Improving Psychiatric Access for Students in Crisis: An Alternative to the Emergency Department

With one in five children and adolescents experiencing a mental health condition in any given year, timely access to high-quality behavioral health services remains an ongoing challenge (1). In the absence of adequate outpatient system capacity, the emergency department (ED) has become the de facto provider of mental health services, particularly for youths in crisis, as reflected in the persistently rising rates of behavioral health ED utilization (2, 3). Many factors contribute to making the ED a suboptimal setting for addressing children’s behavioral health needs; it delivers, at best, variable quality at a high cost. Common issues include a lack of specialized child mental health staffing, limited or absent care coordination, long wait times, a potentially stigmatizing environment, and poor connections to aftercare (1). These issues can set the stage for repeat ED visits or no follow-up at all (4). Common pediatric behavioral health problems referred for evaluation include suicidality, self-injury, anxiety, disruptive behaviors, and substance abuse (5, 6). Many of these ED presentations are avoidable and could be addressed in an ambulatory setting (7); however, outpatient services are often ill equipped to address these issues in a timely fashion or are simply unavailable (8).

Within this context, schools have emerged as one of the largest referral sources to pediatric EDs, referring 30%–40% of the total behavioral health volume, most of which is avoidable (7). Communication between the pediatric ED and the referring school personnel is generally considered inadequate, providing the basis for repeat visits to the ED and incredible stress on families. Moreover, access to follow-up psychiatric care after an ED visit is extremely limited, particularly for children on Medicaid. The current pilot project was initiated because of feedback from parents and educators on how to better meet the needs of students in crisis. The objective was to determine whether a model of same-day, ambulatory psychiatric evaluation could improve access for students in crisis and, by extension, reduce preventable ED utilization.

The Urgent Evaluation Service Model

The urgent evaluation service (UES) model was piloted at a large urban hospital in Brooklyn, New York, beginning in November 2013. The payor mix was approximately 80% Medicaid, with large Latino, Chinese, and South Asian immigrant populations. An examination of pediatric ED referral patterns during the 2012–2013 school year identified high rates of school referral (N=240, 42%) coupled with low admission rates (N=12, 5%), suggestive of preventable patterns of utilization. UES staffing leveraged resources from the existing child psychiatry ED-inpatient consultation team, composed of a dedicated attending child psychiatrist, social worker, and trainee (postgraduate year-2 resident or a social work intern). This core team was physically based in the pediatric psychiatry clinic, with consultations distributed by the attending psychiatrist as they were called in. Given that nearly half of the ED consultations came from schools, it was predicted that the addition of UES responsibilities would not overburden these resources but would instead shift these encounters to the clinic setting. The UES hours, Monday–Friday, 9 a.m. to 3 p.m., matched those of the consultation service and the school day.

Referral criteria focused on safety concerns (suicidal or homicidal ideation) or behavior severely affecting school functioning, absent a recent suicide attempt, acute agitation, severe intoxication, need for medical attention, or any other factor that would preclude safe assessment in an outpatient setting. UES referrals were received at the pediatric psychiatry clinic, at which time the attending physician was notified, reviewed the referral, and triaged the case. If a child presenting to the UES required a higher level of care, he or she could be escorted directly to the pediatric ED, located on the hospital campus. Students who were appropriate for UES would receive same-day assessment, coordination of care with referring school staff, and linkage to outpatient treatment, if warranted. Consistent with clinic and local state policies, students had to be accompanied by a parent or legal guardian who could sign consent for treatment and for communication with school personnel. Students whose parents could not accompany them to the UES would be either referred to the ED or scheduled to be seen in the UES when a parent was available, usually the next business day. If needed, short-term follow-up visits could be provided. A majority of commercial, Medicaid, and managed Medicaid plans were accepted. Uninsured students and those out of network were seen for a single visit free of charge.

For this pilot study, we assessed the impact of the UES on pediatric ED referrals. Approximately 20 “high-utilizer” schools with four or more behavioral health ED referrals in the past school year were identified and invited to attend an information session that introduced the new service. A brief referral form was created to facilitate communication by clearly conveying the reason for the consultation as well as the contact information of the referring school staff. School administrators were also provided with contact information to reach the service’s attending psychiatrist if they had questions about making a referral. Because “inappropriate” (i.e., too acute) UES referrals could easily be escorted to the pediatric ED, this practice eliminated the risk of using the “wrong door” and helped foster school buy-in.

Initial Findings

UES patient characteristics.

From January 1 to June 30 of 2014, a total of 72 students were referred to the UES, of which 70 were evaluated; two students were referred out due to lack of active crisis. (A table showing students’ demographic characteristics is provided as an online supplement.) Only one student seen during this period required referral to a higher level of care, indicating that school staff were largely able to appropriately triage and refer students to the UES and that urgent outpatient evaluation was, in fact, the appropriate level of care. At the same time, having emergency psychiatry services on site helped to ensure that a higher level of care was available if needed. Referrals to the UES largely mirrored school referrals to the pediatric ED, with the largest proportion referred for evaluation of suicidal ideation (N=26, 36%) and self-injurious behavior (N=18, 25%), followed by disruptive behavior (N=11, 15%). Of note, this last grouping refers to students with disruptive behaviors in class but who were not acutely agitated at the time of assessment and could cooperate with the evaluation process. Fifty-eight percent of referred students were female (N=42), and 42% were male (N=30), with the largest proportion of referrals coming from middle school (N=33, 46%), followed by high school (N=23, 32%) and elementary school (N=16, 22%). The high incidence of middle school referrals may speak to the challenges associated with identifying and effectively managing emerging behavioral health issues, particularly depression and anxiety, in this population. Although a majority (N=57, 79%) of students seen in the UES had no prior history of mental health treatment, one striking finding was that 21% (N=15) of those referred already had established outpatient services. It is unclear whether they had first attempted to reach their own provider and were unable or whether they came directly to the UES. Either way, this finding speaks to the potential to improve crisis management and cross-system collaboration between outpatient mental health providers and schools.

Effect on pediatric ED referrals.

A total of 423 pediatric psychiatry visits presented to the pediatric ED in January–June 2013, compared with 313 visits in January–June 2014 during the UES pilot. School referrals represented 44% (N=185) of the total volume of psychiatry visits in 2013, compared with 34% (N=107) in 2014, an absolute decrease of 78 visits. This decrease corresponds approximately to the 71 visits seen in the UES and suggests successful diversion of these visits from the pediatric ED, as was factored into the initial design and staffing of the model. A similar decrease was not observed among other subgroups of ED referral sources, such as home or the pediatrician’s office. The change in the volume of school referrals to the ED was most pronounced during UES hours of operation, dropping from 80 to 45 pre- and postpilot, a decrease of 44%. Simultaneously, the admission rate of school referrals to the ED during the day shift increased from 4% (N=3) in 2013 to 11% (N=5) in 2014, suggesting an increase in the proportion of clinically appropriate ED referrals, with subacute referrals being diverted to the UES. A similar shift was not observed during evenings and weekends.

School and patient engagement.

Although not formally measured, feedback from schools during this pilot was uniformly positive. Unlike the ED, the UES clearly identified a consistent point of contact for schools making a referral. This is of particular value to schools that are highly motivated to obtain assistance in managing crises but typically face a paucity of collaborators from the mental health system. Often, simply reaching out to the referring school to discuss a student had a significant “warming” effect on improving cross-system relationships and driving additional referrals. Schools not included in the initial information session were informed of the UES when they sent a student to the ED for consultation. Several schools also began to make requests for additional staff training, highlighting the opportunity to build on the foundation of accessible crisis services to develop broader “upstream” interventions.

With respect to patient engagement, it has been noted that the ED can be the first point of mental health treatment for many children (9). Unfortunately, given the limitations previously cited, it can, at times, magnify stigma and decrease the likelihood of follow-up. Relocating these evaluations to a specialized ambulatory service sets the stage for a different experience, providing a therapeutic environment that is more conducive to engagement. Assisting families with securing follow-up and bridging care until they are connected similarly adds value that is often missing from an ED visit.

Limitations.

The short duration of the pilot study and the lack of widespread advertising of the UES meant that many schools still sent subacute referrals to the ED, particularly those schools that were not part of the initial outreach session. Conversely, some youths who were appropriately referred to the UES mistakenly presented to the ED, which, because of Emergency Medical Treatment and Labor Act regulations, was unable to triage patients to the UES and still had to conduct an ED evaluation and call for psychiatric consultation. The service impact was also greatly limited by the hours of operation, which were restricted to school hours.

Next Steps

The UES model appears to provide a promising alternative to the ED for school-age children experiencing behavioral health crises. Our experience indicates that the model can provide equivalent or better-quality care at a presumably lower ambulatory price point with improved patient and family experience, all of which are aligned with the goals of the Institute for Healthcare Improvement’s Triple Aim Initiative of improving population health, improving patient experience, and reducing cost. While detailed cost information is not available for this pilot, the national average for a moderate-complexity ED evaluation (99283) is $740 (10) compared with $134 for an outpatient psychiatric evaluation (90792) under Medicaid, an estimated savings of 82% per diverted visit. Since the initial unfunded pilot, UES patient volume has increased from 137 students in the 2014–2015 academic year to 292 students in 2017–2018. This increase was initially sustainable but eventually exceeded the capacity of the pediatric consultation service, with UES evaluations actually exceeding ED consultations at one point in the operation. Also, while the lower cost of a UES visit was inherent to its proposed value, this factor paradoxically added to the challenge of scaling up in a fee-for-service environment while still managing ED coverage. Due to the absence of additional funding, coupled with staff and leadership turnover, the service was suspended in 2019. However, given a sharp uptick in school ED referrals, that decision is being reconsidered. Next steps in developing a UES include expanded afterschool hours and more aggressive advertising and outreach, both of which could help with maximizing service volume and justifying adequate staffing. A unique challenge to staffing a service driven by school volume is the annual drop-off and eventual ramp-up in the summer and fall. Broadening the referral base beyond schools, particularly to pediatrics or other community stakeholders, could buffer these trends.

Although this pilot occurred in a relatively well-resourced setting with a dedicated consultation team, lessons learned can be applied to smaller ambulatory settings to help create effective crisis capacity. Among these lessons are the vital importance of cross-system collaboration and communication with schools and the need to reconfigure existing psychiatric resources to make them more accessible. Because the wait to see a psychiatrist can typically be weeks or even months after intake, leveraging psychiatry hours for students in crisis could maximize their value to the community and larger health systems. Conversely, routine tasks such as noncomplex medication management could be backfilled by an allied health professional (i.e., a psychiatric nurse practitioner) at a potentially lower cost, thereby aiding sustainability. Finally, this urgent evaluation model could be implemented using telepsychiatry in more rural areas.

In summary, we present a new model of rapid assessment for school-age children that has the potential to better utilize scarce pediatric psychiatric resources. Notably, this study was an unfunded pilot that was never fully developed and subsequently discontinued because of leadership changes. Rigorous analysis of a more fully developed UES with attention to utilization; cost-effectiveness; and the experiences and outcomes for patients, families, and school personnel is needed to justify the model.