If you could change one thing to improve the quality of emergency care for deliberate self-harm patients, what would it be? A national survey of nursing leadership

Abstract

Introduction:

Emergency departments increasingly treat patients for deliberate self-harm. This study sought to understand emergency department nursing leadership perspectives on how to improve the quality of emergency care for these patients.

Methods:

Emergency department nursing managers and directors from a national sample of 476 hospitals responded to an open-ended question asking for the one thing they would change to improve the quality of care for self-harm patients who present in their emergency department. We identified and coded key themes for improving the emergency management of these patients, then examined the distribution of these themes and differences by hospital characteristics including urbanicity, patient volume, and teaching status.

Results:

Five themes regarding how to improve care for deliberate self-harm patients were identified: greater access to hospital mental health staff or treatment (26.4%); better access to community-based services and resources (26.4%); more inpatient psychiatric beds readily accessible (20.9%); separate safe spaces in the ED (18.6%); and dedicated staff coverage (7.8%). Endorsement of findings did not differ based on hospital characteristics.

Discussion:

Emergency department nursing leadership strongly endorsed the need for greater access to both hospital- and community-based mental health treatment resources for deliberate self-harm patients. Additional emergency department staff and training, along with greater continuity among systems of care in the community would further improve the quality of emergency care for these patients. Broad policies that address the scarcity of mental health services should also be considered to provide comprehensive care for this high-risk patient population.

INTRODUCTION

Each year approximately three-quarters of a million patients present to US emergency departments (EDs) for the treatment of deliberate self-harm (DSH).¹ Up to a quarter of adults who die of suicide have an ED visit for DSH in the preceding year,² making this the most common final, non-fatal visit to an ED before (suicide).³ EDs treating DSH patients are tasked with treating any physical injury as well as assessing and managing acute risk of future self-harm. Unfortunately, EDs are thought to be inadequately resourced to provide effective behavioral health assessments and management.⁴ An increasing volume of ED visits related to suicidal behavior,^5,6,7 has contributed to patients with psychiatric needs “boarding” (time spent waiting for an inpatient psychiatric bed) longer than other patients in the ED.^8,9 Psychiatric boarding is associated with lower quality of care and overall ED crowding.^9,10 Thus, improving the care of DSH patients has implications not just for these patients, but also for the entire ED. Equally important to time spent in the ED is what happens to patients after discharge from the ED, particularly as most DSH patients are discharged to the community and often do not receive follow-up outpatient mental health care.¹¹ Providing linkages with appropriate inpatient and outpatient mental health care thus also requires continuity of care among EDs, hospitals, and community health care professionals.

Although research has identified some of the shortcomings in the ED care of DSH patients (e.g., lengthy boarding times^9,10 and limited access to mental health specialists,¹² few studies have sought to understand ED nursing leadership perspectives on assessing and managing DSH patients. Among the research focused on how ED providers perceive DSH patients, one study at a large ED composed of semi-structured interviews with 15 providers identified the belief that the ED should focus only on the presenting physical injury and not on the mental health aspect of care required by DSH patients.¹³ This study, along with another survey of 43 emergency nurses from an Australian hospital, found that many ED providers believe that they do not have adequate training or education on how to care for DSH patients and that staff members who have had more specific training on how to work with these patients felt not only better prepared but less judgmental toward them.¹⁴ Other qualitative inquiries with providers have identified a number of obstacles treating DSH patients, including a lack of privacy afforded by the ED setting and insufficient mental health resources in the hospital and the community,^15,16 leaving providers with a sense that the system fails these patients. Although these studies have begun to shed some light on the experiences of ED providers treating DSH patients, all were conducted at one or two hospitals or systems of care with smaller samples of frontline staff, limiting the generalizability of these findings. The purpose of this study was to bridge this gap by providing a representative perspective of nursing leadership on how to improve the quality of emergency care for patients who present with DSH.

METHODS

Study Design

Between May 2017 and January 2018, a management of self-harm survey assessing the availability of key mental health services for ED patients was sent to a random sample of ED nursing leaders. Defining self-harm as an act of nonfatal self-poisoning or self-injury with or without suicidal intent (ICD-9: E950-E959),¹⁷ the survey modified components of the Safety Planning Intervention (SPI), designated as a best practice by the Suicide Prevention Resource Center and American Foundation for Suicide Prevention,^18,19,20 to examine ED safety planning, assessment and discharge practices with DSH patients. Additional questions regarding ED staffing and linkages with mental health care were included, along with an open-ended question designed to elicit the preferred way in which these leaders would improve care for this population: “Given the existing resources, if you could change one thing to improve the quality of emergency care for your deliberate self-harm patients, what would it be?”

Sample

Using national Medicaid claims data received from the Centers for Medicare and Medicaid Services,²¹ we identified hospitals with ≥5 self-harm visits in 2012. From this sampling frame, we selected a nationally representative sample of 665 hospitals that received the survey via mail and/or email. The study had a 77.1% response rate (n=513) and, of these, 93% of respondents completed the open-ended question, yielding a total sample of 476 hospitals. The study team initiated a coordinated recruitment strategy of key leadership staff (nurse directors or managers) at each ED. Participants were sent hard copies of the survey unless they explicitly asked to receive the survey via email – responses were submitted via mail or a web-based version of the survey. Respondents received $100 gift cards upon completion of the survey unless they could not accept incentives because of hospital policy. The study was approved by the University of Pennsylvania Institutional Review Board. This research did not meet the National Institutes of Health definition of a clinical trial and thus, the trial protocol was not registered.

Data analysis procedures

An inductive thematic analysis approach²² identified themes in the open-ended quality of care item. If more than one response was included, only the first answer was included so that there was one response per hospital. All responses were read to gain familiarity with the data before starting initial coding to identify notable concepts. Interview data were coded in stages by two team members, one with knowledge in mental health patient safety and the other with extensive mental health services research expertise. The data were coded separately and then discussed together. Codes were compiled and organized into five unique themes. Inter-coder reliability¹⁷ was assessed during the coding process and discrepancies were resolved with team discussions until consensus was achieved. It was noted that the five themes fell loosely into two broader categories of hospital and community characteristics.

The frequency with which each theme was mentioned was examined alone and then across strata of hospital characteristics, including admission volume, urban/rural, teaching status, ownership, and presence of a psychiatric ED. Using a survey question estimating total annual census, patient volume was split by percentiles into low (<23,000), medium (23,000–64,000) and high (>64,000). Based on data from the national American Hospital Association annual survey of hospitals from 2016 (or the most recent year available if data were missing for 2016),²³ hospitals were coded as: urban or rural; teaching or non-teaching; and private not-for-profit (NFP), private for-profit (FP), or public/government. Finally, presence of a psychiatric ED was determined from the survey question ‘Does your hospital have a separate psychiatric ED that is physically distinct from the medical ED?’ To ensure that respondents were representative of all hospitals in the sampling frame, analyses included survey weights in which hospitals were selected with probability proportional to their self-harm patient volume. Adjusted Wald F-tests, which accommodated this survey design, examined differences in responses by these hospital characteristics at a p<.01 significance level. All statistical analyses were conducted using SAS 9.4, Cary, NC. Also, before initiating the study, we conducted a power analysis, which showed that, with our selected sample, we have 80% power with two-tailed test (alpha=0.05) to detect effect sizes of Cohen’s d=0.31.

RESULTS

Although the survey was typically sent to ED nursing directors,in some cases, they were completed by other hospital staff members. More than three-quarters (78.4%) of surveys were completed with input from nursing leadership. Respondent breakdown was as follows: 59.8% were completed by only ED nursing directors or managers (n=285), 21.2% were completed by more than one staff member in the emergency department (e.g., nursing director and social worker; n=96; 88 of the 96 included nursing leadership); 1.7% were ED medical directors (n=8); 3.1% were social workers (n=15); 8.6% were ‘other’ (e.g., RN, behavioral health director, clinical supervisor; n=41); and 6.5% did not indicate who had completed the survey (n=31). The majority of hospitals were urban (72.9%, n=386) and non-teaching (57.9%, n=215). Hospital ownership was 74.5% private FP (n=348), 14.9% private NFP (n=68), and 10.6% public (n=56). Only 11.2% had a separate psychiatric ED (n=81; Table 1). The results of the thematic analyses revealed five broad categories which, in decreasing order of endorsement, were as follows: greater access to mental health care and staff for the patient while in the ED; more or faster access to community-based resources including outpatient services; greater number of or access to inpatient psychiatric beds; designated or separate space for psychiatric evaluations and treatment; and dedicated staffing for DSH patients in the ED. These categories are described below and illustrated with quotes selected as representative of a common/typical answer or for their unique perspective on a theme. Across the strata of hospital characteristics, there were no statistically significant differences in the identification of: a) hospital factors or community factors overall or b) in the selection of each of the five specific themes (Tables 2 and 3).

Table 1.

Hospital characteristics (n=476)

	N	Weighted %
Volume based on admissions per year
Low Volume (<23,000)	81	26.5
Medium volume (23,000–64,000)	226	48.3
High volume (>64,000)	169	25.3
Urban/rural status
Urban	386	75.7
Rural	86	24.3
Teaching status
Teaching hospital	257	42.1
Non-teaching hospital	215	57.9
Ownership
Private not-for-profit	348	74.5
Private for-profit	68	14.9
Public (government)	56	10.6
Psychiatric Emergency Department (ED)
Has a separate psychiatric ED	81	11.2
No separate psychiatric ED	395	88.8

Table 2.

Distribution of hospital and community factors identified by emergency department (ED) providers, stratified by hospital characteristics (total n=476)

Hospital characteristics	Hospital factors overall	Community factors overall	Adjusted Wald F, (df)*	p-value
Overall % (weighted N)	52.8 (1106)	47.7 (989.5)
Volume			0.51, (2,474)	0.603
Low (%)	49.6	50.4
Medium (%)	52.1	47.9
High (%)	57.3	42.7
Urbanicity			0.01, (1,471)	0.972
Urban (%)	52.6	47.4
Rural (%)	52.9	47.1
Teaching			0.04, (1,471)	0.839
Teaching (%)	52.0	48.0
Non-teaching (%)	53.2	46.8
Ownership			1.26, (2,470)	0.284
Private not-for-profit (%)	50.0	50.0
Private for-profit (%)	61.7	38.3
Government/Public (%)	58.9	41.1
Psychiatric ED			3.99, (1,458)	0.046
Has psychiatric ED (%)	67.8	32.2
No psychiatric ED (%)	50.9	49.0

^*Adjusted Wald F-tests were used to accommodate the weighted, complex survey design

Table 3.

Primary factor identified by emergency department (ED) providers to improve treatment of deliberate self-harm patients, stratified by hospital characteristics (total n=476)

Hospital characteristics	Hospital factors			Community factors
	Greater access to and collaboration with psychiatry, mental health staff or treatment	Distinct, separate or safe space in the ED	Dedicated staff coverage in ED	More, improved, or faster access to community-based resources (e.g., outpatient services)	More or faster access to inpatient psych beds	Adjusted Wald F, (df)	p-value
Overall % (weighted N)	26.4 (552.5)	18.6 (389.5)	7.8 (164.2)	26.4 (989.5)	20.9 (437.0)
Volume						0.61, (8,468)	0.768
Low (%)	25.8	14.9	8.9	25.8	24.6
Medium (%)	24.0	20.8	8.7	27.5	20.3
High (%)	31.5	19.4	5.0	24.7	18.0
Urbanicity						0.02, (4,468)	0.885
Urban (%)	26.2	18.7	7.7	26.8	20.6
Rural (%)	26.5	18.6	7.8	25.2	21.9
Teaching						0.29, (4,468)	0.885
Teaching (%)	25.5	20.0	6.5	28.0	20.1
Non-teaching (%)	26.8	17.7	8.6	25.3	21.5
Ownership						1.85, (8,464)	0.067
Private not-for-profit (%)	28.7	15.3	6.1	27.8	22.2
Private for-profit (%)	16.0	28.3	17.3	17.1	21.2
Government/Public (%)	23.4	29.5	6.1	30.3	10.8
Psychiatric ED						2.28, (4,455)	0.059
Has psychiatric ED (%)	37.4	25.7	4.7	22.1	10.0
No psychiatric ED (%)	25.5	17.6	7.9	27.0	22.0

^*Adjusted Wald F-tests were used to accommodate the weighted, complex survey design

Theme 1: Access to mental health care and staff within the hospital

Many participants wished they had mental health staff in the ED to work with DSH patients. For instance, one respondent at a small, rural non-teaching hospital wanted “more access to mental health professionals. Although we provide the best care we can, we do not have a psychiatrist, psychologist, or psych nurse…employed at our facility.” Others hoped for greater attention from or collaboration with psychiatry within their hospital. One ED manager from a large urban non-teaching hospital wanted to “have psych round more in the ED. It is such a disservice to these patients to sit in the ED for days at a time…” whereas a respondent at a medium-sized urban teaching hospital wished for more collaboration with psychiatry stating, “We medically clear the psychiatric patients and at that time - psych takes over. We are often times left out of the picture in regards to the care and placement.”

Of the respondents seeking greater access to mental health care, approximately 15% thought this should come in the form of additional training, policies, or resources to better prepare ED staff to be able to provide mental health care for DSH patients. One respondent from a medium-sized urban non-teaching hospital would prefer training to “educate nurses on identifying patients at risk and establishing a therapeutic rapport” whereas another respondent from a large urban hospital with a psychiatric ED would rather have systemic policy changes, such as “standardized discharge information. The questions [in the survey] led me to recognize the information that we send patients home with…perhaps should be entered into our formal form to be certain the same information is being covered for all”. Another participant from a medium-sized, urban hospital also expressed a desire for formal processes, such as “clear accreditation guidelines on what is required for interval assessment and documentation on self-harm patients,…guidance and training…if an ED provider is to ever be ‘releasing’ a involuntary psych hold patient before the expiration of that hold. Also if an ED stay awaiting an inpatient bed extends over the 24 hour time frame, is there required expectations of the ED provider”.

Theme 2: Access to mental health care in the community

A respondent from a large urban hospital with a psychiatric ED noted the importance of fast, appropriate follow-up care to avoid a patient’s cyclical return to the ED, “Immediate referral to an OP facility with followup by a case worker/case manager… I feel we do excellent emergent care and work diligently to get patients referred to an appropriate location, but then they are no longer followed by anyone…,until they return to the ED and the cycle begins again.” Others noted a broad range of services needed in the community, including “more availability to inpatient detox, safe shelters for homeless psych patients, access to care for the non-insured”; access to “pre-hospital resources to prevent ED visits - often times the ED setting & delays result in (patient) escalating”; and “a bridge clinic that would allow us to make follow-up appointments prior to discharge from the ED”. Several respondents from large, urban teaching hospitals preferred community-based services to divert DSH patients from coming to the ED, such as an “outpatient crisis center so they aren’t in the ED” or “a true intake center in our City”.

Theme 3: Additional inpatient mental health services

Respondents wanted greater inpatient psychiatric care to reduce ED boarding and patient disengagement. A participant from a medium-sized urban hospital wanted “improved access to adult inpatient facilities. Normally we hold these patients for 3–5 days seeking placement”. Another participant from a small urban non-teaching hospital noted the consequences of lengthy waits, noting that patients “Often…get frustrated and leave while waiting on a facility to accept them for transfer.” Interestingly, almost one-quarter of respondents across all themes made mention of long wait times or expressed a desire for faster disposition times.

Theme 4: Distinct space in the ED for DSH patients

A respondent at a medium-sized, urban non-teaching hospital noted that the ED would benefit from “a designated area that could provide for safety and offer some kind of therapy. Our treatment of the patient borders on inhumane while they are housed for multiple days awaiting placement at a facility that can provide services.” In addition to providing therapeutic care for DSH patients, it was also noted that a safe space would provide “a more controlled environment to prevent runners from escaping the ED. Currently if we can’t chase down and detain the patient they are out…patients…typically wait 3–55 days in our ED.” Wanting a separate psychiatric ED was specifically cited by 7% of respondents. Finally, one respondent from a medium-sized urban non-teaching hospital wished that their ED had “dedicated rooms for mental health patients, but more specifically children. We currently put all mental health patients into 2 rooms and then utilize hallway spaces. We comingle those patients which is less than ideal.”

Theme 5: Dedicated staffing in the ED

Dedicated staffing suggestions ranged from specialized clinicians to address the clinical needs of DSH patients (psychiatric nurses, psychiatrists, and social workers or other behavioral health workers) to nonclinical staff assigned to procure the disposition of DSH patients and the safety of the emergency department (eg, case managers, sitters, and security guards). A respondent from a large urban teaching hospital suggested a “24/7 psychiatrist hospitalist coverage and/or psychiatrist nurse practitioner dedicated for emergency department only” whereas a respondent from a small, urban non-teaching hospital preferred a “MSW available 24/7 to help with screening these patients; creating appropriate safety plans if patients are being discharged; and helping with placement and/or outpatient follow-up”. Respondents from two small, urban non-teaching hospitals preferred staff focused on ensuring safety in the ED, such as a “full time security guard dedicated to ER” or “sitters that are specifically trained to watch behavioral health patients in the ED to ensure there is no self-harm done in the ED”.

DISCUSSION

Despite representing a broad range of hospital types, most respondents prioritized the same areas for improving the quality of care for patients with DSH. ED leadership stressed a desire for more, improved, or easier access to mental health resources both in the hospital and in the community. In contrast, few mentioned a need for improving some basic aspects of clinical management including medical clearance, accuracy of suicide risk assessment, or communication with family members.

A number of models have been suggested to improve access or connections to mental health care for DSH patients presenting in the ED, including creating nurse- or social worker-staffed crisis clinics or dedicated psychiatric emergency service (PES) units.²⁴ These types of units which typically provide outpatient evaluation, treatment and observation to stabilize acute symptoms and minimize psychiatric hospitalization, significantly reduce boarding time²⁵ and improve assessments and access to care.²⁶ Although having a distinct PES may be ideal for any given hospital, many do not have the patient volume or resources to support their own PES, particularly smaller, rural or low income area hospitals.

Other efforts have included shifting the responsibility of mental health care from the hospital to the community at-large, with improvements noted in ED and patient outcomes. One example is the Burke Center, which takes a comprehensive approach to serving residents in 12 rural Texas counties. The Burke Center pairs telepsychiatry with onsite nursing and mental health staff, as well as outpatient services and home visits upon discharge.²⁷ In California, the “Almeda model” expeditiously transfers psychiatric patients from local EDs to a regional psychiatric emergency hospital that provides 24-hour rapid stabilization services. By providing comprehensive psychiatric care, this model has resulted in more than an 80% reduced boarding time for patients awaiting psychiatric care, as well as having helped to alleviate the demand for inpatient psychiatric beds by stabilizing the majority of patients.²⁹ Although these programs create a continuous and unified care model, shifting the responsibility from the hospital-to the community-level requires policy shifts promoting regional cooperation among facilities and changes to billing practices.

IMPLICATIONS FOR EMERGENCY NURSING PRACTICE

Until larger policy shifts address the scarcity of acute inpatient and subacute outpatient mental health services, hospitals could hire dedicated staff in the ED¹² or provide existing staff with relevant training and tools to care for this population more effectively. One example of evidence-based training is ED-SAFE, a brief intervention consisting of a secondary suicide risk screening by an ED physician, safety planning, and post-discharge telephone follow-up by a nurse. An evaluation of ED-SAFE noted 30% fewer suicide attempts in the year after implementation.²⁸ Another alternative, the Safety Planning Intervention (SPI) provides written coping strategies and sources of support during a suicidal crisis, as well as structured follow-up telephone contact. SPI has been found to reduce suicidal behavior and increase treatment engagement.²⁹

Limitations

First, the cross-sectional design does not capture trends over time. Second, non-response or incomplete responses at the survey or item level may bias results. Third, the study sample was derived from EDs with ≥5 self-harm Medicaid-financed patient visits and may limit generalizability. Fourth, we did not use a validated survey instrument. Fifth, responses were predominantly from nursing leadership whose perspective may differ from frontline nurses. Finally, extracting only the first response to the open-ended question could have excluded ideas that may have been equally important to the respondent.

CONCLUSION

In summary, across a wide range of hospital types there is an overarching perception of insufficient access to specialized mental health care inside and outside the hospital, as well as a lack of resources for ED providers. Although innovative and promising models of providing psychiatric emergency services exist, given the prevailing resource constraints, training existing staff in the evaluation and management of DSH may be the most feasible short-term approach to address these pressing service needs. Relying on evidence-based and publicly available resources that provide structured guidelines for decision-making, assessment and treatment of DSH patients,^30,31 is a first-step that all EDs can take with minimal cost and effort. At the same time, advocacy for broader policies that address the scarcity of integrated inpatient and outpatient mental health services at the community and state level should also be pursued, with the goal of providing unified and comprehensive mental health care for this high-risk patient population while also reducing strain on local emergency departments.

Contribution to Emergency Nursing Practice.

• The current literature on the ED management of patients with deliberate self-harm indicates that this is a growing patient population, which is putting increasing strain on emergency department staff, space, and resources.

• This article contributes research findings from nursing leadership—across a broad range of hospital types— who have prioritized similar ways to improve the care for patients who deliberately self-harm and present to emergency departments.

• Key implications for emergency nursing practice found in this article are that additional training, resources, and policies could enable emergency nursing staff to assess, treat, and advocate more effectively for deliberate self-harm patients, resulting in more comprehensive care for these patients.