Stroke Patients’ Knowledge of Stroke
Influence on Time to Presentation
Background and Purpose New treatments for acute stroke will likely have to be given soon after stroke onset. Little is known about stroke patients’ general knowledge about stroke, their interpretation of stroke symptoms, and how these factors influence the timing of their decision to seek medical attention.
Methods We interviewed consecutive stroke patients within 72 hours of stroke onset to define factors influencing time of arrival to the emergency department. Data recorded included demographic information, method of transportation, type of stroke symptoms, the patient’s interpretation of the symptoms, previous stroke, and knowledge of stroke warning signs. Stroke severity was measured with the Barthel Index. Early arrival was defined as within 3 hours of awareness of symptoms.
Results Sixty-seven patients were interviewed; 96% had an ischemic stroke and 4% a cerebral hemorrhage. Although 38% of patients professed to know the warning signs of stroke, only 25% correctly interpreted their symptoms. Patients with prior stroke were more likely to correctly interpret their symptoms (45% versus 16%; P=.03) but were not more likely to present early (19% versus 39%; P=.35). Eighty-six percent of patients presenting more than 3 hours after stroke onset thought that their symptoms were not serious. The 24% (n=16) of early arrivals were more likely to arrive by ambulance (81% versus 38%; P=.003) and had more severe strokes (Barthel Index score of 49 versus 72; P=.01) than late arrivals. Arrival by ambulance was independently associated with early arrival (odds ratio, 5.55; 95% confidence interval, 1.37 to 22.6).
Conclusions Approximately one quarter of stroke patients correctly interpret their symptoms as representing a stroke. This knowledge is not associated with early presentation to the emergency department. Ambulance transport is independently associated with early arrival at the emergency department. Even when patients know that they are having a stroke, most present late because they perceive their symptoms as “not serious.” Widespread public education of stroke-prone individuals may increase the proportion of patients eligible for new acute stroke treatments.
Animal studies of cerebral ischemia suggest that the time window for effective therapeutic intervention may be only 4 hours.1 Research in humans supports these data; the first proven treatment for acute stroke is efficacious when given within 3 hours of stroke onset.2 Despite these observations, the majority of stroke patients do not seek medical attention within 3 or even 6 hours of symptom onset. Although knowledge about stroke undoubtedly influences the time to presentation, no prior study has assessed stroke patients’ knowledge of stroke or how this knowledge affects their decision to seek medical care. Understanding this knowledge base is critical to planning effective public education programs. We undertook this study to better understand how stroke patients’ knowledge of stroke influences their time to hospital presentation.
Subjects and Methods
We interviewed consecutive stroke patients who were admitted to one of the three adult hospitals at Indiana University Medical Center: Wishard Memorial Hospital (a large county hospital), Indiana University Hospital (a university-based tertiary care hospital), and the Roudebush VA Hospital. Patients were identified through regular contact with the ED staff, the neurology admitting teams, and the neurology consultation teams at the three hospitals. One investigator (D.R.) administered a standardized questionnaire within 72 hours of hospital admission. Patients with stroke occurring while hospitalized, comatose patients, and patients with subarachnoid hemorrhage were excluded.
Patients who arrived at the ED within 3 hours of awareness of stroke symptoms were classified as early arrivals; all other patients were classified as late arrivals. Late arrivals were asked why they did not try to come to the hospital immediately. If patients were transferred from outside EDs, time to arrival at the outside ED was used.
All patients were asked the following questions: When did you first know something was wrong? What symptoms did you have? What did you think was wrong? How did you get to the ED? Have you ever had a stroke? Do you know someone who has had a stroke? Do you know the warning signs of a stroke? Do you have a primary care physician? Information from dysphasic patients was obtained from the patient to the extent possible, from family or witnesses, and from the chart; questions pertaining to personal knowledge or thoughts at the time of stroke were not included in the data unless the patient could communicate successfully. Demographic information was obtained regarding patients’ age, sex, race, marital status, education, and presence of health insurance.
If patients awoke with stroke symptoms, the time of awakening was taken as the time of awareness of symptoms. Symptoms were classified in the following categories: (1) motor; (2) sensory; (3) language (dysarthria or dysphasia); and (4) vision (any visual symptom). Distance to the ED was determined with the use of Indianapolis and Indiana maps; distance to the outside ED was used for patients transferred from another ED. The medical record was reviewed to determine the exact time of arrival and the type of insurance. Stroke severity was measured by the BI (range, 0 to 100; normal score=100).
The χ2 test and two-tailed Fisher’s exact test were used to compare categorical variables in early and late arrivals, patients with and without prior stroke, and patients who did and did not arrive by ambulance. Two-tailed Student’s t test was used to compare continuous variables in these groups. Forward logistic regression was performed to identify variables associated with early presentation.
From May through September 1995, 67 patients were interviewed; 64 (96%) had ischemic stroke, and 3 (4%) had intracerebral hemorrhage. Mean age was 64±3 years; the study subjects included 39 men (58%), 42 whites (63%), 24 blacks (36%), and 1 Hispanic (1%). Admissions by hospital were as follows: 38 at Wishard Memorial Hospital (57%), 15 at Indiana University Hospital (22%), and 14 at Roudebush VA Hospital (21%). Four patients were transferred from outlying hospitals. Sixteen patients (24%) were early arrivals; 51 (76%) were late arrivals. The early and late arrivals did not differ significantly in demographic characteristics, distance from the ED, daytime onset of stroke, location of stroke, presence of a primary care physician, presence of health insurance, history of prior stroke, or specific stroke symptoms (Table 1⇓). Anterior circulation stroke occurred in 48 patients (72%); right hemisphere infarction was not more common in late arrivals.
Overall, 38% of patients felt that they knew stroke warning signs, but only 25% correctly assessed their symptoms as representing stroke (Table 2⇓). Of the 49 patients who did not know that their symptoms represented stroke, 71% did not profess to know the cause of their symptoms, and 29% thought their symptoms represented something else. In the late arrival group, the most common reason for the delay in presentation was a lack of recognition of the seriousness of the symptoms (42 of 49 [86%]). Only 7 patients in this group (14%) delayed arrival because of inability to call for help.
To assess the impact of having a prior stroke on subsequent stroke knowledge and behavior, we compared patients with and without prior stroke (Table 3⇓). Patients with a prior stroke were more likely to profess to know stroke warning signs (55% versus 30%; P=.05) and to correctly identify their symptoms as stroke (45% versus 16%; P=.03). In the late arrival group, however, patients with prior stroke were as likely as those without prior stroke to feel that their symptoms were not serious (76% versus 88%; P=.33).
Compared with late arrivals, the early arrivals were more likely to arrive by ambulance (81% versus 38%; P=.003) and had more severe strokes (mean BI, 49 versus 72; P=.01). Forward logistic regression modeling showed only arrival by ambulance to be a significant independent predictor of early arrival (odds ratio, 5.55; 95% confidence interval, 1.37 to 22.6). Although patients who arrived by ambulance had more severe strokes (mean BI, 55 versus 77; P=.007), the addition of stroke severity did not improve the regression model. The distribution of BI scores was skewed, with a higher proportion of near-normal or normal scores in the group that did not arrive by ambulance and a more uniform distribution in the group that did arrive by ambulance (Figure⇓). Although there was a trend toward a higher proportion of patients who knew someone with a stroke in the early arrival group (93% versus 70%; P=.09), prior knowledge about stroke did not influence time to arrival in the ED.
Less than one quarter of our patients sought medical attention in the ED within 3 hours of becoming aware of stroke symptoms. The only factor independently associated with early presentation was arrival by ambulance. Stroke severity was also significantly increased in the early arrivals but did not improve the model for predicting early arrival. Although patients with prior stroke were more likely to correctly identify their symptoms as representing stroke, knowing that they were having a stroke did not prompt patients to seek early medical attention. Even when stroke symptoms were correctly recognized, most patients who delayed interpreted their symptoms as “not serious.”
The delay in presentation that we observed is typical for most stroke patients. Recent studies report average delays of 12 to 24 hours after symptom onset.3 4 5 6 7 8 9 Our data agree with a larger study from Denmark3 in which 25% of 1197 acute stroke patients presented within 3.5 hours. More rapid presentations have been reported from Sweden,10 the United Kingdom,11 and New Zealand.12 The different medical systems and geographic features represented in these studies make them difficult to compare with US populations. Two US studies have found that 50% to 60% of patients present within 3 hours of symptom onset.13 14 The larger of these studies was performed during screening for the National Institute of Neurological Disorders and Stroke 3-hour tissue plasminogen activator study14 ; this method of patient identification likely biases the sample toward earlier arrival by selectively increasing the referral of patients with very short duration of symptoms.
In our patients, the only factor related to early presentation was arrival by ambulance. Barsan and colleagues14 also found that patients who used the 911 system as their first medical contact had earlier hospital arrival. Other authors have found an association between stroke severity and early arrival.3 9 15 When added to the variable “arrival by ambulance,” stroke severity did not improve our model’s ability to predict early arrival. This suggests that arriving by ambulance and stroke severity provide redundant information; patients with near-normal function tend to not call the ambulance, while patients who do arrive by ambulance are more heterogeneous but overall have more severe functional limitations (Figure⇑). It is likely that other patient- and family-specific variables also influence the decision of when to call an ambulance.
Preliminary analysis of 591 patients enrolled in a trial of low-molecular-weight heparinoid in acute ischemic stroke also associated daytime onset of stroke, nonwhite race, and history of TIA with earlier presentation.15 We did not find an association between time of onset or ethnicity with time to presentation. Some investigators have found an association between hemorrhagic stroke and early presentation,4 8 16 but we did not have enough patients with hemorrhagic stroke to make any meaningful comparison between stroke types.
Although critical to planning effective educational programs, few studies have attempted to assess stroke patients’ prior knowledge of stroke. Surprisingly, we found no association between prior experience and knowledge of stroke and early arrival. Patients who had a prior stroke or who knew stroke warning signs did not present earlier than those with no knowledge of stroke symptoms or signs. An especially disturbing finding was that knowing a stroke was occurring did not prompt patients to seek earlier medical care. In a prospective study of 100 stroke patients, Feldmann et al6 found an association between recognition that symptoms signified stroke and early presentation, although this was not an independent predictor. Two reports have associated prior TIA and earlier presentation after stroke.3 15 We did not determine prior TIA in our patients.
Our patients’ overall lack of knowledge about stroke symptoms confirms recent data from a large population survey. A National Stroke Association–sponsored Gallup survey on public awareness of stroke found that 17% of 750 adults could not name one of the five major stroke symptoms.17 This survey also found that most adults do not realize that a previous stroke or heart attack increases the risk of subsequent stroke.
We did not find an association between time to presentation and factors often thought to be barriers to accessing medical care, such as older age, less education, lack of insurance, or increased distance from the hospital. Larger retrospective series also did not find an association between these factors and time to presentation after stroke onset.14 15 However, the Leicestershire study11 and the Copenhagen Stroke Study3 did find a link between living alone and delayed admission after stroke.
No previous study has assessed the absence of health insurance as a potential factor delaying stroke patients’ presentation to the hospital. While we found no association between presence or type of medical insurance and early arrival, it is interesting to note that the shortest reported intervals between stroke onset and presentation to the ED are from countries with free access to medical care.10 11 12
An important aspect of any study is how generalizable the results are to other populations. Our patients were drawn from three hospitals: a tertiary care hospital, a VA hospital, and a large county hospital. Many of our patients came from the counties surrounding Indianapolis; this likely accounts for the 22-mile average distance from the ED at symptom onset. The different hospitals and the diversity of the population sampled may make these results generalizable to other settings.
In summary, the majority of stroke patients do not come to the hospital early enough to receive acute treatment or to participate in acute stroke trials. Most stroke patients are unaware of the warning signs of stroke and present late because they misjudge the seriousness of their symptoms. Even when patients know that they are having a stroke, most do not seek immediate medical attention. Despite an increasing awareness of the brief interval in which effective stroke treatments must be given, this knowledge is not being transmitted to patients at risk for stroke. Widespread public education about stroke symptoms, new stroke treatments, and the time constraints of effective therapy may increase the proportion of patients presenting within the first hours after stroke onset; more patients presenting early is the key to translating efficacious trial results into effective stroke therapy.
Selected Abbreviations and Acronyms
|TIA||=||transient ischemic attack|
Reviews of this manuscript were directed by Associate Editor Marie-Germaine Bousser.
- Received January 11, 1997.
- Revision received February 10, 1997.
- Accepted February 20, 1997.
- Copyright © 1997 by American Heart Association
Jorgensen HS, Nakayama H, Reith J, Raaschou HO, Olsen TS. Factors delaying hospital admission in acute stroke: the Copenhagen Stroke Study. Neurology. 1996;47:383-387.
Alberts MJ, Perry A, Dawson DV, Bertels C. Effects of public and professional education on reducing the delay in presentation and referral of stroke patients. Stroke. 1992;23:352-356.
Feldmann E, Gordon N, Brooks JM, Brass LM, Fayad PB, Sawaya KL, Nazareno F, Levine SR. Factors associated with early presentation of acute stroke. Stroke. 1993;24:1805-1810.
Foulkes MA, Wolf PA, Price TR, Mohr JP, Hier DB. The Stroke Data Bank: design, methods, and baseline characteristics. Stroke. 1988;19:547-554.
Kay R, Woo J, Poon S. Hospital arrival time after onset of stroke. J Neurol Neurosurg Psychiatry. 1992;55:973-974.
Zweifler RM, Taft B, Lyden PD, Rothrock JF. Relationship between time to medical evaluation and ischemic stroke severity: the UCSD Stroke Data Bank. J Stroke Cerebrovasc Dis. 1995;5:111. Abstract.
Harper GD, Haigh RA, Potter JF, Castleden CM. Factors delaying hospital admission after stroke in Leicestershire. Stroke. 1992;23:835-838.
Anderson NE, Broad JB, Bonita R. Delays in hospital admission and investigation in acute stroke. BMJ. 1995;311:162.
Biller J, Patrick JT, Shepard A, Adams HP Jr. Delay time between onset of ischemic stroke and hospital arrival. J Stroke Cerebrovasc Dis. 1993;3:228-230.
Barsan WG, Brott TG, Broderick JP, Haley EC, Levy DE, Marler JR. Time of hospital presentation in patients with acute stroke. Arch Intern Med. 1993;153:2258-2561.
Saver JL, Bruno A, Feldmann E, Libman RB, Clarke WC, Grimsman K, Adams HP Jr. Time from stroke onset to arrival in the hospital. J Stroke Cerebrovasc Dis. 1997;6:146. Abstract.
Fogelholm R, Murros K, Rissanen A, Elmavirta M. Factors delaying hospital admission after acute stroke. Stroke. 1996;27:398-400.
National Stroke Association. Stroke remains a deadly mystery to many Americans. Be Stroke Smart. 1996:13:2.