As in the United States, stroke is the third leading cause of death in Montana, accounting for 6.8 % of the state's deaths in 2003. The 2001 Behavioral Risk Factor Surveillance System survey in Montana indicated that 77% of adults had at least one of five key risk factors for stroke. In 2003, stroke was diagnosed in Montana as a primary or secondary condition in 3,806 inpatients or 3.7% of all inpatients. Of these patients, 56.4% were admitted through the emergency department compared to 38% of all patients.

Ischemic strokes are treatable with tissue plasminogen activator (tPA) if eligible patients receive treatment within three hours of symptom onset. Timely immediate ischemic stroke treatment may result in improved patient outcomes. Many Montana residents do not recognize or may ignore the signs and symptoms and risk factors of stroke, which reduces the opportunity for effective treatment within the recommended 3-hour time frame. These facts, coupled with long travel times from rural areas to stroke treatment facilities in urban areas of the state, create a significant problem of stroke deaths and disabilities in Montana.

Problem Statement: Not enough stroke victims for whom tissue plasminogen activator (tPA) treatment is appropriate are receiving it, due to the following contributing factors:

• Many individuals at high risk for stroke are unaware of the signs and symptoms of stroke and the need to call 911.
  
• Some EMS personnel may not be trained in the assessment of the signs and symptoms of ischemic stroke events eligible for tPA treatment.
  
• Patient travel times to treatment facilities can be long.

The Montana Department of Public Health and Human Services (DPHHS) Cardiovascular Health (CVH) Program has been funded by the Centers for Disease Control and Prevention since 2000. The program conducts statewide activities to promote cardiovascular health and to prevent and control cardiovascular disease and its associated risk factors, such as obesity, high blood pressure, and high blood cholesterol. Other than the American Heart Association, the CVH Program is the only public health-related program in Montana that focuses on heart disease and stroke. Although agencies such as clinics, hospitals, and community health centers provide direct care of individuals with those conditions, the CVH Program uses a population-based approach in its interventions to reach a large number of state residents. One priority of Montana's CVH Program is to increase the public's awareness of signs and symptoms of heart attack and stroke and the need to call 911.

In 2003, Montana's DPHHS CVH Program became a partner with Benefis Healthcare, which has a longstanding tradition of cardiovascular care in Montana. Located in Great Falls, Benefis Healthcare is one of Montana's leaders for stroke treatment and has been involved in clinical trials to test drug treatment other than tPA when patients don't receive tPA within three hours of symptom onset. Benefits Healthcare has also done some small scale stroke outreach programs through the Stroke Unit Task Force. Montana's CVH Program had a previous relationship with Benefis' Stroke Unit Task Force through a Benefis' stroke registered nurse, who was a member of Montana's Cardiovascular Disease Task Force for several years.

The planning team included many organizations, including Benefis Healthcare, MOST of Us (a social norms research organization at Montana State University; http://www.mostofus.org)*, and the City of Great Falls Fire and Rescue as key partners. Its functional structure was organized in a circular fashion with a core planning team that interacted with the other partners as needed to plan and carry out the project.

Interaction Between Partners

DPHHS staff and the consulting physicians communicated weekly via the regular CVH Program staff call and by e-mail.

The Montana DPHHS maintained regular interactions with several of the partners

• The health education specialist was the liaison with MOST of Us and had periodic telephone and e-mail communication during the media campaign development phase.
  
• The staff communicated with Benefis Healthcare staff via e-mail and periodic face-to-face meetings.
  
• Several CVH Program staff communicated with the City of Great Falls Fire and Rescue as needed.
  
• During the implementation phase, the primary prevention specialist periodically distributed brochures and posters to Great Falls pharmacies, senior centers, and parish nurses.

Benefis Healthcare ensured that other providers on the medical team were kept informed about the project.

The planning team examined both quantitative and qualitative research data related to stroke in the Great Falls catchment area, including the results of a telephone survey of residents of two Montana counties to establish a baseline of data on awareness of stroke signs, symptoms, and risk factors. The team also reviewed 16 journal articles related to stroke issues.

The stroke data examined included the following:

1. Year 2000—2001 BRFSS and Montana DPHHS data:

• The estimated number of stroke cases in the Great Falls catchment area (16 counties) (CV-stroke_cases.pdf) (PDF 54K)
  
• The prevalence of stroke among adult Montanans (CV-Prevalence_of_stroke_among_adult_Montanans_2001.pdf) (PDF 34K)
  
• Awareness of stroke signs and symptoms and the need to call 911 among adult Montanans (CV-Awareness_of_stroke_signs_and_symptoms_among_Montana.pdf) (PDF 99K)

2. July 2002—December 2003 data for 382 stroke cases (from Benefis Healthcare):

• Demographic and geographic characteristics of stroke (ischemic and transient ischemic attack [TIA]) (CV-demo_geo_char_of_stroke_patientsGF.pdf) (PDF 90K)
  
• Transportation characteristics of stroke (ischemic or TIA) patients (CV-transp_char_of_stroke_patientsGF.pdf) (PDF 46K)
  
• Discharge destination of stroke (ischemic or TIA) patients (CV-discharge_dest_stroke_patientsGF.pdf) (PDF 30K)
  
• Time from symptom onset to emergency department arrival for ischemic patients meeting the American Heart Association's "Get With the Guidelines" criteria (CV-time_from_symptom_to_arrivalGF.pdf) (PDF 46K)

3. February—March 2004 telephone survey data:

• Random telephone survey of 400 residents, aged 45 years or older, in two rural Montana counties (Yellowstone and Cascade counties), by Northwest Resource Consultants (CV-gf_stroke_presentation_CVD_TF_0704.pdf) (PDF 149K)
• Survey established baseline number of respondents who could identify the following:
  • The warning signs and symptoms of stroke
  • The risk factors for stroke
  • The need to call 911 if a stroke is suspected to be occurring 
• Survey gathered demographic data on the respondents
• Survey results showed many adults at risk for stroke could identify at least two warning signs for stroke, but many do not perceive themselves to be at risk for stroke
• Between one-third and one-half of the respondents with specific risk factors were not aware of their increased risk of a stroke
• Of particular concern was the lack of awareness of risk for stroke in respondents with atrial fibrillation

In addition to these quantitative data sets, the CVH Program staff reviewed the results of formative research done on stroke and performed for the CDC by a nationally recognized research firm, including the following:

• A listing of national stroke-related organizations, including a description, Web site address, prevention and treatment activities, and environmental/policy initiatives (CV-stroke_grid_2_Printable_version.pdf) (PDF 585K)
  
• An interview with a stroke researcher on a stroke intervention ("Sustained Benefit of a Community and Professional Intervention to Increase Acute Stroke Therapy": CV-Dr_Morgenstern_interview-final.pdf) (PDF 87K)
  
• A problem analysis conducted for CDC on stroke-related issues, including messages, audiences, materials/channels, evaluation/desired outcome, and a SWOT analysis (CV-CDC_Deliverable_2_charts_11-25_w_Lit_Review.pdf) (PDF 124K)
  
• A report on interviews with national organizations that carry out stroke communication activities (CV-Interview_Report_National_Orgs_Ogilvy_1103.pdf) (PDF 183K)
  
• A report on interviews with CDC-funded CVH State Programs on their stroke program activities (CV-Interview_Report_State_Programs_Ogilvy_1103.pdf) (PDF 191K)

Although several studies reviewed by the CVH Program staff had documented limited knowledge of stroke warning signs and risk factors, the reviewers found no recent population-based studies that had assessed risk perceptions for a stroke among persons with specific risk factors.

Based on examining the research data and the literature review, the planning team concluded that clinical and public health efforts were needed to increase awareness of the risk for stroke and actions that should be taken in the event of a stroke event among adult Montanans.

The primary priority group affected by the problem was Montana residents at risk for stroke: older adults and those with high risk medical conditions, such as history of heart disease/stroke/transient ischemic attack (TIA), atrial fibrillation, smoking, obesity, high blood pressure, high cholesterol, excessive alcohol use, diabetes, and physical inactivity. The secondary group affected by the problem was first responders, which includes not only medical personnel, but family and other caregivers as well.

Problem Statements:

• Montana residents at high risk for stroke and their caregivers are unaware of the signs and symptoms of stroke and the need to call 911 if they think a stroke event is occurring.
• Some EMS personnel may not be trained in the rapid assessment of the signs and symptoms of ischemic stroke events eligible for tPA treatment.

As noted in Step 1.4, the planning team examined July 2002—December 2003 data from Benefis Healthcare for 382 stroke cases, including the following:

• Demographic and geographic characteristics of stroke (ischemic and TIA) (CV-demo_geo_char_of_stroke_patientsGF.pdf) (PDF 90K)
  
• Transportation characteristics of stroke (ischemic or TIA) patients (CV-transp_char_of_stroke_patientsGF.pdf) (PDF 46K)
  
• Discharge destination of stroke (ischemic or TIA) patients (CV-discharge_dest_stroke_patientsGF.pdf) (PDF 30K)
  
• Time from symptom onset to emergency department arrival for ischemic patients meeting the American Heart Association’s “Get With the Guidelines” criteria (CV-time_from_symptom_to_arrivalGF.pdf) (PDF 46K)

Analysis of these data gave a picture of the issues surrounding travel times for patients and pre-notification issues for effective treatment of tPA-eligible patients.

Also, as noted in Step 1.4, the planning team examined the results of a random telephone survey of 400 residents, aged 45 years and older, in Yellowstone and Cascade counties (CV-gf_stroke_presentation_CVD_TF_0704.pdf) (PDF 149K). The telephone survey gathered demographic data and established a baseline of the number of respondents who could identify:

• The warning signs and symptoms of stroke
  
• The risk factors for stroke
  
• The need to call 911 if a stroke is suspected to be occurring

The results of these two analyses gave a detailed picture of adults with multiple risk factors and of their potential caregivers with respect to their ability to recognize whether a stroke is happening and whether they would call 911 in that event.

The planning team assessed each subproblem identified in Step 1.5 with respect to:

• Degree of importance (more or less important)
  
• Degree of changeability (more or less changeable)

to see the changeability table.

The planning team’s assessment of the EMS personnel issue was that it was more easily changed because there was a smaller, more homogeneous population whose behavior had to be changed. The team felt it would be easier to train a small group of medical personnel in the rapid assessment, transport, and treatment of tPA-eligible patients then it would be to change the behavior of a diverse group of Montana residents in the Great Falls area. It would be easier to reach the EMS staff with the message “use the stroke assessment tool” than it would be to reach the targeted population via media messages of “know the signs and call 911." However, both subproblems were deemed to be equally important.

The planning team considered several factors that could have affected the direction of the project, either positively or negatively. All of these were classified as strengths, weaknesses, and opportunities, and threats (SWOT).

to see the SWOT chart.

The planning team identified three major causes for problems:

The mortality and disability rate due to ischemic stroke should decrease if victims receive emergency medical care within three hours of a stroke event. To bring about this desired state, the following sub-problems needed to be addressed:

• Lack of awareness or poor recognition of warning signs of stroke events by people at risk for stroke and their caregivers
• Lack of motivation by people at risk for stroke to change their behaviors to modify their risk factors for stroke
• Allowing a situation to worsen and not calling a physician or emergency medical services
• Access-to-care issues (e.g., distance to treatment)
• Inconsistency in knowledge, training, and response actions of emergency responders

The planning team developed the following goals for the identified subproblems:

• To increase knowledge of warning signs of stroke and the need to use 911 among those most at risk and their caregivers
• To increase knowledge of what to do when experiencing these symptoms (e.g., what steps to follow and a sense of urgency with time)
• To increase prevalence of positive opinions and attitudes about probable success to prevent and treat stroke
• To increase use of EMS
• To shorten arrival time to hospital emergency departments
• To increase knowledge and consistent application of a stroke protocol by all EMS responders in the Great Falls area

The planning team examined a number of relevant theories, models, and best practices related to these subproblems and concluded that the following models would be helpful in planning potential interventions:

• The Health Belief Model might provide the basis for a message encouraging the target audience to take positive health actions, such as learning the signs and symptoms of a stroke (http://www.tcw.utwente.nl/theorieenoverzicht/Theory clusters/Health Communication/Health_Belief_Model.doc/).*
• The Consumer Information Processing model might guide the design and format for providing information to the target audience http://www.comminit.com/changetheories/ ctheories/changetheories-20.html.
• The PRECEDE-PROCEED planning model whose overriding principle is that most enduring health behavior change is voluntary in nature would guide development of motivational themes (http://hsc.usf.edu/).*
• The Organizational Change Theory model might contribute to developing strategies and processes for new policies and procedures in EMS provider organizations.

The planning team generated an analysis of strengths, weaknesses, opportunities, and threats (SWOT) for the pilot project.

As noted in Step 2.3, the planning team's goals were:

• To increase knowledge of warning signs of stroke and the need to use 911 among those most at risk and their caregivers
• To increase knowledge of what to do when experiencing these symptoms (e.g., what steps to follow and a sense of urgency with time)
• To increase prevalence of positive opinions and attitudes about probable success to prevent and treat stroke
• To increase use of EMS
• To shorten arrival time to hospital emergency departments
• To increase knowledge and consistent application of a stroke protocol by all EMS responders in the Great Falls area

The planning team chose their strategies for implementing the systems change intervention.

By accomplishing the steps in the strategies chart and influencing people to seek prompt emergency medical treatment, the planning team hoped that there would be a decrease in deaths and disability due to stroke in the Great Falls area. Once the systems change was completed in the target area and evaluated, it would be revised as necessary and launched in other counties.

The planning team discussed including a variety of new community partners and resources. They invited the following groups to become partners in implementing the interventions supporting the systems change:

• The regional affiliate of the American Heart Association (AHA)
• Local healthcare providers
• Emergency Medical Services providers
• Local ambulance services
• Pharmacies
• Parish nursing program and faith based organizations
• Clinics
• Work sites
• Senior Centers

Montana DPHHS worked directly with many of these partners, such as AHA and local pharmacies. Benefis Healthcare and the City of Great Falls Fire and Rescue approached and worked with several new community partners, such as the work sites and the private ambulance services. Because there was no local or state stroke organization to act as a partner, the Montana DPHHS CVH Program formed a regional Stroke Work Group with Benefis Healthcare stroke staff and St. Vincent Healthcare and Deaconess Billings Clinic in Billings, Montana.

A cardiovascular health grant from the Centers for Disease Control and Prevention (CDC) provided funding to the Montana DPHHS to develop the systems change intervention. While this was not a sufficient sum to reach the entire state, it supported an effective campaign in a targeted area (i.e., the intervention and comparison counties selected for the pilot project). Partnerships were solidified by first approaching the Great Falls Stroke Unit Task Force with Montana DPHHS' intervention proposal, obtaining their buy-in, and then involving the partners in the planning and implementation phase. Benefis Healthcare and Billings area hospitals were involved in analyzing the pilot project results and modifying the approach for the future Billings campaign and the Great Falls booster campaign.

The planning team examined whether communication would be the dominant intervention for each subproblem or would be used in support of a different intervention, such as policy, community services, or engineering (CV-Definitions.htm). For each intervention, the team listed the possible audiences to be reached through communication, either as the dominant intervention or in a supporting role. The planning team agreed that a systems change supported by two different, but related, interventions were needed to address the subproblems identified in Phase 2 (go to Stroke System Change Example Phase 2).

The two interventions needed were the following:

• A community-based health communication campaign that included media messages and distribution of health education materials in key locations
• An emergency medical services policy change supported by training

to see a chart summarizing their decisions.

The planning team evaluated the priority for targeting each of the groups by answering the following questions:

• Which has the highest priority?
• Which can be most easily reached and influenced?
• Which is most adversely affected?
• Which is a large enough group?
• Which is most vulnerable to health problems?
• Which are most likely to change contributing factors?

The planning team believed that its communication efforts would be accepted by all three subgroups: older adults, individuals with high risk medical factors, and their family and caregivers. These segments were also most adversely affected and most vulnerable to health problems. The planning team also believed it could reach these audiences through the planned communication mix.

The planning team also felt that the emergency responders subproblem would best be addressed through a policy change, supported by a training effort for a priority audience of emergency responders who did not have the knowledge or skills needed to apply the pre-hospital stroke protocol.

The planning team set the following communication goals:

• To educate their audience on the different warning signs of stroke
• To motivate their audience to become more action-oriented and call 911 when recognizing signs and experiencing symptoms
• To motivate their audience to change their behaviors to modify their risk factors for stroke

The planning team used the socio-ecological model for the overall design of the systems change and its supporting policy and communication interventions. The socio-ecological model is based on the belief that certain changes in the social and physical environment will promote positive changes in individuals, and the support of individuals in the population is essential for implementing environmental changes. The interventions were designed to influence individual behavior change at multiple levels—interpersonal, organizational, community and public policy. When trying to change health behaviors of a population, powerful forces, such as social, psychological, and environmental conditioning, can become obstacles, requiring the use of multiple sectors such as work sites, schools, faith-based organizations, and clinics as the channels for change. In addition to targeting individual behavior change through communication, the intervention strategies focused on policy change and improving or integrating systems used in stroke treatment (http://www.hpdp.unc.edu/WISEWOMAN/Chapter6.pdf) (PDF 184K).*

As noted in Step 1.4, the literature review indicated that messages should be positive and stress that some concrete action could be taken by the target audience to decrease their risk of stroke and improve their health outcomes. One of the communication models used by the team was the Health Belief Model, a framework for motivating people to take positive health actions (http://www.med.usf.edu/~kmbrown/HBM_Interactive_Handout.htm).*

The other model used by the team was the Self-Efficacy Model which considers the target audience's confidence in their ability to perform a behavior (http://www.emory.edu/EDUCATION/mfp/self-efficacy.html).*

The Montana DPHHS CVH Program contracted with MOST of Us * to develop the health communication messages and conduct a media campaign (CV-Contractor_tasks_v2.pdf) (PDF 75K). The MSU staff reviewed the results of the literature search provided by the CVH Program staff and then focused in on the particular literature concerning social marketing, reviewing and analyzing journal articles, research briefs, and specialized data sources. They also used university resources such as faculty members, students, and extension agents to gather expertise and build networks for strengthening campaign awareness and participation. The focus of their review and analysis was the following key information:

• What has been done, what did it accomplish, what more could be done?
• What has worked and what could be done differently?
• What is required to allow people to change?

A baseline stroke survey was done in Cascade County prior to the intervention.

The teams developed the following objectives for the communication intervention:

The teams developed the following objectives for the policy intervention (supported by training and communication):

The campaign messages needed to convey that stroke is both preventable and survivable if people know their risk factors, know the warning signs, and call 911 at the first sign of stroke. The information in the messages needed to be packaged in a way that made a personal connection with the audience while giving an action step that anyone could take to avoid or deal with a stroke event. The MOST of Us * staff generated concepts, messages, etc. for the first round of creative briefs for television and radio ads (CV-Draft_media_messages.htm). They also developed a brochure and newspaper ads which they pilot tested in an iterative process using interviews and focus groups (CV-draft_brochure_text.htm).

Development of the creative briefs focused on answering the following questions:

The staff of MOST of Us then met with graphic designers, video and film production people to discuss objectives, concepts, costs, talent, music, etc. Based on the formative research and literature reviews, a decision was made to use the following channels and activities for effectively reaching the target audience:

• Broadcast and cable television: three, 30-second spots
  • Programs and networks that receive the strongest ratings against the target audiences, ensuring maximum reach and frequency at the lowest cost News programs
  • Series of news segments
• Three radio ads
• Poster and tri-fold brochure
• Four newspaper ads
• One newspaper article

Meetings were held between the primary partners to decide who were the appropriate stakeholders and the manner and methods that should be used to contact these stakeholders and elicit their support. The primary partners involved were the Montana DPHHS, Benefis Healthcare, and the City of Great Falls Fire and Rescue. Plans were confirmed with the following additional stakeholders:

After the stakeholders were identified, information was distributed and meetings and discussions were held in the county with the identified stakeholders. During these stakeholder meetings, the design and implementation of the proposed campaign was discussed and evaluated, feedback was received, and information about expectations was obtained regarding the specific intervention implementation.

Because the systems change was supported by both a communication intervention and a policy intervention, the planning team worked with Benefis Healthcare and the City of Great Falls Fire and Rescue to plan an overall project timetable. With so many different organizations and contractors participating in the intervention, flexibility was built into the timetable to allow for changes necessitated by interruptions or delays in the tasks (CV-project_timeline.pdf) (PDF 29K).

As noted in Phase 3, MOST of Us * generated a first round of creative briefs for television and radio ads, a brochure, and newspaper ads at the concept level. These initial briefs were reviewed with the planning team and pilot tested with the target audience in an iterative process using interviews and focus groups. Several versions of each brief were generated before the creative briefs were finalized.

Development of the creative briefs focused on answering the following questions:

Messages were developed based on the results of the creative brief testing and tested with samples of the target audience for effectiveness. The content of the messages was centered around the ideas that stroke is both preventable and survivable if people know their risk factors, know the warning signs, and call 911 at the first sign of stroke (CV-Draft_media_messages.htm).

The planning team considered which settings would attract the largest numbers of the target audience. The settings that were considered to be likely to be the most effective were:

• Automobiles (radio)
• Work sites and businesses
• Pharmacies
• Clinics
• Homes (television, radio)
• Senior centers

The formative research reviewed indicated that a combination of television and radio could be used to reach the target audience for the campaign by purchasing time on programs and networks that receive the strongest ratings against the target audience, ensuring maximum reach and frequency at the lowest cost. It indicated that a high percentage of three target demographics could be reached through the selected television and radio channels: Persons 55+ at 93%, Persons 65+ at 92% and Persons 35-64 at 99% (CVH_Final_Report.pdf) (PDF 306K).

Results of previous communication campaigns supported the distribution of printed materials in senior centers, pharmacies, and work sites as being effective for the target demographics also. Adults aged 65 or older were more likely to report picking up brochures in pharmacies and at senior centers, as well as in clinicians' waiting rooms. Adults 35 or older were found to be more likely to pick up brochures or read the newspapers at their work sites or businesses, so print ads were also planned for Great Falls area newspapers during the campaign period as a reinforcing message channel (CV-Newspaper_Great_Falls_Stroke_Flight_Plan_2004.htm).

Campaign materials were developed and tested with samples of the target audience.

The planning team needed to define the roles and responsibilities of each partner organization in executing the health communication intervention plan. The planning team identified two communication needs, which were 1) to keep the roles and responsibilities of all involved clearly defined and 2) to keep the planning team and program participants informed and involved. The planning team delegated most internal communication responsibilities to the health communication intervention coordinator. The coordinator held regular meetings and communicated via e-mail to keep the planning team members informed.

The planning team used a worksheet for developing an implementation plan for the interventions supporting the systems change.

The planning team identified the following stakeholders as needing to provide input to the evaluation of the interventions:

The planning team invited individuals from several of the stakeholder groups to an initial meeting to discuss roles and responsibilities for the evaluation. After the initial meeting, periodic meetings, conference calls, and e-mail were used to maintain communication.

Using the input received from the stakeholders, the planning team decided to divide the evaluation of the interventions. The City of Great Falls Fire and Rescue would evaluate the policy intervention while Montana DPHHS would conduct the evaluation of the health communication intervention. The goal for the policy intervention was to get stroke victims to treatment quicker by having EMS responders consistently administer the pre-hospital stroke scale to all patients who showed any sign of neurological deficit. The goal for the health communication intervention was to increase awareness about the signs and symptoms of a stroke and the need to seek medical help more quickly.

Because of time constraints and project scope, it was agreed by all the stakeholders that the formal evaluation would focus on final outcomes and not the processes used to develop and implement the interventions. However, stakeholders agreed to hold debriefing meetings at the end of the intervention to discuss both outcomes and process issues and plan for needed changes when the interventions was expanded in the future to other Montana cities and counties.

Implementation The evaluation goal was to compare the actual changes in behavior and EMS response time in the intervention community of Great Falls to the same variables in a comparison community. This initial intervention represented a pilot project to evaluate the response from the target audience(s) to the persuasive messages and the change in identifying stroke victims by EMS personnel. Tracking measures during this phase included the following:

Outputs/Results Target Audience: Whom Would It Reach?

Expected Effects: Intermediate

Expected Effects: Long-term

Through discussions with the stakeholders, the planning team established what evaluation information was needed. This list was then compared to the evaluation plan to make sure that all information needs would be met when the evaluation plan was carried out. The following data and reports were requested:

• Quantitative results on:
  • the change in levels of awareness of symptoms
  • success in reaching the target audience message recall/effectiveness
  • change in use of pre-hospital stroke scale
• A project evaluation report, including recommendations for needed changes for future expansion of the project to other counties

To meet these information needs, the evaluation plan incorporated the following elements:

• Implementation evaluation - responses, strengths and weaknesses
• Exposure/reach evaluation - target audience impressions, degree of understanding
• Effectiveness evaluation - effects at each stage

The following specific intervention standards were written:

• To increase awareness of two or more signs of stroke among adults aged 45 years and older from 74% to 81%, the following interventions were completed:
  
  • The "Warning Signs" and "Modifiable Risk Factors" television ads (CV-Stroke_TV_Scripts_Final.pdf) (PDF 65K) were broadcast. An educational brochure called Prevent/Survive Stroke (CV-FinalBrochure_page1.pdf (PDF 229K) and CV-FinalBrochure_page2.pdf) (PDF 348K) (screen reader accessible versions not available for these resources) was printed.
  • Posters, and other campaign materials were printed.
• Clinics, senior centers, businesses, and pharmacies distributed "Prevent/Survive Stroke" brochures and displayed the posters.
• Health provider partners held interpersonal discussions with patients using the brochures and other materials as points of discussion.
  
  
• To increase awareness of two or more risk factors for stroke among adults aged 45 years and older from 45% to 52%, the following interventions were completed:
  • The "Excuses" radio ad was broadcast (CV-Excuses_with_music_stroke_radio_ad.rm).
  • The "Get the Facts" radio ad was broadcast (CV-Get_The_Facts_with_music_stroke_radio_ad.rm).
  • Clinics, senior centers, businesses, and pharmacies distributed the "Prevent/Survive Stroke" brochure.
  • Health provider partners held interpersonal discussions with patients using the Prevent/Survive Stroke brochures and other materials as points of discussion.

The planning team used volunteers representative of the target audience for the campaign planning, development, and testing stages. MOST of Us * evaluated the results and reported to the planning team.

The planning team also chose the survey method to determine baseline levels of awareness, comprehension, and attitudes/behavior patterns within the target audience in the intervention community and a comparison community for post-testing the same variables after the intervention. The Montana DPHHS epidemiologist and consulting physicians designed the pre-intervention and post-intervention telephone surveys, and a consulting firm was contracted to conduct, analyze and report the results of the surveys. The source of information for the survey was a random, representative sample of the target audience for this campaign (CV-bid_for_stroke_surveys.pdf) (PDF 41K).

The team planned to use data from the EMS responders, Benefis Healthcare's Get With the Guidelines Stroke program (CV-GFstroketbl_1.pdf) (PDF 261KB), and tPA administration data (CV-benefis_stroke_report_0204.pdf) (PDF 1,334K) to measure the change in arrival times to the emergency room and the administration of tPA to eligible patients.

A pretest-posttest control group design was developed for the evaluation of the health communication intervention. The design involved two groups: the intervention community and the comparison community. In both communities, a telephone baseline survey was conducted prior to the intervention (CV-stroke_GF_baseline_phone_survey_0304_Benefis_presentation.pdf) (PDF 115K). A follow up survey was conducted after the intervention (CV-stroke_followup_phone_survey_GFYell_0804.pdf) (PDF 383K). In the survey, participants were asked questions that identified the following:

• Demographic characteristics of age, sex, race, and education
• Knowledge of two or more warning signs for stroke
• Presence of two or more risk factors for stroke
• Proportion of respondents who would call 911 if they thought someone was having a stroke
• Proportion of respondents who would call 911 if they experienced weakness or paralysis that would not go away
• Proportion of respondents who could recall the health communication messages

For evaluating the policy intervention, Benefis Healthcare would provide Get With the Guidelines Stroke data during and after the intervention. The data on transport of patients prior to the intervention would be provided by the City of Great Falls Fire and Rescue, who would also gather and compile the post intervention data.

The evaluations measured knowledge (e.g., risk factors, warning signs and symptoms) and behavioral change (e.g. calling of 911)at the individual level and policy change at the system level. Data were analyzed using standard statistical software. All data gathered during this campaign were provided and analyzed by one of three parties:

• MOST of Us * (media campaign data)
• Northwest Resource Consultants (baseline and follow up telephone survey data)
• Benefis Healthcare (Get With the Guidelines data)
• Montana DPHHS staff and CVH Program staff

Oral and written reports were presented to stakeholders, team members and partners (CV-stroke_intervention_slides_Todd_1004.pdf) (PDF 138K).

The planning team agreed to the following communication strategies with the stakeholders and partners during the evaluation process:

• Regular phone conversations and e-mails to update partners and contractors on progress and issues related to the intervention
• Monthly group meetings of the planning team
• Face-to-face meetings with the Stroke Task Force and Emergency Medical Service providers as needed for project updates
• Meetings with pharmacies, work sites, senior centers to distribute/replenish materials
• Internal DPHHS meetings with CVH Program staff and section supervisor
• Periodic press releases to announce campaign events and successes
• Formal presentations and reports submitted to partners

The project budget was approximately $328,500. Approximately $18,000 of this amount was used for the evaluation component.

Evaluation Timetable
Task	Timeframe
Baseline data collected	March 30, 2004
Follow-up data collected	September 15, 2004
Data analysis	September 24, 2004
Report	December 31, 2004

The planning team summarized and presented the evaluation plan to stakeholders and partners in individual discussions and meetings of the whole group. The information provided included:

• Stakeholder questions (go to Step 5.3 Determine what information stakeholders need and when they need it)
• Intervention Standards (go to Step 5.4 Write intervention standards that correspond with the different types of evaluation)
• Evaluation methods and design (go to Step 5.5 Determine sources and methods that will be used to gather data and Step 5.6 Develop an evaluation design)
• Data analysis and reporting (go to Step 5.7 Develop a data analysis and reporting plan)
• Internal and external communication plan (go to Step 5.8 Formalize agreements and develop an internal and external communication plan)
• Budget and timetable (go to Step 5.9 Develop an evaluation timetable and budget)

During the health communication intervention, several activities were conducted to make the public aware of the campaign. Television interviews were held with local neurologist and campaign materials were distributed to clinics, work sites, pharmacies, faith-based organizations, and senior centers. CVH Program staff met with a point of contact at each of these places and explained suggested uses for the materials and provided contact information for questions, additional materials, etc. Press releases and informational packets were dispersed to all the media detailing the plans, objectives, and goals and points of contact were established to handle questions and media interviews.

Maintaining internal and external communication channels was a challenge. Since the system change required that the two interventions (policy and communication) work in tandem, the key to success was to maintain effective communication throughout the project. The planning team put into place measures to track whether they were meeting their objectives and took action to correct any missteps along the way.

The post-health communication intervention evaluation results showed that there was not a significant difference between the intervention community and the comparison community (CV-stroke_intervention_slides_Todd_1004.pdf) (PDF 138K). In both communities, members of the target audience who saw the ads reported a higher awareness of the warning signs of stroke (CV-stroke_intervention_by_recall_of_ads_1204.pdf) (PDF 53K); however, the ads appeared to have no impact on the audience's likeliness to call 911 as measured in the baseline survey (CV-911.pdf) (PDF 24K).

The planning team considered a number of possible causes:

• Other stroke outreach activities were ongoing in the comparison community during the health communication intervention, making it more difficult to compare the effect of the intervention in the two communities.
• The media ads aired during election primaries, which may have been a competing or distracting factor.
• The health communication messages may not have been clear or effective.
• The target audience's exposure to the messages was too short (three months).
• More messages targeting survivorship and or available effective treatment were needed in multiple channels.
• The target audience's baseline knowledge was already high when the campaign began.
• The design of the intervention did not provide for all the measurements that, in hindsight, turned out to be needed (e.g., a controlled trial to rule out the possibility that persons who were more aware of signs/symptoms and need to call 911 were more likely to watch the messages).
• Stronger communication between patients at risk for stroke and health care professionals was needed.

The intervention as carried out in Great Falls represented a campaign pilot to be replicated in additional communities. The Montana DPHHS conducted a survey of Montana and Northern Wyoming hospitals to determine the availability of diagnostic, treatment, and educational services for stroke patients (CV-Stroke_Assessment_ MT_Hospitals6-25-04.pdf) (PDF 165K). The survey showed that:

• About 50% of the hospitals surveyed had an in-the-field EMS stroke assessment tool.
• Only 20% had a community stroke awareness program.