Super-High-Utilizer Patients in an Urban Academic Emergency Department: Characteristics, Early Identification, and Impact of Strategic Care Management Interventions Free

Patients who visit the emergency department (ED) with high frequency are typically categorized as super users, high utilizers, or super-high utilizers (SHUs) based on the relative frequency of hospital care they require, compared with other ED patients (Chambers et al., 2013; Rinehart et al., 2018). Common characteristics of patients with frequent utilization of emergency services are well described in the literature, including high chronic disease burden, substance abuse, and psychiatric disorders (LaCalle & Rabin, 2010; Sandoval et al., 2008; Szekendi et al., 2015). There has been relatively less focus on the characteristics of high utilizers who are frequently admitted to inpatient services following their ED encounters (Boonyasai et al., 2012; Raven et al., 2011). Healthcare spending attributed to the highest utilizers of the ED, categorized as SHUs, is disproportionately high (Bergenstal et al., 2020; Jiang et al., 2006; Moschetti et al., 2018). In healthcare systems constrained by limited inpatient hospital capacity, SHUs compound important patient safety issues, including ED overcrowding. Further understanding SHU characteristics and exploring patient-centered solutions to this population is critical to the contemporary ED care model.

Multidisciplinary interventions designed for SHUs have been shown to reduce ED visits, hospital admissions, and healthcare resource expenditure, along with improvement in patient satisfaction and quality of life (Althus et al., 2011; Bodenheimer, 2013; Soril et al., 2015). One recent pilot program standardized social service interventions in high utilizers, emphasizing the potential value of a proactive ED social worker role (Adamson et al., 2021). Several other interventions describe an early recognition model to direct limited resources to the most vulnerable patients. Early notification systems have been used to facilitate standardized care delivery from the moment of arrival—a step highlighted as crucial for staff adherence to downstream individualized care plans (Kimmel et al., 2019).

In 2019, a novel SHU Program (SHUP) was piloted at an urban academic quaternary care medical center. This program involved an electronic triage notification system to direct early multidisciplinary ED interventions, initiated from the moment of arrival. Individualized patient-centered care plans were developed by a team of stakeholders for 15 SHU patients whose very high ED utilization resulted in a total of 1,095 total hospital presentations throughout the 2.5-year study period. In this Practice Forum column we describe the enrollment process, development of individualized care plans, electronic triage early notification system, and SHUP’s subsequent impact on ED and inpatient hospital utilization.

INTERVENTION MODEL (METHOD)

NewYork-Presbyterian/Weill Cornell Medicine (NYP/WCM) is an 862-bed quaternary care academic healthcare center located in Manhattan, with over 100,000 ED visits per year. SHUP was developed and led by a team of NYP social workers and case managers in collaboration with WCM emergency medicine physicians and colleagues from case-relevant subspecialty services, such as internal medicine, hematology/oncology, and psychiatry.

An initial analysis of 2018 hospital data demonstrated that a subgroup of 19 patients (each with at least 10 admissions per year) accounted for a total of 1,300 bed days, 1 percent of all hospital bed days. We defined a SHU as any patient with 10 or more ED visits during a calendar year period, with or without inpatient admission. For the pilot project, SHU patients were selected from both that subgroup and frontline ED staff referrals. A study cohort of 15 patients was selected following a screening process by the ED social work and nurse care coordinator team, to determine cases in which targeted, actionable interventions could be designed. For each of these selected patients, the ED social work and nurse care coordinator team performed a detailed case assessment to identify modifiable factors contributing to their SHU status, such as absence of longitudinal care and behavioral issues. Pediatric patients were excluded from the study, as were patients with primary psychiatric or substance abuse diagnoses.

Subsequent care plans were generated by the ED nurse care coordinators and social workers with information gathered from conducting in-person psychosocial assessments and comprehensive chart reviews. These plans were individually tailored to meet each SHU patient’s unique biopsychosocial needs and limitations, including behavioral management. Plan development required collaboration with primary care and relevant specialty service teams providing treatment to these patients and identification of the impact of social determinants of health. This multidisciplinary approach to care plans was designed to encourage appropriate utilization of the ED while managing nonemergent care needs on an outpatient basis. Care plans were presented at a monthly conference incorporating group feedback with revisions prior to final approval by a board-certified emergency physician collaborator. Care plans were then shared with interdisciplinary frontline clinical staff and documented in the electronic medical record (EMR) to ensure 24/7 accessibility and consistent implementation. Periodic plan reassessments were necessary to identify changes to patient needs, which made the plans dynamic and consistently relevant.

The study team collaborated with NYP’s Department of Informatics to design and implement an EMR-based triage notification system. When an enrolled patient registered in the ED the notification system triggered a real-time email, sent to the relevant stakeholders including ED physician and nursing leadership, social work, case management, and relevant specialty service providers. This alert included the patient’s demographics, time since last discharge, and number of total ED visits and admissions in the past 365 days. This notification was instrumental in alerting relevant staff to a SHU ED presentation on arrival, which allowed social services to physically meet the patient and discuss the case with the ED care team early in the evaluation. Clinical teams were expected to refer to the care plans in the EMR to inform their medical decision making. Faculty were in-serviced to the SHU program at regular monthly meetings and were already accustomed to an existing strong collaborative working relationship with the social services team. All team members were advised that the care plans were recommendations and did not supersede the medical judgment of treating physicians should a SHU patient require an alternative plan of care.

A retrospective review was performed of all adult patients over the age of 21 who presented to the NYP/WCM ED from January 1, 2017, to June 30, 2020. A subset of these patients with multiple hospital presentations were selected for closer examination, and a final subset of 15 SHU patients were selected for intervention in the SHUP program. One SHU patient was included for intervention despite very low ED visits as captured by registration, as this patient made regular visits to the ED for secondary gain without medical complaint, which went uncaptured, but the concerning trajectory led the frontline staff to refer this patient. Data were retrieved from the electronic health record system after receiving approval from the Weill Cornell institutional review board. For comparing pre- and postcare management intervention, we computed the ED visits, hospital admissions, and inpatient bed days during this time period for 15 SHU patients. Sociodemographic and health characteristics examined included age, gender, number of comorbidities, housing status, psychiatric disease, and substance abuse history.

RESULTS

The SHU cohort had an average age of 54 years, with 10 men (67 percent) and five women (33 percent). Among this population, six participants (40 percent) had 10 or more comorbidities; 14 participants (93 percent) had five or more comorbidities. Six people (40 percent) were undomiciled. Eleven (73 percent) had known psychiatric disease and eight (53 percent) had documented substance abuse issues. Twelve (80 percent) were independent in their activities of daily living, two (13 percent) were partially dependent, and one person (7 percent) was fully dependent.

Prior to our intervention, this SHU cohort visited the ED a total of 835 times, with a yearly total ED encounter average of 56 visits per annum (median = 36) and a hospital admission rate of 27 percent (227 out of 835), totaling 992 occupied inpatient bed days. Our intervention resulted in a 69 percent (n = 575) reduction in total yearly ED visits, a 74 percent (n = 167) reduction in total yearly hospital admissions, and a 74 percent (n = 730) reduction in total inpatient occupied bed days (see Table 1).

DISCUSSION

This study describes the implementation of a novel budget-neutral SHU program, utilizing an ED-based multidisciplinary care team. Electronic triage early notification system and individualized dynamic care plan interventions targeted 15 patients. Our results for this cohort demonstrate approximately 70 percent reduction in ED visits, hospitalizations, and hospital bed days, highlighting the substantial impact of this pilot intervention over the 2.5-year study period.

The most significant limitation to this work is that EMR recidivism data are restricted to our own institution and as such we are unable to reliably account for encounters outside of our health system, estimated to be as high as 17 percent (Chartier et al., 2021). Further research should leverage any available shared health system data and more comprehensive state CMS data. Another limitation is the imperfect mechanism of email alerts to notify providers and staff of a patient care plan, particularly with off-hours presentations. Greater integration of triage alerts into the EMR with best practice advisories and links to care plans are an anticipated enhancement. In addition, noncompliance with discharge plans and accompanying psychosocial barriers often affect the ability of SHU patients to remain adherent; with this in mind, our team’s focus for the individualized care plans included available resources to be offered on future ED presentations.

This implementation of our SHU care plan initiative promoted interdisciplinary team collaboration, ensuring reduced variability and high-quality patient care during all ED encounters. Next steps include retiring, modifying, and generating new care plans at regular intervals, in addition to a cost savings analysis. We also aim to analyze and define common characteristics among SHUs to develop a predictive model of early identification in which to achieve earlier care interventions, reduce avoidable high resource utilization, and improve health outcomes.

CONCLUSION

Healthcare systems may benefit from focused multidisciplinary interventions for super-high-utilizer patients identified in the ED, including individualized care plans and early notification systems. Such targeted approaches to meeting this cohort’s medical and psychosocial needs may reduce hospital overcrowding and avoidable hospitalizations while improving healthcare for this vulnerable population.

Amos J. Shemesh, MD, is assistant director of clinical services, Department of Emergency Medicine (ED), NewYork-Presbyterian/Weill Cornell Medicine, 525 East 68th Street, New York, NY 10065, USA; email: [email protected].

Daniel L. Golden, MD, is ED resident physician; Amy Y. Kim, PharmD, is pharmacist, Department of Pharmacy;

Yvette Rolon, LCSW-R, ACSW, is program administrator; Lisa Kelly, RN, MSN, CEN, is ED care manager; Seth Herman, LCSW, and Tamara N. Weathers, LCSW, are ED social workers; Daneisha Wright, BAS, is ED administrative assistant; Tim McGarvey, MSN, LMSW, RN, NEA-BC, is director of care coordination and social work; Yiye Zhang, PhD, is assistant professor, Department of Population Health Sciences and Emergency Medicine; and

Peter A. D. Steel, MBBS, is vice chair of clinical services, ED, NewYork-Presbyterian/Weill Cornell Medicine, New York, NY, USA.

References