a b s t r a c t

Background: Research demonstrates that timely recognition and treatment of sepsis can significantly improve pe- diatric patient outcomes, especially regarding time to intravenous fluid and antibiotic administration. Fur- ther research suggests that underlying chronic disease in a septic pediatric patient puts them at higher risk for poor outcomes.

Objective: To compare treatment time for suspected sepsis and septic shock in pediatric patients with chronic dis- ease versus those without chronic disease seen in the Pediatric Emergency Department (PED).

Methods: We reviewed patient data from a pediatric sepsis outcomes dataset collected at two tertiary care pedi- atric hospital sites from January 2017-December 2018. Patients were stratified into two groups: those with and without chronic disease, defined as any patient with at least one of eight chronic health conditions. Inclusion criteria: patients seen in the PED ultimately diagnosed with sepsis or septic shock, patient age 0 to 20 years and time zero for identification of sepsis in the PED. Exclusion criteria: time zero unavailable, inability to deter- mine time of first IVF or antibiotic administration or patient death within the PED. Primary analysis included comparison of time zero to first IVF and antibiotic administration between each group.

Results: 312 patients met inclusion criteria. 169 individuals had chronic disease and 143 did not. Median time to antibiotics in those with chronic disease was 41.9 min versus 43.0 min in patients without chronic disease (p = 0.181). Time to first IVF in those with chronic disease was 22.0 min versus 12.0 min in those without (p = 0.010). Those with an indwelling line/catheter (n = 40) received IVF slower than those without (n = 272), with no sig- nificant difference in time to antibiotic administration by indwelling catheter status (p = 0.063). There were no significant differences in the mode of identification of suspected sepsis or septic shock between those with versus without chronic disease (p = 0.27).

Conclusions: Study findings suggest pediatric patients with chronic disease with suspected sepsis or septic shock in the PED have a slower time to IVF administration but equivocal use of sepsis recognition tools compared to pa- tients without chronic disease.

(C) 2021

1. Background

Pediatric sepsis is a leading cause of morbidity, mortality and healthcare utilization worldwide. Globally, there are 1.2 million cases of pediatric sepsis annually with sepsis accounting for more than 4% of all pediatric hospitalizations in high-income countries [1-6]. The major- ity of pediatric patients who die from sepsis present with or develop re- fractory shock or Multiple organ dysfunction and death occuring within

48 to 72 h of treatment, underscoring the significance of early

* Corresponding author.

E-mail addresses: [email protected] (E. Hegamyer), [email protected] (N. Smith), [email protected] (A.D. Thompson), [email protected] (A.D. Depiero).

recognition and treatment of pediatric sepsis and its role in improving overall outcomes [7-10].

The principle of early recognition and treatment of pediatric sepsis has since become the basis for the development of multiple sepsis rec- ognition tools in use across the United States within Pediatric Emer- gency Departments (PED), which are often centered on Abnormal vital signs suggestive of impending shock or end organ dysfunction [11,12]. Several research studies have demonstrated that timely recognition and treatment of pediatric sepsis with the help of these tools within the PED can significantly improve patient outcomes, in particular in regards to expedient administration of intravenous fluids (IVF) and an- tibiotics [11,12]. Both early administration of antimicrobrials and rapid volume resuscitation strategies within one hour of septic shock recogni- tion have been established as two key factors in the initial treatment of pediatric sepsis influencing patient outcomes, including both mortality

https://doi.org/10.1016/j.ajem.2021.01.026

0735-6757/(C) 2021

rate and hospital length of stay [3,6,7,13-16]. These treatments are often prompted by early recognition of septic shock or organ dysfunction due to sepsis recognition tools demonstrating clinical and vital signs changes [12,17-25].

In addition to early administration of both IVF and antibiotics, under- lying patient risk factors such as high-risk medical conditions are another key variable influencing pediatric sepsis outcomes, often also integrated within electronic health record sepsis alerts along with Vital sign abnormalities. This includes chronic disease, encompassing the pediatric population of individuals with at least one significant chronic condition in two or more body systems or those with a single dominant chronic condition [26]. Overall, comorbid illness was present in 34% of all children hospitalized for sepsis in the United States between 2004 and 2012 with multiple comorbid illnesses identi- fied as a key factor associated with higher odds of both mortality and longer hospital length of stay [2]. Additionally, multiple research studies have shown that chronically ill children have a higher overall burden of disease from sepsis with a higher risk for poor outcomes [3,6,7,13,14,27]. Thus the importance of a prioritized mechanism for the early recognition and treatment of suspected sepsis or septic shock in pediatric chronic disease patients is evident, as it may signifi- cantly impact overall outcomes for these individuals.

The primary objective of this study is determination of the degree to which this optimization of sepsis treatment of patients with chronic dis- ease is currently taking place in the Pediatric Emergency Department. This study compares the initial time to first IV fluid bolus and IV antibi- otic administration for suspected sepsis or septic shock between pediat- ric patients age 0 to 20 years old with chronic disease and those without chronic disease.

1. Methods

This study was a multi-site retrospective study including patients seen in the PED of two pediatric tertiary care center sites. Both sites are suburban free standing Tertiary care children’s hospitals, each with a separate PED, one of which is located in Wilmington, DE with a yearly PED volume of 60,000 patients and the other in Orlando, FL with a yearly PED volume of approximately 40,000.

Data for this study were part of an institution wide pediatric sepsis outcomes data set collected at each site from January 2017 through De- cember 2018. The information within the data set was initially extracted by a researcher whose primary role within both institutions is to obtain patient information for the sepsis outcomes data set from the EHR using predetermined institutional guidelines with specifications of which data points to include. The same EHR system is used at both study sites. The sepsis outcomes data set includes all Pediatric Emergency De- partment patients seen from January 2017 through December 2018 at both study sites, with individual patient encounters initially extracted from the EHR using a set of eight inclusion criteria to capture patients with suspected sepsis or septic shock [28]. These criteria are reflective of the sepsis definitions as determined by a multidisciplinary national expert advisory committee of the Children’s Hospital Association’s Im- proving Pediatric Sepsis Outcomes (IPSO) collaborative and include the following: positive sepsis screen plus treatment, positive sepsis hud- dle, severe sepsis order set use or other infectious disease order set use plus treatment, ICU admission plus treatment, lactate measured plus treatment, pressor plus treatment, or severe sepsis/septic shock ICD- 10 codes or other sepsis ICD-10 codes plus treatment where treatment is defined as IV antibiotics and two boluses or one bolus and pressor within six hours and blood culture within 72 h of antibiotics and boluses (or bolus/pressor) [28]. Any patient meeting at least one set of criteria

was included in the data set.

Data points for each patient encounter included in this study were abstracted from the institution wide pediatric sepsis outcomes data set and include, if applicable: patient age, time of arrival to the Emer- gency Department, sepsis huddle time, time of sepsis screen prompt

in EHR, time of initiation of sepsis order set, time of first antibiotic ad- ministration and first IVF bolus administration and chronic disease sta- tus. Every patient chart was reviewed in detail within the EHR by study staff to evaluate for both inclusion and exclusion criteria and to confirm all documentation within the original sepsis outcomes data set.

For this study, patients were stratified into one of two groups: those with chronic disease and those without chronic disease. The health con- ditions of those patients defined as having chronic disease were predetermined by the IPSO collaborative guidelines set forth for the in- stitutional sepsis outcomes data set and include the following condi- tions: malignancy, asplenia (including all Sickle cell disease patients), bone marrow transplant, indwelling catheter (defined as any central ve- nous line including peripherally inserted central catheters), solid organ transplant, immunocompromised, severe mental retardation cerebral palsy (MRCP) and technology dependence (defined as an individual with tracheostomy, ventriculoperitoneal shunt or Enteral feeding tube). Study participants were classified as those with versus without chronic disease through detailed chart review by study staff, who reviewed the EHR and medical history for each individual to identify un- derlying medical conditions. Each medical condition was then catego- rized into any of the applicable eight chronic disease conditions (e.g. a leukemia patient with a central line falls into the categories of malig- nancy, indwelling catheter and immunocompromised). Any individual with at least one of these eight health conditions was classified in this

study as an individual with chronic disease.

The primary outcome of the study was to examine the time to first IVF bolus and time to first intravenous antibiotic administration from functional time zero of suspected sepsis or septic shock identification in the PED in those with chronic disease versus those without chronic disease. Functional time zero for identification of suspected sepsis or septic shock was defined as the first time in which concern for sepsis or septic shock occurred and for this study includes the following five categories considered to be key markers for process improvement re- garding pediatric sepsis outcomes: sepsis screen prompt in EHR based on the patient’s shock score (prompted by a shock score of >=45 and in- cludes text box with suggestion to consider sepsis in the patient and di- rect link to sepsis order set), sepsis huddle (prompted by a shock score of >=45 and often initiated by the bedside RN, requiring bedside evalua- tion and documentation in the EHR by both the attending physician and bedside RN), sepsis order set initiation (includes laboratory, intra- venous fluid and antibiotic orders), first antibiotic administration or first IVF bolus administration. The shock score for each patient is based primarily on institutional vital sign parameters determined by a multidisciplinary committee across both study sites with five extra points added if patient is labeled to have a “complex medical history” within the EHR. Vital sign parameters and shock score calculation were the same at both study sites. The functional time zero documented for each patient encounter within our study data set was based on which of the aforementioned five categories occurred first in the PED, identified upon chart review within the EHR by study staff as the pre- sumed first moment at which a medical provider suspected sepsis or septic shock.

Secondary outcome measures included goal met of first IVF bolus and antibiotic administration within 60 min of the time at which sepsis or septic shock was suspected (i.e. functional time zero) in those with chronic disease versus those without chronic disease and time to first IVF bolus and antibiotic administration in those with an indwelling catheter versus those without an indwelling catheter. Data were also analyzed to determine if there was a statistically significant difference between the number of underlying chronic medical conditions and the time to first IVF bolus and antibiotic administration after functional time zero.

Inclusion criteria for this study were any patient seen in the PED with suspected sepsis or septic shock based on at least 1 of the 8 criteria as defined by the sepsis outcomes data set [28], in addition to functional time zero for identification of sepsis or septic shock occurring in the PED

and patient age from 0 to 20 years old. Multiple patient encounters for the same individual were included in this study if the appropriate inclu- sion criteria were met.

Exclusion criteria included any patient who bypassed the PED, death within the PED, functional time zero for suspected sepsis or septic shock occurring outside the PED, or inability to determine functional time zero, time of first IVF bolus or first antibiotic administration. Addition- ally, patients transferred from outside facilities were excluded given the inability to reliably confirm documentation of antibiotic and IVF ad- ministration prior to PED arrival. Patients were also excluded if they re- ceived antibiotics in the PED prior to functional time zero, but not if they received intravenous fluids.

Descriptive statistical analysis is presented as medians with inter- quartile ranges for continuous variables. Continuous variables were compared with the Wilcoxon rank sum test and categorical variables

Table 1

Age of Study Participants

Chronic Disease n (%)

Age 0d to <1y	4 (2.4)	32 (22.4)	0.000
1y to <5y	31 (18.3)	37 (25.9)
5y to <13y	86 (50.9)	39 (27.3)
13y to 20y	48 (28.4)	35 (24.5)
Total, N	169	143

^a P-value based on Pearson Chi-Square test

Table 2

Chronic disease subcategory.

No Chronic Disease n (%)

P value^a

with the Pearson Chi Square test. Statistical software used for data anal- ysis is R (version 3.6.1) and SPSS Statistics (version 25). This study was approved by the Nemours Institutional Review Board (IRB) Committee.

1. Results

A total of 368 patients seen in the Pediatric Emergency Department across both study sites were initially identified from the study data set as potential participants in this study. Three hundred and twelve indi- viduals ultimately met inclusion criteria, 169 of whom had chronic dis- ease and 143 without chronic disease (Fig. 1). Table 1 provides age specific demographic information of those patients with versus without chronic disease in this study population, demonstrating that the older the patient the more likely they are to have chronic disease (p = 0.000). Among those patients with chronic disease, the majority (N = 114, 67.4%) had severe mental retardation cerebral palsy (MRCP), technol- ogy dependence or both. Twenty-three percent (N = 40) of patients had an indwelling catheter and thirty-nine patients (23.1%) were im- munocompromised. Twenty-five patients (14.8%) had an underlying malignancy while the rest of patients fell into one of the final three cat- egories, including prior bone marrow transplant, asplenia and/or prior

history of solid organ transplant (Table 2).

Data were examined looking at each of the five categories for func- tional time zero in those with chronic disease versus those without chronic disease. The most frequently identified functional time zero was the sepsis screen EHR prompt based on patient shock score with 35.5% (N = 60) use among those with chronic disease and 32.2% (N = 46) use in those without chronic disease. This was followed by the time of first IVF bolus and time of first antibiotic administration as

Chronic Disease Category Total (N, % of Chronic Disease patients)

Severe MRCP^a 114 (67.4)

Technology Dependent^b 114 (67.4)

Indwelling Catheter^c 40 (23.7)

Immunocompromised 39 (23.1)

Malignancy 25 (14.8)

Solid Organ Transplant 9 (5.3)

Asplenia^d 6 (3.6)

Bone Marrow Transplant 1 (0.5)

^a Severe mental retardation/cerebral palsy

^b Defined as an individual with tracheostomy, ventriculoperitoneal shunt or enteral feeding tube

^c Defined as any Central venous line including peripherally inserted central catheter

^d Includes all sickle cell disease patients

the second and third most commonly used categories of functional time zero, respectively. The sepsis huddle was initiated among 10.7% (N = 18) of patients with chronic disease versus 17.5% (N = 25) pa- tients without chronic disease. Only 9.1% (N = 13) of patients without chronic disease had the sepsis order set initiated versus 11.2% (N = 19) of those with chronic disease. There was no significant difference found between those with chronic disease versus without chronic dis- ease in relationship to any of the five categories of functional time zero (p = 0.27). See Fig. 2.

One of the key areas of focus for this study was the time from func- tional time zero to actual administration of first IVF bolus and antibiotic administration among the two patient groups. The median time to first IVF bolus administration from functional time zero in those with chronic disease was longer (22.0 min) than for those patients without chronic disease (12.0 min), a statistically significant difference (p =

Fig. 1. Participant Inclusion and Categorization.

60 min from the time of identification of sepsis or suspected sepsis within the PED.

With respect to the goal of first IVF bolus administration within 60 min, a higher percentage of patients without chronic disease 87.4% (N = 125) met the IVF bolus administration goal versus the antibiotic administration goal. Among those with chronic disease, 81.7% (N = 138) of patients met the goal for IVF bolus administration with an over- all success rate of 84.3% (N = 263) among all patients. Again, there was no significant difference in meeting the IVF bolus administration goal among those with versus without chronic disease (p = 0.211).

Fig. 2. Functional Time Zero Frequencies.

0.010). The time to first antibiotic administration from functional time zero was shorter (41.9 min) in those patients with chronic disease ver- sus those patients without chronic disease (43.0 min), which was not a significant difference (Table 3).

Among those patients with chronic disease, one specific factor often felt to possibly influence Time to administration of both IVF and antibi- otics in those with sepsis or suspected sepsis is the presence of an in- dwelling catheter. In this study it was found that the time from functional time zero to first IVF bolus administration was longer (20.0 min) for patients with an indwelling catheter versus those with- out an indwelling catheter (16.5 min), although this difference is not significant (p = 0.735). The opposite was noted in regards to first anti- biotic administration from functional time zero, with a median time of

33.0 min to administration in those with an indwelling catheter versus

43.0 min in those without an indwelling catheter (p = 0.063). See Table 4.

When examining the number of medical conditions present for each participant, no significant difference in time to first antibiotic adminis- tration from functional time zero was found (p = 0.254), based on a subdivision of study participants into two comparison groups of those with only one or no chronic medical condition versus those with greater than or equal to two conditions. There was a significant difference among these same patient groups in the time to first IVF bolus adminis- tration from functional time zero with regard to the presence of either one or no chronic medical conditions versus greater than or equal to two chronic medical conditions (p = 0.014). Those with only one or no chronic medical conditions received IVF faster than those with at least two underlying chronic medical conditions. See Fig. 3.

Additionally, we further examined the institutional goal of both first IVF bolus and antibiotic administration within 60 min of functional time zero for patients with suspected sepsis or septic shock. When looking at functional time zero to first antibiotic administration, the study demon- strated having met the goal among those with chronic disease in 68.6% (N = 116) of patients versus 51.4% (N = 55) of those patients without chronic disease. There was no significant difference between meeting the 60 min goal of antibiotic administration between those with versus without chronic disease (p = 0.716). Overall, 67.6% (N = 211) of all study participants met the goal of antibiotic administration within

1. Discussion

This multi-site retrospective study examined identification and treatment of suspected sepsis or septic shock in pediatric ED patients with and without chronic disease, as those patients with underlying chronic disease carry a higher burden of disease from sepsis and argu- ably require at least equivocal if not more aggressive recognition and treatment for sepsis than non-chronic disease patients in the PED [3,6,7,17,18]. Primary results from this study demonstrate a marginally faster antibiotic administration time from functional time zero in the PED among those patients with chronic disease, yet a significantly slower median time to first IVF bolus administration.

One significant factor that may influence the time to administration

of both antibiotics and the first IVF bolus is intravenous access, which is often found to be more difficult among those patients with chronic dis- ease and may in part explain the longer IVF administration time among these patients seen in this study. Another element to consider is the presence of an indwelling catheter, which was present in 23.7% of pa- tients in our study, meaning these individuals had pre-established intra- venous access prior to PED arrival. One might hypothesize that this would allow for faster intravenous access and earlier administration of either IVF or antibiotics among these individuals. Yet IVF administration was actually longer in those patients with an indwelling catheter with no significant differences in either time to antibiotic or IVF administra- tion between those with and without an indwelling catheter. One expla- nation may be the ease with which bedside staff are able to access an indwelling catheter versus placing a peripheral IV, as the process is often a sterile procedure that needs to be performed by trained staff members. Overall, these results suggest that intravenous access in those patients presenting with suspected sepsis or septic shock in the PED, albeit via an indwelling catheter or peripheral access, is a key area for improvement, particularly among those pediatric patients with chronic disease.

The finding that overall time to antibiotic administration was almost twice as long as IVF in both those with and without chronic disease was unexpected. This is presumptively related to the easier availability of IVF versus antibiotics in the PED, the latter in some cases requiring delivery from a location outside the Pediatric Emergency Department, such as a centralized hospital pharmacy. Additionally, the finding may reflect the decision by bedside clinical staff to first administer IVF versus antibi- otics, assuming both were readily available. There was not a standard- ized protocol at either study site regarding the order of IVF versus antibiotic administration, a decision which may also be influenced by

Table 3

Time to Antibiotic and IVF Administration in Patients With versus Without Chronic Disease

	Chronic Disease (N = 169)	No Chronic Disease (N = 143)	P Valuea
Time Zero to Antibioticb
(minutes) Time Zero to IVFb (minutes)	41.9 (17.5-73.0)	43.0 (26.0-81.9)	p = 0.181
	22.0 (1.0-50.0)	12.0 (0.0-29.0)	p = 0.010

^a P-values based on Wilcoxon Rank Sum test.

^b Data presented as median values with interquartile range.

Table 4

Time to Antibiotic and IVF Administration in Patients With Indwelling Catheter versus Without Indwelling Catheter

	Indwelling Catheter (N = 40)	No Indwelling Catheter (N = 272)	P Valuea
Time Zero to Antibioticb
(min) Time Zero to IVF2 (min)	33.0 (11.3-49.8)	43.0 (23.0-81.8)	p = 0.063
	20.0 (0.0-45.5)	16.5 (0.0-39.8)	p = 0.735

^a P-values based on Wilcoxon Rank Sum test

^b Data presented as median values with interquartile range

the presence of one versus multiple sites of intravenous access, data we did not have available for this study. The access to and administration of antibiotics should be an ongoing area for process improvement in the PED, especially in regards to meeting goals for the early treatment of suspected sepsis or septic shock.

Furthermore, we found no significant differences in meeting institu- tional goals for time to administration of IVF and antibiotics under 60 min between those with chronic disease versus those without chronic disease. That being said, the institutional goals for antibiotic and IVF administration within 60 min were met only in 67.6% and 84.3% of all participants, respectively. This again demonstrates contin- ued areas for overall improvement in early identification and treatment of suspected sepsis or septic shock within the PED, particularly in regards to antibiotic administration.

The results of this study show no significant differences between those patients with chronic disease versus those without in regards to the documented mode of identification for suspected sepsis or septic shock (i.e. functional time zero), including a sepsis huddle, sepsis order set, EHR prompt based on shock score or the time to first IVF bolus or antibiotic administration. Thus there does not appear to be one method of identification for suspected sepsis or septic shock that is utilized more frequently for those in the study population with chronic disease. One may also extrapolate that those individuals with chronic disease did not necessarily present with more vital sign abnor- malities, as those modes of identification for suspected sepsis or septic shock based on shock score (i.e. abnormal vital signs) were not used more frequently in a statistically significant manner among this patient population in the study.

Finally, while our study found no significant difference in the time from functional time zero to antibiotic administration for suspected sepsis or septic shock between patients with none or one and those with two or more chronic medical conditions, there was a significant difference in time to IVF administration, with longer time to IVF admin- istration in those with two or more chronic medical conditions. This is reflective of the primary study results in which IVF administration from functional time zero was overall longer in those with chronic dis- ease and suggests the possibility that increased medical complexity and certain conditions may affect these outcomes more than others (e.g. MRCP versus asplenia). We were unable to analyze the latter hypothesis in depth within our particular study given the small number of patients in each category of chronic disease among our participants, but this re- mains an area of potential interest for future studies.

1. Limitations

There were several limitations to this study. One limitation is the in- clusion of both time to IVF bolus and antibiotic administration as a cat- egory of functional time zero, with the concern that the distribution of median times would subsequently be towards lower values. To further evaluate this, we reanalyzed the data without including these two cate- gories as a functional time zero and found no significant changes in the study results in regards to significant differences and overall median values for time to administration of IVF and antibiotics.

An additional limitation to this study was the inability to examine the data by each chronic condition due to small sample size within each category. As referenced above, it may be possible, for example,

Fig. 3. Time to Antibiotic and IVF Administration versus Number of Chronic Disease Conditions *Greater than or equal to two conditions **Less than or equal to one condition.

that the underlying condition of severe MRCP is a more significant factor in time to administration of IVF and antibiotic administration than asplenia.

Finally, this study had a relatively small sample size due to the rela- tively small representation of pediatric sepsis patients in the PED pa- tient population in the two sites during the study’s time frame. This reflects the potential value of further large, multi-center retrospective studies centered on pediatric sepsis patients and outcomes.

1. Conclusion

While pediatric sepsis remains a robust area of research, evidence has continued to demonstrate a more significant burden of disease and poorer outcomes among pediatric patients with underlying chronic conditions who present with sepsis or septic shock [3,6,7,17,18]. While the study results reflect equivocal use of early recognition tools for suspected sepsis or septic shock in the PED among individuals with chronic disease versus those without chronic disease, there remain sig- nificant differences in time to initiation of treatment between these two groups, specifically in regards to administration of IVF. Thus there re- mains a need for ongoing attempts to change such outcomes via contin- ued research studies focused on this high risk population. Additionally, the study results demonstrate a need for ongoing process improvement in more expedient intravenous access for chronic disease patients and faster time to antibiotic administration among all patients presenting to the PED with suspected sepsis or septic shock.

Declaration of Competing Interest

No conflicts of interest or source of funding. No special circumstances.

Acknowledegment

The authors gratefully acknowledge the Children’s Hospital Association’s Improving Pediatric Sepsis Outcomes collaborative, a multi-hospital multiyear collaborative which supported quality im- provement and data analysis for this manuscript. This information is presented with the permission of Children’s Hospital Association.

References

1. Fleischmann-Struzek C, Goldfarb DM, Schlattmann P, et al. The global burden of pae- diatric and Neonatal sepsis: a systematic review. Lancet Respir Med. 2018;6:223-30.
2. Balamuth F, Weiss SL, Neuman MI, et al. Pediatric severe sepsis in U.S. children’s hos- pitals. Pediatr Crit Care Med. 2014;15:798-805.
3. Odetola FO, Gebremariam A, Freed GL. Patient and hospital correlates of clinical out- comes and resource utilization in severe pediatric sepsis. Pediatrics. 2007;119: 487-94.
4. Ruth A, McCracken CE, Fortenberry JD, et al. Pediatric severe sepsis: current trends and outcomes from the pediatric health information systems database. Pediatr Crit Care Med. 2014;15:828-38.
5. Weiss SL, Fitzgerald JC, Pappachan J, et al. Sepsis prevalence, outcomes, and thera- pies (SPROUT) study investigators and pediatric acute lung injury and Sepsis inves- tigators (PALISI) network: global epidemiology of pediatric severe sepsis: the sepsis prevalence, outcomes, and therapies study. Am J Respir Crit Care Med. 2015;191: 1147-57.
6. Schlapbach LJ, Straney L, Alexander J, et al. ANZICS Paediatric study group: mortality related to invasive infections, sepsis, and septic shock in critically ill children in Australia and New Zealand, 2002-13: a multicentre retrospective cohort study. Lan- cet Infect Dis. 2015;15:46-54.
7. Morin L, Ray S, Wilson C, et al. ESPNIC Refractory septic shock definition taskforce the infection systemic inflammation Sepsis section of ESPNIC: refractory septic shock in children: a European Society of Paediatric and Neonatal Intensive Care definition. In- tensive Care Med. 2016;42:1948-57.
8. Schlapbach LJ, MacLaren G, Festa M, et al. Australian & new Zealand Intensive Care Society (ANZICS) Centre for Outcomes & resource evaluation (CORE) and Australian & new Zealand Intensive Care Society (ANZICS) Paediatric study group: prediction of pediatric sepsis mortality within 1 h of intensive care admission. Inten- sive Care Med. 2017;43:1085-96.
9. Weiss SL, Balamuth F, Hensley J, et al. The epidemiology of hospital death following pediatric severe sepsis: when, why, and how children with sepsis die. Pediatr Crit Care Med. 2017;18:823-30.
10. Cvetkovic M, Lutman D, Ramnarayan P, et al. Timing of death in children referred for intensive care with severe sepsis: implications for interventional studies. Pediatr Crit Care Med. 2015;16:410-7.
11. Balamuth F, Alpern ER, Abbadessa MK, et al. Improving Recognition of Pediatric Se- vere Sepsis in the Emergency Department: Contributions of a Vital Sign-Based Elec- tronic Alert and Bedside Clinician Identification. Ann Emerg Med. 2017;70(6): 759-768.e752. doi:7https://doi.org/10.1016/j.annemergmed.2017.1003.1019.
12. Larsen GY, Mecham N, Greenberg R. An emergency department septic shock proto- col and care guideline for children initiated at triage. Pediatrics. 2011;127(6). https://doi.org/10.1542/peds.2010-3513 e1585-1592.
13. Hartman ME, Linde-Swirble WT, Angus DC, et al. Trends in the epidemiology of pe- diatric severe sepsis. Pediatr Crit Care Med. 2014;15:798-805.
14. Kutko MC, Calarco MP, Flaherty MB, et al. Mortality rates in pediatric septic shock with and without multiple organ system failure. Pediatr Crit Care Med. 2003;4: 333-7.
15. de Oliveira CF, de Oliveira DS, Gottschald AF, et al. ACCM/PALS haemodynamic sup- port guidelines for paediatric septic shock: an outcomes comparison with and with- out monitoring Central venous oxygen saturation. Intensive Care Med. 2008;34: 1065-75.
16. Weiss LS, Peters MJ, Alhazzani W, et al. Surviving Sepsis campaign international guidelines for the Management of Septic Shock and Sepsis-Associated Organ Dys- function in children. Pediatr Crit Care Med. 2020;21(2):e52-106.
17. Weiss SL, Fitzgerald JC, Balamuth F, et al. Delayed antimicrobial therapy increases mortality and organ dysfunction duration in pediatric sepsis. Crit Care Med. 2014; 42:2409-17.
18. Paul R, Neuman MI, Monuteaux MC, et al. Adherence to PALS sepsis guidelines and hospital length of stay. Pediatrics. 2012;130:e273-80.
19. Lane RD, Funai T, Reeder R, et al. High reliability pediatric septic shock quality im- provement initiative and decreasing mortality. Pediatrics. 2016;138:e20154153.
20. Balamuth F, Weiss SL, Fitzgerald JC, et al. Protocolized treatment is associated with decreased organ dysfunction in pediatric severe sepsis. Pediatr Crit Care Med. 2016;17:817-22.
21. Cruz AT, Perry AM, Williams EA, et al. Implementation of goal-directed therapy for children with suspected sepsis in the emergency department. Pediatrics. 2011; 127:e758-66.
22. Kortz TB, Axelrod DM, Chisti MJ, et al. Clinical outcomes and mortality before and after implementation of a pediatric sepsis protocol in a limited resource setting: a retrospective cohort study in Bangladesh. PLoS One. 2017;12:e0181160.
23. Long E, Babl FE, Angley E, et al. A prospective quality improvement study in the emergency department targeting paediatric sepsis. Arch Dis Child. 2016;101: 945-50.
24. Workman JK, Ames SG, Reeder RW, et al. Treatment of pediatric septic shock with the surviving sepsis campaign guidelines and PICU patient outcomes. Pediatr Crit Care Med. 2016;17:e451-8.
25. Tuuri RE, Gehrig MG, Busch CE, et al. “Beat the shock clock”: an interprofessional team improves pediatric septic shock care. Clin Pediatr (Phila). 2016;55:626-38.
26. Children”s Hospital Association. 3M Clinical Risk Group Categories Children with Medical Complexity, Children’s Hospital Association. May 2020. Retrieved from: www.childrenshospitals.org/Care/Children-With-Medical-Complexity.
27. Prout AJ, Talisa VB, Carcillo JA, et al. Children with chronic disease bear the highest burden of pediatric sepsis. J Pediatr. 2018;199 194-199.e1.
28. Scott HF, Brilli RJ. Paul R, et al: “evaluating pediatric Sepsis definitions designed for electronic health record extraction and multicenter quality improvement”. Crit Care Med. 2020;48(10):e916-26. https://doi.org/10.1097/CCM.0000000000004505.