In the United States, stroke is the primary cause of severe long-term disability and the third leading cause of death (1). Stroke is also the third leading cause of death and a major cause of disability in New York State (2). In 2003, there were approximately 57,500 hospitalizations attributable to new and recurrent strokes in New York state (2). Although 50% to 70% of stroke survivors regain functional independence, 15% to 30% are permanently disabled, and 20% require institutional care after 3 months (3). Treatment of acute ischemic stroke with tissue-type plasminogen activator within 3 hours of symptom onset is recommended (4). If treatment for stroke is received early, the chance of recovery is improved (5). Ischemic stroke patients treated within 90 minutes of onset of symptoms have the most improvement (6).

A New York State pilot study was designed to determine whether designation of certain health care facilities as stroke centers and transport of stroke patients to these centers would improve care. The finding in that study was that 19.9% of patients arrived at a designated stroke center within 3 hours after the onset of symptoms (7). New York State's findings are similar to other studies, one of which showed that the median prehospital delay time was 2.6 hours, with approximately half of the patients delaying more than 3 hours (8). The delay in time until hospital arrival is a major obstacle for improving outcomes by obtaining treatment within 3 hours of symptom onset (9,10). Factors that contribute to the delay between symptom onset and hospital arrival include identification of stroke symptoms, determination that the symptoms require immediate emergency care, calling 9-1-1, and the time it takes from a 9-1-1 telephone call until hospital arrival (11).

Emergency medical services (EMS) response has not been found to be a major contributor to delay in hospital arrival. Once a call to 9-1-1 is made, it is estimated that the combined mean time of call to EMS arrival, on-scene time, and transport time is approximately 34 minutes (5). Studies show that patients who arrive by ambulance (34%–71% of stroke patients) (7,12-15) arrive sooner after onset of stroke symptoms than do patients who do not use EMS (15). This evidence suggests that most of the delay occurs before the call to 9-1-1 is made.

Research on knowledge of symptoms varies, probably because of methodological differences, including different ways of measuring symptom awareness and intent to call 9-1-1. When respondents were asked to select stroke signs or symptoms from a list, the percentage of people who were able to correctly identify symptoms (16-18) was much higher than when respondents were asked without prompting (19,20). A multistate study using data from the Behavioral Risk Factor Surveillance System (BRFSS) asked respondents to identify which symptoms were stroke symptoms. This study found that awareness of stroke symptoms varied by state, but overall awareness of individual symptoms when prompted was high: 94% for weakness of face, arm, or leg to 61.3% for sudden severe headache (17). In another study, loss of vision was the symptom least often identified, and sudden numbness was the most frequently identified symptom (16). A study that asked open-ended questions without prompting reported a lower percentage of respondents (57%) were able to list at least one of the symptoms (19). Ability to identify all symptoms was low, from 17% (17) to 46% (21) when prompted.

Research that asked respondents what they would do if they thought they were having a stroke suggests that awareness of the need for emergency care for a stroke is high, ranging from 70% (20) to 83% (16). However, when respondents were asked if they would call 9-1-1 when they had a symptom (not specifically said to be a stroke symptom), the percentages were lower (22). Participants reported that they would call 9-1-1 for 34.1% of the stroke symptoms identified. The rest would call a doctor or wait 1 hour or 1 day (22). The multistate BRFSS study found that, when asked to identify stroke symptoms from a list of symptoms, only 17.2% of respondents were aware of all five stroke symptoms and knew to call 9-1-1 for medical assistance when they thought someone was having a stroke or heart attack (17,23). Demographic factors associated with intent to call 9-1-1 are female sex, high income and education levels, and younger ages (21). The findings from these studies suggest that most people would call 9-1-1 if they knew they were or someone else was having a stroke. People recognize stroke as needing immediate care (20). However, the research indicates that people do not recognize the five symptoms of stroke (trouble seeing, facial droop, arm weakness, dizziness or loss of balance, and inability to speak or slurred speech). Therefore, people do not recognize the immediate need to call 9-1-1 for symptoms of stroke (17–20).

The purpose of this paper is to examine factors that are associated with respondents' intent to call 9-1-1 for signs and symptoms of stroke that occur in themselves or another person. The overall goal is to provide a better understanding of the reasons that result in patients not arriving at a hospital emergency department soon enough to receive the greatest benefit from treatment.

The percentage of survey respondents from the four upstate New York counties who reported they would call 9-1-1 for a stroke symptom they identified in themselves or another varied depending on the symptom. The largest proportion of respondents reported that they would call 9-1-1 if they noticed an inability to speak correctly in themselves or someone else. The smallest proportion of respondents reported that they would call 9-1-1 if they or someone else experienced trouble seeing or double vision. One of the most striking observations is that those who delayed getting medical care in the past 6 months were less likely to call 9-1-1 for inability to speak correctly or trouble seeing or double vision than those who had not delayed getting medical care. Although statistical differences between groups existed, little practical difference exists.

The percentage of respondents in this sample who reported intention to call 9-1-1 in response to stroke signs or symptoms was lower than the percentage of BRFSS survey respondents who replied they would call 9-1-1 if they thought someone was having a heart attack or stroke (86%) (17). This finding implies that the awareness of a stroke as necessitating immediate treatment does not transfer to awareness of which signs or symptoms are associated with stroke and the need for emergency care. The study most comparable with ours in how questions were asked was the validation study of the Stroke Action Test (22). The population in our study showed greater intention to call 9-1-1 for four signs or symptoms than the population in the Stroke Action Test: inability to speak correctly (72% vs 47%), facial droop (64% vs 47%), trouble seeing (33% vs 12%), and sudden arm weakness (50% vs 44%). In our study, as in previous ones, sudden visual difficulty was the least recognized symptom of stroke (17,20,35).

A number of factors contribute to the delay in seeking emergency care in response to the signs or symptoms of a stroke (11). Consistent with previous studies, our study found that the respondents' age, sex, and history of hypertension were not related to intent to call 9-1-1 in response to stroke signs or symptoms (11). Personal or family history of hypertension was not associated with higher odds of intention to call 9-1-1. This is an important finding that can inform practice. Because hypertension is a major risk factor for stroke, this finding suggests missed opportunities by health care providers and public health professionals to inform people with hypertension about their risk for stroke.

Prior experience with having a heart attack or stroke or with a family member who had a heart attack or stroke was not associated with greater odds of intent to call 9-1-1 for a stroke sign or symptom. This finding is consistent with the overall findings from the literature review conducted by Moser and colleagues (11) on delay in seeking care, but it is inconsistent with other studies (20). Higher education was associated with intent to call 9-1-1, which is consistent with the review article by Moser and colleagues and a study based on BRFSS findings (23).

This is one of two studies that examined history of delaying care. The other study that examined history of delaying care did not find an association with the intent to call 9-1-1 for stroke symptoms (21). However, that study was conducted among a more diverse population. The difference in findings may be due to insurance coverage, acculturation, or other factors.

This study has several limitations. The response rate (36%) was low, and nonresponse may be associated with knowledge of stroke signs or symptoms and the intent to call 9-1-1. However, the response rate is within the range of response rates commonly achieved in telephone surveys conducted since 2000, and little evidence suggests that a higher response rate would significantly change the findings (36). The potential influence of cell phone use on the response rate is minimal. In 2004, an estimated 1.8% of older adults lived in cell-phone–only households and, in the Northeast, only 3.9% of households have cell phones only (37). Because the study population was primarily white and well educated, the findings are not generalizable to other race or ethnic populations or to a population that is less educated. Because the survey was conducted to evaluate a media campaign, more detailed questions about insurance coverage and knowledge of signs and symptoms were not asked. Respondents were not asked about facial drooping in themselves. This study did not measure actual use of 9-1-1. However, findings from studies on actual 9-1-1 use are difficult to interpret because of the wide range (34%–71%) of stroke patients who arrive by ambulance (7,12-15.)

Despite these limitations, this study has several strengths. It has a large sample size, which improves statistical power and generalizability. The use of unprompted questions designed to assess behavioral intention to call 9-1-1 in response to stroke signs or symptoms is arguably a more accurate reflection of the public's true intent to call 9-1-1. In most studies, when a respondent is presented with a symptom and asked whether it is a symptom of stroke (16,20,23), correct responses are higher than when a respondent is asked to list stroke symptoms (20). Previous research shows that much of the public understands that stroke needs emergency care (20) but that knowledge of stroke symptoms is low (23).

The multivariate modeling takes into account many potential confounders, including demographics and medical history. Perhaps one of the greatest strengths of this study is that it measured history of delay in getting medical care, which has rarely been evaluated in stroke awareness research. Because reasons for delaying seeking treatment for acute coronary syndrome and stroke share some similarities (11), findings from the coronary literature might be relevant to stroke. A delay in seeking immediate medical care may reflect increased use of self-care and problem-solving or reflect the use of defense mechanisms or emotional coping strategies (e.g., minimizing signs and symptoms, denial of serious illness, an unwillingness to trouble others, embarrassment to seek help) (38,39).

The American Heart Association council statement (11) notes that most research addressing the delay in stroke care has focused on decreasing in-hospital treatment delays (11), yet research has suggested that a large portion of the delay in receipt of care is due to prehospital delays (5,15). This study shows that many people do not interpret stroke signs or symptoms as urgent enough to call EMS. Understanding determinants of health care-seeking behavior is important for decreasing the proportion of people who delay seeking emergency care. More research is needed to understand these determinants. The findings from our study suggest that educational efforts should focus on increasing the perception of stroke symptoms as severe and the recognition of the need to call 9-1-1. Studies such as this one, which examined factors that contribute to awareness of the necessity to call 9-1-1, can inform educational campaigns targeting stroke-related behavior change.