A prolonged delay from the stroke onset to hospital arrival decreases the efficacy of time-dependent treatments and results in a greater frequency of stroke-related complications or in-hospital deaths [1, 2, 3]. The prolonged delay is classified into a prehospital delay from symptom onset to arrival at the emergency department (ED) and an in-hospital delay between the patient’s arrival at the hospital and treatment initiation [1, 4]. A prehospital delay (89.2%) constitutes the most significant proportion of total delay time compared to the in-hospital delay [5], despite a 6.0% annual decline of prehospital delay in hours per year among the 123 studies between 1981 and 2007 [2]. Awareness among the patient or bystander that the initial symptom was stroke-related and transportation using an emergency medical service (EMS) were associated with an early arrival to the hospital [6, 7, 8], while living alone or being alone at the onset of symptoms and referral from other hospitals are commonly linked to delayed arrival [7, 8, 9, 10].

After launching the acute stroke quality assessment program as an external audit by the Health Insurance Review and Assessment Service beginning in 2007, defect-free stroke care quality has improved in Korea [11]. Korean patients can choose their health care providers and institutions because of the low barrier to medical service use. Among acute stroke patients, about half of them are directed to primary care clinics or other healthcare professionals without using EMS [12]. Knowing the trends in prehospital visits using a representative and extensive database is necessary to aid in policy-making to reduce prehospital delays. Several studies have been conducted via a patient survey or medical records review [13, 14, 15], but did not directly examine annual trends.

Using a national database, this study aimed to evaluate trends and characteristics in the prehospital visit of newly-diagnosed stroke patients using a national database.

In this Korean study of claims data, 21.3% of acute stroke patients were found to have visited one or more hospitals before diagnosis. The annual rates of prehospital visits decreased over time. However, more than 10,000 patients yearly still visited a different medical institution to treat their acute symptoms, and the number of patients with prehospital visits did not reduce because of increased stroke incidence.

A systematic review of 123 papers published between 1981 and 2007 reported a wide range in prehospital delay times (mean hours, 1.2–98.8 hours; arrivals within three hours, 6%–92%) [2]. In the Report of Assessment for the Quality of Acute Stroke Care in Korea 2018, which analyzed data from stroke patients defined by ICD-10 codes, the median interval from onset to ED admission was 214 minutes (interquartile range [IQR], 65–844 minutes) [17]. According to EMS usage, the prehospital delay time showed a significant gap, i.e., 118.5 (IQR, 49–382) minutes for EMS users vs. 525 minutes (IQR, 163–1507.5) minutes for non–EMS users [17]. Although we could not measure the time from symptom onset to the ED admission in this study, a delay in arrivals could stem from not contacting the EMS, considering the rate of patients who visited more than one day before final admission for treatment (20.41%) in 2018 according to previous research [18].

Several studies of prehospital visits among stroke patients have focused on patient-related factors (e.g., recognized symptoms, awareness brought on by public education or campaigns, or prior experiences) [18, 19], system-related factors (i.e., triage by EMS or accessibility of EMS services, including due to geographical [e.g., distance from a stroke care units], temporal [e.g., traffic in rush hour], or socioeconomic factors [e.g., level of insurance coverage]) [8, 15, 20, 21]. Prehospital delays are multifactorial, triggered by both patient- and system-related factors [20]. This study added evidence of system-related factors, suggesting the cause of delay was due to visiting at least one hospital before ED admission.

Fortunately, the rate of prehospital visits has decreased over time. This gradual decline every year may result from public health education to increase stroke awareness by highlighting common symptoms and the time-critical nature of the disease [7]. The implementation of public education focused on teaching about stroke symptoms to facilitate the seeking of emergency care where appropriate has been recommended by the American Heart Association/American Stroke Association since 2013 [22]. About 69% of urban people aged at least 50 years old and 79% of people voluntarily participating in a nationwide stroke educational program performed by the Korean Stroke Society responded that they would call EMS to visit early as soon as possible [23, 24]. The Korean general population’s early stroke symptom recognition rate improved from 51.2% in 2017 to 61.7% in 2019. However, it decreased recently (54.2% in 2021) [25]. This study did not directly compare with the pre-and post-period of the COVID-19 outbreak. When the recent trends of prehospital visits were evaluated, we need to consider various factors related to the COVID-19 outbreak, such as changes in stroke incidence, awareness of disease, or EMS use [26]. During the COVID-19 outbreak, the proportion of direct visits to the ED of acute ischemic stroke patients through public EMS services increased. However, their prognosis was worse than that of pre-COVID-19 patients with delays at the prehospital delay [26]. Even with improvements in the public’s knowledge of the definition of stroke and its warning signs, there are several reasons that the rate of prehospital visits remains high. First, a relatively low barrier to medical use might have influenced the rate of prehospital visits to Korean medical services. Patients unable to identify their symptoms might contact their local or community doctor first rather than go directly to the ED [27]. Second, 59.2% of patients used EMS, according to the Report from the Epidemiology Research Council of the Korean Stroke Society, which was lower than the rates of other Organization for Economic Co-operation and Development countries (58%–79%) [28]. Third, the type of symptom(s) being experienced might influence the prehospital delay. For example, vague symptoms, such as sensory disturbances, vertigo, or dizziness, were associated with a prehospital delay, while the presence of more severe symptoms, such as confusion, decreased consciousness, or aphasia, was associated with early arrival [19, 29]. The American Heart Association “Face, Arm, Speech, Time to Call 911” campaign is based on recognizing weakness and aphasia or dysarthria [30]. The use of EMS may be affected by the nature of the symptoms recognized; in other words, there may be more significant EMS usage among patients with severe symptoms, such as weakness, aphasia, or an altered level of consciousness [31]. Although we did not survey the subjects’ initial symptoms, the high prehospital visit rate in young participants can be explained by an inability to predict stroke with no severe symptoms or to select an appropriate hospital.

It is interpreted that patients with higher income status showed a greater rate of prehospital visits in 2010, which can be explained that medical expenses incurred by prehospital visits are not a problem for high-income individuals [32]. On the other hand, low income was associated with a significant increase in prehospital delays in Nepal, which has not yet established a public health insurance system [33]. Previous research found no association between income status and prehospital delay in the middle- and high-income countries [6, 14]. Therefore, the unique association between income status and the rate of prehospital visits in Korea may have had an impact. A difference in the residential area may be associated with the rate of prehospital visits because patients with residences in low urbanization areas contacted the EMS less frequently (p = 0.009) and more frequently made prehospital visits (p = 0.020) without a different awareness of stroke or distance to the hospital [21]. We found a reversal in the trend of prehospital visits since 2016 according to sex. The reversed proportion of prehospital visits in female patients may have been due to the higher participation in public education or campaigns than male patients [23].

Prehospital visits in Korean stroke patients decreased from 25.1% in 2010 to 17.8% in 2019. However, more than 10,000 patients still visited other medical institutions before admission to treatment. Targeted policies to control prehospital delay by reducing the rate of prehospital visits among young patients and those who reside in low urban areas is needed.

ED, the emergency department; EMS, emergency medical service; NHID, National Health Information Database; NHIS, National Health Insurance Service; ICD-10, International Classification of Diseases, 10th revision; CT, computer tomography; MRI, magnetic resonance image.

Raw data were generated at the National Health Information Database from the National Health Insurance Service in Korea. Derived data supporting the findings of this study are available from the corresponding author on request.

These should be presented as follows: JS, HK, JL, JM, and YK designed the research study. JS, HK, HJJ, JL, JM, and YK performed the research. KPK provided help and advice on methodology. HJJ analyzed the data. JS, HK, JL, JM, KPK, and YK wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

The present study was conducted following the principles of the Declaration of Helsinki. This study was approved by the institutional review board of the Clinical Research Ethics Committee of Konkuk University, Seoul, Korea (approval no. 7001355-202105-E-140), which also waived the requirement for informed consent because this study did not include identifying individuals and employed previously collected data gathered from the general public. It was also approved in a review by the internal committee of NHIS (NHIS study no. NHIS-2021-1-479).

We thanks to the National Health Insurance Sharing Service for the data.

This research received no external funding.

The authors declare no conflict of interest.