Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 6.
Published in final edited form as: Suicide Life Threat Behav. 2023 Jul 11;53(4):702–712. doi: 10.1111/sltb.12975

Demographic predictors of emergency service utilization patterns in youth at risk of suicide

Mira A Bajaj 1, Holly C Wilcox 1,2, Leslie B Adams 2, Alan L Berman 1, Mary Cwik 3, Christopher Kitchen 4, Leslie Miller 1, Paul S Nestadt 1, Eric P Slade 5,6, Emily E Haroz 3
PMCID: PMC10916713  NIHMSID: NIHMS1971020  PMID: 37431982

Abstract

Objective:

To explore demographic predictors of Emergency Department (ED) utilization among youth with a history of suicidality (i.e., ideation or behaviors).

Methods:

Electronic health records were extracted from 2017 to 2021 for 3094 8–22 year-old patients with a history of suicidality at an urban academic medical center ED in the Mid-Atlantic. Logistic regression analyses were used to assess for demographic predictors of ED utilization frequency, timing of subsequent visits, and reasons for subsequent visits over a 24-month follow-up period.

Results:

Black race (OR = 1.45, 95% CI = 1.11–1.92), Female sex (OR = 1.59, 95% CI = 1.26–2.03), and having Medicaid insurance (OR = 1.71, 95% CI = 1.37–2.14) were associated with increased utilization, while being under 18 was associated with lower utilization (<12: OR = 0.38, 95% CI = 0.26–0.56; 12–18: OR = 0.47, 95% CI = 0.35–0.63). These demographics were also associated with ED readmission within 90 days, while being under 18 was associated with a lower odds of readmission.

Conclusions:

Among patients with a history of suicidality, those who identify as Black, young adults, patients with Medicaid, and female patients were more likely to be frequent utilizers of the ED within the 2 years following their initial visit. This pattern may suggest inadequate health care access for these groups, and a need to develop better care coordination with an intersectional focus to facilitate utilization of other health services.

Keywords: emergency services, service utilization, youth

INTRODUCTION

Suicide is a major public health issue, particularly for youth. In the United States, suicide was the second leading cause of death in youth ages 10–14 years and the third in youth ages 15–24 years in 2020 (Centers for Disease Control and Prevention, 2020). Youth suicide rates have increased over time since 2007 (Curtin, 2020; Curtin et al., 2022). This increasing trend is especially pronounced for racial and ethnic minority youth, including a rise in suicide rates for Black youth (Bridge et al., 2018; Ramchand et al., 2021; Sheftall et al., 2022).

The suicide death rate among Black youth is increasing faster than in any other racial/ethnic group (Boyd, 2020). The 2021 National Youth Risk Behavior Survey of high school students found that Black female students had an increased prevalence of suicide attempts compared with their White counterparts (Gaylor et al., 2023). The COVID-19 pandemic intersects with rising rates of youth depression, anxiety, and suicidal thoughts and behaviors along with long-standing social determinants of health, including discrimination, stigma, and racism. Worsening mental health trends in association with the COVID-19 pandemic have been observed in youth of color (Hawrilenko et al., 2021). A recent analysis comparing suicide rates from 2020 to 2019 found that while suicide rates decreased for non-Hispanic White and Asian populations, increased rates of suicide were observed for Hispanic, non-Hispanic Black, and non-Hispanic American Indian and Alaska Native populations, particularly among males (Curtin & Hedegaard, 2021).

The Emergency Department (ED) is a key setting for identifying and treating youth at risk for suicide, as pediatric ED visits for mental health reasons have increased over time (Hoffmann et al., 2019; Lo et al., 2020; Rogers et al., 2017). In 2021, the American Academy of Pediatrics declared a “National State of Emergency in Children’s Mental Health” (American Academy of Pediatrics, 2021) and the surgeon general issued an advisory on youth mental health crises (Office of the Surgeon General, 2021). Given the high frequency of interaction with health facilities and associated potential for identifying youth at risk of suicide, screening for suicide risk in healthcare settings is recommended by the 2012 National Strategy on Suicide Prevention (Office of the Surgeon General (US) & National Action Alliance for Suicide Prevention (US), 2012) and recognized as a Patient Safety Goal by the Joint Commission (The Joint Commission, 2019). Recent work has supported the accuracy of suicide risk screening as well as other strategies such as predictive models that aim to identify and classify risk (DeVylder et al., 2019; Haroz, Kitchen, et al., 2021; King et al., 2009). Positive responses to suicide screening questions have been shown to be associated with subsequent psychiatric hospitalizations and repeat ED visits (Ballard et al., 2013).

Though the ED plays a key role in identifying and treating youth at risk of suicide, high ED utilization rates are costly to the healthcare system. ED utilization rates are also considered a proxy metric of access to and quality of non-emergency services, with high utilization rates being suggestive of challenges with accessing care in outpatient or community settings (Merrick et al., 2010). Youth of color have faced particular disparities in their access to and utilization of care. For example, only approximately one-third of Black youth who have died by suicide have a documented history of mental health treatment (Sheftall et al., 2022). These youth are not accessing mental health care at the same rate as their White peers (Freedenthal, 2007). Structural racism impacts the availability and utility of this care for Black youth (Alvarez et al., 2022). When youth are not accessing outpatient care for suicidality, they may frequent other emergency services instead.

The objective of this study was to understand demographic predictors of ED utilization patterns among youth who screened positive for suicide risk and/or presented with a suicide-related chief complaint in the ED. We evaluated the frequency, timing and reasons for subsequent ED utilization over a 24-month follow-up period. Understanding emergency service utilization patterns for youth at risk of suicide can facilitate identification of potential gaps in service access and more targeted interventions for priority groups who are chronically utilizing the ED (Haroz, Sarapik, et al., 2021).

METHODS

Data source

We conducted a retrospective cohort study of patients seen at a large urban academic center emergency department in the Mid-Atlantic. The medical record review and analysis included in this study were approved by the Johns Hopkins Institutional Review Board. In 2017, our institution implemented a universal screening approach with the Ask Suicide Questionnaire (ASQ) (Horowitz et al., 2012), which is comprised of four questions that assess current suicide risks. A positive screen occurs if the respondent endorses any of the four questions. If the screen is positive the patient is then assessed for acuity with a fifth question about current suicidal ideation.

Using electronic health record data, we extracted encounter-level and patient-level data for patients 8–22 years old seen in the emergency department between January 1, 2017, and December 30, 2021. Data were selected starting from January 2017 to align with the implementation of the ASQ. We selected patients who had at least one encounter with potential suicidal ideation or behavior, defined by either (1) positive ASQ screen, (2) diagnosis of suicidal ideation or attempt associated with the encounter, defined using a set of ICD-10 codes grouped together by the Clinical Classifications Software Refined database (Agency for Healthcare Research and Quality, 2022) or (3) chief complaint of suicidal ideation or attempt (denoted “high-risk index encounter”). It should be noted that young adult patients (>18) who presented for a mental health concern were relocated to an adult psychiatric wing which did not routinely screen with the ASQ, so this subgroup was identified through diagnosis and/or chief complaint only. For patients who met any of these criteria, we extracted data on their subsequent visits to the emergency department within a 24-month period. To ensure each patient had 24 full months of follow-up, index encounter data were restricted to occurring between January 1, 2017 and December 30, 2019.

Baseline data extracted from index encounters included age, sex, race, ethnicity, and insurance type. Patient race and ethnicity data in the EHR are collected at the time of registration via self-report by either youth or their guardian. These variables were selected a-priori as key baseline demographic variables which could influence utilization. Outcome data extracted from the follow-up period included the total number of ED visits in the observation period, timing of ED encounters following the index encounter, and chief complaint for subsequent ED encounters.

Analysis

Descriptive analyses focused on understanding the characteristics of patients with a history of suicidal ideation/behavior with high utilization of emergency services. Based on the distribution of the total number of ED visits per patient, three categories of utilization were created: low utilizers (1–4 ED visits), medium utilizers (5–9 ED visits), and high utilizers (10+ ED visits) during the 24-month follow-up time period.

Univariate analyses used logistic regression to assess the association between patients’ baseline index-visit data (age, sex, race, ethnicity, insurance type) and their subsequent ED utilization frequency (low utilizer, medium utilizer, high utilizer) during the follow-up period. Multivariable logistic regression analyses were conducted with all of the baseline index-visit variables as predictors. Poisson analyses were utilized as a secondary sensitivity analysis, with the outcome variable defined as the number of subsequent ED visits.

We conducted additional exploratory analyses to assess how our main findings held up within different subgroups and to further explore the contribution of significant predictor variables. Because there may be differences in the validity of suicidal screening in younger children compared to adolescents, we repeated the above analyses in the subgroup of patients aged 12 and older. Furthermore, because the ASQ was not universally administered to young adults, we repeated our analysis excluding those over 18. Hierarchical regression analysis was used to confirm the unique contribution of potentially correlated predictor variables (race, insurance status, sex). We also checked for interaction effects.

Secondary analyses focused on understanding the timing of and chief complaint for ED visits following the index visit. To assess the timing of subsequent encounters for each patient, binary variables captured whether patients had at least one subsequent ED visit within 30 and 90 days after their index high-risk visit. To categorize chief complaints of subsequent visits, chief complaints from encounters within the dataset were extracted, along with their frequency of occurrence. For chief complaints with a frequency of 10 or greater, we manually categorized the complaint as either “mental health crisis,” “other mental health concern,” and “non-mental health concern” (see Table S4 for full list of complaints with a frequency of 10 or greater). The following chief complaints were categorized as a “mental health crisis”: emergency petition, suicidal ideation, suicide attempt, or suicidal. The following chief complaints were categorized as “other mental health concern”: psychiatric evaluation, aggressive behavior, mental health problem, mental health visit, altered mental status, depression, anxiety, alcohol intoxication, behavior problem, drug/alcohol assessment, behavior problems, eating disorder, manic behavior, withdrawal, addiction problem, anorexia, drug overdose, low mood, or paranoid. All other chief complaints were categorized as “non-mental health concerns.” At the patient level, binary variables captured whether each patient had any subsequent visits that were mental health crises, other mental health concerns, or non-mental health concerns.

Multivariable logistic regression analyses assessed the association between information from the baseline index-visit and the presence of a subsequent encounter within 30 and 90 days. The association between baseline index-visit data and subsequent encounter chief complaint category (mental health crisis, other mental health, non-mental health) was assessed using multivariable logistic regressions.

Analyses were conducted using R version 4.2.0 software, and statistical significance was evaluated at a level of p < 0.05.

RESULTS

The distribution of the total number of ED visits per patient within the study period is depicted in Figure 1a. Baseline demographic and clinical variables are depicted in Table 1 by utilization group. This study included 3094 patients who had an index high-risk ED visit within the time frame specified. Among these, 2678 (86.55%) were included in the low utilization category, 307 (9.92%) were included in the medium utilization category, and 109 (3.52%) were included in the high utilization category.

FIGURE 1.

FIGURE 1

Open in a new tab

Distribution of Total Number of Emergency Department Visits per patient over a 24-month follow-up period after an initial visit to the emergency department at a high-risk for suicidality, (a) total sample and (b) grouped by race. High-Risk is defined as either (1) positive Ask-Suicide-Questionnaire screen, (2) diagnosis of suicidal ideation or attempt associated with the encounter, or (3) chief complaint of suicidal ideation or attempt.

TABLE 1.

Univariate analyses of Emergency Department (ED) utilization category by baseline demographic variables at the index visit.

Characteristic Low ED utilizers: 1–4 visits (N = 2678) Medium ED utilizers: 5–9 visits (N = 307) High ED utilizers: 10+ visits (N = 109) Medium versus low ED utilizers High versus low ED utilizers
N % N % N % OR 95% CI OR 95% CI
Age (years)
 <12 525 19.6 46 15.0 13 11.9 0.38 0.25–0.58 0.24 0.12–0.47
 12–18 1913 71.4 206 67.1 71 65.1 0.47 0.34–0.65 0.36 0.22–0.57
 ≥18 240 9.0 55 17.9 25 22.9 1.00 ref 1.00 ref
Sex
 Male 1030 38.5 81 26.4 31 28.4 1.00 ref 1.00 ref
 Female 1648 61.5 26 73.62 78 71.6 1.74 1.34–2.27 1.57 1.03–2.40
Race
 White or European American 728 27.2 59 19.2 19 17.4 1.00 ref 1.00 ref
 Black or African American 1594 59.5 223 72.6 80 73.4 1.73 1.28–2.33 1.92 1.16–3.20
 Other/Unknown race 356 13.3 25 8.1 10 9.2 0.87 0.53–1.41 1.08 0.49–2.34
Ethnicity
 Hispanic or Latino 201 7.5 21 6.8 7 6.4 0.89 0.56–1.42 0.83 0.38–1.81
Insurance type
 Commercial 1394 52.1 130 42.4 33 30.3 1.00 ref 1.00 ref
 Medicaid 1130 42.2 168 54.7 73 67.0 1.59 1.25–2.03 2.73 1.80–4.15
 Other/Unknown 154 5.7 9 2.9 3 2.7 0.63 0.31–1.26 0.82 0.25–2.71
Open in a new tab

Note: Bolded values are statistically significant (p < 0.05).

Table 1 depicts the results from univariate logistic regressions with ED utilization category as the outcome and index-visit characteristics as predictor variables. Because the univariate findings were relatively consistent across the medium and high utilization categories, these categories were collapsed to increase power for multivariable analyses.

Table 2 depicts the results of the multivariable regression for utilization category (low vs. medium/high utilization) with predictor variables of age, sex, race, ethnicity, and insurance status. Being under 18 at the index visit was associated with a lower odds of frequent utilization when compared to patients over 18 (<12: OR = 0.38, 95% CI = 0.26–0.56, p < 0.001; 12–18: OR = 0.47, 95% CI = 0.35–0.63, p < 0.001). Female patients showed an increased odds of frequent utilization when compared to male patients (OR = 1.59, 95% CI = 1.26–2.03, p < 0.001). Patients who identified as Black showed an increased odds of frequent utilization relative to White patients (OR = 1.45, 95% CI = 1.11–1.92, p = 0.008). The distribution of the total number of ED visits per patient grouped by race is depicted in Figure 1b. Patients insured by Medicaid showed an increased odds of frequent utilization relative to those with commercial insurance (OR = 1.71, 95% CI = 1.37–2.14, p < 0.001).

TABLE 2.

Multivariable adjusted odds ratios for medium/high versus low Emergency Department utilization.

Odds ratio 95% CI p-Value
Age (years)
 <12 0.38 0.26–0.56 <0.001
 12–18 0.47 0.35–0.63 <0.001
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 1.59 1.26–2.03 <0.001
Race
 White or European American 1.00 (ref) – –
 Black or African American 1.45 1.11–1.92 0.008
 Other/Unknown race 0.73 0.44–1.18 0.204
Hispanic ethnicity 1.51 0.89–2.51 0.121
Insurance type
 Commercial 1.00 (ref) – –
 Medicaid 1.71 1.37–2.14 <0.001
 Other/Unknown 0.58 0.30–1.04 0.085
Open in a new tab

Note: Bolded values are statistically significant (p < 0.05).

Multivariable Poisson regression analyses using total number of ED visits as the outcome variable were consistent with these findings (results in Table S1): being over 18 at the time of the index visit, female sex, Black race, and having Medicaid insurance were all associated with a higher number of ED visits over the 2-year period. In the Poisson regression, Hispanic ethnicity was also associated with a higher number of high-risk ED visits. Exploratory sensitivity analyses confirmed concordant findings when excluding either those under 12 or those over 18. No significant interactions were found between covariates in our analysis. Hierarchical regression analyses revealed that at stage 1, insurance type contributed significantly to the regression model (χ2 (1) = 37.78, p < 0.001) with a McFadden pseudo-R2 of 0.015. In step 2, adding race led to an increase of the pseudo-R2 to 0.023, which was significant (χ2 (1) = 18.16, p < 0.001). In step 3, adding sex led to an increase of the pseudo-R2 to 0.032, which was significant (χ2 (1) = 21.54, p < 0.001). Together, this analysis suggests that each of these predictors uniquely contributes to the model. Overall, these exploratory sensitivity analyses are concordant with the findings of our original analysis.

Secondary analyses focused on the timing of subsequent ED visits following the initial high-risk index visit. Baseline characteristics of the sample which returned to the ED for any reason at each time frame of interest (30, 90 days) are described in Table S2. Table 3 depicts the results of multivariable regression analyses examining the association of predictor variables of age, sex, race, ethnicity, and insurance status with timing of subsequent visits within 30 and 90 days. Being under 18 was associated with a lower odds of return to the ED within 30 days relative to the young adult age group (<12: OR = 0.66, 95% CI = 0.45–0.98, p = 0.037; OR = 0.53, 95% CI = 0.39–0.74, p < 0.001). The following subgroups were associated with a greater odds of returning to the ED within 90 days relative to their respective reference groups: female patients (OR = 1.23, 95% CI = 1.02–1.48, p = 0.029), Black patients (OR = 1.29, 95% CI = 1.05–1.61, p = 0.019) and patients with Medicaid insurance (OR = 1.28, 95% CI = 1.07–1.54, p = 0.007). Being under 18 was associated with a lower odds of returning to the ED within 90 days (<12: OR = 0.68, 95% CI = 0.50–0.92, p = 0.014; 12–18 OR = 0.54, 95% CI = 0.42–0.71, p < 0.001).

TABLE 3.

Multivariable logistic regression for risk of subsequent ED visit within 30 and 90 days following a high-risk index visit.

Odds ratio 95% CI p-Value
Within 30 days
Age (years)
 <12 0.66 0.45–0.98 0.037
 12–18 0.53 0.39–0.74 <0.001
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 1.20 0.94–1.53 0.139
Race
 White or European American 1.00 (ref) – –
 Black or African American 1.11 0.84–1.48 0.457
 Other/Unknown race 1.34 0.86–2.07 0.188
Hispanic ethnicity 0.85 0.49–1.42 0.534
Insurance status
 Commercial 1.00 (ref) – –
 Medicaid 1.29 1.02–1.64 0.031
 Other/Unknown 0.57 0.28–1.04 0.087
Within 90 days
Age (years)
 <12 0.68 0.50–0.92 0.014
 12–18 0.54 0.42–0.71 <0.001
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 1.23 1.02–1.48 0.029
Race
 White or European American 1.00 (ref) – –
 Black or African American 1.29 1.05–1.61 0.019
 Other/Unknown race 1.21 0.85–1.70 0.294
Hispanic ethnicity 1.06 0.71–1.58 0.761
Insurance status
 Commercial 1.00 (ref) – –
 Medicaid 1.28 1.07–1.54 0.007
 Other/Unknown 0.79 0.51–1.19 0.267
Open in a new tab

Note: Bolded values are statistically significant (p < 0.05).

Secondary analyses also examined the chief complaint for ED visits following the index high-risk visit. Baseline characteristics of the sample which had follow-up visits for mental health crises, other mental health concerns, and non-mental health concerns are depicted in Table S3. Table 4 depicts the results of the multivariable regression models. Patients with Medicaid showed a greater odds of returning to the ED with a mental health crisis when compared to patients with commercial insurance (OR = 1.57, 95% CI = 1.29–1.91, p < 0.001). Patients in the 12–18 age group showed a greater odds of returning with a mental health crisis when compared to those in the 18+ age group (OR = 1.48, 95% CI = 1.07–2.11, p = 0.023). Patients under the age of 18 showed a greater odds of returning to the ED with a non-crisis mental health encounter when compared to young adults (<12: OR = 2.10, 95% CI = 1.33–3.40, p = 0.002; 12–18: OR = 1.64, 95% CI =1.09–2.56, p = 0.022). Patients with Medicaid showed a greater odds of returning to the ED with a non-crisis mental health encounter when compared to patients with commercial insurance (OR = 1.44, 95% CI = 1.15–1.81; p = 0.001). Patients under the age of 18 showed a lower odds of returning to the ED with a non-mental health encounter when compared to young adults (<12: OR = 0.41, 95% CI = 0.31–0.55, p < 0.001; 12–18: OR = 0.47, 95% CI = 0.37–0.60, p < 0.001). Patients with unknown/other insurance status had lower odds of returning for non-mental health related concerns compared to those with commercial insurance (OR = 0.64, 95% CI = 0.44–0.91; p = 0.014). Patients within the following subgroups showed a greater odds of returning to the ED with a non-mental health encounter when compared to their respective reference groups: Female patients (OR = 1.50, 95% CI = 1.28–1.76, p < 0.001), Black patients (OR = 1.80, 95% CI = 1.50–2.17, p < 0.001), Hispanic patients (OR = 1.85, 95% CI = 1.31–2.61, p < 0.001), and patients with Medicaid insurance (OR = 1.37, 95% CI = 1.17–1.61, p < 0.001).

TABLE 4.

Multivariable logistic regression for type of ED services following index high-risk visit controlling for age, gender, insurance status, ethnicity, and race.

Odds ratio 95% CI p Value
Mental health crisis
Age (years)
 <12 1.43 0.98–2.13 0.069
 12–18 1.48 1.07–2.11 0.023
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 1.03 0.85–1.26 0.743
Race
 White or European American 1.00 (ref) – –
 Black or African American 0.89 0.71–1.11 0.311
 Other/Unknown race 0.88 0.61–1.27 0.506
Hispanic Ethnicity 0.99 0.65–1.52 0.996
Insurance status
 Commercial 1.00 (ref) – –
 Medicaid 1.57 1.29–1.91 <0.001
 Other/Unknown 1.05 0.65–1.62 0.841
Other Mental Health Concern
Age (years)
 <12 2.10 1.33–3.40 0.022
 12–18 1.64 1.09–2.56 0.022
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 0.99 0.80–1.24 0.917
Race
 White or European American 1.00 (ref) – –
 Black or African American 1.08 0.83–1.41 0.581
 Other/Unknown race 0.87 0.55–1.33 0.521
Hispanic Ethnicity 1.46 0.90–2.33 0.120
Insurance status
 Commercial 1.00 (ref) – –
 Medicaid 1.44 1.15–1.81 0.001
 Other/Unknown 0.80 0.44–1.37 0.438
Non-mental health concern
Age (years)
 <12 0.41 0.31–0.55 <0.001
 12–18 0.47 0.37–0.60 <0.001
 ≥18 1.00 (ref) – –
Sex
 Male 1.00 (ref) – –
 Female 1.50 1.28–1.76 <0.001
Race
 White or European American 1.00 (ref) – –
 Black or African American 1.80 1.50–2.17 <0.001
 Other/Unknown race 1.08 0.79–1.46 0.694
Hispanic ethnicity 1.85 1.31–2.61 <0.001
Insurance status
 Commercial 1.00 (ref) – –
 Medicaid 1.37 1.17–1.61 <0.001
 Other/Unknown 0.64 0.44–0.91 0.014
Open in a new tab

Note: Bolded values are statistically significant (p < 0.05).

DISCUSSION

Our study aimed to characterize emergency service utilization patterns in youth at elevated risk for suicide. We examined youth who presented to the ED with positive suicidality screens and/or suicide-related behaviors and assessed for demographic predictors of subsequent ED utilization patterns including frequency, timing and reasons over a 24-month follow-up period. Among our sample, patients who were Black, young adults, female, or insured by Medicaid were more likely to be frequent utilizers of the emergency department in the 2-year follow-up period. These demographic factors were also associated with greater odds of returning to the ED within 90 days. Patients over the age of 18, female patients, and Black patients were more likely to return for a non-mental health concern, while patients with Medicaid were more likely to return for both mental and non-mental health reasons.

Our findings shed light on which patient populations are frequent utilizers of emergency services, potentially because their needs are not being met elsewhere in the healthcare system. High ED utilization is thought to represent not only a lack of access to or quality of care in other healthcare settings, but also an underlying structural effect due to limited resource distribution or other social determinants of health for marginalized groups. Previous work has shown that emergency visits for suicide attempts were lower in those with increased access to other mental health services (Gentil et al., 2020). A system of care for youth at risk of suicide is needed to help ensure their needs are met and suicide risk is adequately managed (Haroz, Sarapik, et al., 2021).

Previous findings have shown that patients who identify as Black are more likely to utilize the ED and less likely to have a primary care doctor than their white counterparts (Parast et al., 2022). This is consistent with our own findings showing higher utilization particularly among Black patients who have a history of suicidal ideation or behaviors. This finding also questions whether the healthcare system is adequately providing care for these youth. One study found that only 22% of Black parents of a child with mental health challenges sought care within 30 days of identifying the need for mental healthcare, with concern of involuntary hospitalization being a key reason for deferring care (Richmond et al., 2022). Moreover, another study found that Black preadolescents who reported to the ED with behavioral health concerns were more likely to be discharged rather than admitted compared to their non-Black peers (Vidal et al., 2023). Another study of adolescent patients who presented to the ED at high risk for suicide found that Black patients were less likely to have received outpatient therapy in the month prior to their presentation relative to their white counterparts (King et al., 2020). Coupled with these findings, our study reaffirms the critical need for enhanced care quality and linkage support for this population. The VA suicide risk identification program has been paired with more intensive resources to manage both in-patient and outpatient care. Recent studies found that this approach was associated with less ED utilization (Berg et al., 2018) indicating a possible solution to reducing care disparities (Haroz, Sarapik, et al., 2021).

In our study, young adults were more likely to utilize the emergency department than younger patients, particularly for non-mental health concerns. As patients transition out of pediatric care, they become responsible for coordinating their own care, which could be one possible explanation for increased emergency services utilization (Fortuna et al., 2010).

Patients with Medicaid were consistently found to utilize the ED often, within 90 days post initial ED visit, and for both mental health crises and non-mental health concerns. This is consistent with previous findings showing that those with Medicaid insurance tend to be higher utilizers of emergency services compared to those with other types of health insurance, suggesting a disparity in access to and quality of care in this population (Vinton et al., 2014). High patient-rated quality and access to primary care have previously been shown to be associated with lower ED utilization for children with Medicaid (Brousseau et al., 2009). Similarly, female patients were found to utilize the ED more often. This is consistent with some previous studies (Milbrett & Halm, 2009), but differs from other recent findings from survey data (Schlichting et al., 2017), though this previous work focused on ED utilization patterns more broadly rather than within patients with suicidality.

Interventions to reduce emergency service utilization may help to decrease healthcare costs and reduce health disparities. The current study provides unique and important insights into how to prioritize and tailor these types of interventions. Mobile crisis services have been shown to be an effective and cost-advantaged intervention in reducing emergency visits for behavioral health reasons (Fendrich et al., 2019; Fields & Shannahan, 2016). Based on the current findings, emergency department staff could raise awareness about this service particularly among patients who are young adults, are females, are Black, or insured by Medicaid. Other work has suggested that increased case management, care coordination, or access to a psychiatric social worker could help reduce frequent utilization (Blonigen et al., 2018; Enard & Ganelin, 2013; Hearld & Alexander, 2012; Kumar & Klein, 2013); when the resources are available but limited, these patients could be prioritized. Qualitative research and patient-centered design may be needed to tailor programs and outreach, as well as train providers, to address their unique mental health concerns, specific barriers to care, and overlapping identities (e.g., a black young adult with Medicaid insurance). Ultimately, once risk is identified, health systems should help to promote a system of care, particularly for subgroups that are at elevated risk with more limited access to outpatient and community-based care (Haroz, Sarapik, et al., 2021). Existing systems-based interventions highlight cost-effective and evidence-based models of suicide prevention (Baker et al., 2018; Jobes et al., 2018; Thomas et al., 2022). Ultimately, payers should be incentivized to fund such a systems-based approach given the opportunity to avoid high-cost encounters such as emergency services visits and inpatient hospitalizations.

Limitations

This study has some notable limitations. Our analyses rely on data from one hospital, so we could be missing any patient visits to other emergency departments. Patients who had a positive experience at the index visit may choose to return to the same hospital, whereas those who did not may have chosen one of many other local EDs in the city and would then artificially appear to be low utilizers. A single setting may limit the generaliz-ability of our findings to other populations and settings dissimilar to that of a large urban academic center. The ASQ was not universally collected in our young adult population, which could have led to under-detection of eligible patients in this age group, though exploratory analyses excluding this age group were consistent with our main findings. Additionally, the odds ratios for the significant predictors of our outcomes were modest, which may not translate to a clinically significant difference. Our dataset overlaps with the start of the COVID-19 pandemic, which could have affected utilization patterns (Baugh et al., 2021; Duncan et al., 2022). Finally, we do not have any information on subsequent deaths (including suicide or overdose) for the patients in our sample. These limitations necessitate further research to assess whether these findings are confirmed in prospective samples and to explore the potential cost–benefit analysis of interventions to address the disparities found in this work.

Conclusion

Among youth who were identified as being at an elevated risk for suicide, young adults, patients with Medicaid, patients who identify as Black, and female patients showed a greater odds of being frequent utilizers of the emergency room within the 2 years following an initial high-risk visit. Our findings illustrate that racially minoritized groups with intersectional needs may require interventions to coordinate connections to non-emergency-based care as a way to reduce health disparities among those at risk for suicide. Ultimately, providing support for connecting to outside services may facilitate better care for patients, reduce costs, and decrease strains on emergency health services.

Supplementary Material

suppl tables

ACKNOWLEDGEMENTS

E.E.H. and LBA are supported by grant number K01MH116335 and K01MH127310 from the National Institute of Mental Health (NIMH), respectively. L.M. receives grant support from PCORI, the National Network of Depression Centers, and the Once Upon a Time Foundation. PSN is supported by the American Foundation for Suicide Prevention (YIG-0-093-18) and NIDA (K23DA055693).

Funding information

National Institute of Mental Health, Grant/Award Number: K01MH116335

Footnotes

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to report.

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are not available due to privacy concerns and HIPAA.

REFERENCES

  1. Agency for Healthcare Research and Quality. (2022). Clinical classifications software refined (CCSR). Healthcare Cost and Utilization Project (HCUP). https://www.hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp [Google Scholar]
  2. Alvarez K, Polanco-Roman L, Samuel Breslow A, & Molock S (2022). Structural racism and suicide prevention for Ethnoracially minoritized youth: A conceptual framework and illustration across systems. The American Journal of Psychiatry, 179(6), 422–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. American Academy of Pediatrics. (2021). AAP-AACAP-CHA Declaration of a National Emergency in Child and Adolescent Mental Health. https://www.aap.org/en/advocacy/child-and-adolescent-healthy-mental-development/aap-aacap-cha-declaration-of-a-national-emergency-in-child-and-adolescent-mental-health/
  4. Baker ST, Nicholas J, Shand F, Green R, & Christensen H (2018). A comparison of multi-component systems approaches to suicide prevention. Australasian Psychiatry: Bulletin of Royal Australian and New Zealand College of Psychiatrists, 26(2), 128–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ballard ED, Horowitz LM, Jobes DA, Wagner BM, Pao M, & Teach SJ (2013). Association of positive responses to suicide screening questions with hospital admission and repeated emergency department visits in children and adolescents. Pediatric Emergency Care, 29(10), 1070–1074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Baugh JJ, White BA, McEvoy D, Yun BJ, Brown DFM, Raja AS, & Dutta S (2021). The cases not seen: Patterns of emergency department visits and procedures in the era of COVID-19. The American Journal of Emergency Medicine, 46, 476–481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Berg JM, Malte CA, Reger MA, & Hawkins EJ (2018). Medical records flag for suicide risk: Predictors and subsequent use of care among veterans with substance use disorders. Psychiatric Services, 69(9), 993–1000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Blonigen DM, Manfredi L, Heinz A, Bi X, Suarez P, Nevedal AL, Vashi AA, Timko C, & Wagner T (2018). Reducing frequent utilization of psychiatric emergency services among veterans while maintaining quality of care. Psychiatric Services, 69(4), 438–447. [DOI] [PubMed] [Google Scholar]
  9. Boyd RC (2020). 2.4 Report from the congressional black caucus taskforce for black youth suicide and mental health. In Journal of the American Academy of Child & Adolescent Psychiatry, 59(10), S269. 10.1016/j.jaac.2020.07.557 [DOI] [Google Scholar]
  10. Bridge JA, Horowitz LM, Fontanella CA, Sheftall AH, Greenhouse J, Kelleher KJ, & Campo JV (2018). Age-related racial disparity in suicide rates among US youths from 2001 through 2015. JAMA Pediatrics, 172(7), 697–699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brousseau DC, Gorelick MH, Hoffmann RG, Flores G, & Nattinger AB (2009). Primary care quality and subsequent emergency department utilization for children in Wisconsin Medicaid. Academic Pediatrics, 9(1), 33–39. [DOI] [PubMed] [Google Scholar]
  12. Centers for Disease Control and Prevention. (2020). Web-based Injury Statistics Query and Reporting System. https://wisqars.cdc.gov/data/lcd/home
  13. Curtin SC (2020). National Vital Statistics Reports. https://www.cdc.gov/nchs/data/nvsr/nvsr69/NVSR-69-11-508.pdf
  14. Curtin SC, Garnett M, & Ahmad F (2022). Provisional numbers and rates of suicide by month and demographic characteristics: United States, 2021. National Center for Health Statistics (U.S.). 10.15620/cdc:120830 [DOI] [Google Scholar]
  15. Curtin SC, & Hedegaard H (2021). Provisional numbers and rates of suicide by month and demographic characteristics: United States, 2020. National Center for Health Statistics (U.S.). 10.15620/cdc:110369 [DOI] [Google Scholar]
  16. DeVylder JE, Ryan TC, Cwik M, Wilson ME, Jay S, Nestadt PS, Goldstein M, & Wilcox HC (2019). Assessment of selective and universal screening for suicide risk in a pediatric emergency department. JAMA Network Open, 2(10), e1914070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Duncan A, Herrera C-N, Okobi M, Nandi S, & Oblath R (2022). Locked down or locked out? Trends in psychiatric emergency services utilization during the COVID-19 pandemic. Journal of Health Services Research & Policy, 28(2), 80–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Enard KR, & Ganelin DM (2013). Reducing preventable emergency department utilization and costs by using community health workers as patient navigators. Journal of Healthcare Management/American College of Healthcare Executives, 58(6), 412–427; discussion 428. [PMC free article] [PubMed] [Google Scholar]
  19. Fendrich M, Ives M, Kurz B, Becker J, Vanderploeg J, Bory C, Lin H-J, & Plant R (2019). Impact of Mobile crisis services on emergency department use among youths with behavioral health service needs. Psychiatric Services, 70(10), 881–887. [DOI] [PubMed] [Google Scholar]
  20. Fields S, & Shannahan R (2016). Services in Support of community living for youth with serious behavioral health challenges: Mobile crisis response and stabilization services. The National Technical Assistance Network for Children’s behavioral health. https://theinstitute.umaryland.edu/media/ssw/institute/national-center-documents/Mobile-Crisis-Response-&-Stabilization-Services-May-2016.pdf [Google Scholar]
  21. Fortuna RJ, Robbins BW, Mani N, & Halterman JS (2010). Dependence on emergency care among young adults in the United States. Journal of General Internal Medicine, 25(7), 663–669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Freedenthal S (2007). Racial disparities in mental health service use by adolescents who thought about or attempted suicide. Suicide & Life-Threatening Behavior, 37(1), 22–34. [DOI] [PubMed] [Google Scholar]
  23. Gaylor EM, Krause KH, Welder LE, Cooper AC, Ashley C, Mack KA, Crosby AE, Trinh E, Ivey-Stephenson AZ, & Whittle L (2023). Suicidal thoughts and behaviors among high school students – youth risk behavior survey, United States, 2021. MMWR Supplements, 72(1), 45–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gentil L, Huỳnh C, Grenier G, & Fleury M-J (2020). Predictors of emergency department visits for suicidal ideation and suicide attempt. Psychiatry Research, 285, 112805. [DOI] [PubMed] [Google Scholar]
  25. Haroz EE, Kitchen C, Nestadt PS, Wilcox HC, DeVylder JE, & Kharrazi H (2021). Comparing the predictive value of screening to the use of electronic health record data for detecting future suicidal thoughts and behavior in an urban pediatric emergency department: A preliminary analysis. Suicide & Life-Threatening Behavior, 51, 1189–1202. 10.1111/sltb.12800 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Haroz EE, Sarapik LM, Adams LB, Nestadt P, Athey A, Slade E, Cwik M, Berman AL, & Wilcox H (2021). A Cascade of care model for suicide prevention. 10.31234/osf.io/g8xfh [DOI] [PMC free article] [PubMed]
  27. Hawrilenko M, Kroshus E, Tandon P, & Christakis D (2021). The association between school closures and child mental health during COVID-19. JAMA Network Open, 4(9), e2124092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hearld LR, & Alexander JA (2012). Patient-centered care and emergency department utilization: A path analysis of the mediating effects of care coordination and delays in care. Medical Care Research and Review: MCRR, 69(5), 560–580. [DOI] [PubMed] [Google Scholar]
  29. Hoffmann JA, Stack AM, Samnaliev M, Monuteaux MC, & Lee LK (2019). Trends in visits and costs for mental health emergencies in a pediatric emergency department, 2010–2016. Academic Pediatrics, 19(4), 386–393. [DOI] [PubMed] [Google Scholar]
  30. Horowitz LM, Bridge JA, Teach SJ, Ballard E, Klima J, Rosenstein DL, Wharff EA, Ginnis K, Cannon E, Joshi P, & Pao M (2012). Ask suicide-screening questions (ASQ): A brief instrument for the pediatric emergency department. Archives of Pediatrics & Adolescent Medicine, 166(12), 1170–1176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jobes DA, Gregorian MJ, & Colborn VA (2018). A stepped care approach to clinical suicide prevention. Psychological Services, 15(3), 243–250. [DOI] [PubMed] [Google Scholar]
  32. King CA, Brent D, Grupp-Phelan J, Shenoi R, Page K, Mahabee-Gittens EM, Chernick LS, Melzer-Lange M, Rea M, McGuire TC, Littlefield A, Casper TC, & Pediatric Emergency Care Applied Research Network (PECARN). (2020). Five profiles of adolescents at elevated risk for suicide attempts: Differences in mental health service use. Journal of the American Academy of Child and Adolescent Psychiatry, 59(9), 1058–1068.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. King CA, O’Mara RM, Hayward CN, & Cunningham RM (2009). Adolescent suicide risk screening in the emergency department. Academic Emergency Medicine: Official Journal of the Society for Academic Emergency Medicine, 16(11), 1234–1241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kumar GS, & Klein R (2013). Effectiveness of case management strategies in reducing emergency department visits in frequent user patient populations: A systematic review. The Journal of Emergency Medicine, 44(3), 717–729. [DOI] [PubMed] [Google Scholar]
  35. Lo CB, Bridge JA, Shi J, Ludwig L, & Stanley RM (2020). Children’s mental health emergency department visits: 2007–2016. Pediatrics, 145(6), e20191536. 10.1542/peds.2019-1536 [DOI] [PubMed] [Google Scholar]
  36. Merrick EL, Perloff J, & Tompkins CP (2010). Emergency department utilization patterns for Medicare beneficiaries with serious mental disorders. Psychiatric Services, 61(6), 628–631. [DOI] [PubMed] [Google Scholar]
  37. Milbrett P, & Halm M (2009). Characteristics and predictors of frequent utilization of emergency services. Journal of Emergency Nursing, 35(3), 191–198; quiz 273. [DOI] [PubMed] [Google Scholar]
  38. Office of the Surgeon General. (2021). U.S. surgeon general issues advisory on youth mental health crisis further exposed by COVID-19 pandemic. US Department of Health and Human Services. https://www.hhs.gov/about/news/2021/12/07/us-surgeon-general-issues-advisory-on-youth-mental-health-crisis-further-exposed-by-covid-19-pandemic.html [Google Scholar]
  39. Office of the Surgeon General (US), & National Action Alliance for Suicide Prevention (US). (2012). 2012 National Strategy for suicide prevention: Goals and objectives for action: A report of the U.S. surgeon general and of the National Action Alliance for suicide prevention. US Department of Health & Human Services (US). [PubMed] [Google Scholar]
  40. Parast L, Mathews M, Martino S, Lehrman WG, Stark D, & Elliott MN (2022). Racial/ethnic differences in emergency department utilization and experience. Journal of General Internal Medicine, 37(1), 49–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Ramchand R, Gordon JA, & Pearson JL (2021). Trends in suicide rates by race and ethnicity in the United States. JAMA Network Open, 4(5), e2111563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Richmond J, Adams LB, Annis IE, Ellis AR, Perryman T, Sikich L, & Thomas KC (2022). Rapid and deferred help seeking among African American parents of children with emotional and behavioral difficulties. Psychiatric Services, 73, 1359–1366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Rogers SC, Mulvey CH, Divietro S, & Sturm J (2017). Escalating mental health care in pediatric emergency departments. Clinical Pediatrics, 56(5), 488–491. [DOI] [PubMed] [Google Scholar]
  44. Schlichting LE, Rogers ML, Gjelsvik A, Linakis JG, & Vivier PM (2017). Pediatric emergency department utilization and reliance by insurance coverage in the United States. Academic Emergency Medicine, 24(12), 1483–1490. [DOI] [PubMed] [Google Scholar]
  45. Sheftall AH, Vakil F, Ruch DA, Boyd RC, Lindsey MA, & Bridge JA (2022). Black youth suicide: Investigation of current trends and precipitating circumstances. Journal of the American Academy of Child and Adolescent Psychiatry, 61(5), 662–675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. The Joint Commission. (2019). Suicide Prevention. https://www.jointcommission.org/resources/patient-safety-topics/suicide-prevention/
  47. Thomas GV, Chapman MR, & Grumet JG (2022). Investing in suicide prevention: Zero suicide in a pediatric system of care. In Ackerman JP & Horowitz LM (Eds.), Youth suicide prevention and intervention: Best practices and policy implications (pp. 135–143). Springer International Publishing. [Google Scholar]
  48. Vidal C, Ngo TL, Wilcox HC, Hammond CJ, Campo JV, O’Donnell E, & Ryan LM (2023). Racial differences in emergency department visit characteristics and Management of Preadolescents at risk of suicide. Psychiatric Services, 74(3), 312–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Vinton DT, Capp R, Rooks SP, Abbott JT, & Ginde AA (2014). Frequent users of US emergency departments: Characteristics and opportunities for intervention. Emergency Medicine Journal: EMJ, 31(7), 526–532. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

suppl tables
NIHMS1971020-supplement-suppl_tables.docx (41.5KB, docx)

Data Availability Statement

The data that support the findings of this study are not available due to privacy concerns and HIPAA.

RESOURCES