“These Are Our Kids”: Qualitative Interviews with Clinical Leaders in General Emergency Departments on Motivations, Processes and Guidelines in Pediatric Sepsis Care

Abstract

Objective:

Sepsis is a leading cause of pediatric death requiring emergency resuscitation. Most children with sepsis are treated in general Emergency Departments (EDs), yet research has focused on pediatric EDs. We sought to identify barriers and facilitators to pediatric sepsis care in general EDs, including care processes, the role of guidelines and incentivized metrics.

Methods:

In this qualitative study, we conducted semistructured interviews with key informant physician and nurse leaders overseeing pediatric sepsis in general EDs in 2021, including medical directors, nurse managers, quality coordinators. Interviews were audio recorded, transcribed, and coded using deductive domains based on steps of sepsis care, pediatric readiness, and structural dynamics. Domains were analyzed across interviews in matrices, using thematic analysis within domains.

Results:

21 clinical leaders representing 26 hospitals, including Trauma Levels 1-4, were interviewed. Themes included (1) motivation to improve pediatric sepsis care based on moral imperative and location; (2) need for actionable pediatric sepsis guidelines; (3) children’s hospitals’ role in education, protocols, transfer, consultation; and (4) mixed feelings about reportable metrics, particularly in low-pediatric-volume EDs. Sepsis care process challenges included diagnosis, intravenous access and antibiotic delivery, but varied among hospitals.

Conclusions:

Leaders in general EDs were motivated to provide high-quality pediatric sepsis care, but did not agree whether reportable metrics would drive improvements. They universally sought more direct support from their nearest children’s hospitals, and more actionable guidelines. Efforts to address pediatric sepsis quality in general EDs should prioritize guideline design, responsive pediatric transfer and consultation systems, and locally-specific process improvement.

Introduction

Background

Pediatric sepsis is a time-sensitive emergency, and a leading cause of pediatric morbidity and mortality, affecting >70,000 US children yearly.^1,2 Preventing sepsis deaths through early recognition and treatment is a global and national public health priority.³ Policy-makers have supported incentivized performance metrics and legislative mandates to spur quality improvement in pediatric sepsis, and many hospitals have participated in voluntary pediatric sepsis quality improvement collaboratives.^4-6 However, the majority of improvements in pediatic sepsis care and outcomes have been reported in children’s hospitals.^5,7,8 The overall pediatric preparedness of general Emergency Departments (EDs) is a point of vulnerability in emergency care for children.⁹ Although successful programs have improved preparedness in terms of equipment and personnel, they have identified pediatric guideline implementation as an emerging priority in pediatric preparedness.^10,11

Importance

Sepsis requires immediate, first-hour resuscitation that cannot be delayed for transfer to a specialized pediatric center.¹² In fact, most children seeking emergency care in the US first arrive to general EDs, which treat both adults and children, including 65-80% of children with sepsis.^13-15 Studies suggest that pediatric sepsis outcomes are worse at lower-volume pediatric centers, even after policies to improve pediatric sepsis care have been enacted.^14-16 Thus, while some of the greatest opportunities to improve pediatric sepsis treatment and outcomes exist in non-children’s hospitals, little is known about the strengths and needs around sepsis care in these settings. Clinical leaders in pediatric and sepsis care in general EDs are likely to have insights into the delivery of high-quality pediatric sepsis care in general settings, yet are least likely to be represented on task forces or guideline committees dedicated to a pediatric specialty area such as sepsis.^12,17

Goals of this Investigation

This study aimed to elicit the perspective of clinical leaders in general EDs on the motivations, barriers and facilitators to providing high-quality pediatric sepsis care in these settings. In particular, we sought to characterize the response of leaders to tools such as legislative mandates, pediatric sepsis guidelines, and incentivized metrics to improve pediatric sepsis care.

Methods

Study Design and Setting

We conducted and analyzed semi-structured qualitative interviews with key informants, using emergent thematic analysis, following the Standards for Reporting of Qualitative Research guidelines.^18,19 This study was reviewed and determined to be exempt research by the Colorado Multiple Institutional Review Board.

Selection of Participants

Purposive sampling was used to identify key informants who were physician and nurse leaders overseeing pediatric sepsis in general EDs. General EDs were defined as EDs that routinely provided medical care to acutely ill adults and children in the ED and/or inpatient setting. Participants were eligible if they identified themselves as being a physician or nurse in a leadership role in their institution that included oversight of pediatric sepsis care, such as a program manager or medical director in pediatrics, emergency medicine, or quality and safety. Physicians and nurses who worked in a hospital treating only children were excluded. No more than three participants from a single health care system were interviewed.

We placed study announcements in regional professional sepsis and pediatric newsletters, with a distribution list of 366 recipients in varied clinical and administrative roles in health care, and emailed invitations to clinicians who had attended conferences and lectures on pediatric sepsis or who had contacted the children’s hospital with questions related to sepsis. Additional candidates were subsequently recruited with snowball sampling.

Data Collection

We developed the interview guide (Web Appendix 1) with consultation from practicing pediatric and emergency physicians, qualitative experts, and based on the literature. Interview guides were pilot tested with internal sepsis operations and clinical leaders to improve clarity and consistency of questions. Participants were asked to describe current practices and needs around pediatric sepsis preparedness, diagnosis, treatment, transfer, and quality improvement in their hospitals. They were also asked to assess the acceptability and feasibility of external motivations, like state mandates and payor incentives, for pediatric sepsis preparedness and care.

All interviews were conducted virtually, using audio/video communication (Zoom, San Jose, CA) by a pediatric emergency physician trained in qualitative methods (H.S.). Participants were aware from the recruitment email that the interviewer practiced at a children’s hospital and she began the interview with a statement that the goal was to better understand how pediatric sepsis care worked at their hospitals, was not intended to evaluate their sepsis care, and responses would be anonymous. Some participants may have attended lectures by the interviewer in the past and were explicitly aware that she was a researcher focused on pediatric sepsis care, which she explained at the start of all interviews.

The interviewer repeated back to the key informant their descriptions and invited additional clarification and elaboration. Each interview was audio recorded, professionally transcribed verbatim, and spot-checked for accuracy. After each interview, the interviewer completed a debriefing form, which was discussed with the analytic team after interviews. Interviews were conducted until thematic saturation was reached. The analytic team made a determination that thematic saturation had been reached when there was no new content being identified for at least three interviews in a row, with repetitious interviews.^20,21

Data Analysis

Potentially identifying details of participants, patients, and hospitals were removed from transcripts. Transcripts were coded in Atlas.ti, version 8 using deductive domains derived from the interview guide to capture key points. To achieve consensus in coding and matrix summary, interview transcripts were independently analyzed by two investigators (H.S., A.D.) and iteratively compared until consensus was reached. Consensus was reached after three transcripts, and two more transcripts were independently double-coded and summarized to ensure consensus had been achieved and and 20% double-coding was achieved (Web Appendix 2). One investigator coded and summarized the remaining interviews (A.D.), with a third investigator and qualitative methodologist (B.D.H.) spot-checking summaries. Matrix analysis with emergent thematic analysis was used to summarize and analyze codes across interviews, by domain. Two members of the research team (H.S., A.D.) independently identified key themes from the matrix and then compared themes. A third investigator (B.D.H.) reviewed and reconciled the themes generated; discrepancies were resolved by consensus. The chacteristics of the qualitative team are described in Web Appendix 3.

Results

Characteristics of Study Subjects

We conducted 19 hour-long interviews with 21 participants, representing 26 hospitals or EDs located in Colorado, Idaho, Missouri, Oregon, and Washington (Table 1). Although all participants were invited to participate in individual interviews, the study team accommodated two pairs of participants who worked closely together and who requested to be interviewed together, resulting in two dyad interviews. Rural and urban hospitals and Trauma Levels 1-4 were included (Table 1). In interviews where informants were able to characterize yearly ED visit numbers, children comprised 10-20% of ED volumes except for one participant who reported that 1% of visits were pediatric. In 11 of the 19 interviews, the participants reported that pediatric sepsis order sets or protocols were in active use in their health system. The roles and job titles of physician and nursing leaders whose responsibilities encompassed pediatric sepsis quality varied (Web Appendix 4). At smaller hospitals, a single nurse or physician had oversight of processes, quality and education for all clinical care areas, while at other hospitals, responsibilities were more narrowly defined in areas of pediatrics, emergency medicine or sepsis. Among the physicians interviewed, all of whom worked in EDs staffed mostly by general emergency physicians, four were either Pediatric Emergency Medicine-trained or dual-trained in Pediatrics and Emergency Medicine. All functioned in some capacity as a pediatric or sepsis content expert in their general ED.

Table 1.

Participant characteristics and characteristics of their primary ED site. Characteristics of additional clinical sites are presented for those who worked in more than one Emergency Department.

Participant Characteristics (N=21)	n (%)
Clinical Training
Nurse	12 (57%)
Emergency Medicine Physician	3 (14%)
Pediatric/Emergency Medicine Physician	4 (19%)
Pediatric or Medicine-Pediatric Physician	2 (10%)
Nursing Leadership Position
Emergency Department – Overall Leader	1 (5%)
Emergency Department – Pediatric Care Leader	2 (10%)
Health System/Hospital – Overall	2 (10%)
Health System/Hospital - Pediatric	5 (24%)
Sepsis (Adult and Pediatric)	2 (10%)
Physician Leadership Position
Emergency Department – Overall Leader	1 (5%)
Emergency Department – Pediatric Care Leader	3 (14%)
Health System/Hospital – Overall	2 (10%)
Health System/Hospital – Pediatric	2 (10%)
Sepsis (Adult and Pediatric)	1 (5%)
Characteristics of Primary ED Site of Participants (N=14)	n (%)
Trauma Designation
Level 1	3 (21%)
Level 2	3 (21%)
Level 3	4 (29%)
Level 4	4 (29%)
Pediatric Inpatient Capacity	11 (79%)
Rural	3 (21%)
Critical Access Hospital	3 (21%)
Emergency Department with designated pediatric rooms	3 (21%)
Characteristics of Additional ED Sites of Participants (N=12)	n (%)
Free-Standing ED	6 (50%)
Hospital-Based ED	6 (50%)
Trauma Designation
Level 2	1 (8%)
Level 3	3 (25%)
Level 4	1 (8%)
Non-trauma	1 (8%)

Main Results

Theme 1: Moral and Community Motivations:

Participants described strong intrinsic, emotional motivations to be prepared to provide high-quality care for pediatric sepsis (Table 2). Many compared pediatric sepsis to other rare emergencies in which hospitals invest significant resources to maintain readiness, such as gunshot wounds or emergent childbirth in the ED. In some cases, participants described a specific case of a child who died of sepsis as a galvanizing case for their hospital system.

Table 2.

Motivations in pediatric sepsis preparedness and quality identified by key informants: location and moral imperative.

Theme	Explanation/Subtheme	Quotes (Participant Number)
Location	Community connection to the hospital motivates pediatric sepsis preparedness	When we take care of peds patients, these are our kids, our friends' kids, you're gonna see 'em playing basketball, you're gonna see 'em running around the grocery store. I think the reason that we have such a strong focus on patient care is predominantly because of delivering rural medicine. (18)
		We get a huge support of finances from the community, both in oil and gas taxes, and then also in the bonds… as a result, they feel the need to be really responsive to the community about what their needs are and taking care of the community and putting patient care very high up… they’re very enthused on helping us participate in that and get our pediatric readiness. (17)
		That was a big push, was to really invest more in children and pediatrics and make that a priority… We get the foundation involved, so I think—people are always like, “Oh, of course, kids, peds, yes, whatever we need.” (14)
		We provide to a lot of northern Colorado and even some of the surrounding, like Wyoming, south Wyoming. You know, that whole area care. I know our hospital system really wants to be the best at what they do and they're very interested in patient care and quality care and everything. (15)
		We take care of our neighbors. We are a tight-knit community. We all think that's how we'd want our kid taken care of. (13)
		It’s inner city and so they want to focus on those and give them the best care that they can get so that they can have the resources to succeed later on in life. [Our hospital] just provides a lot of external resources as well, and so I think that we treat them as the family versus just the kid or just an adult. (16)
	Proximity to a dedicated children’s hospital, decreases motivation for pediatric preparedness	There’s always gonna be some anxiety around the sick kid because just don’t see a lot, especially within the metro area and [the children’s hospitals] really absorb so much of that that we will always have some anxieties built around the sick kid. (8)
		Other campuses have such a strong children’s hospital presence that we do not have, that they don’t necessarily—it’s not that they don’t prioritize pediatric patients, but they don’t have to worry about it as much because they have the experts there and available. (11)
		All of the people that bring us kids, they appropriately triage away from us because they know that we don’t have really the capabilities for PICU. I think they take sick kids on to either [two children’s hospitals] appropriately. (7)
Moral imperative	There is a moral and emotional imperative to being prepared for pediatric sepsis, no matter how rare	We saw more pediatric sepsis than we did childbirths in the ER, and you're prepared for childbirth in the ER. You still want to make sure that, if someone comes in pushing, you know what to do… I've worked in little Podunk hospitals, and we have something for those crazy situations, the hemorrhages, the—we don't get that many gunshot wounds, but we still have something in place in case we get them. I feel like, with pediatric sepsis, it's the same thing. (5)
		I definitely feel like there are cases that I’ve seen that have been mismanaged and resulted in fatality that, if the family knew, they maybe would publicize it more if they knew that more could have been done, if they knew. I have just a very vivid memory of trying to help over the phone while they were resuscitating a newborn with sepsis who—they didn’t even get IV access for over an hour, so no fluids and no antibiotics. He had a heart rate. He had a heart rate when he came in, and he died. That really affected me. It was what really pushed me to try and highlight to our leadership kids are dying at a higher rate in our community hospitals than they are in our pediatric hospital. (3)
		We had a case three years ago of a kid that was inpatient that ended up passing, sadly. Totally playing dinosaurs. We were thinking about discharging and two hours later he was needing to be intubated. (13)

Location and community connectedness motivated hospital leaders to prioritize pediatric sepsis care. Participants from rural communities described awareness that their hospital was a resource on which their entire community relied and described the importance of pediatric care to the boards and foundations which supported their local hospital. A sense of community was important to pediatric preparedness in urban and suburban areas as well, with participants in larger cities describing the importance of being prepared on their “side of town.”

At times, location decreased motivation to prepare for pediatric emergencies. Hospitals situated close to a dedicated children’s hospital described low pediatric patient volume, with parents and emergency medical services bringing most severely ill children to the children’s hospital, and easy transfer out of pediatric patients. To these hospitals, pediatric preparation was less of a priority.

Theme 2: Pediatric Sepsis Guidelines: Specificity is Helpful:

Participants were eager for guidance from pediatric-specific sepsis tools and guidelines, and had mixed opinions of the utility of current guidelines (Table 3). Although not asked directly about particular guidelines by name, many participants discussed the role that either the American College of Critical Care Medicine (ACCM) guidelines or Pediatric Surviving Sepsis guidelines played in sepsis care at their hospital.^12,17 While some participants reported that the guidelines were not specific enough to operationalize, particularly those who had sepsis protocols in place already, while others reported the opposite, especially from health systems that had not yet developed sepsis protocols.

Table 3.

Key informants’ perceptions of the role of specific tools and structures that they identified as relevant to improving pediatric sepsis care: published guidelines, a regional children’s hospital, incentivized or mandated metric reporting.

Theme	Explanation/Subtheme	Quotes (Participant Number)
Role of Published Pediatric Sepsis Guidelines	Guidelines were not as specific as participants would like	We did… review the new pediatric sepsis guidelines, and I felt like the biggest problem with those is that they were very non-descriptive. All of them were like, maybe do this, but maybe also consider this. (19)
		There just isn’t a clear if you meet x, y, and z criteria like you do with the neonates and adults that you automatically do x, y, and z. It’s a much greyer area. (16)
	Guidelines were specific and helpful	With Surviving Sepsis, coming out with a little bit more concrete guidelines, recommendations, it has allowed us the opportunity to kickstart a more standardized treatment and care for our pediatric population. (6)
		We were halfway through our project when we read about like, oh, this performance bundle and this intervention bundle. It was like it perfectly captured what we were trying to do and would have been much nicer from the get-go if we had just—someone had given us a checklist, been like, ‘Hey, you need a screening protocol. It should have this, this, and this. You need an intervention bundle. It should have this, this, and this,’ which that did. (5)
Importance of regional children’s hospital	Formal and informal relationships and information-sharing	In a rural facility, we don't have an education department, we don't have an evidence-based practice department. We really try to work with some of the systems in the city to do the right thing and implement that here. (18)
		A lot of our staff members work PRN at the children’s hospital. That’s interesting too because then they’re bringing all of those really great things back here to help to continue to improve the care. (19)
	Ease or barriers in transfer and real-time consultation with the children’s hospital	I think people want to take care of people, and they want to do the right thing. Being kind and receptive on the receiving end of things, I think, goes a long ways. I think… if someone's calling you for help, they probably need help, and just be nice with it. (10)
		I think if they were [starting vasopressors], they would probably be in consultation with either one of our pediatric hospitalists or the accepting ICU physician on the other line. I think they would definitely be doing it in consultation with another physician. (7)
“Double-edged sword” of reportable metrics	Mandates and reportable metrics could improve pediatric sepsis care	I think a standardized measurement would certainly be beneficial and would allow us to gauge how we’re doing, very similar to our sepsis measures for adults. Having pediatric standards would be immensely helpful, I’m sure. When there was money backing or potential money loss for nonperformance, I think that gets the attention of high-level senior leaders who help support those decisions. There’s no doubt that if there were state-level regulations, that we would certainly have to comply. (6)
		The squeaky wheel gets the grease, so then if it were mandated, just like CMS [Centers for Medicare and Medicaid Services] mandates for adults, then it would be like, okay, now we really need to spend time doing this because now it's a priority. (12)
		I think that peds-specific incentives, it would be noticed, and it would be progress, and people would—it would give people that need the extra boost we need to advocate to the business side of the health care system. (3)
	Concerns about sepsis metrics, financial incentives	Being reimbursed for pediatric sepsis is really not getting reimbursed for things. It's like oh, we're gonna reimburse you, but if you actually don't do it, you're not getting paid… I just don't think the payment incentive stuff is very medicine friendly. It's just I feel like it's just an excuse for people not to pay. (15)
		I mean there's no way the incentives gonna be huge. The only way for it to be meaningful is gonna be based on volume. (8)
	Concerns based on adult mandates	I mean it's just people that you know don't have sepsis or whatever, but they sorta fall into that category and end up getting unnecessary antibiotics. It's like well, you gotta give 'em because it's part of the measure. (15)
		Yeah, forcing unnecessary fluid administration. That's been a big issue with adults. As you know, if the lactate is above 4 and the blood pressure is high, and the patient has CHF, and they missed their dialysis, and their x-ray looks like… they're fluid-filled. Fluid still should be initiated. It will be a fallout, even if the provider says this is why we're not giving it. That's very, very disheartening. Then it tends to lend itself to a process of, "Well, that's a stupid measure, I'm not doin' it." Providers get these emails. They say, "I did the right thing for the patient. I'm just gonna throw all these CMS measures in the trash 'cause I don't care." (9)
		I think politics and medicine always—people have the right goals, but I think sometimes, it's hard to try and put a one size fits all on care… I think, with adult sepsis care, it makes us feel like our souls are dying when you give a 30 cc per kg bolus to somebody who you know is already volume overloaded. (10)
		I think those circumstances where it’s much less frequency, people tend to crave more guidance, whereas if they feel like they’re an expert in something like an adult sepsis that they see a lot, they’re much less likely to want so much specific guidance. (19)
	Metrics not meaningful if case numbers are small, hard to identify	For other things, we just run a report, exactly how many patients with CHF or how many patients have sepsis. With the kids, it was difficult 'cause we shipped 'em out. We don't know if they got diagnosed with sepsis. (5)
		Well, I guess I would say our group would certainly be willing to do what we had to do if it was required by the government. Of course, would we be thrilled about it? I would say no because our volume is so low. (9)
	Mixed Feelings	I love quality. That's why I do it here. I think that unless you're focused on a project, I think it goes away. I think anytime that you can identify meaningful measures that really impact the bedside, those are the best measures. That's why we try—even though we don't have a lot of statistical significance, we still have data that says, this is where we're at. I hate quality measures that are like, did you check the box? Because I don't think that that always provides benefit, but if you can come to me and say, patient outcomes or morbidity and mortality is decreased if you do this… we would partake in that kind of a program. (18)
		I can see it as a double-edged sword, where it would be good because sometimes we—we're a rural hospital. We don't know what we're doing sometimes… those mandates would be nice to give us guidance, but at the same time, how do we adapt that to a rural hospital? (13)

Theme 3: The role of a local children’s hospital:

Many participants used sepsis tools, algorithms, pocket cards and education from a children’s hospital in their region. These relationships were both formal and informal, including recurring visits for education or pathway dissemination, and personal relationships that facilitated knowledge sharing. Participants described an appreciation for pediatric-specific protocols that children’s hospitals shared with them.

Participants also reported frustrations when they encountered challenges trying to transfer a child with sepsis to a children’s hospital. They described delays in the transfer process, such as a physician accepting the transfer, or waiting for a team capable of pediatric critical care transport. They felt that pediatric subspecialists consulting via telephone were sometimes supportive, but sometimes did not understand the constraints of their general emergency department. The quality of support from a children’s hospital in transfer and real-time phone consultation was described as an important factor in the quality of pediatric sepsis care, both positively and negatively.

Theme 4: Mandates and metrics are “a double-edge sword”:

No study participants practiced in healthcare systems participating in incentivized or mandated pediatric sepsis quality reporting, although all participated in adult sepsis quality metric reporting/incentives. Participants’ responses to the idea of such measures were mixed. Some participants felt positively about their potential impact in their hospital, while others had concerns about unintended negative consequences, or that reporting requirements would strain already overburdened quality improvement personnel.

Some participants thought legislative mandates would increase resources and personnel for pediatric sepsis in hospitals in which children were often not the focus. Others worried that metrics based on averages or percentages would not be based on adequate numbers of cases to accurately reflect their care. Participants were also concerned that in the context of a larger general hospital, a per-case incentive for a pediatric condition was seen as unlikely to be financially relevant. One participant described the importance of learning from in-depth review of individual cases rather than reporting averages in settings with few pediatric sepsis cases, and feared that mandated one-size-fits-all reporting would detract from a useful approach that was already in place for their hospital.

Concerns about potential unintended negative clinical consequences of mandated or incentivized pediatric sepsis metrics included antibiotic overuse, allergic reactions, unnecessary testing, alert fatigue, and whether requirements would be adequately adjusted to pediatric needs at varied ages. Smaller EDs were concerned that pediatric sepsis requirements might be designed for children’s EDs and not adaptable for smaller, general EDs. Additionally, the participants who described negative experiences with adult sepsis CMS requirements expressed concerns about pediatric requirements. They described a disheartening effect of adult regulations on providers who felt that certain adult measures did not always represent good medical care for specific patients, leading them to discount the entire bundle for all patients.

However, others reconciled their negative feelings about adult sepsis requirements with more positive hope for pediatric requirements. Leaders who described being well-supported in quality goals and working in settings where pediatrics was prioritized did not feel they needed incentives or mandates to improve care. However, they were interested in the medical guidance that mandates might provide. In settings where leaders felt pediatric care was not a priority, they were supportive of the leverage that an external incentive might apply to focus on pediatric sepsis quality improvement.

Other findings:

Participants commented on additional, specific aspects of the processes of pediatric sepsis care in addition to the main themes identified above. Responses about specific processes of care were varied, with some respondants reporting challenges in the areas of pediatric sepsis recognition, pediatric intravenous access, and timely antibiotic delivery (Web Appendix 5). Participants described intravenous bolus administration and vasopressor administration as areas of strength. The unifying themes that emerged from analysis focused on motivations and facilitators, rather than specific processes or forms of education.

Limitations

This qualitative study sought to understand pediatric sepsis care processes, needs and motivations as reported by clinical leaders in general EDs, but did not add quantitative data to their report. While we believe that understanding the values and needs of organizational leaders is an important element of quality improvement and implementation science, we cannot assess whether their perceptions correlate with the interventions most likely to impact patient outcomes. Our sampling technique was not proportionally representative of all general EDs, although it was successful in including a broad range of hospital and ED location, size, and trauma level, which likely enhances transferability. We recruited through a regional sepsis email distribution list, in order to reach a broad range of hospitals. The invitation to participate was circulated to 366 individuals, however we do not know how many non-respondants met inclusion criteria and reasons for non-participation, which limits our ability to assess the representativeness of the included participants. Additionally, by relying on voluntary responses to an emailed announcement, it is likely that our sample overrepresented hospitals that were engaged with sepsis care, and may have biased our results to reflect more motivation and engagement with pediatric sepsis quality improvement than is true of all general EDs.

The interviews may have been affected by being conducted by a pediatric emergency physician from a children’s hospital, an approach that was chosen to allow content expertise that was helpful in understanding clinical terminology and asking appropriate follow-up questions, but may also have introduced social desirability bias causing participants to describe more motivation and interest in pediatric sepsis care. Included hospitals were located primarily in the Western US, a region where distances to pediatric-ready EDs are longer than other regions, and findings may not generalize to other regions where travel distances to pediatric centers are shorter.²² Notably, no EDs were in a state with a current pediatric sepsis reporting mandate or incentive, which is reflective of the vast majority of US states currently, so the informants were only able to comment on the perceived impact of metrics, not their own experience.

Discussion

The emergency care of pediatric sepsis, particularly in general EDs, has emerged as a public health and policy focus in the last decade. Theories of implementation science emphasize the importance of understanding the “outer and inner settings,” or the “external environment” and “organizational characteristics,” in order to develop interventions that are likely to be effective, reach patients and improve outcomes.^23,24 In this qualitative study, we sought to describe the environment and setting of pediatric sepsis care in general EDs, particularly motivations, barriers and desired interventions.

We found that most leaders described high intrinsic motivation to ensure their ED provided high-quality pediatric sepsis care. When pediatric sepsis was not a priority, the primary reason given was proximity to a major children’s hospital and consequently, exceedingly low pediatric volume. Our study suggests pediatric sepsis quality improvement approaches that would be welcomed by general EDs and might consequently be likely to be adopted. Specifically, participants appreciated and sought guidelines that were specific and actionable, and cited the quality of support from their local pediatric hospitals as major factors in their quality of care. Pediatric-focused quality improvement efforts such as the Improving Pediatric Sepsis Outcomes collaborative have developed these kinds of clinically-actionable tools, but these have not been well-disseminated to general emergency physicians, and these were not mentioned by informants in this study.²⁵

Some participants reported that existing pediatric sepsis guidelines did not feel actionable. There are limitations in the clinical actions that a formal guideline can recommend when it is following the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology, which require recommendations be made only as strong as the supporting evidence base.^12,26 Of the 77 pediatric Surviving Sepsis 2020 guideline statements, only 7 were underpinned by moderate or high-quality evidence, leading to appropriate limitations in the strength of its recommendations.¹² The formal evidence assessment for GRADE-based guidelines fills an important need in academic and clinical medicine. However, our study suggests that it also would be beneficial to focus on accompanying user-friendly tools for the clinical frontline, including simplified suggestions for order sets, metrics, and pathways. Guidelines are usually designed by content experts for use by clinical leaders such as those in our study, for whom pediatric sepsis is only a small part of their scope of work.^12,17 Needs and preferences of generalist clinicians should be incorporated in order to promote usability of guidelines and related tools by non-content experts.

Pediatric sepsis leadership and quality oversight in general EDs was often in the purview of physicians and nurses who cared for both adults and children, and participants frequently referred to their experience with adult sepsis protocols. In some cases, sepsis quality improvement work in adults promoted parallel work for pediatric sepsis, however other informants described negative experiences with adult sepsis metrics and mandates that colored their reactions to the idea of pediatric metrics. While aligning adult and pediatric guidelines in structure and metrics has potential to improve efficiency and dissemination in general EDs, the mixed responses of informants suggested that such alignment could potentially hurt uptake by general ED clinicians who had a negative reaction to adult sepsis quality metrics or mandates.

In addition to ongoing, successful programs to improve overall pediatric preparedness of general EDs, there has been separate legislative and health policy interest in improving care of pediatric sepsis specifically.^4,11 One approach to drive improvement has been creating mandates and incentives, which have shown success in improving outcomes in adult sepsis.²⁷ Participants in this study had mixed responses to this approach. Some felt that pediatric reporting mandates could force prioritization of pediatric needs in a general hospital system where children were a small proportion of overall patients. However, others worried that requirements might be a ‘checkbox’ with limited impact, or that measures would be unreliable in hospitals with very small pediatric volume. This concern aligns with findings from New York state where bundle compliance was lowest at hospitals with low pediatric volumes, even after legislated pediatric sepsis mandates.¹⁶

The clinical leaders interviewed for this study described the importance of the quality of their interactions with their nearest children’s hospitals to the quality of pediatric sepsis care. They sought and reported positive experiences with not only education, but specifically sepsis protocols, ordersets and tools such as ‘pocket cards’ developed and shared by children’s hospitals in their region. Their expressed needs and positive interactions with children’s hospitals to support pediatric emergency care were in alignment with literature repoted on pediatric emergency preparedness in other regions. For example, the University of California Davis has reported on their partnership between a regional children’s hospital and two rural emergency medical services agencies, in which they shared training, education, pediatric quality improvement interventions, telemedicine and guidelines.²⁸ This aligns with the types of interventions described as helpful by the participants in this study, and replicating this approach in a promising direction for pediatric sepsis care dissemination.

When transfer or real-time consultation was sought, the participants in our study described mixed experiences of support in the care of critically ill children in their center. Friendly and supportive transfer and real-time consultation are opportunities for pediatric subspecialists to enhance engagement with general EDs, thus improving pediatric sepsis care and knowledge dissemination for their region.

It is important to note areas where common themes did not emerge in these interviews. Participants identified differing process challenges, including diagnosis, intravenous access and timely antibiotic administration. The organizational structures around children and pediatric sepsis care differed within this group of ‘general EDs.’ Pediatric sepsis care fell within the purview of sepsis or pediatric clinical processes, and the staffing, physical structure, hours, and pediatric content expertise supporting care of children in a general ED differed greatly as well. Thus, while the broad themes identified here can shape national initiatives, implementing change also requires deeper understanding of local organizational characteristics and needs.

The themes that emerged in this study can inform future work to implement evidence based pediatric sepsis care in general EDs. Leaders described being motivated to provide high-quality pediatric sepsis care, particularly in hospitals geographically distant from children’s hospitals. Although the idea of using incentives or mandates to improve pediatric sepsis care has gained support from health care leaders and legislators, clinical leaders in general EDs reported mixed feelings about whether incentivized metrics and legislative mandates would improve care.

The sepsis care processes where leaders most sought improvement included diagnosis, intravenous access and antibiotic delivery, but varied among hospitals. General ED leaders universally expressed a desire for more support, directly from their nearest children’s hospitals, as well as more specific and actionable guidelines. These findings suggest that pediatric sepsis improvement efforts most likely to be adopted in general EDs would focus on guideline design and dissemination, require locally-specific process improvement, and support transfer and realtime medical consultation.