Factors associated with refusal of transport to hospital among patients experiencing suicidal crisis in South Korea

Abstract

Introduction

Preventing suicide requires identifying psychosocial characteristics and risk factors of people who do not go to the hospital. This study examined factors associated with refusal of transport to the hospital after reporting suicidal ideation or suicide attempt to emergency medical services.

Methods

This retrospective observational study involved reviewing emergency medical service dispatch records collected from January 2019 to December 2020 in Gyeonggi Province, South Korea on patients seeking emergency assistance for suicidal ideation or attempt. Univariate and multivariable regression analyses were employed to identify factors associated with refusal of transport to the hospital. Additionally, predictors of refusal stratified by suicidal ideation and attempt were examined.

Results

A total of 3449 cases were included in the analysis. Patients with suicidal ideation and no psychiatric or medical issues were more likely to refuse transport. In the suicide attempt group, certain patient characteristics were linked to a higher probability of refusal, including age 25 to 44 or 45 to 64, male gender, and/or not intoxicated by alcohol.

Conclusions

The pre-hospital phase is a critical link between patients and mental health services. Therefore, identifying the characteristics of patients who refuse transport to the hospital and developing intervention measures for them is essential.

Highlights

• •Suicidal ideation was associated with not being transported to a hospital.
• •Having a psychiatric history was associated with being transported to a hospital.
• •Suicide attempt was associated with being transported to a hospital.
• •The probability of being transported to a hospital varied based on suicide method.

1. Introduction

Besides taking the life of one person, suicide causes lasting suffering for those left behind [1]. According to a World Health Organization report, approximately 700,000 people worldwide commit suicide every year, making it the 17th leading cause of death in 2019 [2]. In the United States, suicide was ranked as the 11th leading cause of death in 2021, with a rate of 14.5 deaths per 100,000 population [3]. Suicide is also a serious problem in South Korea, where the suicide rate has been considerably higher than the rate in other countries for several years [4]. Suicide in South Korea was the 5th leading cause of death in 2020 [5], and the nation had the highest suicide rate among Organization for Economic Cooperation and Development countries from 2019 to 2021 [6].

Suicide attempts are an important risk factor for suicide [7]. Suicide attempts are repetitive, with individuals estimated to make up to 20 suicide attempts before committing suicide [8], and studies in Europe have shown that suicide attempt rates are 10–40 times higher than completed suicides [9]. Suicide attempters should be provided with psychosocial assessments to reduce the repetition of this behavior [10]. Several countries have strengthened mental health systems to support people at risk of suicide [11,12]. However, less than half of people at risk of suicide seek help, and only 40 % of adults aged 18 and older report seeking help [13]. Therefore, healthcare settings play an important role in identifying people at risk of suicide.

Emergency departments (ED) can be a starting point for managing people at risk for suicide. Registering and managing people who have attempted suicide in the case management service is known as an effective way to reduce suicide mortality [14]. In South Korea, case management services are offered to suicide attempters who visit the ED because it has been shown that a significant proportion of people who die by suicide visit the ED before they die [[15], [16], [17], [18]]. However, these services cannot be provided if the person does not arrive at the ED.

Suicidal crisis intervention should be actively implemented for not only patients transferred to medical institutions but also people who cannot be referred to medical institutions. Therefore, this study aimed to identify the characteristics of patients who presented to an EMS with suicidal ideation or attempt but were not transported to the hospital. In particular, we wanted to determine whether the type of suicidal behavior (suicidal ideation or attempt) is associated with transport refusal.

2. Materials and methods

2.1. Ethical approval statement

The Institutional Review Board of Hallym University Dongtan Sacred Heart Hospital approved the study protocol (IRB no. HDT 2023-02-013). Data were accessed on February 23, 2023 for research purposes. Patients’ personal information was not collected and informed consent was waived.

2.2. Study design and setting

This retrospective observational study was conducted based on EMS records in Gyeonggi province, South Korea. Gyeonggi province has an area of 10,172 km² and a population of 13,500,000, as estimated in 2021. The EMS system in Gyeonggi is government-based and available to anyone for free. As of 2021, the Gyeonggi EMS system consisted of two headquarters, 35 fire stations with 263 ambulances, and 1912 emergency medical technicians (EMTs) [19].

EMTs prioritize treatment of accompanying injuries among patients with psychiatric emergencies, including suicide [20]. In addition, it is mandatory to evaluate the patient's history of mental illness, whether they are causing harm to themselves or those around them, and whether they are intoxicated by alcohol or drugs and must be transferred to a medical institution. If transport is not possible despite attempting all feasible means, such as patient stabilization, physical restraint, and connection to mental health crisis intervention services, transport may be denied under direct medical supervision.

2.3. Study population

This study analyzed EMS pre-hospital dispatch records collected from January 2019 to December 2020. We included patients aged 10 years and older who sought help from EMS for suicidal ideation or attempt. We excluded patients with non-suicidal self-injury [21]. Patients with a pain response or those who were unconscious were also excluded from the study due to their inability to express a refusal of transfer [22]. Patients with missing variables, such as age, gender, suicidal behaviors, time of EMS call, alcohol consumption, and location were also excluded.

2.4. Variables

As self-report measures of suicidal ideation/attempt have been shown to be a reliable source of data, we categorized suicide ideation and attempts through patient self-report [23]. Referring to studies that analyzed the characteristics of patients who declined participation in suicide prevention treatment, we identified the following variables and potential risk factors: demographic characteristics, time-related variables, and factors related to presenting suicidal behavior [[24], [25], [26], [27], [28]]. Demographic characteristics included age group (categorized as 10–24, 25–44, 45–64, 65 and older), gender, and nationality (Korean, non-Korean). Time-related variables included the COVID-19 period, season, and EMS call times, considering their potential impact on transportation outcomes. Factors associated with suicidal behavior included examining the individual's mental state at the scene, the location, and the plan or method of suicide. Suicide methods were categorized as hanging, gas inhalation, jumping, drug overdose, cutting, ingestion of pesticides/caustics, and others [28]. If more than one method was used, we categorized them into the “two or more” category.

The outcome variable was refusal of transport to a hospital. We analyzed factors associated with patients who refused transport and were ultimately not transported to the hospital.

2.5. Statistical analysis

Patient demographics and clinical characteristics were compared between the transported group and the group that refused transport. Categorical variables in each group were compared using the chi-squared test or Fisher's exact test. Univariate and multivariable regression analyses were performed to explore the association of possible risk factors with refusal of transport. A multivariable regression model examined the association between transport refusal and suicidal behavior, adjusting for age, gender, COVID-19 period, season, time of report, past medical history, psychiatric history, and alcohol consumption. Moreover, we also evaluated risk factors, stratified by suicide incidents and attempts. Statistical analysis was performed using R ver. 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). P values were based on a significance level of 0.05.

3. Results

There were 5014 suicide-related dispatches analyzed during the study period. We excluded 414 cases of self-harm and 1069 dispatches due to cardiac arrest or altered mental status at the scene. We also excluded 82 patients with missing variables. The final analysis included 3449 patients (Fig. 1).

Fig. 1.

Study flowchart. EMS = emergency medical service.

3.1. Comparison of characteristics between transported and not transported groups

The characteristics of patients who were transported (“no refusal”) and not transported to the hospital (“refusal of transport”) are presented in Table 1. Approximately 30 % of patients in the study refused to be transported to the hospital. No significant differences were observed between the two groups based on age or nationality. More patients refused transport during COVID-19 and during nighttime hours. By contrast, the refusal group was less likely to have medical comorbidities or psychiatric diseases. In addition, the proportion of suicidal ideation was significantly higher in the refusal group.

Table 1.

Baseline characteristics of the patients.

Variable	Total	No refusal	Refusal of transport	p
	N = 3449	n = 2400	n = 1049
Age group (years)				0.08
10–24	754 (21.9)	543 (22.6)	211 (20.1)
25–44	1430 (41.5)	963 (40.1)	467 (44.5)
45–64	1102 (32.0)	775 (32.3)	327 (31.2)
≥65	163 (4.7)	119 (5.0)	44 (4.2)
Gender				0.001
Female	1672 (48.5)	1210 (50.4)	462 (44.0)
Male	1777 (51.5)	1190 (49.6)	587 (56.0)
Non-Korean	51 (1.5)	39 (1.6)	12 (1.1)	0.36
With medical comorbidity	502 (14.6)	432 (18.0)	70 (6.7)	<0.001
With psychiatric disease	438 (12.7)	370 (15.4)	68 (6.5)	<0.001
Alcohol consumption	1205 (34.9)	887 (37.0)	318 (30.3)	<0.001
COVID-19 period	1335 (38.7)	862 (35.9)	473 (45.1)	<0.001
Season				0.09
Spring	797 (23.1)	569 (23.7)	228 (21.7)
Summer	901 (26.1)	622 (25.9)	279 (26.6)
Fall	917 (26.6)	612 (25.5)	305 (29.1)
Winter	834 (24.2)	597 (24.9)	237 (22.6)
Call time				0.04
6 a.m. to 6 p.m.	1462 (42.4)	1046 (43.6)	416 (39.7)
6 p.m. to 6 a.m.	1987 (57.6)	1354 (56.4)	633 (60.3)
Suicidal behavior				<0.001
Ideation	793 (23.0)	441 (18.4)	352 (33.6)
Attempt	2656 (77.0)	1959 (81.6)	697 (66.4)

Note. COVID-19 period: March 1, 2020 to December 31, 2020; Spring: March–May; Summer: June–August; Fall: September–November; Winter: December–February. The values presented are number (%) and were compared using the chi-squared test or Fisher's exact test.

The clinical characteristics of patients by suicidal ideation or attempts are presented in Table 2. In the case of suicidal ideation, a significant difference was observed between the no refusal group and the refusal group in the planning of suicide. Furthermore, for patients who attempted suicide, the methods of hanging, jumping, cutting, and others were more prevalent in the refusal group.

Table 2.

Clinical characteristics by suicide behavior.

Variable	Total	No refusal	Refusal of transport	p
Suicidal ideation	N = 793	n = 441	n = 352
Mental status				0.58
Alert	791 (99.7)	439 (99.5)	352 (100.0)
Verbal	2 (0.3)	2 (0.5)	0 (0.0)
Location of call				0.54
Home	586 (73.9)	333 (75.5)	253 (71.9)
Group living	21 (2.6)	9 (2.0)	12 (3.4)
Public place	87 (11.0)	46 (10.4)	41 (11.6)
Outdoors	99 (12.5)	53 (12.0)	46 (13.1)
With suicide plan	331 (41.7)	157 (35.6)	174 (49.4)	<0.001
Suicide attempt	N = 2656	n = 1959	n = 697
Mental status				<0.001
Alert	2358 (88.8)	1670 (85.2)	688 (98.7)
Verbal	298 (11.2)	289 (14.8)	9 (1.3)
Location of call				0.95
Home	1982 (74.6)	1463 (74.7)	519 (74.5)
Group living	92 (3.5)	70 (3.6)	22 (3.2)
Public place	232 (8.7)	170 (8.7)	62 (8.9)
Outdoors	350 (13.2)	256 (13.1)	94 (13.5)
Suicide methods				<0.001
Hanging	348 (13.1)	211 (10.8)	137 (19.7)
Gas inhalation	400 (15.1)	301 (15.4)	99 (14.2)
Jumping	271 (10.2)	150 (7.7)	121 (17.4)
Drug overdose	763 (28.7)	646 (33.0)	117 (16.8)
Cutting	359 (13.5)	249 (12.7)	110 (15.8)
Pesticide/caustic	208 (7.8)	185 (9.4)	23 (3.3)
Two or more	125 (4.7)	116 (5.9)	9 (1.3)
Others	182 (6.9)	101 (5.2)	81 (11.6)

The values presented are number (%) and were compared using the chi-squared test or Fisher's exact test.

3.2. Results of univariate and multivariable logistic regression for refusal of transport

Table 3 shows the association between potential predictors of refusal of transport to the hospital. The results showed that refusal of transport was significantly associated with suicidal ideation compared to suicide attempt. In addition, patients were also more likely to refuse if they were young adults or men. However, patients with a previous medical history (adjusted odds ratio [aOR], 0.31; 95 % confidence interval [CI], 0.23–0.41), history of psychiatric illness (aOR, 0.31; 95 % CI, 0.18–0.32), or current alcohol consumption (aOR, 0.70; 95 % CI, 0.59–0.83) were less likely to refuse transport.

Table 3.

Logistic regression analysis for refusal of transport to the hospital.

Variable	Univariate		Multivariable
	cOR (95 % CI)	p	aOR (95 % CI)	p
Suicidal behavior
Ideation	Reference		Reference
Attempt	0.45 (0.38─0.53)	<0.001	0.31 (0.26─0.38)	<0.001
Age group (years)
10–24	Reference		Reference
25–44	1.25 (1.03─1.52)	0.03	1.26 (1.03─1.55)	0.03
45–64	1.09 (0.89─1.33)	0.43	1.18 (0.94─1.48)	0.16
≥65	0.95 (0.65─1.38)	0.80	1.39 (0.91─2.12)	0.12
Gender (male vs. female)	1.29 (1.12─1.50)	<0.001	1.27 (1.08─1.48)	0.003
COVID-19 period	1.47 (1.26─1.70)	<0.001	1.48 (1.27─1.73)	<0.001
Season
Spring	Reference		Reference
Summer	1.12 (0.91─1.38)	0.29	1.12 (0.90─1.39)	0.33
Fall	1.24 (1.01─1.53)	0.04	1.30 (1.05─1.62)	0.02
Winter	0.99 (0.80─1.23)	0.93	1.12 (0.88─1.41)	0.36
Call time
6 a.m. to 6 p.m.	Reference		Reference
6 p.m. to 6 a.m.	1.18 (1.01─1.36)	0.03	1.25 (1.07─1.47)	0.01
Medical comorbidity (yes vs. no)	0.33 (0.25─0.42)	<0.001	0.31 (0.23─0.41)	<0.001
Psychiatric disease (yes vs. no)	0.38 (0.29─0.49)	<0.001	0.24 (0.18─0.32)	<0.001
Alcohol consumption (yes vs. no)	0.74 (0.63─0.87)	<0.001	0.70 (0.59─0.83)	<0.001

Note. cOR, crude odds ratio; aOR, adjusted odds ratio; CI, confidence interval; COVID-19 period: March 1, 2020 to December 31, 2020; Spring: March–May; Summer: June–August; Fall: September–November; Winter: December–February.

3.3. Stratification by suicidal ideation or suicide attempt

Fig. 2 shows the association between potential predictors of refusal of transport stratified by patients who attempted suicide versus those who thought about it. No significant association was found for age group or gender in the suicidal ideation group, but in the suicide attempt group, individuals aged 25–44 or 45–64 and men were more likely to refuse transport. In both groups, the likelihood of refusal of transport increased during the COVID-19 period, but a particularly strong association was observed in the suicidal ideation group. Moreover, in the suicide attempt group, no seasonal association was found, but nighttime hours were a risk factor for transport refusal.

Fig. 2.

Association between potential predictors of refusal of transport stratified by suicidal ideation and suicide attempt. COVID-19 period: March 1, 2020 to December 31, 2020. Spring: March–May. Summer: June–August. Fall: September–November. Winter: December–February. aOR = adjusted odds ratio; CI = confidence interval.

4. Discussion

This study found that about one-third of cases reported to EMS for suicide attempts or suicidal ideation were not transported to the hospital. Patients were more likely to refuse transport when they had no prior medical history or psychiatric disease. Whether a person had attempted suicide or only thought about it was a significant predictor of refusal of transport. Furthermore, the predictors of refusal of transport had different associations depending on whether the person had attempted suicide.

Health care workers in clinical settings, such as hospitals, clinics, and doctor's offices, often encounter patients' refusal of treatment after a suicide attempt. Similar situations also arise in the pre-hospital setting [29]. Although there is substantial research on the proportion of patients who refuse treatment and are voluntarily discharged against medical advice from the ED, relatively little research is available on the proportion of patients not transported to the hospital in the pre-hospital phase [30,31]. The results of this study showed that a significant number of patients do not transferred to the hospital after suicidal ideation and suicide attempts; therefore, it is important to identify the characteristics of patients who refuse to be transported to the hospital in order to provide mental health services to patients with suicidal crises and reduce suicides.

This study identified factors associated with patients with a history of suicidal ideation or suicide attempts refusing to be transported to the hospital. Suicidal ideation was associated with fewer hospital transports compared to suicide attempts. The study also found that patients with lower clinical severity were less likely to be transported to the hospital by EMS, in particular, the effect size was larger for the absence of psychiatric disorders in the suicidal ideation group than for the absence of psychiatric disorders in the suicide attempt group. Similarly, previous research has shown that patients with less serious attempts and fewer medical and mental issues were more likely to decline suicide prevention case management programs after a suicide attempt [26]. However, suicidal ideation is a risk factor for suicide attempts and requires an in-depth clinical assessment, as do suicide attempters [32,33]. In addition, according to previous studies, not only do people without mental illness attempt suicide, but more than 50 % of people who die by suicide do not have a mental illness [34,35]. Therefore, EMTs should not accept a patient's refusal of transport based on the clinical characteristics of being a low-risk patient.

The role of EMTs in helping patients experiencing suicidal crisis access mental health care services is pivotal [29]. This role becomes even more important when considering that EMS could be the only contact patients have with the healthcare system before attempting suicide [36]. EMTs typically perceive life-threatening medical crises as emergencies and tend to view mental illness as secondary to the assessment and treatment of medical disorders [37,38]. This view of suicidal behavior by EMTs in emergency care can be fatal [39]. Furthermore, their limited understanding of mental illness underscores the necessity for ongoing competency development [39]. Therefore, there is a need to enhance training programs for EMTs to effectively intervene in suicidal crises. Such training should emphasize the recognition that every suicidal behavior is an emergency and must be taken seriously. Continuous training will further enhance their skills in assessing and managing patients experiencing psychiatric crises.

Intervention in suicide crisis should begin at the pre-hospital level, with the essential involvement of EMS. Previous research indicates that one-third of psychiatric emergency patients and approximately 44 % of suicidal patients arrive at the ED via ambulance [[40], [41], [42]]. The number of suicidal crisis patients encountered by EMTs in the field is even higher, including those not transported to the hospital, as observed in this study. Despite the potential importance of EMS in the continuum of care for suicidal patients, pre-hospital protocols often prioritize scene safety, hospital transport, and conditions necessitating restraints or involuntary hospitalization [43]. DeCou et al.'s study found that only half of the EMS protocols addressing suicidal or self-harming patients included suicide risk assessment, and even fewer mentioned alternatives to hospital-based care [43]. Therefore, there is a critical need to develop comprehensive suicide management protocols for EMS, integrating evidence-based suicide screening and assessment tools [44]. Support should also be provided to those who decline hospital transport by connecting them to community resources such as online mental health providers or on-call psychiatric crisis lines [43]. This ensures that patients at risk of suicide can access assistance even without visiting the hospital.

This study has the advantage of analyzing characteristics associated with patients in suicide crisis who were not transported to the hospital. However, retrospective analysis of EMS dispatch records introduced several limitations. First, due to the nature of the study involving previously collected data, 82 cases with unknown variables were excluded, but they accounted for only 2 % of the total 3449 cases included in the final analysis. Additionally, among the excluded cases, 34 patients were transported to the hospital while 48 were not, suggesting no significant systematic bias based on transport status. Second, information, such as socioeconomic factors, family history, previous suicide attempts, and the presence of stressful situations, was not included in the analysis because it could not be collected in the field. Third, clinical severity assessments of the patient's suicidal behavior were lacking as pre-hospital data were utilized. Lastly, EMS records are case-based rather than patient-based, which makes it impossible to track repeated calls from the same individual.

5. Conclusion

This study found that a notable proportion of patients in suicidal crisis refuse hospital transport, with a higher likelihood of refusal among those expressing suicidal ideation or without comorbidities. All suicidal behaviors are risk factors for suicide and should be thoroughly evaluated in the pre-hospital setting. Moreover, given the critical role of EMS in linking patients to mental health services, it is essential to establish protocols for referring individuals to crisis intervention.

Human or animals rights

Not Applicable.

Informed consent and patient details

The Institutional Review Board of Hallym University Dongtan Sacred Heart Hospital approved the study protocol (IRB no. HDT 2023-02-013). Data were accessed on February 23, 2023 for research purposes. Patients’ personal information was not collected and informed consent was waived.

Data availability statement

The data will be made available on request.

Funding

This research did not receive any specific funding.

CRediT authorship contribution statement

Hang A. Park: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Conceptualization. Choung Ah Lee: Writing – review & editing, Supervision, Methodology.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.