Stabilizing Pediatric Patients During Psychiatric Boarding: A Quality Improvement Project

For youth who visit the emergency department (ED) for mental health evaluation and who require psychiatric hospitalization, a lack of inpatient psychiatric beds results in boarding in the medical hospital, often with little treatment. The Joint Commission defines boarding as “the practice of holding patients in the ED or another temporary location after the decision to admit or transfer has been made” and recommends boarding not exceed 4 hours.¹ However, in practice, medically cleared youth may board for hours, days, or weeks. A systematic review of studies before the COVID-19 pandemic found that the average boarding time for youth in the ED varied between 5 and 41 hours, and youth on a medical unit were found to board for an average of 24 to 36 hours.² In a survey of pediatric providers within children’s and community hospitals during the pandemic, 84.4% of respondent hospitals disclosed increased frequency of pediatric boarding, with 75.3% reporting increased duration.³ Another study found that the average boarding time more than doubled from 2.1 days to 4.6 days within their institution during the first year of the pandemic.⁴ 

Boarding has implications for children and caregivers, as well as for hospital systems and communities. For children, boarding ensures safety. Providers and caregivers share concerns about whether boarding may worsen mental health symptoms⁵ and struggle with seeing youth wait for needed treatment without being able to expedite it.^6,7 Medically cleared patients who are boarding strain inpatient medical units as hospitals work to manage patient flow and balance a holding pattern.^8,9 Although it is difficult to estimate the cost of boarding for health care institutions, estimates based on single-center studies place the average cost of each boarded patient at $4269¹⁰ and the hourly cost at around $219.¹¹ 

Our institution worked to manage the COVID-19 crisis in parallel with the continued inundation of behavioral health needs. Limited psychiatric beds in the community led to long lengths of stay in the medical hospital for patients awaiting placement and treatment. Because we faced both high patient volume and acuity, community behavioral health services were also strained. With longer wait times for partial hospitalization programs (PHP), discharge planning was difficult, and some children were admitted because timely outpatient services could not be located. Examining data from our hospital, we found that the average boarding time for patients in 2020 was 40 hours with some patients waiting up to 3 weeks.

Children presenting in acute mental health crisis have similar needs from the time they arrive for assessment to the time they board on medical units. Literature supports the notion that increasing contact with clinical staff and introducing interventions in the medical setting can lead to the stabilization of youth boarding for mental health treatment. In one study conducted at a large, urban pediatric hospital, 90% of patients boarding received brief therapy and 50% received medication management supports, resulting in a number of youth showing improvement in symptoms while boarding (33%), some showing decline (12%), and the rest with no change in presentation.¹² Evidence-based interventions such as the Family Based Crisis Intervention^13–15 and the Family Intervention for Suicide Program ED¹⁶ support the concept that youth in mental health crises can stabilize while in the ED before being safely discharged home with close follow-up.

Given the literature outlining safe and promising evidence-based interventions, along with our clinical team’s experience working with youth in the ED and those boarding for psychiatric care, our team implemented a quality improvement (QI) project to address increased boarding in our ED by using a stabilization protocol that included additional resources such as staffing and brief evidence-based interventions for youth. Through the creation of a stabilization protocol for patients experiencing boarding in our ED observation unit, we aimed to decrease the percentage of mental health encounters resulting in admission to inpatient psychiatric hospitals. A secondary aim was decreasing the length of stay (LOS). This article was written according to Standards for Quality Improvement Reporting Excellence 2.0 guidelines.¹⁷ 

The project was implemented at a 600-bed free-standing, urban pediatric hospital with more than 100 000 patients presenting to the ED annually. About 4.5% of these patients have behavioral health needs. Our team’s focus for this project was a 14-bed short stay ED observation unit, the emergency department extended care unit (EDECU). The EDECU was originally created for patients requiring a medical stay of less than 36 hours; over the last 5 years, however, this unit has primarily been used for patients who are boarding as they await psychiatric placement. Before this project, there were no psychiatrists, social workers, or child life specialists assigned specifically to the EDECU. Patients would have brief interfaces with ED psychiatry for daily mental status examinations, and the remainder of their day would be unstructured, with family visits, check-ins from other providers, and mealtimes. Social workers tasked with locating treatment beds would work with families to find placement.

This project was undertaken in late February 2021 by a multidisciplinary team comprised of ED physicians, advanced practice providers, nursing, psychiatrists, social workers, safety observers, child life specialists (CLS), and advisers from our hospital’s quality assurance department. The project team first met with hospital leadership to describe the increased boarding rates and LOS in the EDECU and secure support to plan the interventions. The multidisciplinary team met twice a week for 2 weeks to develop a key driver diagram (Fig 1) to identify areas for change and plan interventions. Implementation involved clinical considerations, like treatment interventions and daily structure, and operational considerations like additional staffing, changes to the unit environment, and enhancements of the electronic health record (EHR). The team created a process map to illustrate the new care model incorporating the stabilization protocol (Fig 2). These change ideas were implemented on March 1, 2021.

The new care model included operational and clinical changes. Operationally, providers were designated to the specific boarding patients in the EDECU. At the onset of the intervention period, 2 clinically licensed social workers and a psychiatrist were redeployed from the main ED to the EDECU and shifted roles after business hours to provide support to the ED. This reconfiguration of staffing ensured presence on the unit each weekday. The CLS revamped the unit playroom, and considering the needs of psychiatric patients, added activities that could be completed throughout the day or after hours (arts and crafts, puzzles, games, journals). An electronic order was developed to indicate the patient was on the stabilization protocol. In August 2021, approval was given to hire 2 full-time social workers whose primary roles would be in the EDECU, continuing to work on the pilot intervention and supporting all patients experiencing psychiatric boarding. Nursing and CLS were assigned to the unit for continuity of care.

Clinically, the psychiatrist, social worker, CLS, safety observers, and the medical providers met each morning in the EDECU to discuss patients on the protocol, areas of progress and challenge, and the plan for the day. Patients on the protocol benefited from more therapeutic time with the clinical team. Social workers taught patients cognitive behavioral therapy (CBT) skills, including problem solving, behavioral activation, and relaxation strategies (such as guided imagery and breathing exercises). Social workers engaged patients in brief interventions involving safety planning¹⁸ and enhancing family communication. Caregivers were provided psychoeducation on mental health diagnoses, the system of care, and lethal means reduction. Social workers met with the patient and family separately and together and remained in communication with the psychiatrist to discuss progress.

Patients with behavioral health needs were assessed in the ED, and those recommended for inpatient psychiatric care were eligible to board in the EDECU if they were aged at least 8 years, able to complete Activities of Daily Living independently, and did not require violent restraints at the time of transfer to the unit. Patients with diagnoses of development delay or intellectual disability, patients presenting with severe aggression, and patients with active eating disorders are not admitted to the EDECU and are instead treated elsewhere throughout the hospital. Those patients experiencing boarding in the EDECU with an initial disposition of inpatient psychiatric treatment due to suicide risk were considered for the stabilization protocol. All patients boarding at baseline and during the intervention period met the same, broader eligibility of the EDECU.

Inclusion criteria for the stabilization protocol consisted of the patient presenting with suicidal thoughts, a caregiver available for intervention, and a voluntary admission status (patient/parent agreeable to psychiatric hospitalization). To determine which patients to place on the protocol, psychiatry and social work met to review each admission, focusing on the reason for presentation, history of mental health treatment, and family supports. This behavioral health team focused on a group of 3 to 4 patients at a time for the protocol given the time needed to conduct the evidence-based interventions for each patient.

Pilot patients were defined as patients who underwent the stabilization protocol during the intervention period, whereas nonpilot patients received care as usual in the EDECU while boarding for psychiatric placement. With the initiation of the pilot project, we designated CLS and music therapy on the unit, allowing structured recreation time and encouraging more engagement among caregivers, children, and staff. Nursing staff and CLS were instrumental in motivating children to engage in activities. Additionally, CLS scheduled daily, individual sessions with each child during which they would facilitate either a developmental or therapeutic play intervention, allowing children to process, practice coping strategies, and model a healthy and therapeutic relationship. Throughout the intervention period, nonpilot patients continued receiving care as usual, consisting of daily psychiatric assessments and social work involvement focused on inpatient placement. The extra recreation and music therapy time and a daily schedule with consistent staff engagement were available to all patients in the EDECU, including pilot and nonpilot patients.

Data were extracted from the EHR from January 2019 through February 2021 (baseline period) and from March 2021 to May 2022 (intervention period). Data included disposition, designation of pilot or nonpilot, LOS in the EDECU, and return visits to the ED. The team monitored the interventions in monthly meetings to review data for the previous month with attention to disposition plans—whether admitted to the inpatient medical unit, transferred to an inpatient psychiatric facility, or discharged home with other treatment recommendations. For youth discharged home, services recommended ranged from a PHP, an intensive outpatient program, wraparound services, or outpatient therapy.

The primary outcome measure was the percentage of EDECU mental health encounters resulting in the patient being admitted to inpatient psychiatric services and the percentage of encounters in which the patient was discharged to home. Encounter is defined as a period of hospital admission, meaning 1 encounter could represent several days. A secondary measure was the LOS as defined by the time from when the patient arrived in the EDECU to when the patient was discharged from the EDECU. A balancing measure of the percentage of encounters admitted from the EDECU to our inpatient medical unit was used to ensure that any improvements seen in the outcome measures were not a result of an increase in utilization of our inpatient medical services. Another balancing measure included the proportion of encounters with ED revisits for behavioral health concerns within 30 days of EDECU transfer or discharge.

Statistical process control (SPC) charts were used to analyze our outcome and balancing measures. Over time, when comparing baseline with intervention periods, applying Shewhart’s rules for special cause variation recommends a centerline shift if there are 8 consecutive points above or below the baseline average.¹⁹ We used χ² tests to compare outcomes between pilot and nonpilot encounters with suicidal ideation during the intervention period. Analyses were conducted using R statistical software (version 4.22, Vienna, Austria).

This boarding stabilization project was reviewed with our hospital’s Institutional Review Board guidelines to make the determination for whether data collection could be considered QI, and the project met criteria to be considered QI.

During the baseline period, there were 758 encounters for which patients were admitted to the EDECU for inpatient psychiatric boarding (representing 12.3% of ED encounters with behavioral health concerns). During the intervention period, there were 761 such encounters (12.4% of ED encounters with behavioral health concerns). The stabilization protocol was implemented in 164 encounters, involving 159 unique patients; this accounted for 22% of total encounters in the EDECU during the intervention period. The nonpilot protocol group accounted for 597 total encounters during the intervention period, with 446 of those encounters having suicidal ideation as part of their presentation and 151 encounters having a psychiatric concern not related to suicidal ideation. Twenty patients had at least 1 pilot encounter and 1 nonpilot encounter during the intervention period. Characteristics of patients in the baseline and intervention period are shown in Table 1.

During the baseline period, 65.7% (n = 498) of encounters in which patients boarded in the EDECU resulted in the patient being admitted to an inpatient psychiatric facility (Fig 3). A centerline shift occurred during the intervention period with this measure decreasing to 49.0% (n = 373). The balancing measure of admission to inpatient medical unit from the EDECU occurred in 16.9% (n = 128) during the baseline period compared with 8.7% (n = 66) of encounters during the intervention period (Fig 4). During the intervention period, 1.8% (3 of 164) of pilot encounters were admitted to the medical unit compared with 8.1% (36 of 446, P = .000) of similar nonpilot encounters (Table 2). ED revisits in 30 days occurred in 1.3% (n = 10) during the baseline period compared with 2.1% (n = 13) during the intervention period, and 3.7% (6 of 164) of pilot encounters compared with 1.6% (7 of 446, P = .122) nonpilot encounters (Fig 5).

Comparing pilot with similar nonpilot encounters with suicide risk during the intervention period, 58.2% (261 of 446) of those nonpilot encounters resulted in admission to an inpatient psychiatric facility compared with 22.6% (37 of 164, P = .000) of pilot encounters (Table 2). Pilot encounters (75.0%) were more likely than nonpilot encounters with suicide risk (31.4%, P = .000) to result in the patient being discharged home and had a median EDECU LOS of 95.5 hours (IQR 47.4–144.0) compared with 66.4 hours (IQR 36.6–129.1, P = .000) for nonpilot encounters (Table 2). The LOS during the baseline period was 50.5 hours compared with 101.7 hours during the intervention period (Fig 6).

The stabilization protocol shows promise in reducing boarding experienced by pediatric patients awaiting inpatient psychiatric hospitalization. Patients who received this protocol were more likely to be discharged than patients who did not receive the pilot interventions. These findings align with aspects of the limited literature on mental health stabilization of pediatric patients boarding.¹² Patients who received the stabilization protocol had a longer LOS than nonprotocol patients; however, they had a lower rate of inpatient medical hospitalization, which freed beds in the medical units during times of high capacity in our hospital. In addition, we found that patients undergoing the stabilization protocol in the EDECU were more likely to remain in the EDECU and engage in interventions to facilitate discharge home, whereas individuals less likely to be able to discharge home would more frequently be admitted to the medical unit for prolonged boarding, leading to higher EDECU LOS in protocol patients.

In the months leading up to this QI project, the percentage of EDECU encounters admitted to an inpatient psychiatric facility was decreasing and the EDECU LOS was increasing (Figs 3 and 4). During this time, fewer patients were transferred to inpatient psychiatric facilities from the EDECU largely because of limited bed availability in the community from the strain of the COVID-19 pandemic, leading to longer boarding times. Also, some patients with a recommendation for PHP were temporarily boarding in the EDECU because of unsafe wait times for these services and limited services to bridge to care. These trends helped bring our multidisciplinary team together to advocate for and start this QI project.

Having dedicated members of the behavioral health team present allowed for several important improvements in care. The psychiatrist was able to spend greater time with each boarding child and family, allowing the chance for reassessments when clinically indicated. In addition, psychiatrists could dedicate time to conceptualizing cases along with medical and social work teams in a multidisciplinary meeting focused on patients who were boarding. Psychiatrists could also effectively initiate and manage medications as indicated for these patients. Social work clinicians could engage more intensively with families. In addition to the treatment interventions from the behavioral health team in the EDECU, having additional staff and a daily structure to create a therapeutic environment was crucial and benefited all patients, not just those receiving the stabilization protocol. The approval to hire 2 full-time social workers in August 2021 was a critical move, illustrating the hospital’s commitment to support this initiative. After the QI intervention period, the protocol became standard treatment for patients that are boarding in the EDECU presenting with suicidal thoughts.

In 2019, a group of mostly child psychiatrists and emergency medicine physicians formed a panel to discuss the behavioral health crisis among pediatric hospitals and to create a series of best practice recommendations for pediatric patients boarding in medical settings.²⁰ When asked about staffing, participants on the National Pediatric Boarding Consensus Panel identified access to social work as the most important factor related to staffing and provider consistency. Our project affirms that these factors are critical in caring for pediatric patients while they board and may be important when considering necessary staffing for stabilizing patients.

For children requiring medical intervention, the most acutely ill children receive care first. For children requiring psychiatric services, the severity of their illness does not translate to faster access to care. This concept of reverse triaging⁹ represents issues within the health system, but medical hospitals can meet these challenges by changing the way they care for patients boarding for psychiatric care. There is limited research on interventions specifically targeting youth who are boarding, and little is known about how boarding impacts the psychiatric presentation of youth in mental health crisis. Given that youth typically receive little to no mental health treatment while boarding, our team anticipated that a number of youths would respond well to additional staffing supports and skill-building, which is what occurred.

There are several important limitations to note. Often, each of the 14 beds on the unit were occupied by a patient experiencing boarding, making it necessary to determine which patients would receive the stabilization intervention when several were eligible. Patients who underwent the stabilization protocol had a caregiver present as part of the inclusion criteria; children with family support present may have been more likely to stabilize at baseline. Social work practitioners used different approaches within the CBT framework that varied based on clinical style. Patients were accepted and transferred to inpatient psychiatric hospitals often sporadically and without warning, interfering with ongoing stabilization in the EDECU. Our project relied on increased staffing, which may be difficult to translate to settings with fewer resources. However, we believe institutions should consider the overall impact, cost, potential savings, and value to their patients, staff, and systems of care.

Because of the rapid implementation of this project, we were unable to elicit formal feedback from stakeholders, including caregivers whose children received this intervention or ED providers who worked on this unit before and during this project. Anecdotally, numerous caregivers expressed positive feelings about the time spent engaged with the team and seeing their children stabilize for discharge. ED staff also provided positive feedback regarding the increased behavioral health supports and expressed optimism that we were working toward solutions for the boarding crisis, showing promise in addressing the moral distress providers have previously associated with psychiatric boarding.⁵ 

Patients awaiting inpatient psychiatric hospitalization in a medical setting are an often-overlooked patient population. This study shows that by working as a multidisciplinary team, reallocating resources, and dedicating staff to initiate treatment for patients experiencing boarding, we can decrease the need for inpatient psychiatric care. Retaining clinicians specifically tasked to work with families experiencing boarding is key to providing better care and creates opportunities for intervention. Although these efforts are not meant to replicate or replace inpatient psychiatric treatment, they offer a pathway to stabilization and intensive community-based services for youth who might otherwise be waiting with no treatment.

The authors would like to thank the clinical teams who provide excellent care to families in the emergency department. We thank Megan Deetscreek and Amanda Nomie for their thoughtful contributions to the preparation of this article. We thank Kristin McNaughton for editing assistance. We thank Anik Jhonsa, Karen White, Jane Lavelle, and Evan Fieldston for their continued support of the pilot stabilization program.

ED

emergency department

EDECU

emergency department extended care unit

CBT

cognitive behavioral therapy

EHR

electronic health record

LOS

length of stay

PHP

partial hospitalization program

QI

quality improvement

SPC

statistical process control