment. We present a new sepsis CDS EHR workflow that accommodates diagnostic uncertainty to improve adherence and timeliness of guideline-directed treatment and improve documentation.
Methods
Our team met with clinical leaders across different inpatient levels of care to study different provider workflows when alerted to a patient with possible sepsis. We then analyzed ways to best match the clinical workflows with available EHR tools. Based on those findings, we redesigned our EHR sepsis alert and created a new CDS tool to facilitate structured documentation and treatment when appropriate in cases where sepsis was possible or suspected.
Results
First, we redesigned our sepsis alert to present relevant data for the clinician to decide on a diagnosis of sepsis and automatically document the choice in the chart. In the alert, we introduced a diagnosis option of “Possible Sepsis.” Then, if the provider selected a confirmed or possible diagnosis of sepsis, they would be brought to the CDS step of the workflow. In that window, we ask providers to address the three aspects of guideline-directed treatment by placing orders from the same window, or to defer for now. The “defer” and “possible” flags also create a soft reminder alert for providers to re-evaluate the patient and either rule-in or rule-out sepsis and ensure all components of diagnosis, treatment, and documentation are addressed.
Conclusion
What distinguishes the design of this workflow is its flexibility and option for a “possible” diagnosis status after the alert. By providing a variety of diagnostic decisions and treatment options right at the point of sepsis detection, the workflow can support the provider longitudinally as the clinical picture evolves, even if sepsis is ultimately ruled-out. Importantly, the process supports earlier recognition and treatment for sepsis cases that seem borderline or unclear at presentation. This workflow can be adapted to other clinical pathologies that may be equally challenging to diagnose in the beginning stages, such as heart failure exacerbations. Now that the workflow has been implemented, further analysis involves calculating the effect of the workflow on timeliness and adherence to guideline-directed treatment.
References
1. Amland RC, Sutariya BB. An investigation of sepsis surveillance and emergency treatment on patient mortality outcomes: An observational cohort study. JAMIA Open. 2018 May 15;1(1):107–14.
2. Winterbottom F, Seoane L, Sundell E, Niazi J, Nash T. Improving sepsis outcomes for acutely ill adults using interdisciplinary order sets. Clin Nurse Spec. 2011;25(4):180–5.
3. Gatewood MO, Wemple M, Greco S, Kritek PA, Durvasula R. A quality improvement project to improve early sepsis care in the emergency department. BMJ Qual Saf. 2015 Dec 1;24(12):787–95.