While disagreements remain about the value of strict compliance with CMS-mandated sepsis bundles, they are the closest things to best practices we have. Studies have shown compliance saves lives, reduces complications of hospitalization, and reduces readmission rates. A July 2019 study in the Journal of the American Medical Association (JAMA) found that “Mandated protocolized sepsis care in New York State was associated with a significantly greater decline in risk-adjusted mortality in New York compared with a group of control states that did not implement mandated protocolized sepsis care.”
Additional motivation for compliance stems from the financial and reputational penalties for noncompliance, as well as reduced readmissions and lengths of stay. Preventing treatment delays using nurse-led sepsis protocols and screening tools can save $2 million annually for a typical 300-bed hospital, $5 million for a five-hospital system, and $7.5 million for a regional system.
Nevertheless, full bundle compliance, including both treatment and timeline adherences, remains a considerable challenge in many settings.
Automated surveillance captures cases sooner
In a hospital, centralized automated surveillance of all known septic patients facilitates higher compliance with CMS and local sepsis treatment protocols. When combined with readily accessible sepsis-specific order sets, detailed policies and procedures tailored to each clinical setting, analysis of performance metrics, and the sharing of those metrics with all key personnel, real-time sepsis surveillance makes bundle compliance considerably easier.
Failing to identify septic patients early not only compromises bundle compliance but the delay can lead to irreversible clinical declines. A central monitoring approach that deploys a proven automated sepsis surveillance solution enables sepsis coordinators to zero in quickly on proper sepsis treatment in ways that less focused approaches do not.
Not all surveillance solutions are created equal
While robust studies comparing focused, AI-based sepsis surveillance solutions to EHR-provided solutions are not yet widely available, the greater specificity of sepsis alerts that focused surveillance solutions provide and the smoother workflows enabled by enhanced interoperability have shown improved outcomes on mortality, morbidity, hospital stay, and costs. A 2020 systematic review and meta-analysis published in Internal Care Medicine showed that individual machine learning models could accurately predict sepsis onset ahead of time on retrospective data. An article published in the Journal of the American Informatics Association (JAMIA) in January 2017 concluded, “A program consisting of change management and electronic surveillance with highly sensitive and specific decision support delivered to the point of care resulted in a significant reduction in deaths from sepsis.” In the latter study, the improved clinical surveillance relied on natural language processing, a type of AI that enriched the tool’s diagnostic power by incorporating clinician notes into its risk assessment, achieving sensitivity of 95% and specificity of 82% for detecting sepsis. With those levels of accuracy, sepsis coordinators and their teams are better able to:
-
Identify decompensating patients faster and notify rapid response teams and administrative personnel in a variety of ways for rapid care and if needed transfer to the ICU.
-
Track therapeutic implementation while integrating hospital assignment systems into algorithms that can send targeted reminders to clinical staff to perform those elements of sepsis bundles that haven’t been completed.
-
Document cases where comorbidities may warrant appropriate deviation from sepsis protocols, as with reduced IV fluid administration.
In short, providing clinicians with timely alerts and evidence-based guidance is crucial to reduce variation in care and to improve financial and clinical outcomes. In fact, the JAMIA study found:
-
Mortality decreases of more than 53%
-
30-day readmission declines of nearly 31%
Based on these and other studies, it’s clear that a best-in-class centralized monitoring program for patients with sepsis is an imperative, especially if it includes an automated surveillance solution that offers:
-
Early and accurate alerts that go beyond initial sepsis identification to include alerts that capture variance from compliance with expected care
-
Alerts that integrate seamlessly into clinicians’ workflow
-
Alerts that can be configured to appear for all appropriate care team members
-
Alerts that include clinical decision support recommendations to reduce variability in care
-
Program-specific analytics and change management support to encourage adoption and ongoing improvement
Saving lives and costs through sepsis surveillance
Centralized delivery of early, highly accurate, patient-specific alerts for patients with sepsis give care teams the tools they need to proactively identify potential problems before they occur or shortly after. And when the right person receives alerts at the right time, your team can consistently deliver the evidence-based care of CMS sepsis bundles and your own internal protocols.
That’s why identifying a sepsis solution with most or all of the characteristics described above is so important. Doing so leads to more lives saved and fewer dollars lost to what no longer needs to be an intractable problem for hospitals and health systems.
