Background
In patients with acute stroke, intravenous thrombolysis with recombinant tissue plasminogen activator and endovascular treatment are effective, but only within 6 h after stroke onset [
1,
2]. Unfortunately, many patients with acute stroke do not seek immediate medical help [
3,
4]. This knowledge has prompted several public campaigns to increase public awareness of stroke symptoms. However, these campaigns have limited or no sustained effect on the proportion of patients eligible for acute stroke treatments [
5‐
8].
It has been presumed that in case of a stroke situation the patient or bystander has to recognize stroke symptoms, attribute these symptoms to acute stroke, but also has to recognize the importance to seek immediate help [
9]. Other factors of importance in the process of help seeking are age, stroke severity and ethnicity [
10‐
12]. The interactions between these factors are unknown. We explored the process of help seeking behavior in patients with acute stroke, evaluating knowledge about stroke symptoms, socio-demographic and clinical characteristics, and onset-to-alarm time (OAT).
Discussion
In our prospective observational study, a shorter OAT after stroke was associated with correct recognition of one-sided weakness and/or sensory loss, correct interpretation of the stroke situation by the patient, stroke characteristics and logistics of stroke care. This is concordant with other studies [
9,
16,
17]. In a qualitative study different aspects of interpretation influenced action or no action [
9]. One of these aspects was the presence and influence of another person at stroke onset. Our study showed no relation between the presence and influence of another person and OAT. In the study of Faiz et al. stroke severity, transport by ambulance and lower age were significantly associated with earlier admission [
17]. We did not find an association between age and OAT. In contrast, another study reported that older patients were more likely to call the emergency medical number [
10]. The crucial role of ambulance services has been described [
16]. Not only transport by ambulance, but also items as pre-notification to the receiving hospital and telemedicine-based interaction between the hospital and ambulance are suggested for further reduction in treatment time.
Knowledge about stroke symptoms was not associated with shorter OAT. This is in contrast with other studies [
3,
18‐
20]. In a quantitative study of 113 patients awareness of stroke symptoms and signs was associated with earlier hospital arrival [
20]. However, awareness was investigated with a structured questionnaire. In our study knowledge was investigated open-ended, for patients could name any symptom they thought about. We showed that recalling stroke symptoms is much easier instead of open-ended questioning. The definition of knowledge about stroke symptoms depends on the research method. In another study with 150 stroke patients, different cognitive aspects were associated with delay [
3]. Not only poor knowledge of stroke symptoms, but not realizing the importance of these symptoms resulted in delay. It seems contradictory that knowledge was not associated with OAT in our study, while a correct interpretation of the stroke situation was. However, the interpretation in the setting of an acute stroke situation is different from evaluating knowledge by reproducing single stroke symptoms in the office.
Our findings show that help seeking behavior after stroke is not a fixed process starting with having knowledge about stroke symptoms in general, followed by correct recognition of one’s own stroke symptoms, interpretation as a stroke and finally action. Nearly half of our patients with a stroke undertook no action within the first hour after symptom onset, despite having knowledge or recognizing symptoms. On the other hand, many patients still undertook action without knowledge, incorrect recognition and incorrect interpretation (shown in Fig.
1). A qualitative study proposed a model of help seeking behavior at the time of stroke [
9]. Patients seemed to follow fixed steps from knowledge of stroke symptoms towards action in the process of help seeking. Our study does not support this hypothesis. Additional factors, for instance fear, ignorance, ideas about seriousness or access to medical services, may play a role in the decision to seek help [
21]. Many studies focused on some isolated factors responsible for patient delay. Demographic factors [
22,
23], logistic factors [
16,
17] and social or psychological factors [
24‐
27] were studied. All these elements are of influence and may interact with each other, but a complete insight in the process of help seeking behavior has not been found yet. We believe a more complete insight is needed for more successful stroke campaigns. Differentiating the focus and strategies of these campaigns could reach and educate more future stroke patients. Reaching different age groups or people with different education and interest in information should probably be addressed to in different ways. Changing behavior of people with fear for hospitals or denial to be ill are other important elements to be taken into account.
Our study had several limitations. Firstly, our study was performed within a randomized clinical study [
13]. The inclusion criteria used in this study led to selection bias. Patients had to be included within 24 h after symptom onset. This will lead to underrepresentation of patients with very long delays in help seeking behavior. Some patients are waiting for days and sometimes even for weeks before consulting a doctor [
4,
17]. Secondly, included patients were interviewed after stroke onset, which may have caused a bias towards a better knowledge of stroke symptoms. Recall bias may cause overestimation of stroke symptoms knowledge. Studies have been performed in control populations to assess knowledge of stroke [
28,
29]. These studies give insight in knowledge, but no information about the process leading to help seeking in a real stroke situation, in which social and emotional factors may play a decisive role [
11,
22,
30,
31]. Patients with aphasia or a language barrier were not included, possibly leading to selection bias. Due to investment of extra time for completing the questionnaire and other logistical factors, not all patients eligible for this sub-study could be interviewed. However, we believe that our results are still generalizable to the overall stroke cohort, because the patients were randomly included. Finally, we investigated help seeking action within the first hour after stroke onset. Over the past 5 years, much effort has been made in shortening the onset to treatment time [
16,
32,
33].
Conclusion
A shorter OAT after stroke is associated with correct recognition of one-sided weakness and/or sensory loss, correct interpretation of the stroke situation by the patient, stroke characteristics and logistics of stroke care, but not by knowledge of stroke symptoms. Our data support the assumption that help seeking behavior after stroke is not a fixed process, starting with having knowledge and followed by correct recognition and interpretation. This process is complex and seems influenced by many factors. Future research must focus on the type of behavior and its motivation when experiencing a stroke. This is crucial to design appropriate stroke campaigns for changing patient behavior after stroke and thereby reducing patient delay optimally.
Acknowledgements
We thank all our colleagues of the Neurology department for their inclusions in the PASS study and completing the questionnaire of this sub-study.