The triage nurse has alerted you that he just put a febrile 4-year-old in Bed 6 and the patient’s vitals have triggered the sepsis alert flag. This flag was instituted by your ED as part of a quality improvement initiative to better your compliance with sepsis identification and management. You go into the room to assess the patient and it is apparent that this patient is the real deal. You order your institution’s sepsis bundle.
The medical student is full of questions. What is in the sepsis bundle? Why bundle things together rather than order them individually? Is there evidence that the elements ordered as a group are somehow more effective than the individual components?
There is a lot of variability between physicians in their approach to the diagnosis and management of common medical conditions. For years, guidelines have been periodically released to try to reduce this variability and get individual practitioners to follow evidence-based best practices. That has not always been successful.
A recent study by Evans and associates published in JAMA looked at the effectiveness of an entirely different approach in New York where a statewide sepsis mandate was instituted in 2013 after a pediatric sepsis death. The mandate was based on the conclusion that an institutional protocol was the best way to get all providers on the same page. The directive was designed to change behavior at the level of the hospital rather than the individual provider.
New York’s mandate included a one-hour bundle of blood cultures, antibiotics and a 20 mL/kg fluid bolus. The study looked retrospectively at 59 state hospitals using the New York State Department of Health database. Exposure was completion of the entire bundle within one hour and the primary outcome was all-cause in-hospital mortality.
In the study 1,179 patients were included, of whom 294 completed the bundle within one hour and 885 did not. The authors found that mortality was least in the group that completed the bundle. If the bundle was completed in up to three hours there was still some decrease in mortality. Completion of the bundle was also associated with a shorter hospital length of stay. Interestingly, the association of individual elements in the bundle and outcome did not reach statistical significance.
The New York study also looked at the characteristics of the hospitals that were most likely to complete the bundle. Those hospitals cared for more pediatric patients and were more likely to have PICUs and pediatric sub-specialty care.
In 2016, the Children’s Hospital of Philadelphia published an article looking at its outcomes based on an instituted sepsis protocol. They developed a computer physician order set to deliver elements in a bundle in the ED. Next they looked at whether the use of the standardized order set was associated with the resolution of organ dysfunction by hospital day two. Of 189 patients analyzed, 121 (64%) got treated via the protocol and 68 did not.
They found that patients who received the protocol were more likely to be free of organ dysfunction by the second hospital day, had shorter PICU stays and shorter hospital length of stay. They were also less likely to be transferred to a higher level of care. However, there was no difference in mortality between the two groups. The differences seen could not be explained by differences in measures of the severity of illness or by pre-existing conditions. Again, as in the JAMA study, individual elements of the bundle were not significantly different between the groups.
It is interesting to compare these two studies as regards to the rate of compliance with the bundle. Although the Philadelphia numbers were smaller, 64% of their subjects received the bundle as opposed to about 25 % of the New York patients. Clearly it is much easier to change one institution’s management strategy compared to multiple, diverse institutions.
The American College of Critical Care Medicine issued clinical guidelines for the management of pediatric shock in 2002 and 2007, and updated again in 2014. As with the state of New York, the thrust of the newest recommendations was an emphasis on providing sepsis care at the institutional level rather than by the individual practitioner. The guideline promoted institutional protocols that allowed for a rapid recognition tool that would alert staff to potential sepsis in a pediatric patient, resuscitation and stabilization bundle based on best practices and a performance bundle for monitoring and improving compliance with these practices. These protocols and bundles were meant to reduce incomplete compliance with guidelines for sepsis management.
How is this supposed to work? Your hospital develops a trigger tool to alert you when a child might be septic, using a combination of vital sign abnormalities as well as possibly lab abnormalities and/or information about pre-existing conditions. When the sepsis alert is triggered and the patient is deemed at risk for sepsis, a bundle is ordered. The bundle is an order set that varies by institution, but commonly contains the following elements: a blood culture, broad spectrum antibiotics and fluid bolus(es). The goal of this bundle is to provide the elements rapidly, ideally within one hour.
So why is it that the bundle was associated with improved outcomes in both studies, but the individual elements were not? The JAMA article proposed some possible explanations. Perhaps there are biological and/or physiological factors that create some synergy between the three elements. The presence of a bundle may serve as a surrogate for an environment of heightened awareness of the treatment team. And maybe a delay in completing the bundle was due to some competing event such as difficulty in establishing venous access or intubation that had an effect on outcome.
These two studies clearly demonstrate that the sepsis bundle, as a package of timely interventions, performs better than any one of the components individually. They also give further support to the development of hospital protocols that get everyone involved.
A “change the system, not just the provider” approach delivers better sepsis care to children.
 Evans IVR, et al. Association between the New York sepsis care mandate and in-hospital mortality for pediatric sepsis. JAMA. 2018;320(4):358-367.
 Balamuth et al. Protocolized treatment is associated with decreased organ dysfunction in pediatric severe sepsis. Pediatr Crit Care Med. 2016;17(9):817-822.
 Carcillo JA, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock. Crit Care Med 2002;30:1365-1378.
 Brierley J, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock:2007 update. Crit Care Med 2009;37:666-688.
 Davis AL, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock: 2014 update. Crit care Med 2017;45:1061-1093.