Sepsis remains a leading cause of pediatric morbidity and mortality, hospitalizing 40,000 children in the U.S. and killing 5,000 annually. The condition is treatable when recognized early, but it can encompass a broad range of underlying causes and present in different ways, making it difficult to identify. Common indicators of sepsis, such as fever and tachycardia, are prevalent in children without sepsis, and vital sign changes may not be reliable markers for sepsis in medically complex patients.
These factors make sepsis easy to miss, but technology can reduce the risk by creating a shared mental model of potential sepsis, monitoring specific risk factors, and alerting frontline workers. This can be done with an automatic system that reliably screens patients at regular intervals within the electronic health record (EHR). Yale New Haven Children’s Hospital in New Haven, Connecticut, uses an electronic screening tool based on age-adjusted vital signs as well as clinical markers of perfusion. The tool is integrated with the hospital’s EHR to access a broad range of clinical information to identify rapid changes in the physiologic status of hospitalized children, providing increased opportunities for earlier intervention.
When a patient screens positive for potential sepsis, a best practice alert (BPA) appears in the patient’s chart, which includes sepsis pathways and standardized order sets. The team also uses the EHR software to perform a case-by-case review to reveal missed cases and perform a root cause analysis to adjust the measurements for sepsis indicators.
Over a two-year period (2017-2019), the rate of hospital-onset severe sepsis at Yale New Haven Children’s Hospital decreased from a baseline rate of 2.05% to 0.38% following the implementation of the screening trigger tool, as well as other initiatives such as an order set, dashboard, sepsis simulation, and education. Mortality related to sepsis decreased from 16% to 5% over two years, and median length of stay decreased from 18 to 11 hospital days.
Though implementing these interventions is simple, there are some challenges. There must be buy-in from the frontline teams using the electronic interface who already experience a high burden of alarms. To minimize alarm fatigue, it is crucial to regularly evaluate the screening trigger tool and validate the model in pediatrics.
Importantly, it takes time for evidence-based medicine and new practice guidelines to be integrated and adopted into current clinical practice—research suggests that it can take up to 20 years. The Surviving Sepsis Guidelines are updated every few years, which means hospitals must be agile to incorporate the recommendations as the standard of care at all levels of patient care.
While technology cannot replace the physical exam skills and critical thinking of frontline providers, it can serve as an invaluable tool to improve early recognition of sepsis. Of course, continued education will be critical to optimize the use of the technology as it continues to evolve.
Despite these challenges, our single-center experience demonstrates that low-cost technological interventions can successfully reduce the rate of hospital-onset severe sepsis without increasing lengths of stay or intensive care admissions at children’s hospitals.