Characteristics of Adult Rapid Response Events in a Freestanding Children’s Hospital

This study was performed at an inner-city, freestanding, level-1 pediatric trauma and burn center with an annual census of over 80 000 visits and 11 000 inpatient admissions from the emergency department (ED) annually. The children’s hospital has a large pediatric residency program and 11 fellowship training programs across a range of subspecialties. The children’s hospital is adjacent to several adult hospitals as part of a large medical organization.

We conducted a descriptive retrospective review of patients ≥18 years of age who received care during an adult RRE between January 1, 2011 and December 31, 2020. Patients ≥18 years of age and already hospitalized at our institution at the time of the event or receiving care in the ED were excluded. The study was approved by our Institutional Review Board.

We defined a rapid response event (RRE) as an event where an individual with an acute medical need was assessed by a designated RRT. The RRT was comprised of a physician leader (chief resident during “business hours” defined as 8:00 am to 5:00 pm Monday through Friday or designated senior resident during nights and weekends), a respiratory therapist, and a pediatric nurse. Additional pediatric residents may participate in the RRT, but the chief or senior resident was the established team leader for the RREs. The RRT was dispatched by an overhead institutional paging system and can be activated by anyone in the hospital. A paper-based record of the RRE was created at the time of the event. This nurse assigned to carry the RRT pager is responsible for recording on a templated paper and the physician leader is responsible for signing the paper documentation at the end of the event. Following the event, the record was stored in a centralized institutional administrative office maintained by the Director for Quality and Patient Safety. Events that were escalated from an RRE to a code blue or trauma code are reported in the results section but were excluded from analysis. We excluded these because the code team is comprised of physicians with specific adult cardiac life support (ACLS) and advanced trauma life support training and is much different than the standard RRT. The code blue is initiated for hemodynamically unstable patients or patients with respiratory instability. The team is led by a pediatric intensive care fellow with adult training, and an anesthesiologist with adult training and subspecialty pediatric training responds to the activation. The RRT policy states that the physician leader must assess whether a patient is stable for transport directly via the connecting interhospital tunnel to the adult ED. If a patient is determined as “unstable” based on respiratory, cardiovascular, or neurologic criteria that are defined in the policy or for other reasons at the discretion of the physician leader, the patient should be taken to the pediatric ED.

For this study, paper records during the study time frame were reviewed to identify eligible patients. All RRE data were collected on a paper form, with documentation of chief complaint, staff participants, and patient identifiers such as name, date of birth, and gender. Patient data were extracted and then crossreferenced with the institutional electronic medical record (EMR) that contains demographic and clinical information on all patients from their entry into the hospital system since 2006. Patient demographic information collected and analyzed from the paper record and the EMR included: age, gender, reason for being in the hospital, event timing, and event location. We also analyzed the patient’s medical history, medications, interventions performed, and patient disposition following the RRE. If an encounter was opened in the EMR (any location in the Detroit Medical Center Hospital System) on the date of the RRE it was reviewed. Patients who were not registered with an encounter (eg, patients who refused care after the RRE or required no additional medical care) had only the RRE record reviewed. All data were abstracted by 2 research assistants who were trained by the principal investigator using a standardized data collection tool. Data were then independently verified by the principal investigator for accuracy.

We report categorical variables by numbers and percentages. The normality of continuous variables was tested by the Shapiro-Wilk test. We reported normally distributed continuous variables by mean and SD, whereas non-normally distributed continuous variables were reported using median and inter quartile range. Pearson’s χ² test was used to analyze the distribution of categorical variable by groups, provided no expected frequency was less than 1, and no more than 20% of the cell had an expected frequency of less than 5, otherwise Fisher’s Exact test was used. Comparisons for normal continuous variables were conducted using Student’s t test, whereas non-normally distributed continuous variables were compared using the Wilcoxon rank sum test. Comparisons for normal continuous variables were conducted using independent samples Student’s t test, whereas non-normally distributed continuous variables were compared using the Wilcoxon Rank Sum test. We used RStudio (version 1.3.1093, RStudio Team, Boston, MA) to perform the statistical analyses. Significance level was set at 0.05. Subgroup analysis was performed comparing patient and event characteristics for patients who were taken immediately to the pediatric ED compared with those who were not treated further or left against medical advice to try to identify factors that were associated with a “higher acuity” adult RRE, defined as patients taken directly to the pediatric ED and/or admission to the ICU. This analysis excluded patients transferred directly to the adult ED because final disposition was missing for some patients and the final disposition was the outcome of interest for this subgroup.

We reviewed a total of 1769 RREs over the specified 10-year period. A total of 429 events met inclusion criteria for analysis. Adult RREs were stable over time and averaged 48.5 events per year before 2020. During the coronavirus disease 2019 pandemic, events declined to 21 events in 2020. The majority of patients were female (n = 295, 69%) and the median age was 37 (interquartile range, 28–49) years. In this cohort, 3% (n = 13) of patients were already established patients within the institution and continued to receive pediatric subspecialty care as adults but were not inpatient at the time of RRE. We found that 49% (n = 210) of patients were family members of a pediatric patient hospitalized at the time of the event, whereas 28% (n = 121) were staff members at our institution (Table 1). As shown in Figure 1, 37% of patients had at least 1 pre-existing medical diagnosis.

Most events occurred in inpatient floors (n = 131, 30.5%), lobby or cafeteria (n = 74, 17.2%), or ICU areas (n = 71, 16.6%) during business hours (Table 2). The most common presenting complaints included syncope or dizziness (n = 156, 36%), chest pain (n = 49, 11%), and shortness of breath or dyspnea (n = 36, 8%) (Table 2). The majority of events did not require any medical intervention during the adult RRE. Among patients requiring any intervention, 18% received a point of care laboratory testing and <1% required an emergent intervention defined as administration of epinephrine (Table 2). Only 3% of events were upgraded to a medical or trauma code after initial assessment by the RRT. Notably, 14% (n = 60) of events were associated with a report of grief or receiving bad news.

At the completion of the adult RRE, 1 patient was admitted directly to the pediatric ICU, 43% (n = 183) of patients were transferred to an adult ED, and 46% (n = 196) of patients were taken directly to the pediatric ED. After assessment in the pediatric ED, 148 (35%) of patients were transferred to an adult ED, 1 patient died, and 3 were admitted to the pediatric ICU (Fig 2). The patient who died was brought into the hospital in extremis and though an RRT was called when a code blue would have been more appropriate, the patient was quickly moved to the pediatric ED. The 3 patients who were admitted to the pediatric ICU were patients well known to our subspecialists.

Subgroup analysis of the 196 patients taken directly to the pediatric ED compared with the 49 patients who were discharged or left against medical advice directly from the adult RRE showed that among the patients transferred to the pediatric ED, more were female, aged 26 to 45 years, and were not established patients (Table 3). Chief complaints were not different between the 2 groups. Half of the patients were noted to have comorbidities, with the most common being neurologic (34%), cardiac (31%), and respiratory (22%). Patients taking multiple home medications were more likely to be transferred to the pediatric ED. All other patients who received any intervention during RRE were transferred to the pediatric ED. (Table 3).

In this study, we report patient and event characteristics for 429 adult RRE in a freestanding children’s hospital. Unlike previous reports, this cohort included only adult nonhospitalized RRE.^8–12  The aims of this study were to describe our institutional experience with nonhospitalized adult RREs and to examine the relationship between various demographic and clinical factors with the final patient disposition. In this study, we identified key patient and event characteristics that could aid in prevention of adult RREs, team structure and composition of RRTs, and educational needs for a RRT caring for adults in a freestanding children’s hospital.

Most adult RREs were for females in the 26 to 45 year age range. This could be explained by a combination of a larger female family presence and a female staff predominance in our hospital. Most adult RREs were in nonpatients who were on home medication and had existing medical conditions. Although these patient related variables are likely not modifiable, additional support for female caregivers with medical conditions could be an area of focus for hospitals. As previously reported, we did find that most events occurred during “business hours.”^9,11  This information could have implications for staffing of the RRT, especially during the current era.

When considering event characteristics, there were 3 major chief complaints; syncope or dizziness, anxiety or panic attack, and seizure-like events. Importantly, 14% of events were associated with the delivery of bad news or grief response, and the majority of events occurred in inpatient floors and ICUs. All physicians working in a hospital setting have delivered bad news at some point in their careers, but unfortunately training is variable. The Accreditation Council for Graduate Medical Education requires programs to evaluate competency in resident and fellow communication with patients and families and suggests that programs may teach this skill through direct clinical experience, simulation, or other means of active learning. Whereas our data may over or underreport the true prevalence of a grief response causing activation of RRT, with validation from additional studies, this may be an area where focused intervention could lead to prevention of adult RREs. For example, dedicating a staff member, such as hospital clergy or social work, to provide support and services during and after relaying bad news could be a better use of resources then activation of the RRT.

Most adult RREs required no intervention or noncritical care interventions, including laboratory testing, electrocardiogram, medication delivery, or food and drink. In the entire cohort, 46% of the patients were taken immediately to the pediatric ED because they were judged to be too unstable for initial transfer to the adult ED. Of these patients taken to the pediatric ED, 76% were subsequently transferred to the adult ED. Patients taken to the pediatric ED had some patient characteristics similar to the entire cohort. Most were female, aged 26 to 45, and were a parent or family member or staff. Unlike the larger cohort, patients taken to the pediatric ED were more likely to have a cardiac or neurologic past medical history. This finding suggests that RRT education for adult scenarios should focus on the patient characteristics described in this cohort and process improvement initiatives could target the most common patient and event characteristics.

Intervention during the adult RRE including laboratory testing, electrocardiogram, delivery of oxygen, intravenous fluid or medications was associated with transfer to the pediatric ED. Whereas most of these patients were ultimately transferred to the adult ED, 1 patient died in the pediatric ED, and 3 were admitted to the pediatric ICU. Although rare, critical adult RREs do occur. The average age of adult RREs is outside of the typical pediatric provider’s pediatric acute life support training. Most pediatric patients are transitioned to adult care providers between 18 and 21 years of age.¹³  RRT team leaders would benefit from adult specific training for common adult RRT complaints, including ACLS for rare but life-threatening events. Although the Accreditation Council for Graduate Medical Education does not require ACLS for pediatric residents, this should be considered for hospitals where pediatric residents play a significant role in RREs. Our data do not suggest that ACLS training should be a requirement for all members of the RRT. Standardized care algorithms for adult RREs may be higher yield and have been associated with improvement in RRE outcomes.⁷  Being unfamiliar or uncomfortable with adult emergencies, and without the aid of standardized algorithms, could limit the potential for the pediatric trained provider to manage an adult RRE efficiently and effectively. We suggest that providers with ACLS training should be available to members of the RRT for the rare critical scenario and focused education based on the most common adult RRE scenarios should be provided to RRT participants. Further studies to assess interventions used and time to disposition, resource use, and patient outcome are needed.

This study has limitations inherent to a retrospective chart review, including dependence on the quality of documentation. All RRT data were collected via paper chart with variable levels of completion and some missing data. Another limitation was the composition of the rapid response team with various levels of experience and changing personnel based on time of day. In addition, since there were no standardized algorithms for adult acute care management or formal education for team members regarding adult evaluation (outside of review of the adult policy), substantial variability in practice likely exists. The RRT makeup is different across institutions so outcomes and dispositions of patients may vary by team experience. Finally, this study was limited to a single, urban centered institution, which could affect its generalizability.

Adult medical emergencies are common in a freestanding children’s hospital. Both patient and event characteristics provide valuable information for hospital resource utilization, educational intervention, and development of treatment algorithms. In our study, adult RREs were more likely to occur in females aged 26 to 45 with a pre-existing medical condition and most events were noncritical in nature and associated with chief complaints of syncope or dizziness, anxiety or panic attack, and seizure-like events. Almost half of patients were taken directly to the pediatric ED and, although rare, some patients required emergent intervention. For these infrequent but severe circumstances, adult critical care management should be available to the rapid response team. Future studies should focus on development of targeted education and treatment algorithms and quality improvement initiatives focused on resource utilization for adult RREs.