Pediatric patients with behavioral needs are frequently admitted to the hospital for medical care; when behavioral crises occur, patients and staff are at risk for injury. Our aim was to implement a behavior response team (BRT) to increase the days between employee injury due to aggressive patient interactions on the inpatient medical units from 99 to 150 over 1 year.
A multidisciplinary team used quality improvement methods to design and implement the BRT system that includes 2 options: huddle to proactively plan for patients exhibiting early signs of escalation and STAT for immediate help for patients with imminent risk of harm to self or others. Using run and statistical process control charts, we tracked events per month, days between Occupational Safety & Health Administration-recordable events, and violent restraint use over time for 1 year after implementation. Staff pre and postimplementation surveys were compared to assess staff perception of safety and support provided by the BRT intervention.
The BRT was implemented across the inpatient system in July 2020, with an average number of 13 events per month. Days between Occupational Safety & Health Administration-recordable events remained stable with a maximum of 134 days. Restraint use remained stable at 0.74 per 1000 patient days. The perception of behavioral support available to staff increased significantly pre to postsurvey.
The implementation of a BRT can improve staff perception of support and confidence in safely caring for patients with behavior needs on the inpatient medical unit, although additional provider- and system-level improvements are needed to prevent employee injuries.
Pediatric mental health needs are increasing in the United States, and resources are inadequate to support children in crisis.1,2 Patients at risk for harm to self or others may require admission to a psychiatric facility. However, limited bed availability often necessitates holding in emergency departments (EDs) or admission to inpatient acute care settings while awaiting psychiatry services. The number of patients “boarding” in these areas has increased,3 with an estimated average duration of up to 41 hours in the ED4 and a median of 48 hours in the inpatient acute care setting.5
Additionally, pediatric patients with disruptive and agitated behaviors, usually secondary to underlying psychiatric disorders (eg, anxiety, oppositional defiant disorder), challenges with sensory processes and communication (eg, autism spectrum disorder), or medical causes (eg, delirium, drug ingestion, central nervous system infection) are frequently admitted to the hospital for the treatment of medical conditions.6 These behaviors include aggression, obstacles to participation in necessary medical procedures, and attempts to leave the hospital that can pose self-harm or staff injury risk. However, the medical inpatient unit may not be adequately resourced to provide preventive strategies, optimal environmental conditions, and support necessary to avoid these behaviors during routine care or procedures.7,8 At our hospital, a Behavior Safety Service was created in 2010 to provide behavioral support through inpatient medical unit consultation, with the primary goal of ensuring that every patient receives necessary medical care. The Behavior Safety Service is composed of staff from diverse mental health backgrounds (psychology, counseling, social work, special education) who have extensive de-escalation training and patient care experience; our hospital utilizes therapeutic crisis intervention (TCI)9 training across the organization. The service is consulted for patients with a history of active behavioral needs, such as aggression, flight risk, or difficulty participating in medical procedures.
Despite the safety recommendations provided by the Behavior Safety Service, aggressive patient interactions continued to occur without reliable methods to anticipate and prevent events. Several events led to patient and staff injury, due to the lack of identification of high-risk patients and situations, recognition of signs of escalation, and appropriate response to these behaviors. After these events, a multidisciplinary team convened to perform a failure modes and effects analysis (FMEA)10 to systematically identify top areas for improvement. The FMEA team identified the need to (1) refine risk prediction abilities by sharing safety plans across encounters and sites of care and (2) develop a more robust system to proactively prevent and respond to escalating patient behaviors, with consideration of expanded Behavior Safety Service resources and de-escalation training. In this report, we focus on the second improvement area.
Using quality improvement (QI) methods,10 we designed and implemented a Behavior Response Team (BRT) to augment established efforts in our system and to more effectively respond to behavior events on the medical inpatient units. Our aim was to increase the days between Occupational Safety & Health Administration (OSHA)-recordable employee injuries due to aggressive patient behaviors from 99 to 150 over 1 year. Our secondary aim was to maintain or lower the use of violent restraints at a rate of 0.74/1000 patient days over 1 year. We also sought to assess staff satisfaction with behavioral support at the bedside.
Methods
Context
This improvement study took place at Cincinnati Children’s Hospital Medical Center, a free-standing, multiple-campus quaternary care system with 530 medical beds and 140 (110 inpatient, 30 residential) psychiatric inpatient beds in Cincinnati, Ohio; psychiatric beds are located at a separate campus from the base hospital. The study was conducted at the base campus, which housed 13 acute care medical inpatient units; the PICU was also included. Hospital admissions totaled nearly 29 000 in fiscal year 2020. In addition to nurses and physicians, many patients with agitation are supported by bedside attendants (patient care assistants or behavior health specialists with de-escalation training).
Interventions
A multidisciplinary study team included pediatric hospitalists, Behavior Safety Service leaders, nursing managers and directors, a psychiatry nursing project manager, a human factors research expert, pediatric resident physicians, and bedside nurses. We solicited family member input from parents who have children with behavioral support needs who are frequently admitted to the hospital before and during the implementation process to ensure a patient-centered approach. The team mapped the existing behavioral response process and identified key drivers (Fig 1). Interventions were designed to address these drivers, with emphasis on learning from process failures.
Expansion of De-Escalation Training
In preparation for creating a team available to respond at all hours, we presented the findings of the FMEA investigation to hospital leaders to garner financial support to hire additional Behavior Safety Service staff. We also expanded TCI training (January 2020) across roles to include nursing managers, protective services, and patient care attendants, which facilitated more team members available to approach events with shared language and TCI certification in de-escalation and restraint.
Standardization of a Behavior Response Effort Across the Inpatient System
Team Design
Bedside team input is critical to the success of a BRT because the bedside attendants, nurses, and physicians caring for patients with behavioral risk are experts in their medical history and care needs. To compose the BRT group, the bedside team members are joined by additional TCI-trained staff: the on-shift manager of patient services (system-level nursing safety leader), a protective services officer, and a Behavior Safety Service member in the team leader role (Table 1).
BRT Process Development (March 2020)
The BRT process includes 2 options: BRT huddle and BRT stat. Huddles are called to mitigate signs of early escalation through proactive planning, with the intent to prevent behavior escalation; responders arrive within 15 minutes to align with existing medical response team expectations. BRT stat is called for patients with imminent risk of harm to self or others; responders arrive as quickly as possible, aligning with Code Blue expectations.
For both options, the bedside team calls the central hospital operator, who alerts BRT responders with the patient’s location and huddle versus stat description. The bedside team (nurse or charge nurse) calls the physician/advance practice provider (APP) team caring for the patient if the team is not already at the bedside.
When the BRT arrives at the unit, the Behavior Safety Service member identifies him/her/themself as the BRT leader. Responder roles are delineated in Table 1. The leader utilizes a script to elicit bedside team concerns and collaborates with the team to develop and execute an appropriate de-escalation plan using TCI principles (Table 2). If the team arrives to find a patient actively harming themself or others, the plan is made quickly with rapid role assignment and in-room intervention.
BRT Process Support
Given that BRT huddle and stat events involve rapidly forming teams and may occur on units that do not frequently encounter patients with agitation or aggression, several support materials and system enhancements were developed:
Materials to guide roles and standard workflow, including a script that outlines each BRT step (Table 2).
The most commonly used oral and intramuscular medications were placed in the dispensing machine on every unit for easy access so the care team did not need to wait for medications to arrive from the pharmacy (June 2020).
Each unit was supplied with a BRT backpack, placed in a central location (usually near the medical code cart), which included the script, behavioral protective personal equipment (eg, Kevlar arm sleeves, gloves, etc), ligature removal tool, restraints, and impact cushion (September 2020).
After a BRT event, the physician/APP caring for the patient documents an acute event note in the electronic health record (EHR). The QI team partnered with EHR leaders to create a BRT-specific section in the same area in which medical events are documented. The section includes common orders, including medications with doses and violent restraints by age, as well as a brief note template (September 2020).
Process Buy-In and Shared Ownership
The QI team included members from stakeholder groups to elicit feedback from their broader groups as the process was developed. Beginning in May 2020, the QI team led several mock BRTs, first in “tabletop” format to work through the general process and later in simulated scenarios on the units. Bedside staff feedback informed iterative process improvement. Educational materials were later distributed during staff meetings, institutional meetings, and via e-mail.
We obtained provider buy-in via education sessions with pediatric residents, Hospital Medicine attending physicians/APPs (medical service that most frequently cares for patients with agitated behaviors), and institutional medical director meetings for spread to other specialists who provide inpatient service.
Feedback and Shared Learning
To elicit feedback, we developed a short electronic survey (August 2020) to be completed by the team leader after each event, with pertinent information regarding event details, whether medications or restraints were needed, if injuries occurred, and any learnings. Family feedback was sought by the team informally after events. The BRT leadership team reviewed events in real-time, soliciting additional information from unit leaders, and met monthly to review the number of events, any injuries, and process improvement ideas.
Leadership Support to Team Members
Given the risk of behavioral safety work causing emotional and physical injury, team member support was incorporated into the process. After each stat event, the team leader conducted a debrief session to provide a forum for open discussion, which included ensuring that any injuries were reported to our employee safety hotline and that any necessary medical care was sought immediately. Team members were invited to share what went well and areas for improvement and confirmed that all were emotionally ready to return to their roles. Postevent, unit leaders followed up with team members and if necessary, referred them to our local peer-to-peer support or employee assistance programs.
Study of the Intervention
We used plan-do-study-act cycles to design and learn from process adjustments.10 We documented the number of huddle and stat events monthly. Baseline numbers of OSHA-recordable employee injuries and restraint frequency were collected for 1 year (July 2019 to June 2020) before BRT implementation and then prospectively for 1 year after implementation. A modified survey developed by Zicko et al11 was used to assess behavioral resources and support available to the same bedside staff before and after the process had been live for >6 months. The survey was administered to nurses, Behavior Safety Service members, bedside patient attendants, pediatric residents, and Hospital Medicine attending physicians/APPs.
Measures
Primary measures included (1) monthly BRT huddle and stat events to follow utilization of the new process over time and (2) the outcome measure of days between OSHA-recordable employee injury events due to aggressive patient behaviors. Secondary measures included (1) the rate of violent patient restraints per 1000 patient days on medical inpatient units and (2) employee satisfaction pre and postimplementation. The rate of violent restraints also served as a balancing measure to ensure restraints were not recommended at a higher rate during BRT events. We also documented the number of patients cared for by the Behavior Safety Service monthly.
Analysis
The primary outcome measure of days between OSHA-recordable events was depicted on a statistical process control T chart. All OSHA-recordable employee injury events were reviewed by members of the study team (AS and MS) to determine if the events were potentially preventable by a BRT (ie, signs of patient escalation that could have been addressed earlier through the efforts of the multidisciplinary team to proactively prevent the injury, but a BRT was not called). BRT events over time were followed on a run chart, and the violent restraint rate was depicted on a statistical process control U chart, with special cause delineated by points outside control limits (defined as 3 SDs greater or less than the mean).12
For survey analysis, descriptive statistics were used to summarize sample demographics and survey items. A paired t test was performed to compare baseline and post-intervention survey scores. All data were analyzed in SAS v9.4 (SAS Institute, Inc, Cary, NC) with a 2-sided significance level of .05.
Ethical Considerations
With this work, we aimed to improve the local quality of care and not to generate generalizable knowledge. Therefore, it was not considered human subjects research and did not require institutional review board review.
Results
The number of BRT events in the year after BRT implementation averaged 13 per month (total 167 events), with 55% huddles and 45% stat events; each event was counted separately, although some patients had >1 event during hospitalization (Fig 2A). There were 6 OSHA-recordable employee injuries due to aggressive patient interactions, with a maximum of 134 days between events (Fig 2B); 2 of these events were deemed BRT-preventable, whereas the other 4 events occurred in the context of an unexpected escalation during routine medical care. No specific interventions were associated with the length of time between events as the BRT process was implemented. Enhancements to the process were made based on feedback during events, such as the addition of the BRT backpack as a resource for easy access to items like restraints in times of crisis and modifications to medication access (EHR centralized orders, placement of common medications in the unit dispensers). The need for violent restraints remained unchanged at 0.74 per 1000 patient days (Fig 3). During the study period, the overall hospital census remained relatively stable month to month, although it was lower than previous years because of the effects of the coronavirus disease 2019 pandemic. In comparison with the previous year, Behavior Safety Service followed an average of 1 more patient per month across medical units (11 patients per month compared with 10 patients per month in the past), suggesting an overall higher behavioral patient acuity in our system over the first year in which the BRT was implemented.
A total of 326 providers and staff (response rate 20.5%) completed both the pre- and post-survey, with findings summarized in Table 3. Several areas significantly (P < .0001, 1–10 point Likert scale) improved pre to post BRT implementation, including the level of confidence in caring for escalating patient behaviors (6.2 to 6.83), knowing who to call (6.67 to 7.74), and support provided by the responding team (7.82 to 8.27). The effectiveness of interventions suggested by BRTs and the plans created were rated highly in the ability to address patient and staff needs (9.26 and 9.16, respectively).
Discussion
Using the Model for Improvement,10 we implemented a Behavior Response Team to support patients at risk for aggressive interactions and respond to events to increase safe medical care without risk of injury to patients or staff. There was an average of 13 BRT events per month. After implementation, there was a maximum of 134 days between OSHA-recordable events, which did not meet our goal. Two of the 6 events were deemed potentially preventable by a BRT, meaning that a BRT may have been helpful but was not called. Although the BRT process provided additional real-time crisis support to staff, its implementation was not sufficient alone to sustainably decrease OSHA-recordable events. The use of violent restraints remained the same throughout the study period. Staff reported feeling more supported with a more robust safety system in place.
We implemented a new BRT system to supplement our previous Behavior Safety Service consultation model. The BRT facilitated the ability to collaborate at the system level and increased TCI-trained staff expertise availability for timely and reliable event response. Our new BRT process was well used by inpatient teams, as evidenced by its immediate and consistent uptake, with one benefit being the impact on unit staff. Caring for patients with disruptive and agitated behaviors poses a threat to providers’ psychological and physical safety. Violence in the health care workplace has been highly studied throughout the world over the last decade; nurses remain the highest reporters, and the ED has been a primary area of focus in the literature.13–16 Although we were unable to sustainably decrease OSHA-recordable events due to aggressive patient interactions, our pre post survey indicated that our teams felt more confident caring for their patients, had knowledge on how to obtain help, and felt supported by additional resources. Having a functioning system in place to easily access timely help is important for the wellbeing of not only our patients but also the staff members who care for them.
Although similar behavior emergency response teams have been developed at other institutions,17,18 only Zicko et al11 studied and showed a decrease in injury events. We identified several explanations for our inability to increase the number of days between OSHA-recordable events due to aggressive patient interactions. First, although historically injuries have been underreported in workplaces,13,16 the systemic focus on safety in the institution could have increased reporting of events during this time period. Additionally, the increased prevalence of mental health needs1 and higher rates of hospitalization6 across the country create a higher risk for aggressive patient interactions in the inpatient setting, as noted at our institution with Behavior Safety Service consultations on more patients monthly than in the past. The coronavirus disease 2019 pandemic further exacerbated mental health crises in children and highlighted the need for more accessible mental health services.19–21 Given these challenges, it is impossible to know how many events were prevented by the implementation of the BRT in our system.
In addition to the BRT process, overall safety requires a multifaceted approach that incorporates additional improvements at both the individual provider and the system levels. Widespread de-escalation training with regular interval recertification, as well as individual accountability to follow safety procedures at all times, are imperative. During our team’s examination of the causes of the OSHA-recordable events that occurred in our study, most were due to system-level factors that would not have been addressed or preventable by the BRT. For example, factors such as environmental safety checks not being completed or a patient unexpectedly escalating during routine nursing care can lead to harm that is unable to be mitigated through the BRT process. Although the BRT can assist in reminding team members of safe practices, its reach is limited and should be considered in the context of a larger safety system, which relates back to our initial FMEA work identifying the need for refined risk prediction practices across the system; this aspect was not addressed by the new process. In the future, our team hopes to increase the number of TCI-trained providers, reinforce best practices on every shift with every patient interaction, and continue to standardize the process of risk identification.
We did not note a decrease in violent restraint use with BRT implementation, but our overall rate was already low; this measure also served as a balancing measure to ensure our BRT process did not lead to higher restraint use. Although difficult to directly compare, the rate is similar to a study by Dalton et al,22 which revealed a rate of 1% of behavioral health patient admissions during which a physical restraint order was placed for violent behavior after interventions to reduce restraint use. Through the BRT process, we aimed to ensure patient and employee safety; although this approach encourages the use of de-escalation techniques and pharmacological interventions as indicated, it also includes team discussion and agreement for using physical holds and violent restraints.
This QI initiative was completed in a single institution with extensive resources and may not be generalizable to other organizations. The implementation process required financial support from institutional leadership, chiefly in hiring additional Behavior Safety Service team members to provide 24/7 coverage. Additionally, the Behavior Safety Service is a service that was well-established before our project, and whose expertise was instrumental to the response team implementation; although a similar service may not be available at other hospitals, the evaluation of resources to seek and implement behavioral support is foundational to the development of a response team. Our overall survey response for both pre and post assessments was low, which may introduce bias and affect the interpretation of the survey results; additionally, we were unable to determine clinical or psychological significance to assess the true impact on staff. For future work to determine the impacts of BRT-like interventions, we recommend more sophisticated measurements of psychological safety and related outcomes, such as nursing turnover. Our study is also lacking an assessment of patient and family perspectives on the BRT process, although anecdotal feedback from families was overwhelmingly positive.
Conclusions
We implemented a Behavior Response Team to respond to aggressive patient interactions on medical inpatient units, with options for a proactive planning huddle or stat to respond to urgent needs. The new process was highly used by bedside staff with survey results indicating an improved perception of resource support and confidence in providing care to patients at risk for injury to self or others. However, we were unable to increase the number of days between OSHA-recordable injury events, indicating the BRT system alone is not adequate to ensure employee safety, and additional work in this area is critical. Future steps include expanding de-escalation training, encouraging teams to proactively call early BRT huddles to avoid escalation to stat events, considering how the BRT system can better augment and support other interventions that address system-level complexities to facilitate a more robust safety culture, and assessing psychological safety impacts on staff.
Acknowledgments
The authors wish to acknowledge Lindsay Boehl, Tina Brooks, Joelene Kammer, Joseph Staneck, Alexander Sullivan, and the Behavior Safety Service and Protective Services teams at our institution for their support in the development and implementation of BRTs.
Dr Statile conceptualized and designed the study, drafted the initial manuscript, collected and analyzed data, and reviewed and revised the manuscript; Mr Schweer, Drs Herrmann, Warniment, Daraiseh, and Simmons, Ms Duncan, Ms Demeritt, Ms Keehn, Mr Edwards, and Mr Brown participated in study design, analyzed data, and critically reviewed the manuscript for important intellectual content; Ms Whitesell and Ms Lin collected data, conducted analyses, and critically reviewed the manuscript for important intellectual content; Ms Muth, Dr Sorensen, and Mr Hill critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Dr Simmons’ current affiliation is Department of Value and Clinical Excellence, Children’s Minnesota, Minneapolis, MN.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest relevant to this article to disclose.