Improving Care at Emergency Department Discharge for Pediatric Patients with Anaphylaxis Using a Quality Improvement Framework

Abstract

Introduction:

Anaphylaxis is a potentially fatal systemic reaction that requires prompt recognition and targeted treatment. Despite international consensus and national guidelines, there is often incomplete care for pediatric patients discharged from the emergency department (ED) with a diagnosis of anaphylaxis. Our institution experienced wide variability in discharge planning for patients with anaphylaxis. The goal of our study was to improve care at ED discharge for pediatric patients with anaphylaxis using a quality improvement framework. The specific aims were to increase the frequency of patients diagnosed with anaphylaxis who receive an anaphylaxis action plan at ED discharge from 0% to 60% and to increase referrals to an allergy clinic from a baseline of 61%–80% between October 2020 and April 2021.

Methods:

Targeted interventions included revisions to the electronic health record system, forging interdisciplinary partnerships and emphasizing provider education. Outcome measures were the proportion of patients receiving an anaphylaxis action plan and an allergy clinic follow-up. The balancing measure was the ED length of stay.

Results:

The study showed an increase in anaphylaxis action plans from 0% to 34%. Allergy clinic referral rates improved from 61% to 82% within the same period. The average length of stay of 347 minutes remained unchanged.

Conclusions:

Revising the discharge instructions to include an anaphylaxis action plan and reinforcing provider behaviors with educational interventions led to an overall improvement in discharge care for patients with anaphylaxis. Future work will focus on electronic health record changes to continue progress in additional clinical settings.

INTRODUCTION

Anaphylaxis is a life-threatening, systemic allergic reaction that is often underrecognized and undertreated.¹ Although consensus exists among allergy and emergency physicians on the diagnostic criteria for anaphylaxis, there remains wide variation among adult and pediatric emergency department (ED) providers in the treatment and discharge planning for patients with anaphylactic reactions.^1,2 International consensus advises that all patients with anaphylactic reactions who are discharged from an ED setting receive a written anaphylaxis emergency action plan, an epinephrine autoinjector or prescription with teaching on its use, and a referral to an allergist.^3,4

Despite these published guidelines, adherence to these recommendations has remained poor, leading to incomplete discharge care in pediatric and adult populations.^1,5–7 Studies show that approximately 60% of patients discharged from the ED setting receive an epinephrine autoinjector prescription and less than half receive a referral for follow-up with an allergist.^1,8 To combat this issue, some institutions have implemented clinical care guidelines and quality improvement (QI) strategies to standardize discharge care. Interventions have included discharging patients from the ED with an epinephrine autoinjector and implementing clearly defined algorithms with specific instructions for intramuscular epinephrine use and an allergy clinic referral.^9,10

An ED anaphylaxis order set based on the American Academy of Allergy, Asthma and Immunology (AAAI) guidelines has been available in our electronic health record (EHR) since 2012 and was revised in 2019 before the start of our project. At baseline, this order set was utilized in 86% of patients with anaphylaxis, which led to compliance with American Academy of Allergy, Asthma and Immunology recommendations for management of acute symptoms; however, the discharge had more opportunities for improvement. Although epinephrine autoinjector prescriptions were given to 77% of anaphylactic patients before initiation of this project, and patients received general discharge instructions about anaphylaxis and autoinjector use, they did not receive an action plan to guide response to future symptoms. Allergy clinic referral was also inconsistent, with only about half of patients receiving one. This represented missed opportunities for a specialist to obtain a detailed allergy history, coordinate allergy testing, tailor management strategies, reinforce epinephrine autoinjector training, and counsel the patient on avoidance of allergy triggers.

Due to the large volume of patients who presented to our ED with anaphylactic reactions, we set out to improve anaphylaxis care at ED discharge. The specific aims were to increase the percentage of patients with anaphylaxis receiving an anaphylaxis action plan at ED discharge from 0% to 60% and to increase referrals to an allergy clinic from a baseline of 61% to 80% between October 2020 and April 2021. Given the success that previous pediatric hospitals had using revised clinical pathways in the EHR to standardize care for commonly treated disease processes, we determined that adapting this implementation strategy in addition to educational initiatives would work best for our hospital.^11–13

METHODS

Context

This QI study was conducted at a freestanding, academic, tertiary care pediatric hospital with two ED locations. One location is an ED within a primary freestanding children’s hospital, and the second is a pediatric ED located within a community hospital. Each has 24-hour staffing by pediatric emergency medicine physicians, general pediatricians, physician assistants, and nurse practitioners 7 days a week. Together, these sites have 130,000 patient visits per year, including approximately 340 patients who meet anaphylaxis criteria.

Ethical Considerations

The project received approval from the institutional review board for research purposes. The QI project required individual chart review for data collection. Collected data were deidentified before analysis.

Interventions

A multidisciplinary team of ED nurses, physicians, and pharmacists conducted this project from May 2019 onward. The team was later expanded to include clinical informatics team members. In May 2019, a brief EHR inquiry regarding ED anaphylaxis discharge care was conducted by two pediatric emergency medicine physicians in the group to determine variations in practice and areas for improvement. All patients with an International Classification of Diseases, tenth edition (ICD-10) code of allergic reaction, anaphylaxis, and allergy were included in the initial data collection. A study team member then completed a manual chart review on each patient to determine whether patients met anaphylaxis criteria as defined by the World Health Organization.³ The final population of interest was any ED patient who met the clinical criteria for anaphylaxis and was discharged home from the ED. Exclusion criteria included other allergic reactions (not anaphylaxis), admission to the hospital, or death.

We created an ED process map for anaphylactic patients in May 2019 to analyze the current process and inform future interventions. Supplemental Digital Content 1, http://links.lww.com/PQ9/A417, which shows the process for patients presenting to the ED with anaphylaxis before the interventions of this QI project illustrates the process for anaphylactic patients presenting to the ED before the start of the project. The QI team determined that the discharge process for patients with anaphylaxis varied widely among providers, especially regarding written discharge instructions and allergy clinic follow-up. A key driver diagram was created to identify steps in the process leading to this variance and interventions to improve the process (Fig. 1). From May 2019 to November 2019, the team met monthly to review current literature, analyze the baseline data, and update the key driver diagram. The team worked with the allergy clinic and members of the clinical informatics team to revise the current anaphylaxis order set within the EHR. Revisions to the order set included recommendations to discharge patients with a prescription for intramuscular epinephrine and follow-up with an allergist within 6 weeks. The updated order set was available for all ED providers beginning in January 2020. An electronic anaphylaxis decision tool was added in the EHR to properly identify patients with anaphylaxis.

Fig. 1.

Key driver diagram of the ED anaphylaxis discharge improvement initiative. This figure illustrates the factors contributing to achieving the QI project’s aim and proposed change ideas.

The main intervention involved the creation of a revised anaphylaxis discharge packet that included the Food Allergy Research & Education (FARE) Food Allergy & Anaphylaxis Emergency care plan and recommendations for allergy clinic follow-up. (See Supplemental Digital Content 2, http://links.lww.com/PQ9/A418, which illustrates the FARE Food Allergy & Anaphylaxis Emergency care plan.)¹⁴ This discharge packet resulted from a thorough literature review and discussions with the hospital’s allergists regarding best care practices for anaphylaxis discharge. A final electronic version of the revised anaphylaxis discharge packet was published to the EHR and made available for all providers starting in October 2020. Use of the revised discharge instructions emphasized using educational initiatives, announcements at departmental meetings, and posters displayed in the ED, as annotated in Figures 2 and 3. The QI team received feedback from providers that the revised discharge instructions were difficult to find while searching the EHR, among other generic allergy discharge instructions that did not contain the FARE guidelines; therefore, the discharge instructions were renamed in August 2021 to make them easier to locate for providers.

Fig. 2.

Patients receiving anaphylaxis action plan at discharge. A p-chart of patients receiving anaphylaxis action plan at discharge with QI interventions.

Fig. 3.

Allergy referrals for patients presenting with anaphylaxis. A p-chart of patients receiving referral to allergy clinic at discharge with QI interventions.

Measures

The population of interest, which defined the denominator for our measures, included patients who met anaphylaxis criteria and were discharged home after a period of observation in the ED. Anaphylaxis was defined as all patients presenting to the ED with acute onset of symptoms of generalized involvement of the skin and/or mucosal tissue and at least involvement of one other system (ie, respiratory compromise, reduced blood pressure/associated symptoms of end-organ dysfunction, or severe gastrointestinal symptoms). Patients admitted to the hospital for anaphylaxis have a different discharge process than the ED and, therefore, were excluded from the study population. In addition, patients with isolated allergy symptoms (ie, urticaria in absence of other symptoms) that did not meet the criteria for anaphylaxis were excluded.

The primary outcome measure was the proportion of patients receiving allergy referrals at discharge. This measure was defined as the number of patients with anaphylaxis who received an allergy referral or had established allergy clinic follow-up at discharge divided by the total patients discharged from the ED that met the criteria for anaphylaxis. The second outcome measure was the proportion of patients receiving an anaphylaxis action plan at discharge. This measure was defined as the number of patients with anaphylaxis that received discharge instructions with an anaphylaxis action plan at ED discharge divided by the total of patients discharged from the ED that met the criteria for anaphylaxis. The outcome measures were also considered process measures for our project. Our balancing measure was ED length of stay for patients who presented to the ED with anaphylaxis.

Data were collected monthly and reviewed at the team meetings. The QI team assessed the impact of the interventions using standard process control (SPC) charts to determine whether the interventions led to special cause variation. Special cause variation was determined by the standard rules of Shewhart chart analysis as defined in The Health Care Data Guide.¹⁵

Analysis

SPC charts were reviewed with the team at monthly meetings to analyze postintervention data, discuss findings in monthly data review, and review the relationship of data changes to past interventions. Although our EDs are spread out over two sites, the number of patients per month that presented to our community ED site with anaphylaxis was too small to interpret with SPC charts by location. Because of this, the data from our primary ED and community sites were combined.

RESULTS

A total of 425 patient encounters that met the inclusion criteria were reviewed for this study. From this total, 102 encounters determined the baseline data. The baseline data showed that 0% of patients received an anaphylaxis action plan, 86% of providers used the anaphylaxis order set in the EHR, and 61% received allergy clinic referrals.

Our data showed special cause variation. The primary outcome measures were the proportion of patients receiving an anaphylaxis action plan at discharge and the proportion of patients receiving allergy referral at discharge. The proportion of patients with anaphylaxis who presented to the ED and received an anaphylaxis action plan at discharge showed special cause variation with a centerline shift from 0% to 34% with sustained change following the intervention (Fig. 2). The proportion of patients who received an allergy referral at discharge increased from 61% to 82% with a centerline shift on the SPC charts (Fig. 3). The balancing measure, ED length of stay, showed no change during the interventions (Fig. 4).

Fig. 4.

Balancing measure. X bar and S chart of the average length of stay in the ED for patients with anaphylaxis before and after the QI interventions.

DISCUSSION

We demonstrated that targeted interventions involving revisions in the EHR emphasized by interdisciplinary partnerships and provider education resulted in measurable improvement in discharge coordination for patients with anaphylaxis. Our study improved patient care by providing home-based anaphylaxis action plans and allergy outpatient referrals for discharged anaphylactic ED patients. Although additional steps were added to the discharge process, our interventions did not prolong the overall ED length of stay (ie, 347 minutes), which is a critical measure of overall ED throughput and efficiency.

Our results showed that the most significant improvements occurred with automated changes in the EHR and increased faculty awareness and buy-in following educational initiatives. Specifically, there were notable increases in allergy referrals following announcements regarding allergy clinic referral rates during our departmental meetings. This was reinforced with email notifications and poster reminders in the ED. Implementing the FARE anaphylaxis plan in the discharge instructions was the most successful initiative to increase the proportion of patients with anaphylaxis receiving anaphylaxis action plans. This was likely because most providers were unaware that anaphylaxis plans existed for use in the ED setting and that before the revision of the discharge instructions, they were difficult to find.

The importance of patient and family recognition of anaphylaxis symptoms and clarity for when to administer epinephrine is critical. Prompt administration of epinephrine is potentially life-saving and may reduce biphasic symptoms if given early. Contemporary studies have estimated that approximately 5% of patients experience biphasic symptoms that may occur after ED discharge, and reexposure to an allergen in the future is common.⁴ By emphasizing the recommendations for allergy follow-up, families may be more likely to obtain specialist care and access education and testing as indicated. In addition, action plans provide clear, individualized, patient-centered information to help families rapidly assess the need for epinephrine in the event of allergic symptoms, noting the dose of epinephrine and providing graphics and text describing how to use various brands of autoinjectors. In collaboration with the allergy department, we chose to use the same action plan they provide to patients and families with anaphylaxis, which provides consistency in follow-up and ongoing care.

Previous studies have shown that patients frequently do not receive optimal care after ED discharge, with about 60% receiving epinephrine autoinjector prescriptions and less than half receiving a recommendation for follow-up with an allergist.^8,16 There is little in the literature specifically documenting rates of patients receiving an anaphylaxis action plan at discharge.

Allergy referral rates occur in less than half of ED discharges for anaphylactic patients, and follow-up rates are even lower. In one study of children insured by Medicaid, only 17% of those treated for anaphylaxis in the ED followed up with an allergist in the subsequent year.¹⁷ In a similar study in privately insured patients, 43% followed up with an allergist, and 6.4% had a return ED visit.¹⁸ A strategy to improve care is to leverage the EHR to include evidence-based orders. One study in a pediatric ED demonstrated improvements in many aspects of care of anaphylactic patients after the implementation of an order set.⁹ It also showed decreased admission rates and improvements in discharge with epinephrine. Still, the percent receiving allergy referrals remained stable at approximately 60%, suggesting that the coordination of follow-up care for anaphylaxis requires strategies that differ from the rest of the discharge process.

Limitations

This study has several limitations. A main limitation was the reliance on provider behavioral change via education and awareness rather than automation or technological interventions. The current EHR does not allow integration of the discharge instructions and allergy referrals into the order set, so even if the order set is used, providers must remember to manually enter allergy clinic follow-up and select the correct discharge instructions with the anaphylaxis action plan when discharging a patient. We relied on education and increasing awareness, which are valuable but do not reach all clinicians. Specifically, residents in the ED were not included in the primary educational efforts in this project. Because residents prepare a large proportion of the discharge paperwork and are not dedicated ED staff, many were not aware of the specific discharge process and paperwork for the anaphylaxis population, leading to some incorrect or absent allergy referrals and missing anaphylaxis action plans. These errors were not always identified and corrected by the supervising providers. Continued efforts to streamline discharge paperwork and referrals are in place, but they will require changes to the EHR. In addition, in-person nursing education was impacted due to COVID-19–related, hospital-wide safety measures, and reduction of nurse educator positions within the department for staffing reasons, coupled with the hiring of new nursing staff who may have missed initial education efforts.

The COVID-19 pandemic concurrent with this study had several impacts. Early in the project, patient volumes were atypically low, followed by an unusually high patient census coupled with a significant nursing shortage. Despite these patient volume fluctuations, ED length of stay did not change throughout the study.

CONCLUSIONS

In conclusion, we improved discharge care for patients with anaphylaxis treated in the pediatric ED through multidisciplinary educational efforts and EHR changes. Revising anaphylaxis discharge instructions and educating providers increased the frequency of anaphylactic patients receiving an evidence-based anaphylaxis action plan at discharge, and an increase in allergy referrals for patients who met anaphylaxis criteria. The next phase of the project will include directly dispensing epinephrine autoinjectors in the ED, as it was demonstrated in a recent study at our institution that the epinephrine autoinjector fill rate from ED discharge prescriptions was only 54.8%.¹⁹ Areas for further study include monitoring for changes in allergy clinic utilization and epinephrine autoinjector fill rates. This QI project model has potential for use in anaphylactic patients discharged in additional clinical settings.

ACKNOWLEDGMENTS

Assistance with the study: Josh Heffren, PharmD, and Amanda Troger, BSN, RN.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.