Background
Given the evidence of effectiveness of thrombolysis for acute stroke and of stroke units, stroke should be treated as a medical emergency in the same way as myocardial infarction [
1‐
3]. However, many patients are seen too late to benefit from early treatment, often because of a lack of knowledge or awareness of stroke symptoms, or lack of emergency response to them, on the part of both the public and professionals [
4]. Other factors such as a belief that the symptoms will subside or that nothing can be done may also play a part in delay to presentation at hospital [
5]. Reviews of studies assessing levels of stroke knowledge and awareness among the general public, stroke patients and those at increased risk of stroke [
6,
7] have concluded that knowledge of stroke symptoms is generally poor and although most recognise the need for an emergency response this may not translate into action. A recent study reported that an adequate knowledge of stroke symptoms (i.e. three correct signs) is not associated with the intention to call emergency services in response to stroke [
8].
A number of different interventions to improve knowledge of stroke symptoms and the appropriate action have been tested, for example, community stroke screening events, patient education programmes and mass media campaigns. A drawback of screening events and patient education programmes is they often target small numbers of people from a specific group such as those who have suffered, or are at risk of, a stroke when it is argued there is a need to target wider demographic groups [
9] and those who may be a witness to a stroke [
10]. Mass media interventions, although more costly, have the potential to reach a much larger audience.
Mass media interventions have been successful in reducing the use of tobacco [
11], in improving road safety by reducing drink and driving [
12], and increasing the use of safety belts [
13]. However, in other health related areas they have had a small to moderate impact on behaviour change [
14]. Considering acute myocardial infarction - similar to stroke in that the event should be regarded as an emergency - the success of mass media campaigns in changing behaviour at the onset of symptoms was mixed. Two reviews concluded that mass media interventions had little impact on reducing delay to presentation at hospital and the findings are difficult to interpret as most studies were methodologically flawed [
15,
16]. Neither review commented on the quality of, or theoretical base for, intervention development.
The evidence base for mass media campaigns more widely is itself limited; a Cochrane systematic review included 15 studies evaluating mass media campaigns designed to increase health service utilisation, all interrupted time series, and of variable methodological quality [
17]. The authors concluded that, despite limitations, there was evidence that such campaigns may have "an important role in influencing the use of health care interventions" but that "further research... is needed on whether mass media coverage brings about appropriate use of services in those patients who will benefit most". It has been argued that adhering to the principles of effective campaign design has led to an increase in the success of mass media campaigns over the years [
18]. The major principles of good design are as follows: gain an understanding of the target audience in terms of the problem behaviour, their preferred message and the most effective means of delivering that message, through exploratory research; use theory to identify the focus of the campaign message; to achieve maximum effectiveness, segment the audience to create groups with similar message preferences; design the message from the findings of the exploratory research and choose the channels most widely viewed by the target group; evaluate and monitor the process of campaign activities; use a rigorous design to evaluate the intervention such as time series and controlled before and after designs [
18].
In England, as part of the national stroke strategy [
19], the Department of Health recently implemented a national mass media campaign to promote public awareness of stroke symptoms and of the need for emergency response using the FAST (Face, Arm Speech, Time) test [
20]. Implementation of this campaign begs an important question about the evidence base for the effectiveness of mass media campaigns in this specific area in changing knowledge of stroke (signs and symptoms) and, more importantly, behaviour (calling an emergency ambulance to ensure rapid access to treatment).
In light of the recent national stroke awareness campaign in the UK and the continued use of mass media campaigns in other countries it is timely to review the effectiveness and development of these interventions. Other reviews in the area of stroke education have focused on stroke prevention [
21] or have included a combination of different types of interventions [
22]. We were unable to identify any reviews examining the effectiveness and design of mass media interventions to improve knowledge of stroke symptoms and awareness of the need for an emergency response. In terms of reviewing the development of the intervention, we believe mass media campaigns are complex interventions - where often the aim is to change behaviour and improve knowledge - and should adhere to the structured development and evaluation as suggested by the MRC Framework [
23].
The aims of this study are to:
-
conduct a systematic review to assess the effectiveness of mass media campaigns in changing knowledge (stroke symptoms/signs and need for emergency access), behaviour (access to emergency services) or early treatment with thrombolysis.
-
examine the methods and theoretical basis for development of the interventions using the MRC Framework guidance.
Methods
Search strategy
The Cochrane Stroke Group search terms for stroke [
24] were used along with other terms developed, tested and then agreed by the study team and adapted for each database (Additional File
1). Searches were conducted in ten electronic databases (MEDLINE, EMBASE, CINAHL, Web of Knowledge, CSA Ilumina - ASSIA, Sociological Abstracts -, PsycInfo, ZETOC, AgeInfo and FRANCIS) from 1980 to 2010, the Cochrane Library (1980-2010), EPPI-Centre database and National Research Register. Manual searches through the reference lists of papers were also carried out.
Inclusion and exclusion criteria
Primary studies in English evaluating the effectiveness of mass media interventions were included; where relevant, related papers were obtained to gather information on the methods of intervention development. Studies were selected according to the following criteria: (a) targeted groups: the general public aged 18 or over; (b) outcomes: knowledge of stroke symptoms and, awareness of the need for an emergency response, rates of acute stroke treatments, and time to presentation at hospital; (c) design: randomised controlled trials (RCTs), quasi-experimental studies, controlled (CBA) and uncontrolled before and after (BA) studies and interrupted time series (ITS). Interventions aimed solely at health professionals were excluded, but those that targeted both the general public and health professionals were included.
Titles and abstracts were screened to identify studies of likely relevance and full papers obtained. A structured form was used to determine study inclusion. Two reviewers (JL, HR) extracted data from the final papers into structured tables. The results are presented as a narrative synthesis as the interventions varied in their format and presentation, were evaluated using different methods and outcomes, and included a range of study populations.
Review of intervention development using the MRC Framework
Intervention development for each study was classified using the five key phases suggested in the MRC Framework for evaluating complex interventions [
25] as follows.
Preclinical phase to identify the evidence to support the type of intervention: this could be from a systematic review or identifying or developing relevant theory
Phase 1
Modelling the processes and outcomes of the intervention
Phase 2
Exploratory trial to test out the intervention and outcome measures
These phases were entered into a matrix. Each study was examined, using the information from all relevant published articles, to determine how the intervention was developed and evaluated; a summary of the process was recorded into the matrix under the appropriate phase. This enabled the team to examine across studies the extent to which the intervention development and evaluation was in line with the guidance recommended in the MRC Framework.
Limitations
There are limitations to this review as a result of the range of methods and study designs employed in the included studies. Only three studies used a controlled before and after design [
30,
31,
34], limiting the conclusions that can be drawn. Of the ten studies only the latter three would be accepted as valid for inclusion in a systematic review by the Cochrane Effective Practice and Organisation of Care Group [
37]. Therefore it is not possible to state with any degree of confidence whether any changes identified were attributable to the intervention or to other factors. Few studies had pre-specified primary outcome measures and several used multiple analyses. None of the studies used qualitative methods (such as interviews with participants) to help understand their findings [
38]. There was very little reporting of the theoretical or empirical basis of interventions, or of their development methods, as recommended by the MRC Framework for development and evaluation of complex interventions; it appears that few studies have built upon earlier, potentially promising, interventions.
One published study not included in this review reported the development and piloting of an educational kit (Stroke Heroes Act FAST animation, brochure and poster) designed to improve knowledge of the symptoms of stroke and the need for emergency action using the FAST acronym [
39]. The materials from this educational kit have been used in 28 countries. In Massachusetts they have been used in mass media campaigns and evaluated through telephone surveys [
40] and claim to have significantly improved knowledge of symptoms and the need for emergency response, but these results do not appear to have been published in peer reviewed journals.
Conclusions
In conclusion, although some studies showed increases in symptom awareness and awareness of need for emergency response, and increased use of emergency transport, none show a full picture of increased awareness, increased use of emergency response, shorter time to arrival and increased use of thrombolysis following a mass media campaign. There is clearly a need for more robust evaluation of such campaigns using studies with at least a controlled before and after design, and including qualitative methods to support understanding of any demonstrated impact and to help unpick the elements of any campaign that might be important. The nationwide campaign in England implemented in February 2009 offered an opportunity to do this, but unfortunately there is no such robust evaluation in progress, despite considerable investment of public resources [
20]. The Department of Health website reports a 55% increase in emergency calls for stroke following campaign implementation but no further details are provided [
41]. When campaigns are evaluated, such as those using the Stroke Heroes Act FAST materials [
40], there is a need for the results to be published in peer reviewed journals. This would inform others of the efficacy of the intervention and robustness of the evaluation, and add to the existing body of knowledge about improving response to stroke.
Finally, the reported studies not only have limited methodological evaluation, but also little evidence of theoretically grounded development and piloting of the interventions. With the exception of the Morgenstern et al. study [
34], there is no evidence of such structured development. Mass media campaigns clearly can be successful in improving knowledge and changing behaviours in other fields of health and safety promotion. Unlike other interventions, such as stroke patient education and community stroke screening programmes, mass media campaigns have the potential to improve knowledge and awareness and change the behaviours of a large number of people. We would urge future developers of mass media campaigns to consider a more structured approach, such as that recommended in the MRC Framework for the Development and Evaluation of Complex Interventions [
23].
Competing interests
Authors JL, MM, HR, MW and RT have no competing interests. GF has consultancy relationships with Boehringer Ingelheim and Lundbeck, grants/grants pending from Servier Pharmaceuticals, Lundbeck and Mitsubishi, payment for development of educational presentations including service on speakers' bureaus from Boehringer Ingelheim and travel/accommodations expenses covered or reimbursed by Boehringer Ingelheim that might have an interest in the submitted work in the previous 3 years.
Authors' contributions
JL conducted the searches and electronic sifting of titles and abstracts. JL and HR independently reviewed the retained papers and extracted data. JL and RT wrote the first draft of the manuscript. All authors commented on the first draft and all revisions.