Introduction
Mental health problems in children and adolescents are emerging as a major public health concern worldwide (1). According to the 2022 Korean Domestic Child and Adolescent Mental Health Survey, the lifetime prevalence of mental disorders among children and adolescents reached 16.1%, whereas only 6.6% used mental health services (2). This treatment gap may delay timely intervention, leading to emergency department (ED) visits during acute psychiatric crises, including suicide attempt (SA) or non-suicidal self-harm (NSSH) (3,4). A steady increase was noted in ED visits for mental health problems among children and adolescents in Korea (5,6), highlighting the growing role of EDs in pediatric psychiatric care (7,8).
Pediatric psychiatric patients presenting to EDs are a heterogeneous population with diverse clinical presentations. Some patients present after SA or NSSH, requiring treatment for both physical injury and psychiatric risk assessment. Others present with suicidal ideation (SI) without physical NSSH, or with other psychiatric problems (OPP) such as anxiety, agitation, aggression, depressive symptoms, or exacerbation of preexisting psychiatric disorders. Although previous studies have mainly focused on the SA and NSSH (9-11), these non-injury or non-self-harm presentations also require timely ED-based intervention (12). In Korea, suicide has remained the leading cause of adolescent death (13), and the rate of ED visits related to SA or NSSH has increased annually (5). While domestic studies have primarily focused on SA or NSSH, there remains a scarcity of studies on the full spectrum of pediatric psychiatric ED presentations (6,9,14).
Therefore, we aimed to examine 11-year annual trends in pediatric psychiatric ED visits and to compare clinical characteristics across 4 clinical presentation groups: SA, NSSH, SI, and OPP, thereby characterizing the full spectrum of psychiatric emergencies.
Methods
We conducted a single-center retrospective study at the regional emergency medical center of Gyeongsang National University Hospital, located in Jinju, Korea, from January 2013 through December 2023. The ED provides emergency care and psychiatric consultation as a regional referral center for acute conditions on a 24/7 basis. Psychiatric consultations were conducted by psychiatric residents or board-certified psychiatrists. This study was approved by the institutional review board of Gyeongsang National University Hospital with a waiver for informed consent given its retrospective nature and minimal risk to study participants (IRB no. 2024-04-14).
We included all pediatric patients (≤18 years) who presented to the ED with psychiatric complaints and received formal psychiatric interview. Data were collected from the ED records and psychiatric consultation notes. Patients with incomplete psychiatric evaluation were excluded, including those involving refusal of such interviews by the patients or legal guardians, absconding, discharge against medical advice, or death. All eligible visits, including repeat visits, were included in the annual trend analysis. For the comparative patient-level analysis, only the first visit by each patient was included to ensure statistical independence. Given its clinical implications, repeat visit patterns were analyzed at the patient level according to the clinical presentation groups (see definitions of terms in Appendix 1 [https://doi.org/10.22470/pemj.2026.01627]).
To characterize the full spectrum of psychiatric ED presentations, patients were classified into 4 mutually exclusive clinical presentation groups, according to the main clinical presentations during the index visit: SA, NSSH, SI, and OPP (Appendix 1). This classification was performed by 2 independent investigators, with any discrepancies resolved through discussion and consensus. When multiple symptoms were documented, the primary reason for the ED visit was determined based on the records. For patients in the OPP group, we additionally reviewed the specific reasons for the visits and the principal psychiatric diagnoses.
Variables related to baseline characteristics included age, sex, living with both parents, presence of siblings, family history of psychiatric disorders, presence of family members at ED arrival, previous NSSH or SA, and details of psychiatric treatment (currently in treatment, treatment discontinued, or no history of treatment). Variables related to clinical characteristics were the Korean Triage and Acuity Scale levels 1–3, injury mechanisms (drug ingestion, cutting, fall from a height, hanging/strangulation, or drowning), psychosocial stressors (family, social/interpersonal, academic, psychosocial trauma [documented experiences of sexual violence, domestic child abuse, or school violence], or individual/psychiatric), time to psychiatric interview, ED length of stay (EDLOS), and ED disposition (outpatient referral, hospitalization, or transfer-out).
All analyses were performed using IBM SPSS software ver. 21.0 (IBM Corp.). Annual trends in psychiatric ED visits were analyzed using simple linear regression. Regression coefficients (B), 95% confidence intervals, and P values were reported to indicate the estimated annual change in the number of ED visits. For the characteristics, the normality of continuous variables was assessed using the Shapiro-Wilk test. Continuous variables were expressed as medians with interquartile ranges and compared among the 4 groups using the Kruskal-Wallis test, followed by the Dunn test with Bonferroni correction for post hoc pairwise comparisons. Categorical variables were compared using the chi-square tests or Fisher exact tests. Post hoc analyses were conducted using adjusted standardized residuals, with an absolute value of >1.96 considered significant. A P <0.05 was considered significant.
Results
During the study period, 730 pediatric patients who visited the ED for psychiatric complaints were identified. Excluding patients undergoing incomplete psychiatric evaluation, 688 patients were included in the annual trend analysis, of whom 465 were included in the comparative patient-level analysis (Figure 1). The annual number of pediatric psychiatric ED patients increased significantly over the period by the simple linear regression (Figure 2; see numerical data in Appendix 2 [https://doi.org/10.22470/pemj.2026.01627]). The total number increased by an estimated 8.79 patients per year, with the largest annual increase observed in the SA group at 2.99 patients per year.
The baseline characteristics of the study population are summarized in Table 1. Although median ages differed significantly among the clinical presentation groups, the statistical significance was not considered clinically meaningful. The other significant differences were observed in the sex, living with both parents, presence of siblings, presence of family members at ED arrival, previous NSSH or SA, and previous psychiatric treatment. Although female patients predominated in all groups, the female proportions were higher in the SA and NSSH groups than in the other 2 groups. Patients in the SA group were less likely to live with both parents or have arrived at the ED with their family members. Previous NSSH was more common in the SA and NSSH groups than in the other groups, whereas previous SA was more frequent in the SA and SI groups than in the other groups. Current psychiatric treatment was more common in the NSSH and SI groups than in the other groups.
Clinical characteristics of the groups are summarized in Table 2. The SA group had the highest proportion of a Korean Triage and Acuity Scale 1–3. Among the patients in the SA or NSSH group, drug ingestion was the most common method (60.0%). Cutting was more frequent in the NSSH group, whereas highly lethal methods, including falls from a height, hanging/strangulation, and drowning, were observed exclusively in the SA group.
Psychosocial stressors differed across the groups (Table 2). Family-related, social/interpersonal, and academic stressors were most frequently reported in the NSSH group, followed by the SA group. Psychosocial trauma-related stressors were most frequent in the SI group, whereas individual/psychiatric ones were so in the OPP group.
Median values of the time to psychiatric interview and EDLOS were longest in the SA group (Table 3). ED disposition differed among the groups: the SA group had the highest rates of hospitalization or transfer, whereas the NSSH group had the most frequent outpatient referral.
Among the 465 patients included in the first-visit analysis, 115 patients (24.7%) had at least 1 repeat ED visit during the period, with a total of 223 repeat visit events after the index visits (Table 4). The proportion of patients with such visits was highest in the SI group (51.0%). Among the patients with at least 1 repeat visit, 53.9% returned with the same presentations as their index visits, while 36 of 84 repeat visitors initially presenting with NSSH, SI, or OPP (42.9%) returned with more severe presentations. Subsequent or recurrent SAs occurred in 33.0%.
Details of the OPP group are presented in Appendix 3 (https://doi.org/10.22470/pemj.2026.01627). The most common primary reason for the visits was internalizing symptoms, including anxiety, panic, or depressive symptoms (33.9%), followed by agitation/aggressive behavior, somatic symptoms related to psychiatric distress, and psychotic symptoms. Psychotic disorders were the most common category of the principal psychiatric diagnosis.
Discussion
This study evaluated the full spectrum of pediatric psychiatric ED visits over the 11-year period at a Korean regional emergency medical center. Overall, such visits increased more than threefold during the period. Although self-harm and suicidality-related presentations accounted for a large proportion of visits, OPP represented the largest group, indicating that psychiatric emergencies in the ED extended beyond the spectrum of SA, NSSH, and SI. Distinct psychosocial and clinical characteristics were observed across the 4 clinical manifestation groups.
The increase in psychiatric ED visits in this study (Figure 1) is consistent with trends reported in Korea and other countries (5,15). In particular, marked increases in the visits related to SA and NSSH groups are clinically concerning (16), as such patients require both treatment of physical injuries and psychiatric risk assessment. Patients with SI or OPP also require timely evaluation, safety planning, or acute symptom stabilization depending on their presentations. These findings support the need for ED-based systems that provide timely psychiatric assessment, risk stratification, and linkage to psychiatric follow-up care.
This full-spectrum approach highlights the clinical importance of patients with OPP who are often underrepresented in studies of pediatric psychiatric emergencies. In the present study, the OPP represented the largest group and showed characteristics distinct from those of the self-harm/suicidality-related groups (Table 1). The relatively higher male proportion in the OPP group was presumed to be partly related to differences in psychiatric symptom expressions (5,14-16). Adolescent females are more often represented in suicide and self-harm behaviors, whereas males may more commonly present with externalizing symptoms such as agitation or aggressive behavior (17,18). The additional analyses of the OPP group further support the clinical relevance of the manifestations, showing diverse acute psychiatric symptoms and heterogeneous principal psychiatric diagnoses. These findings suggest that OPP should not be regarded merely as a residual category, but as an important distinct group within psychiatric ED presentations. The OPP group also had a relatively low rate of current psychiatric treatment (Table 1). Taken together with the diverse manifestations and diagnoses observed in the group, this finding suggests that EDs may serve as an initial point of psychiatric assessment for such manifestations. However, this interpretation should be made cautiously because the present study did not evaluate prior unmet needs for mental health services or barriers to outpatient psychiatric care.
The methods of suicide or self-harm have been reported as commonly involving drug ingestion and cutting (3,19), which were observed in our study. In particular, the finding that drug ingestion was the most common injury mechanisms suggests the importance of assessing access to household or prescription medications during the ED stay. Although the NSSH means the absence of suicidal intent by definition (20), no highly lethal methods (e.g., falls from a height) were observed in the NSSH group (Table 2). However, NSSH is associated with subsequent suicidal behavior and requires careful risk assessment and follow-ups (21,22). Indeed, the highly lethal methods, which were observed only in the SA group, require immediate medical and psychiatric intervention (23).
The repeat visit analysis also supports the importance of follow-ups after psychiatric ED visits. Patients initially presenting with SI had the highest repeat visit rate, and some initially presenting with NSSH or SI subsequently returned with SA (Table 4). These findings suggest that non-fatal or non-injurious presentations may still indicate ongoing or escalating risk in a subset of patients (24), indicating the importance of structured suicide risk assessment, safety planning, or linkage to follow-up care after ED discharge.
The underlying factors associated with the psychiatric presentations differed among the groups. The higher frequency of family-related stressors in the SA or NSSH groups (Table 2) is consistent with previous studies showing the associations between family factors and adolescent suicidal behavior (25). These findings support the importance of considering familial context during ED assessment of adolescents presenting with self-harm or suicidality. The prominence of academic stress in the NSSH group is also consistent with previous reports linking school-related stress to adolescent NSSH (26), supporting the need to assess the stress in EDs. Psychosocial trauma-related stressors were most prominent in the SI group, supporting the need to assess the medical history of trauma in adolescents presenting with SI.
The SA group showed the highest proportions of high acuity, the longest median values of the time to psychiatric interview and EDLOS, and the highest hospitalization rate among the groups (Tables 2, 3). These suggest that patients presenting after SAs may require more intensive medical and psychiatric management (27). The findings related to the time intervals may reflect the need for medical stabilization, psychiatric risk assessment, disposition planning, and arrangement of inpatient or transfer care. Considering that EDs may be overstimulating places for psychiatric patients (28), these findings support the need for timely psychiatric assessment, safe observation spaces, and efficient disposition pathways.
This study has limitations. First, the excluded patients who did not complete psychiatric evaluation might include patients with severe symptoms who were unable to wait for evaluation, or those with relatively minor symptoms who left early at the discretion of their guardians. This exclusion may have affected the proportion of a variable of each group, as well as the EDLOS or disposition. Second, the repeat visit analysis did not include visits to other institutions. Although our hospital is the only institution in the region that provides psychiatric consultation in the ED, the recurrence of psychiatric crises or progression to more severe presentations might have been underestimated. Third, given the inherent features of retrospective chart review, the categorization of the clinical manifestations largely depended on documented suicidal intent and reports from patients or guardians. Thus, misclassification may have occurred when suicidal intent was unclear or incompletely documented. Fourth, the study lacked an analysis of the potential impact of changes in ED operations, referral patterns, psychiatric consultation availability, documentation practices, or social awareness of pediatric mental health problems during the study period. Finally, the single-center setting may limit the generalizability of our study findings to other settings.
In conclusion, pediatric psychiatric patients presenting to EDs represent a heterogeneous population across a broad spectrum of clinical presentations, including SA, NSSH, SI, and OPP. The distinct clinical characteristics observed across the clinical manifestations suggest that psychiatric patients who visit EDs should be managed by assessment tailored for each manifestation.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
