Analysis of onset-to-door time and its influencing factors in Chinese patients with acute ischemic stroke during the 2020 COVID-19 epidemic: a preliminary, prospective, multicenter study

This study was a multicenter, large-sample, prospective, and observational study.

A total of 3495 patients with AIS were recruited from nine hospitals (including Fushun Central Hospital, Wuzhou Workers’ Hospital, Huaihua First People’s Hospital, Inner Mongolia Autonomous Region People’s Hospital, the First Affiliated Hospital of Shaoyang Medical College, Xiangxi Tujia and Miao Autonomous Prefecture People’s Hospital, Affiliated Hospital of Yan’an University, Yueyang Central Hospital, and Zhuzhou Central Hospital) certified as “stroke centers” [20] by the China National Stroke Prevention Project Committee Commission from January to June 2022.

The data collection and entry personnel in all subcenters had professional knowledge of stroke and were trained by the project manager. AIS was diagnosed according to the guidelines [21], and intracranial hemorrhage was excluded using head computed tomography or magnetic resonance imaging [22]. The stroke subtype was based on the trial of Org 10,172 in acute stroke treatment (TOAST) classification of stroke [23].

The variables studied included sex, age, educational level, residence status, medical insurance, wake-up stroke, first symptom of AIS, distance between onset location and first-visit hospital, transfer method for patients, whether an inter-hospital transfer was performed, medical history, pre-onset modified Rankin scale (mRS) score, stroke severity (according to the National Institutes of Health Stroke Scales [NHISS] first score after onset, moderate and severe stroke have NHISS score 5–14 and NHISS score 15–42 respectively) [24], patient’s knowledge about AIS, and TOAST classification. The division of patients based on ODT is crucial for stratifying stroke care and predicting outcomes [25]. Therefore, patients with AIS were categorized into the ODT ≤ 3 h group and ODT > 3 h group accordingly.

In this section, we outline a comprehensive understanding of various symptoms encountered in AIS cases, ranging from common manifestations, such as vomiting or unconsciousness, to more specific indicators, including diplopia or dysarthria. The initial symptoms of AIS are defined as follows [26, 27]: (1) Vomiting: Involuntary expulsion of stomach contents through the mouth or nasal cavity. (2) Unconsciousness: Lack of response to external stimuli, coma, or other non-alert states. (3) Paralysis: Complete loss of voluntary motor function, which may affect specific body parts or one side. (4) Diplopia: Simultaneous perception of two images of the same object. (5) Aphasia: Loss or impairment of the ability to express or understand language, characterized by difficulties in speaking, expressing oneself, or understanding others. (6) Dysarthria: Unclear speech or difficulty in pronouncing words due to impaired neuromuscular control. (7) Drooping of the angle of the mouth: Noticeable drooping of one side of the mouth corner, resulting in an asymmetrical facial expression. (8) Headache: Persistent pain or discomfort experienced in the head. (9) Paresthesia: Abnormal sensations felt on the skin, such as numbness, tingling, or burning, without obvious stimulation. (10) Vertigo: Sensation of spinning or movement of the surrounding environment or oneself, often accompanied by balance disorders. 11) Other symptoms included visual disturbances that are difficulty to classify within the categories mentioned above. Detailed symptom information can be provided upon entry of specific data.

In this study, we typically documented the precise time when the patient or a witness first noticed stroke symptoms, such as sudden weakness, speech difficulties, or visual disturbances. Alternatively, when the onset time of symptoms was unclear, the following methods were used: (a) When the patient woke up with symptoms, the time before sleep when the last symptom-free period was confirmed was considered as the onset time of symptoms [28, 29]. (b) When the exact time of symptom onset cannot be determined, the time of the last confirmed symptom-free period was considered as the onset time of symptoms [28, 29]. Subsequently, the time of the patient’s arrival at the outpatient department or emergency room, specifically at the triage entrance [30], was documented. The ODT was calculated by subtracting the recorded onset time of stroke symptoms from the time of arrival at the healthcare facility’s door. For instance: (1) Unconsciousness: If the patient lost consciousness, the time when symptoms started was determined based on witness accounts or when the patient was found. This time was considered as the onset time, and then the time when the patient arrived at the hospital was recorded to calculate ODT. (2) Headaches: For localized headaches, the time when the patient or witnesses noticed the headache starting was considered the onset time. Following this, the time of arrival at the hospital was recorded to calculate ODT.

All statistical analyses were performed using IBM SPSS Statistics (version 26.0; IBM Corp., Armonk, N.Y., USA). The measured data with normal distribution are expressed as the mean ± standard deviation, and the independent-sample t-test was used for between-group comparisons. When data does not follow a normal distribution, quartiles are used to describe the data. Categorical variables are presented as counts and percentages, and the differences between the two groups were analyzed using the chi-squared test. Firstly, the differences between the ODT ≤ 3 h group and the ODT > 3 h group were analyzed using single-factor analysis; subsequently, the variables with significant differences were included in the binary logistic multivariate analysis, which typically yields confidence intervals for parameter estimates and conducts multicollinearity tests on the variables within the multivariable model. All statistical tests were two-sided, and the threshold for statistical significance was set at P < 0.05.

A total of 3495 patients with AIS were recruited, and the average is 1,698.88 min (range, 9–10,062 min), and the median is 883 min (Q1 = 181 min, Q3 = 2555 min). There were 763 patients (21.83%) with ODT ≤ 3 h and 2732 patients (78.17%) with ODT > 3 h. Specifically, the number of patients with ODT 3–6 h, 6–12 h, 12–24 h, 24–72 h, and > 72 h accounted for 12.26%, 13.18%, 19.02%, 23.10%, and 10.61% of all patients with AIS, respectively (Fig. 1). There were 2,317 patients (66.29%) with ODT ≤ 24 h and 1178 patients (33.71%) with ODT >24 h.

Living in rural areas (OR: 1.478, 95% CI: 1.024–2.146) and existing interhospital transfer (OR: 7.479, 95% CI: 2.548–32.337) were risk factors for ODT (Table 2).

Distance between the onset location and first-visit hospital ≤ 20 km (OR: 0.355, 95% CI: 0.236–0.530), transportation of patients by EMSs (OR: 0.346, 95% CI: 0.216–0.555), history of atrial fibrillation (OR: 0.375, 95% CI: 0.207–0.679), moderate stroke (OR: 0.644, 95% CI: 0.462–0.901), and severe stroke (OR: 0.506, 95% CI: 0.285–0.908) were protective factors for ODT.

This study showed that the median of ODT was 852 min (range, 215–2459 min), and 21.83% of patients had ODT ≤ 3 h. A multicenter study in the United States showed that 21–40% of patients with AIS reach the hospital within 3 h of symptom onset [31]. In a 2006 study that included 62 subcenters in China that showed similar results, the median ODT was 15 h [13]. Our findings reveal that ODT has not shown much reduction after more than 10 years and is still 3–6 h longer than that in developed countries [30]. Furthermore, a study in 2012–2013 indicated that patients in China experienced more pre-hospital delays compared to those in the United States (1318 min vs. 644 min) [14]. There is a significant difference between ODTs in China and those in developed countries [9, 15, 32, 33].

The World Health Organization’s MONICA manual provides standardized guidelines for registering stroke events [26]. These guidelines ensure that stroke cases are consistently defined and registered, facilitating accurate comparisons across different populations and regions. The MONICA project has played a crucial role in standardizing the registration of acute stroke events, enabling uniform data collection and analysis [26]. Research has demonstrated that adherence to the MONICA criteria for stroke registration is essential for quality control and accurate event validation [34]. The protocols established by the MONICA project have been widely adopted in various studies for registering stroke events, highlighting the broad acceptance and utility of these guidelines [35]. The standardized approach to stroke event registration outlined in the MONICA manual is critical for ensuring the accuracy and consistency of data collected across diverse populations and periods.

Over time, the accuracy of patients’ and witnesses’ recollection of the onset time may diminish, posing a challenge in determining the ODT accurately. We have therefore implemented the following measures to address recall bias in our study design: utilizing standardized questionnaires and interview methods. Additionally, we observed that most patients experience an ODT of less than 1 day. Therefore, we argue that including all patients in the primary analysis, even those with an onset of illness exceeding 24 h, can offer a more comprehensive depiction of the actual situation.

This study showed that living in rural areas was a risk factor for ODT. Only 8.18% of patients in rural areas in China reached the hospital within 3 h [36], while 45.8% of patients in urban areas reached the hospital within 3 h [37]. Compared with patients in urban areas, those in rural areas are typically older and have lower levels of education, poor housing conditions, and high poverty rates.

China’s economic and healthcare service development has been uneven. Medical and health services supply in China has obvious differences in spatial distribution [19], and the eastern region has the highest medical and health services supply level, followed by the western and central regions. Patients in some parts of China experienced pre-hospital delays owing to poor economic and sanitary conditions [3]. In rural areas, insufficient medical resources, low levels of medical care, and fewer medical staff members make it extremely difficult to meet the needs of patients with stroke. In addition, rural residents have limited access to medical knowledge about first aid; this often results in patients missing the optimal stroke treatment time [38]. The coverage and reimbursement rates of medical insurance in rural areas are lower than those in urban areas, and the frequency of rural patients visiting hospitals is also low [39, 40], which may also result in longer ODTs in rural areas compared with in urban areas.

Our study showed that the distance between the onset location and the initial hospital ≤ 20 km was associated with shorter ODT. Long distances are an important factor delaying patient transport. Improving transport efficiency is a solution that the EMS plays a crucial role in achieving. EMS most closely affects ODT [6]. When there is an optimal EMS, the median ODT can be reduced to 151 min, and the proportion of patients reaching the hospital within 3 h can be increased to 54% [6]. However, EMS usage adds to medical costs; therefore, EMS construction is not feasible in some areas. The awareness of patients regarding EMS usage is also relatively low, and the proportion of patients with AIS using it in China is extremely low, as shown in this study; our results are also consistent with the findings of Wang et al. [14].

Our findings showed that patients with moderate or severe stroke were more likely to reach the stroke center within 3 h after onset. Similar results have been reported by Iversen et al. [41]. Patients with moderate or severe stroke were more likely to arrive at the hospital promptly and receive reperfusion therapy. The more serious the stroke, the more it is likely to attract patients’ and bystanders’ attention; this was associated with a higher probability of using EMS. Our research suggests that patients with mild strokes often experience more delays, which can be attributed to several factors: (1) Atypical Symptoms: Mild strokes may manifest with subtle or non-specific symptoms that patients may not immediately recognize as indicative of a stroke. (2) Minimization of symptoms: Patients with mild strokes may diminish the severity of their symptoms or attribute them to other less serious conditions, delaying their decision to seek medical attention. (3) Fear or denial: Some patients may experience fear or denial about the possibility of experiencing a stroke. This psychological barrier can prevent them from promptly seeking medical care. (4) Neglect: Patients with mild strokes may perceive their symptoms as less urgent and may prioritize other obligations over seeking immediate medical attention. Thus, we emphasize the importance of concentrating on patients with mild stroke and the significance of timely referrals.

Bystanders are more likely to notice typical stroke symptoms such as limb weakness, speech disturbance, and walking difficulties [42]. It has been reported that living alone increases admission delay, and the recognition of symptoms by bystanders may shorten it [43]. The onset of symptoms can influence a patient’s decision-making. When dysarthria or decreased muscle strength were the first symptoms, the rate of hospital visits increased significantly within 4.5 h (P < 0.01) [6, 44, 45]. The more prominent the impact of the first symptom on daily living, the easier it is to attract the attention of patients and their families, the stronger the desire to seek medical attention, and the shorter the ODT. Only 53.8% of patients with posterior circulation stroke reach the hospital within 3 h, compared to 68.4% of patients with anterior circulation stroke [7]. Compared with dysphagia and limb weakness caused by posterior circulation stroke, posterior circulation stroke often presents with non-specific symptoms such as dizziness, vertigo, and nausea, which are easily attributed to poor rest, anxiety, and failure. Cryptogenic stroke is common in young people [46]. However, young patients often ignore the possibility of stroke onset, which leads to a pre-hospital delay. Patients often choose self-observation when stroke occurs and only visit the hospital if the symptoms persist or worsen because of the inability to identify stroke in an accurate and timely manner [13]. Recognizing symptoms of stroke is an independent factor associated with early arrival [47].

Most patients in our study chose to go to the hospital by themselves, which increased the probability of inter-hospital transfer and caused pre-hospital delays. There are two main specific situations of interhospital transfer: (1) Patients who independently seek medical attention may arrive at a hospital without a stroke center, necessitating their transfer to one. (2) In our study, patients who utilized an ambulance were directly transported to a stroke center. However, interhospital transfers may occur for these patients if the initial hospital cannot administer mechanical thrombectomy treatment. Only one in eight patients with stroke in China arrived at the hospital via EMS [48], compared to 59.6% in the DASH II study [49]. Our study showed that patients who visited the hospital via EMSs (ambulances) had shorter ODTs. Moreover, many studies have demonstrated a reduction in pre-hospital delays via EMS [50, 51]. The 2019 AHA/ASA guidelines indicate that patients with stroke who use EMS arrive at the emergency department earlier, and more eligible patients receive IVT [27]. A stroke emergency map (an intelligent EMS that can guide ambulances to transport patients more effectively) in China has effectively shortened the ODT and improved the thrombolysis rate [52].

Our study demonstrated that patients with AF are not prone to pre-hospital delays, given that strokes resulting from atrial fibrillation tend to be more severe [53, 54]. Cardiac stroke typically occurs abruptly with evident symptoms. Patients often experience obvious discomfort, which helps in raising the alert faster, causing them to seek medical attention in time. The multivariate regression analysis revealed that AF was an independent factor associated with early arrival [47]. AF and a history of coronary artery disease accelerated the presentation to the hospital [13]; sudden onset of symptoms, loss of consciousness, recognition of symptoms as stroke, and feelings of fear and panic were associated with hospital arrival within 3 h.

The COVID-19 pandemic has markedly impacted patients with stroke, affecting different aspects of stroke care. Studies have demonstrated a decrease in hospital admissions for transient ischemic attacks and mild to moderate stroke during the COVID-19 era [55, 56]. Additionally, the pandemic has disrupted the chain of acute stroke care, resulting in potential risks such as decreased thrombectomy rates [57, 58] and modifications in the acute stroke care pathway [59]. Furthermore, the pandemic has caused a delay in patients with AIS seeking treatment at stroke centers [4]. Both pre- and post-hospital delays have been considerably prolonged, and the number of patients receiving intravenous thrombolysis treatment has decreased [60].

Moreover, a higher occurrence of severe strokes and an increased in-hospital mortality rate have been observed during the COVID-19 pandemic [61]. The pandemic has also raised concerns about the collateral damage on stroke emergency services, as well as the necessity to reorganize stroke networks in order to provide optimal care while mitigating the risk of transmission [55]. In conclusion, the COVID-19 pandemic has had a multifaceted impact on patients with stroke, affecting various aspects of stroke care, including hospital admissions, acute stroke care pathways, delayed presentation, and treatment.

The stroke population recruited in each subcenter of this study had certain regional characteristics; therefore, generalization of the research conclusions was affected to some extent. This study lacks detailed information on imaging, timing of EMS notification, and adjustment for socioeconomic factors. These limitations may have specific implications for interpreting and inferring research results: (1) Lack of imaging information: Inability to accurately assess disease severity and progression. (2) Lack of information on EMS notification time: Inability to determine the timeliness of patient medical assistance and the absence of a reference for optimizing emergency response systems. (3) Failure to adjust for socioeconomic factors: Socioeconomic status may influence patients’ healthcare-seeking behavior. Neglecting socioeconomic factors may introduce bias, potentially leading to an overestimation or underestimation of the impact of certain factors. Lastly, the COVID-19 pandemic presented significant challenges for this study. Lockdown restrictions and safety concerns led to limited data collection, resulting in a smaller sample size.