INTRODUCTION
Suicide represents a major global public health crisis. In 2021 alone, approximately 727,000 people worldwide died by suicide, exceeding the combined number of deaths caused by malaria, HIV, breast cancer, war, and homicide [
1]. This figure corresponds to 1.1% of all deaths globally, indicating that more than one in every 100 deaths was attributable to suicide. Reflecting the severity of this issue, the reduction of suicide mortality has been incorporated as a key indicator within the United Nations Sustainable Development Goals, underscoring suicide prevention as a critical priority for the international community.
The manifestations of this crisis are also clearly observed in emergency care settings. In the United States, emergency department (ED) visits related to suicide attempts and intentional self-harm markedly increased from 1.43 million in 2011 to 5.37 million in 2021, imposing a substantial burden on the emergency care system [
2]. The ED serves as a primary point of initial contact with the healthcare system for individuals who attempt suicide; therefore, it plays a pivotal role in suicide prevention strategies and initiatives to reduce suicide-related mortality.
The situation in Korea is even more concerning. Korea has consistently recorded the highest suicide rate among member countries of the Organization for Economic Co-operation and Development (OECD), a pattern that has persisted for more than a decade [
3]. The COVID-19 pandemic has been associated with worsening mental health conditions and an increased risk of suicide [
4], and instability within the healthcare system following the government’s 2024 policy to expand medical school enrollment, accompanied by the withdrawal of resident physicians, has placed additional strain on the overall capacity of emergency medical services [
5]. Such structural challenges within the healthcare system are likely to compromise the ability to respond adequately to psychiatric emergencies, including suicide and self-harm.
Although a previous study analyzed changes in ED utilization related to suicide and self-harm from 2018 to 2022, the time-dependent nature of suicide epidemiology necessitates continuous monitoring and reassessment of these trends. In particular, 2023, corresponding to the post–COVID-19 recovery phase, and 2024, during which structural changes in the emergency care system became pronounced due to healthcare workforce attrition [
5], may have been characterized by distinct patterns of ED utilization and clinical outcomes among patients presenting with suicide attempts and self-harm compared with those observed in earlier periods. Accordingly, the present study extends the analysis period to more precisely evaluate the impact of recent environmental and healthcare system changes on suicide attempt– and self-harm–related indicators in ED settings.
METHODS
Ethics statement
This study was approved by the Institutional Review Board of the National Medical Center, Korea (No. NMC-2023-08-094). The requirement for informed consent was waived due to the retrospective, observational, and anonymous nature of the study.
Study population
This retrospective study used data from Korea’s National Emergency Department Information System (NEDIS). NEDIS is a nationwide ED registry that has been managed by the National Emergency Medical Center since 2003 and contains administrative, clinical, and diagnostic information for all patients presenting to EDs. We analyzed patients who visited EDs nationwide between 2020 and 2024.
Patients with self-harm or suicide attempts were identified based on intentionality classification at ED presentation, specific initial triage classification codes recorded in the Korean Triage and Acuity Scale (KTAS), or discharge diagnoses using Korean Classification of Diseases (KCD) codes X60–X84 (excluding X65). The initial KTAS classification codes refer not to KTAS levels (1–5) but to standardized five-character alphanumeric codes assigned during the initial triage process. Patients were classified as having self-harm or suicide attempts if their initial triage code corresponded to one of the following categories: ABECA, BBECA, ABECB, BBECB, ABECD, BBECD, AABCC, BABCC, AABCD, or BABCD.
KCD code X65 (intentional self-poisoning by and exposure to alcohol) was excluded because alcohol-related poisoning recorded in ED administrative data is frequently subject to misclassification as accidental intoxication and may lack specificity for suicidal intent. Exclusion of X65 was intended to improve the specificity of case identification.
Patients who were dead on arrival were excluded. To evaluate age-specific characteristics, patients were categorized as pediatric (<18 years), adult (18–64 years), or older adult (≥65 years).
Statistical analysis
Age- and sex-standardized ED visit rates per 100,000 population were calculated using the 2020 mid-year estimated population of Korea as the standard population. Age- and sex-standardized mortality rates per 100,000 ED visits were calculated using the number of ED visits recorded in NEDIS from 2020 to 2024 as the standard population. Temporal trends were assessed using linear regression models with calendar year included as a continuous variable, and P-values for trend were reported. Standardized incidence and mortality rates in 2023 and 2024 were compared using the z-test, and visit-based in-hospital mortality was compared using the chi-square test. All statistical analyses were performed using SAS ver. 9.4 (SAS Institute Inc) with a two-sided significance level of 0.05.
KEY OBSERVATIONS FROM NEDIS DATA
A total of 209,746 ED visits related to self-harm and suicide attempts were included in the analysis (
Table 1). Overall, women accounted for a higher proportion of visits than men (63.2% vs. 36.8%). The proportion of women exceeded the overall average in the pediatric and adult groups, whereas men predominated in the older adult group. Most patients were treated at higher-level emergency medical centers, including regional emergency medical centers (42.6%) and local emergency medical centers (54.0%), and direct visits were more common than interhospital transfers (94.1% vs. 5.4%). Transport by the National Fire Agency ambulance service (119 ambulance
) was the most frequent mode of arrival (64.0%).
The median ED length of stay was 3.7 hours (interquartile range [IQR], 1.9–7.7 hours), with the longest length of stay observed in the older adult group at 4.6 hours (IQR, 2.4–10.3 hours). The overall hospital admission rate was 29.1%, and more than half of older adult patients (51.2%) required hospitalization. Notably, the proportion of intensive care unit (ICU) admissions among patients with self-harm or suicide attempts was 15.5%, substantially exceeding the ICU admission rate among all ED patients (3.1%) [
6].
Five-year trends in age- and sex-standardized ED visit rates, mortality rates, and visit-based in-hospital mortality stratified by age group are presented in
Fig. 1. Age- and sex-standardized ED visit rates fluctuated year to year in the overall population and across all age groups; however, no statistically significant linear trend was observed over the 5-year period (P for trend >0.05). Nevertheless, nonlinear patterns were noted in the adult and pediatric groups, with increases after 2021 followed by a decline in 2024.
Age- and sex-standardized mortality rates in the overall population showed a gradual upward tendency after 2020, although the 5-year linear trend did not reach statistical significance (P for trend=0.308). In the adult group, an increasing trend in mortality was observed (P for trend=0.099), whereas no clear temporal changes were identified in the pediatric or older adult groups. Visit-based in-hospital mortality also increased in 2024 compared with 2020 in the overall population; however, the linear trend across years was not statistically significant (P for trend=0.163).
To evaluate the impact of healthcare system disruption, outcomes in 2023 and 2024 were compared (
Table 2). Age- and sex-standardized ED visit rates decreased significantly in 2024 compared with 2023 in the overall population and across all age groups (z-test, all P<0.05). In contrast, age- and sex-standardized mortality rates increased significantly in 2024 in the overall population and the adult group (all P<0.05), whereas no significant differences were observed in the pediatric or older adult groups.
Visit-based in-hospital mortality also increased significantly in 2024 in the overall population and the adult group (
Table 3). No statistically significant differences between years were observed in the pediatric or older adult groups.
INTERPRETATION AND IMPLICATIONS
This study analyzed the characteristics and temporal trends of patients presenting with self-harm and suicide attempts between 2020 and 2024 using data from NEDIS. During this period, Korea experienced a series of unprecedented healthcare crises, including the COVID-19 pandemic in 2020, a subsequent recovery phase from 2021 to 2023, and a major disruption to the healthcare system in 2024 related to the medical workforce crisis. By examining long-term changes in the incidence and outcomes of self-harm– and suicide attempt–related ED visits across these overlapping crises, this study provides meaningful insights into the epidemiology of psychiatric emergencies under sustained healthcare system stress.
Age- and sex-standardized ED visit rates increased gradually after 2020 and then declined sharply in 2024, a pattern that was particularly pronounced in the adult age group. In a previous study analyzing ED data from a tertiary hospital in Incheon during the COVID-19 period, COVID-19 itself was identified as an independent risk factor for suicide attempts, and statistically significant increases were observed in suicide attempts with lower rescue potential as well as shifts in socioeconomic motivations among patients presenting after suicide attempts [
4]. These findings suggest that large-scale social crises, such as infectious disease outbreaks, may amplify suicide risk, which is consistent with the increasing incidence observed after 2020 in the present study.
When 5-year incidence trends were examined by age group, incidence rates in the older adult group remained relatively stable over time. In contrast, both the adult and pediatric groups demonstrated a noticeable increase in age- and sex-standardized incidence rates from 2020 to 2022, followed by a decline in 2024. Although the overall 5-year linear trend did not reach statistical significance, the upward pattern during the pandemic and recovery period was evident in these younger age groups. These findings are consistent with those of several previous studies comparing the pre-pandemic and pandemic periods [
4,
7]. Notably, one of the most striking findings of the present study was the pronounced increase in self-harm and suicide attempts among adolescents. This subgroup was estimated to be predominantly composed of adolescents, as the mean age was 15.3±1.7 years. This aligns with the World Health Organization statistics indicating that suicide ranked as the third leading cause of death among individuals aged 15–29 years in 2021, after road traffic injuries and interpersonal violence [
8]. In addition, a prior study analyzing nationwide ED data in Korea from 2016 to 2020 reported a sustained increase in ED visit rates due to suicide attempts and self-harm among young adults aged 19–29 years [
3]. Increasing incidence of suicide attempts and self-harm among adolescents and young adults has been reported globally [
9], highlighting the need for more proactive and targeted policy interventions focused on these age groups.
In the older adult group, incidence rates did not change substantially over time; however, hospitalization rates, ICU admission rates, and mortality were markedly higher than in other age groups. The hospitalization rate among older adults was 51.2%, substantially exceeding that of the overall study population, and the proportion of ICU admissions was also considerably higher than that observed among all ED patients. Age- and sex-standardized mortality rates were highest in the older adult group, consistent with previous studies reporting a substantially higher likelihood of death following suicide attempts among older individuals than among younger age groups [
3]. Suicide attempts in older adults are more likely to involve planned behaviors associated with physical suffering or social isolation and are therefore more frequently fatal [
10]. In addition, older patients often have multiple medical and surgical comorbidities [
11], which may exacerbate the clinical consequences of self-harm or suicide attempts and contribute to mortality from causes beyond the index event [
12,
13]. These findings suggest the need for policy approaches tailored specifically to older adults, distinct from those addressing the increasing trends observed among adolescents.
In the context of these increasing trends, the sharp decline in incidence observed in 2024 is difficult to interpret as a true reduction in risk. Instead, it may reflect reduced access to emergency medical services during the healthcare system disruption, such that individuals requiring urgent care for self-harm or suicide attempts were less likely to be captured within the existing emergency care safety net.
In addition, analyses of fire service transport records have shown that a substantial proportion of emergency patients, including those with potentially severe conditions, are transported to facilities outside designated emergency medical center level institutions and that regional transport patterns differ considerably from national averages [
14]. These findings suggest that administrative ED databases such as NEDIS may not fully capture all emergency transports at the national level. Accordingly, the observed decrease in incidence may not indicate a true decline in occurrence but may instead reflect that a subset of patients was not represented in the available data.
Meanwhile, although most patients with self-harm or suicide attempts in this study were treated at center-level emergency medical facilities, this finding may partially reflect limitations of the NEDIS data. In some institutions, information on intentionality may not have been transmitted, potentially resulting in undercapture of eligible patients, and it is unlikely that all individuals with self-harm or suicide attempts entered the emergency care system through NEDIS-participating institutions. In particular, during periods such as the COVID-19 pandemic or the healthcare system disruption, patients may have been redistributed to nonparticipating facilities, including outpatient clinics or emergency care posts. Under these circumstances, patients who did present to NEDIS-reporting institutions may have represented a relatively higher-acuity population, which could provide one plausible explanation for the concurrent observation of decreased incidence and increased mortality in this study.
The increase in standardized mortality observed in 2024 may suggest excess mortality; however, this interpretation warrants caution given the concurrent decrease in age- and sex-standardized incidence rates. This pattern raises the possibility that patients with less severe self-harm or suicide attempts were not captured in the ED data. Such lower-acuity patients may have been diverted to healthcare settings outside the emergency care system, thereby reducing the number of ED presentations and increasing the overall severity among patients who did present to EDs, which in turn could manifest as an apparent increase in mortality. To determine more clearly whether excess mortality occurred, further studies incorporating data sources outside the emergency medical system, such as national fire service dispatch records, will be required.
Finally, the findings of this study suggest that many longitudinal investigations limited to ED populations may not fully represent the broader population of individuals who die by suicide. Notably, fewer than 16% of individuals who attempt suicide are reported to receive care from mental health professionals, and this proportion has been shown to decline by an average of 9.1% annually [
2]. These findings indicate that hospital-based interventions alone are insufficient for effective suicide prevention and underscore the importance of strategies aimed at early identification and sustained support of high-risk individuals in the community who remain disconnected from the emergency medical care system.
NOTES
-
Author contributions
Conceptualization: YC, JJ; Formal analysis: all authors; Investigation: all authors; Methodology: all authors; Supervision: JJ, MIC; Writing–original draft: YC; Writing–review & editing: all authors. All authors read and approved the final manuscript.
-
Conflicts of interest
The authors have no conflicts of interest to declare.
-
Funding
The authors received no financial support for this study.
-
Data availability
Data analyzed in this study were obtained from the National Emergency Medical Center (NEMC) under the Korean Ministry of Health and Welfare, and were used under license for the current study. Although the data are not publicly accessible, they are available from the corresponding author upon reasonable request with permission from the NEMC.
Fig. 1.Five-year trends. (A) Age- and sex-standardized emergency department (ED) visit rates per 100,000 population. (B) Age- and sex-standardized mortality rates per 100,000 ED visits. (C) In-hospital mortality among patients with suicide attempt or self-harm. P for trend was calculated using linear regression.
Table 1.Baseline characteristics and outcomes of patients presenting with suicide attempt or self-harm
Table 1.
|
Characteristic |
Total (n=209,746) |
Pediatric group (<18 yr) (n=22,640) |
Adult group (18–64 yr) (n=163,296) |
Older adult group (≥65 yr) (n=23,808) |
|
Age (yr) |
38.1±19.2 |
15.3±1.7 |
35.7±13.4 |
76.0±7.5 |
|
Sex |
|
|
|
|
|
Male |
77,191 (36.8) |
4,780 (21.1) |
59,419 (36.4) |
12,991 (54.6) |
|
Female |
132,555 (63.2) |
17,860 (78.9) |
103,877 (63.6) |
10,817 (45.4) |
|
Time from the symptom onset to ED arrivala)
|
1.6 (0.9–4.6) |
2.2 (1.0–8.8) |
1.5 (0.9–4.1) |
1.6 (0.9–4.4) |
|
Type of ED |
|
|
|
|
|
Level I |
89,436 (42.6) |
9,978 (44.1) |
68,326 (41.8) |
11,131 (46.8) |
|
Level II |
113,314 (54.0) |
12,043 (53.2) |
89,446 (54.8) |
11,825 (49.7) |
|
Level III |
6,996 (3.3) |
619 (2.7) |
5,524 (3.4) |
852 (3.6) |
|
Route of arrival |
|
|
|
|
|
Direct visit |
197,459 (94.1) |
20,894 (92.3) |
154,743 (94.8) |
21,822 (91.7) |
|
Transfer from other hospital |
11,328 (5.4) |
1,540 (6.8) |
7,856 (4.8) |
1,931 (8.1) |
|
Referral from outpatient clinic |
882 (0.4) |
200 (0.9) |
634 (0.4) |
48 (0.2) |
|
Other |
30 (0.0) |
2 (0.0) |
23 (0.0) |
5 (0.0) |
|
Unknown |
47 (0.0) |
4 (0.0) |
40 (0.0) |
2 (0.0) |
|
Transport |
|
|
|
|
|
119 Ambulance |
134,231 (64.0) |
8,115 (35.8) |
106,724 (65.4) |
19,392 (81.5) |
|
Other medical institution ambulance |
1,173 (0.6) |
91 (0.4) |
811 (0.5) |
271 (1.1) |
|
Other ambulance |
5,469 (2.6) |
321 (1.4) |
3,827 (2.3) |
1,320 (5.5) |
|
Police or official transport |
1,929 (0.9) |
143 (0.6) |
1,713 (1.0) |
73 (0.3) |
|
Air transport |
223 (0.1) |
19 (0.1) |
155 (0.1) |
49 (0.2) |
|
Other transport |
65,467 (31.2) |
13,712 (60.6) |
49,097 (30.1) |
2,658 (11.2) |
|
Walk-in |
1,125 (0.5) |
220 (1.0) |
868 (0.5) |
37 (0.2) |
|
Other/unknown |
129 (0.1) |
19 (0.1) |
101 (0.1) |
8 (0.0) |
|
Length of stay (hr)b)
|
|
|
|
|
|
Mean±SD |
6.5±8.6 |
5.7±7.0 |
6.2±8.1 |
8.7±12.0 |
|
Median (IQR) |
3.7 (1.9–7.7) |
3.4 (1.8–6.6) |
3.6 (1.8–7.6) |
4.6 (2.4–10.3) |
|
0–6 |
141,725 (67.6) |
16,300 (72.0) |
111,209 (68.1) |
14,216 (59.7) |
|
6–12 |
37,548 (17.9) |
3,550 (15.7) |
29,535 (18.1) |
4,462 (18.7) |
|
12–24 |
23,159 (11.0) |
2,208 (9.8) |
17,500 (10.7) |
3,451 (14.5) |
|
≥ 24 |
7,233 (3.4) |
577 (2.5) |
4,985 (3.1) |
1,671 (7.0) |
|
Unknown |
81 (0.0) |
5 (0.0) |
67 (0.0) |
8 (0.0) |
|
ED disposition |
|
|
|
|
|
Discharge |
133,008 (63.4) |
15,883 (70.2) |
108,782 (66.6) |
8,343 (35.0) |
|
Admissionc)
|
61,000 (29.1) |
5,758 (25.4) |
43,048 (26.4) |
12,193 (51.2) |
|
General ward |
28,348 (13.5) |
4,205 (18.6) |
20,162 (12.3) |
3,980 (16.7) |
|
Intensive care unit |
32,547 (15.5) |
1,520 (6.7) |
22,822 (14.0) |
8,205 (34.5) |
|
Transfer |
8,267 (3.9) |
717 (3.2) |
6,108 (3.7) |
1,442 (6.1) |
|
Comfort care discharge |
41 (0.0) |
2 (0.0) |
26 (0.0) |
13 (0.1) |
|
Death |
5,458 (2.6) |
203 (0.9) |
3,519 (2.2) |
1,736 (7.3) |
|
Other/unknown |
1,972 (0.9) |
77 (0.3) |
1,813 (1.1) |
81 (0.3) |
|
Hospital disposition |
|
|
|
|
|
Discharge |
185,430 (88.4) |
21,281 (94.0) |
147,017 (90.0) |
17,131 (72.0) |
|
Transfer |
13,369 (6.4) |
971 (4.3) |
8,948 (5.5) |
3,450 (14.5) |
|
Comfort care discharge |
67 (0.0) |
3 (0.0) |
42 (0.0) |
22 (0.1) |
|
Death |
8,129 (3.9) |
254 (1.1) |
4,897 (3.0) |
2,978 (12.5) |
|
Other/unknown |
2,751 (1.3) |
131 (0.6) |
2,392 (1.5) |
227 (1.0) |
Table 2.Comparison of age- and sex-standardized visit and mortality rates per 100,000 population among patients with suicide attempt or self-harm in 2023 and 2024
Table 2.
|
Variable |
2023
|
2024
|
z |
P-value |
|
Rate |
SE |
Rate |
SE |
|
Incidence rate (per 100,000 population) |
|
|
|
|
|
|
|
Total |
94.11 |
0.43 |
72.38 |
0.38 |
37.60 |
<0.05 |
|
Pediatric group (<18 yr) |
93.49 |
1.14 |
66.19 |
0.96 |
18.37 |
<0.05 |
|
Adult group (18–64 yr) |
141.96 |
0.98 |
107.57 |
0.85 |
26.57 |
<0.05 |
|
Older adult group (≥65 yr) |
56.45 |
0.77 |
46.87 |
0.69 |
9.29 |
<0.05 |
|
Mortality rate (per 100,000 ED visits) |
|
|
|
|
|
|
|
Total |
21.03 |
0.50 |
24.98 |
0.61 |
–5.01 |
<0.05 |
|
Pediatric group (<18 yr) |
4.36 |
0.55 |
5.34 |
0.69 |
–1.10 |
0.27 |
|
Adult group (18–64 yr) |
22.42 |
0.70 |
28.89 |
0.89 |
–5.74 |
<0.05 |
|
Older adult group (≥65 yr) |
15.64 |
1.08 |
16.70 |
1.14 |
–0.67 |
0.50 |
Table 3.Comparison of visit-based in-hospital mortality among patients with suicide attempt or self-harm in 2023 and 2024
Table 3.
|
Group |
Visit-based in-hospital mortality (%)
|
P-value |
|
2023 |
2024 |
|
Total |
3.70 |
4.72 |
<0.05 |
|
Pediatric group (<18 yr) |
1.04 |
1.42 |
0.104 |
|
Adult group (18–64 yr) |
2.87 |
3.89 |
<0.05 |
|
Older adult group (≥65 yr) |
12.13 |
12.55 |
0.540 |
REFERENCES
- 1. World Health Organization. Suicide worldwide in 2021: global health estimates. World Health Organization; 2025.
- 2. Bommersbach TJ, Olfson M, Rhee TG. National trends in emergency department visits for suicide attempts and intentional self-harm. Am J Psychiatry 2024;181:741-52.
- 3. Lee KS, Sung HK, Yoo SY, Min HS. Temporal trends and characteristics of adult patients in emergency department related to suicide attempt or self-harm in Korea, 2016-2020. J Korean Med Sci 2023;38:e40.
- 4. Kim H, Durey A, Kang S, et al. Factors affecting patients who attempted suicide in the emergency department due to the prolonged pandemic of COVID-19. Clin Exp Emerg Med 2023;10:418-25.
- 5. Yoon JH, Kwon IH, Park HW. The South Korean health-care system in crisis. Lancet 2024;403:2589.
- 6. Park HA, Kim T, Kim HJ, Han SH. Epidemiological trends in emergency department visits by age group: a report from the National Emergency Department Information System (NEDIS) of Korea, 2020-2024. Clin Exp Emerg Med 2025;12:405-13.
- 7. Lee J, Kim D, Lee WJ, Woo SH, Jeong S, Kim SH. Association of the COVID-19 pandemic and low-rescue suicide attempts in patients visiting the emergency department after attempting suicide. J Korean Med Sci 2021;36:e243.
- 8. Knipe D, Padmanathan P, Newton-Howes G, Chan LF, Kapur N. Suicide and self-harm. Lancet 2022;399:1903-16.
- 9. Hawton K, Bale L, Brand F, et al. Mortality in children and adolescents following presentation to hospital after non-fatal self-harm in the Multicentre Study of Self-harm: a prospective observational cohort study. Lancet Child Adolesc Health 2020;4:111-20.
- 10. Jung KY, Kim T, Cha WC. Self-harm characteristics of younger-old and older-old adults admitted to emergency departments: a nationwide study. Signa Vitae 2023;19:159-66.
- 11. Kim S, Kang H, Cho Y, et al. Emergency department utilization and risk factors for mortality in older patients: an analysis of Korean National Emergency Department Information System data. Clin Exp Emerg Med 2021;8:128-36.
- 12. Goldman-Mellor S, Olfson M, Lidon-Moyano C, Schoenbaum M. Association of suicide and other mortality with emergency department presentation. JAMA Netw Open 2019;2:e1917571.
- 13. Goldman-Mellor S, Olfson M, Schoenbaum M. Trends in mortality after emergency department presentation for suicidal behaviour in California. Inj Prev 2025;31:594-6.
- 14. Park J, Jung J, Seo B, et al. [Analysis and policy recommendations for problems in the emergency transport system]. Research Report No. 2024-17. Korean Medical Association Research Institute for Healthcare Policy; 2025.
