Stroke awareness and knowledge in an urban New Zealand population

BACKGROUND
Stroke is the third most common cause of death and a major cause of chronic disability in New Zealand. Linked to risk factors that develop across the life-course, stroke is considered to be largely preventable. This study assessed the awareness of stroke risk, symptoms, detection, and prevention behaviors in an urban New Zealand population.


METHODS
Demographics, stroke risk factors awareness, symptoms, responsiveness, and prevention behaviors were evaluated using a structured oral questionnaire. Binomial logistic regression analyses were used to identify predictors of stroke literacy.


RESULTS
Although personal experience of stroke increased awareness of symptoms and their likeliness to indicate the need for urgent medical attention, only 42.7% of the respondents (n = 850) identified stroke as involving both blood and the brain. Educational attainment at or above a trade certificate, apprenticeship, or diploma increased the awareness of stroke symptoms compared with those with no formal educational attainment. Pacific Island respondents were less likely than New Zealand Europeans to identify a number of stroke risk factors. Māori, Pacific Island, and Asian respondents were less likely to identify symptoms of stroke and indicate the need for urgent medical attention.


CONCLUSIONS
The variability in stroke awareness and knowledge may suggest the need to enhance stroke-related health literacy that facilitates understanding of risk and of factors that reduce morbidity and mortality after stroke in people of Māori and Pacific Island descent and in those with lower educational attainment or socioeconomic status. It is therefore important that stroke awareness campaigns include tailored components for target audiences.

Stroke, a preventable non-communicable disease (NCD) associated with potentially modifiable lifestyle factors (1), is associated with long-term disability and is the second-leading cause of death worldwide (2). New Zealand has the second highest age-adjusted incidence of stroke among developed countries (3). When stroke occurs, early detection and treatment are critical in reducing the risk of disability and death (4). Early recognition and urgent transfer to stroke centres to ensure patients receive medical treatment within hours of stroke onset significantly improves survival rates (5). Although dramatic improvements in medical care in recent years has resulted in improved outcomes (6), increasing rates of survival and a trend towards greater indirect costs following stroke (7) could have significant economic impact for lower income and developing countries (2).
Although the incidence of stroke in the total New Zealand population is high (3, 8, 9 151), within New Zealand, significant variability exists in rates of stroke-related risk, morbidity and mortality (6,10,11).
Increased burden of modifiable risk factors for stroke, including NCDs such as hypertension, hypercholesterolemia, metabolic syndrome and diabetes, overweight and obesity (12), are apparent in non-European ethnicities, low socioeconomic status (SES) populations (13) and in those with lower levels of educational attainment (14). Higher proportions of Māori and peoples of Pacific descent, whom live in the most deprived areas of New Zealand (15,16), have greater incidence of a number of NCDs.
Māori adults are more than twice as likely to smoke as non-Māori, and have increased NCD burden including cardiovascular disease, diabetes, hypertension, obesity, and poor dietary habits (16). Peoples of Pacific descent have an increased incidence of diabetes, obesity, poor dietary habits, and greater prevalence of hazardous drinking patterns compared with non-Pacific peoples (17). These factors could contribute to increased incidence of NCDs, including stroke, in these populations. Māori have increased incidence of stroke compared with New Zealanders of European origin (18), sustain greater loss of health from cerebrovascular disease compared with non-Māori, and greater mortality compared with non-Māori in individuals below the age of 65 (35.6% vs. 6.6%, respectively) (19,20). In New Zealand,  Pacific descent have greater hospitalisation rates from stroke compared with non-Pacific   peoples of the same gender and age, and are more likely to die from cerebrovascular disease compared with non-Pacific peoples (17).
Despite ongoing campaigns to increase stroke awareness in New Zealand, and with a recent focus on Pacific populations (21), public awareness of stroke in New Zealand has not been assessed to determine where knowledge inadequacies exist. New Zealand data suggests that although awareness of stroke as a disease is high (99%), only 53% of respondents have adequate knowledge of stroke, defined by an ability to relate stroke to the brain and blood vessels (22). Knowledge of risk factors of stroke, and knowledge of appropriate responses to stroke signs or symptoms have not been assessed. A survey of the public awareness of stroke in New Zealand could inform the development of community health literacy and population-appropriate stroke intervention programs in order facilitate effective knowledge translation. Identifying inconsistencies in public awareness of stroke within a representative sample of the New Zealand population could inform future public awareness campaigns. The purpose of this study was to assess public awareness and knowledge of stroke in an urban adult population with a demographic distribution representative of New Zealand.

Ethical Approval
The University of Auckland Human Participants Ethics Committee (Reference Number 2011/7772) approved this study.

Subject Sampling
Adult members of the general public were approached by trained research assistants working in five convenience-based sample sites outside major retail stores on public streets in the Auckland area. Ethnic diversity within the group of research assistants was reflective of those represented within the target population. Sampling sites were identified following the method established for a similar survey of public awareness of aphasia in New Zealand (22). Sampling was carried out on a variety of days and occurred throughout the morning and afternoon on any single collection day. Shoppers who appeared to be over the age of 16 years were approached as they entered the store. Collection continued until data was broadly representative of the demographic diversity found in the New Zealand adult population according to the most recent New Zealand census data available at the time of data collection (2012).

Study Design
A brief explanation of the study was given and verbal consent obtained prior to participation.
Demographic data, including age, gender, highest level of educational qualification, and occupation were collected. Participants were asked to state which ethnic groups they identified with and were able to state as many as were applicable. A brief face-to-face, interview-based questionnaire (see Supplement 1) was administered to ascertain awareness and knowledge of stroke. Questions were designed to assess stroke literacy factors and to enable comparison to similar international studies (23).
Questions investigated stroke awareness and knowledge of stroke, stroke risk factors, stroke symptoms, and immediate actions required in response to observation of stroke symptoms.
Open-ended questions were used to enable respondents to answer 'in their own words', and allowed researchers insight into stroke literacy (24,25). Response recording was completed by the interviewer at the time of the interview using a combination of verbatim written records and common response check-lists. The common response list (see Supplement 1) was not visible to the interviewee but allowed the interviewer to rapidly record responses. Common response lists were established prior to data collection from similar international studies and were confirmed following analysis of responses from the first 50 respondents in this study. Written information about stroke awareness was offered to all participants on conclusion of the interview.

Coding and Statistical Analyses
Responses were coded using the prioritised coding schedule outlined by the New Zealand Department of Statistics (26). Although use of prioritisation is no longer recommended, it allows statistical identification of minority groups about whom policy decisions are made and is relevant given the high incidence of stroke in Māori and Pacific populations (6). Socioeconomic status was estimated using the New Zealand Socioeconomic Index (NZSEI) 1996 (27). The NZSEI-96 combines educational Coding of responses to questions examining stroke awareness and knowledge followed the previously defined methods (22). This defined stroke awareness as responding positively to the question: 'Have you heard of stroke?' Adequate stroke knowledge was defined as knowing that stroke involved both 'blood' and 'brain', (or a variation of these terms such as 'vascular' or 'nervous system') and not stating additional incorrect factors (e.g. 'it is a heart attack'). Correctly identified symptoms of stroke, as defined by the F.A.S.T mnemonic, which is active in New Zealand via the Stroke Foundation (28) were coded according to the National Institute of Neurological Disorders and Stroke (NINDS) symptom classification (29).
Responses to questions examining knowledge of stroke risk factors were coded according to nonmodifiable and modifiable risk factors. Classifications were based on those described in similar studies (22,30,31), and inter-coder consistency was verified throughout the study. Participants' personal experience of stroke was coded depending on whether the person they knew who had experienced a stroke was (a) themselves, (b) family and friends with whom the participant would likely have had close personal involvement (e.g. spouse, parent, child, sibling, cousin, aunt/uncle, grandparent/greatgrandparent, friend, extended family), or (c) patients (if participant worked in the healthcare field). In cases where work colleagues, or acquaintances had experienced a stroke, this was not coded as personal experience as it was not possible to assess the likelihood of this being a relationship close enough to involve close, personal involvement with the affected person. Responses to questions assessing knowledge of appropriate actions to take when stroke symptoms are observed were coded according to whether the need for urgent medical attention had been identified. Incorrect responses were coded according to whether they involved non-urgent medical attention (e.g. 'go see the GP' (non-emergency medical practitioner)) or self-management (e.g. 'have a lie down').
Statistical analyses used were consistent with the stroke literacy literature to enable comparison.
Binomial logistic regression analyses were performed to estimate odds ratios (OR) with a 95% (CI), comparing gender, age, ethnicity, socioeconomic status, educational attainment and personal experience with stroke. Potential for impact of the variables listed in Table 1 was explored. The frequency of correct identification of stroke symptoms, as determined by the F.A.S.T. campaign, were analysed with Cochran's Q Test.

Demographics
A total of 850 respondents completed the survey. The sample was broadly representative of the Auckland, New Zealand population (32) ( Table 1).

Stroke Awareness and Responsiveness
Data presented represents that in which significant variance is apparent. Of the total respondents, 97.5% (n=829) had previously heard of stroke, 42.7% could correctly define a stroke as involving both blood and the brain, and 82.4% identified the need for urgent medical attention.
The ability to define a stroke as involving both blood and the brain varied by ethnicity (Figure 1), educational attainment, age and personal experience of stroke (Table 2). Respondents with educational attainment at or above a trade or apprenticeship certificate, or a diploma were significantly more likely to define a stroke as involving both blood and the brain compared with respondents with no formal educational attainment. Respondents identifying as Māori, Pacific Island and Asian were significantly less likely to be able to define a stroke as involving both blood and the brain. Respondents aged from 50-59 years were more likely to be able to define a stroke as involving both blood and the brain compared with 70+ year-old respondents. Respondents who had personal experience with stroke were significantly more likely to be able to define a stroke as involving both blood and the brain and to suggest the need for urgent medical attention. Of those who could not identify stroke as relating to blood and brain, 17.9% (n=150) of total respondents defined a stroke as being a 'heart attack'.
While knowledge that urgent medical attention was required in response to a stroke was high (82.4%), this varied by SES, ethnicity, educational attainment, age and personal experience of stroke (Table 2).
Respondents from SES quintile 4 were significantly less likely to suggest self-management of stroke compared with quintile 5 (lowest SES). Respondents identifying as Māori, Pacific Island and Asian were less likely to suggest the need for urgent medical attention compared with respondents identifying as New Zealand European/European. Respondents identifying as Pacific Island and Asian were significantly more likely to suggest actively seeking non-urgent medical attention (i.e., seeing a nonemergency medical practitioner) compared those identifying as New Zealand European/European, whilst Asian respondents were less likely to suggest self-management. Compared with 70+ year-old respondents, 60-69 year-old respondents had significantly greater stroke awareness. Respondents who had personal experience with stroke were significantly more likely to suggest the need for urgent medical attention.

Identification of Risk Factors for Stroke
Whilst MELAA/Other respondents were significantly more likely to suggest diabetes as a risk factor for stroke compared with New Zealand European/European respondents, Asian respondents were Island respondents were significantly less likely to suggest family history of stroke, high cholesterol, obesity, and smoking as risk factors for stroke, but significantly more likely to suggest stress as a risk factor for stroke compared with New Zealand European/European respondents (Table 3).
Respondents in SES quintile 4 were significantly more likely to identify lack of exercise and poor dietary habits as risk factors for stroke compared with respondents in quintile 5 (lowest quintile) ( Table   4).

Identification of Factors that could Reduce Risk of Stroke
MELAA/Other respondents were significantly more likely to suggest a healthy lifestyle as a modifiable factor that could contribute to reduced risk of stroke, whilst Māori respondents were significantly more likely to suggest controlling diabetes as a factor which could reduce the risk of stroke compared with New Zealand European/European respondents. Pacific Island respondents were significantly less likely to suggest controlling blood pressure, not smoking, and moderating alcohol intake as modifiable factors that could contribute to reduced risk of stroke, whilst Asian respondents were significantly less likely to suggest controlling blood pressure, and not smoking as modifiable factors compared with New Zealand European/European respondents (Table 5).
Female respondents were significantly less likely to identify managing stress (0.7 (0.4-0.9), p<0.05) as a modifiable factor that could contribute to reduced risk of stroke compared with males.

Discussion
This is the first published study that examined baseline awareness of stroke symptoms, risk factors and response in a large sample of an ethnically diverse New Zealand population. This study has shown that while there is a moderate overall knowledge of stroke, this varies depending on ethnicity with particularly low levels of awareness in people of Māori, Pacific and Asian descent. The burden of stroke falls disproportionately on these people, many of whom reside in low SES communities (9,15,33).  (34). Stroke, an NCD for which developmental programming in the womb impacts on risk in adulthood (35)(36)(37)(38), requires a life-course approach to reducing incidence and must, thus, address lifestyle factors which are associated with low SES and deprivation. Improved public understanding of factors throughout the life-course, which contribute to risk of stroke, should be targeted to communities where the double burden of increased risk and reduced accessibility to resources are found.
Educational attainment below the level of a trade certificate, apprenticeship or a diploma below a Bachelor's degree was associated with reduced knowledge of stroke. Although, generally, the population sampled was an appropriate representation of Auckland based on the 2006 census, educational attainment was greater in the study population. As greater educational attainment is associated with improved stroke literacy (39), this could influence results, suggesting that this study may over-estimate stroke knowledge and awareness for the wider Auckland area. Limited knowledge of stroke in a significant proportion of New Zealanders, as a function of lower levels of educational attainment, could have a pervasive impact on the identification of stroke symptoms and the ability to respond appropriately. Improved public stroke awareness may increase the ability to identify stroke, reduce pre-hospital delay (40), and reduce the likelihood of death and disability (4).
Worldwide, public awareness of stroke is poor and is associated with lower educational attainment (23).
Whilst a low level of basic literacy is associated with barriers to preventative healthcare and reduced public awareness of stroke symptoms (41), low health literacy regarding stroke could be a primary impediment to improving health behaviours which may reduce the incidence of stroke. Health literacy refers to the motivation and ability to access, understand and use information relating to health-related issues in decision making at a personal, community or societal level (42,43). While generic dispositions and capabilities are associated with health literacy, it is also contextual and situated. Individuals and communities experience variable levels of access to information, and relevant knowledge of and about different contexts such as stroke. Reduced knowledge and awareness of stroke is associated with delay in obtaining urgent medical attention (40,44,45), which in turn is associated with negative clinical outcomes (4,5). Inability to identify and respond to stroke could delay medical intervention and have far-reaching economic impact on the direct and indirect costs (7,46,47)  identified in this study may suggest that current stroke awareness campaigns are not adequately meeting the needs of low SES, Māori, and Pacific populations within New Zealand. It is therefore important that population-wide awareness campaigns address access issues and include tailored components for target audiences, a strategy that has been shown to be effective in New Zealand populations (49).
In conclusion, this study indicates that, as in many developed nations, knowledge and awareness of stroke in an urban New Zealand population is poor. Whilst recognizing and controlling modifiable risk factors for stroke could reduce the risk of stroke, the ability to identify and respond appropriately in the instance of stroke has the potential to improve long-term health outcomes. The development of targeted campaigns that recognise and respond to the needs of different sectors within New Zealand could improve health outcomes and reduce the future burden of disease and disability at the individual, community, and national level. The results of this study provide compelling support for the development of an educational programme to enhance stroke-related health literacy, employing an approach that facilitates understanding of life-course risk reduction, symptomatology of stroke and appropriate responses, for current and future generations.        ***p<0.001; **p<0.01; *p<0.05