Layout 1 Introduction Using qualitative data to develop health interventions Qualitative data can be useful in all stages of health intervention development including the design, delivery, and evaluation stages. Guidance exists on how to develop complex health interventions (Craig et al., 2008), and an expert consensus panel of intervention developers recom- mends qualitative methods in addition to quantitative methods in designing complex health interventions (O'- Cathain 2019). Specifically, the panel advises that re- searchers can take a target population centered approach by performing qualitative interviews of the intended tar- get population to inform intervention development. An- other approach suggested by the panel is to partner with stakeholders to co-design the intervention, obtaining stakeholder input through qualitative interviews and sur- veys. Qualitative data are not only useful in developing de novo interventions, but also in adapting existing inter- ventions for a new target population (Duggleby et al., 2020). Despite the use of qualitative research methods by health intervention researchers, there is a dearth of studies Using qualitative data to inform the adaptation of a stroke preparedness health intervention Mellanie V. Springer,1 Tiffany Hodges,1 Cristi Lanning,2 Michael Tupper,3 Lesli E. Skolarus1 1Stroke Program, Department of Neurology, University of Michigan Medical School, Ann Arbor, Michigan; 2Eastern Michigan University, Ypsilanti, Michigan; 3University of Michi- gan Medical School, Ann Arbor, Michigan, United States ABSTRACT Qualitative research methods are often used to develop health interventions, but few researchers report how their qual- itative data informed intervention development. Improved com- pleteness of reporting may facilitate the development of effective behavior change interventions. Our objective was to describe how we used qualitative data to develop our stroke ed- ucation intervention consisting of a pamphlet and video. First, we created a questionnaire grounded in the theory of planned behavior to determine reasons people delay in activating emer- gency medical services and presenting to the hospital after stroke symptom onset. From our questionnaire data, we identi- fied theoretical constructs that affect behavior which informed the active components of our intervention. We then conducted cognitive interviews to determine emergency department pa- tients’ understanding of the intervention pamphlet and video. Our cognitive interview data provided insight into how our in- tervention might produce behavior change. Our hope is that other researchers will similarly reflect upon and report on how they used their qualitative data to develop health interventions. Correspondence: Mellanie V. Springer, Stroke Program, Depart- ment of Neurology, University of Michigan Medical School, 1500 E. Medical Center drive, Ann Arbor, Michigan 48109-5855, United States. Tel.: 734.936.9075; Fax: 734.232.4447. E-mail: mvsprin@med.umich.edu Key words: Qualitative data; stroke; health intervention; behavioral intervention. Contributions: Contributed to the conception or design of the work: MS, LS; Drafted the manuscript: MS; Revised the manuscript for important intellectual content: MS, TH, CL, MT, LS; Acquired, an- alyzed, or interpreted the data: MS, TH, CL. Conflict of interest: The authors declare no conflict of interest. Funding: This work was supported by the National Institutes of Health/ National Institute of Neurological Disorders and Stroke K01NS117555. The funding source had no role in the study design, collection or analysis of data, writing of the report, or decision to submit the manuscript for publication. Availability of data and materials: A subset of the data generated or analyzed during this study are included in this published article. Ethics approval and consent to participate: The required Ethics Committee approved this study (IRB 1635357). The study is con- formed with the Helsinki Declaration of 1964, as revised in 2013, concerning human and animal rights. All patients participating in this study signed a written informed consent form for participating in this study. Informed consent: Written informed consent was obtained from participants for anonymized patient information to be published in this article. Received for publication: 25 May 2022. Revision received: 15 November 2022. Accepted for publication: 6 December 2022. Publisher’s note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher. ©Copyright: the Author(s), 2022 Licensee PAGEPress, Italy Qualitative Research in Medicine & Healthcare 2022; 6:10639 doi:10.4081/qrmh.2022.10639 This article is distributed under the terms of the Creative Commons Attribution-NonCommercial International License (CC BY-NC 4.0) which permits any noncommercial use, distribution, and reproduc- tion in any medium, provided the original author(s) and source are credited. [page 68] [Qualitative Research in Medicine & Healthcare 2022; 6:10639] Qualitative Research in Medicine & Healthcare 2022; volume 6:10639 No n- co mm er cia l u se on ly that report in detail how qualitative data were used to in- form intervention design (Wigginton et al., 2020). Often, researchers simply report the outcome of their qualitative studies, but there is methodological value in researchers’ reflecting on and reporting how their quali- tative data are used to develop interventions (Sandelowski & Leeman, 2012; Wigginton et al., 2020) and how quali- tative findings are used to interpret intervention results (Lewin et al., 2009). There are significant advantages to thorough qualita- tive methodology reporting. First, this practice can help other health intervention researchers understand how qualitative data can be used to develop interventions and interpret intervention results. Second, researchers might identify the theory-based active components of an inter- vention through reflecting on their qualitative data. To achieve improved completeness of intervention re- porting, researchers can use existing tools. The CONSORT statement for randomized controlled trials of non-pharma- cologic treatments states that researchers should describe the components of their intervention (Boutron et al., 2017). In behavior change intervention research, taxonomies have been developed to facilitate uniform description of the ac- tive components of behavior change interventions. Greater specification of the link between theory and selection of the active components of an intervention has also been recom- mended (Michie et al., 2018). When designing behavior change interventions, tables identifying links between the- oretical constructs that affect behavior, called “mechanisms of action” (MoA), and active components of the interven- tion that enable behavior change, called “behavior change techniques” (BCT), can be used to select the appropriate BCT or identify the underlying MoA (Carey et al., 2019). Reflection on qualitative data may lead to the identification of MoAs, particularly when qualitative research methods are grounded in theory. In this article, we attempt to overcome shortcomings in reporting the outcome of qualitative research for inter- vention development by adding to the few existing studies that describe the use of qualitative data in intervention de- velopment (Fjeldsoe et al., 2012; Yardley et al., 2015). Specifically, we focus on the use of qualitative data in the design stage of health interventions, reporting on how our qualitative data (i) were used to inform development of a stroke education intervention, (ii) revealed MoAs of our intervention components, (iii) and informed how our in- tervention might produce positive behavior change. We begin by describing prior research that supports the need to develop a stroke education intervention. Next, we provide an overview of our qualitative research meth- ods, presenting our qualitative and quantitative data and the ways that qualitative data were used to inform devel- opment of our intervention linking MoAs to behavior change techniques. We then describe how our qualitative data informed how our intervention might produce posi- tive behavior change. Our study consisted of a survey of stroke patients to inform adaptation of our intervention and cognitive inter- views of patients to assess their understanding of the adapted intervention materials to determine whether fur- ther adaptation was needed. We present the methods and results separately for both parts of the study. Background on the need for the intervention Stroke is a leading cause of disability (Katan & Luft, 2018). Tissue plasminogen activator (tPA) is a treatment for stroke that reduces disability and can only be given at the hospital within a limited time after stroke symptom onset (National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group, 1995). Black Americans are less likely to receive tPA than White Americans (Aparicio, 2015; Boehme, 2014; Faysel, 2019; Hsia et al., 2011a; Johnston, 2001; Nasr et al., 2013), at least partly due to Black Americans’ delay in hospital arrival after stroke symptom onset (Hsia et al., 2011b; Lacy et al, 2001; Lichtman et al., 2009; Siegler et al., 2013; Springer et al., 2017). Randomized controlled trials of stroke pre- paredness interventions for Black Americans that teach stroke symptoms and the importance of activating emer- gency medical services (hereafter referred to as “calling 911”) are few in number and have occurred in settings such as schools and churches (Williams, et al., 2008; Williams, et al. 2019). The Stroke Ready brief intervention was a randomized controlled trial of a community stroke preparedness inter- vention among people at a community health center con- sisting of an oral presentation of a pamphlet on stroke symptoms, stroke treatment, and the importance of calling 911 upon stroke symptom recognition. The brief interven- tion was found to increase intent to call 911 immediately after the intervention (Robles et al., 2020). Our objective was to adapt the Stroke Ready brief intervention for pa- tients presenting to an emergency department of a predom- inantly Black American community. Adaptation of the intervention is necessary because of the unique features of the emergency department environment such as its unpre- dictability, rapid pace, and possible anxiety felt by the pa- tients (Wei & Camargo, 2000). In future research, we plan to test the ability of the adapted intervention to increase stroke symptom recognition and intent to call 911. We chose the emergency department setting to increase the reach of the intervention to Black Americans who are at high risk for stroke and those who are frequent users of the emergency department (Colligan, et al., 2017).29 Intervention materials The Stroke Ready intervention materials consisted of a pamphlet and a video, both of which describe stroke symptoms, stroke treatment, and the importance of rapid activation of emergency medical services. Both the pam- phlet and the video are grounded in the theory of planned [Qualitative Research in Medicine & Healthcare 2022; 6:10639] [page 69] Article No n- co mm er cia l u se on ly behavior, which is described in detail below. The video is approximately three minutes long and consists of local ac- tors singing about stroke symptoms and the importance of rapid activation of emergency medical services. The video also contains a skit depicting a person who develops stroke symptoms and a witness calling 911 upon stroke symptom recognition. The pamphlet includes a definition of stroke, the consequences of stroke, an explanation of stroke symptoms, how to check for stroke symptoms, the importance of calling 911 when stroke symptoms start, and a description of stroke treatment. The purpose of our study was to adapt the pamphlet and intervention video, as they were originally designed without the input of stroke survivors. We therefore surveyed stroke patients who delayed in hospital arrival after stroke symptom onset to inform adaptation of the intervention materials. This study was approved by the hospital’s institutional re- view board and was conducted in agreement with the Helsinki declaration. Informed participant consent was obtained. Survey of stroke patients to inform intervention adaptation Materials and Methods Setting The study setting was a stroke inpatient ward of a hos- pital located in Flint, Michigan. The city of Flint, Michi- gan has a population of 54% Black Americans, 38% White Americans, and other racial groups in smaller pro- portions; 51.5% of Flint residents are female, and 12% of the population has a Bachelor’s degree or higher educa- tion (U.S. Census Bureau, 2021).30 Participants Participants included in the study were patients hos- pitalized with stroke who were 18 years of age and older, English speaking, and had arrived at the hospital three hours or more after stroke symptom onset. Patients who awoke with stroke symptoms were excluded. Survey instrument We developed a survey instrument (questionnaire) to determine reasons that stroke patients delay in hospital ar- rival after stroke symptom onset and grounded the ques- tionnaire in the theory of planned behavior. The theory of planned behavior states that a person’s intent to engage in a particular behavior is influenced by their attitude toward the behavior, subjective norm/social pressure to perform the behavior, and perceived behavioral control/self-effi- cacy to perform the behavior (Ajzen, 1991). A person’s attitude towards the behavior is influenced by their be- havioral beliefs (i.e., expectations about the outcome as- sociated with the behavior). The subjective norm around a behavior, in turn, is informed by a person’s normative beliefs (beliefs about the social expectations around per- forming the behavior). Finally, perceived behavioral con- trol is shaped by control beliefs regarding how easy or difficult it is to perform the behavior (Figure 1). The behavior of interest was calling 911 upon stroke symptom onset. We included questions that asked: i) rea- sons for delay in hospital arrival (in order to determine control beliefs and behavioral beliefs); ii) actions per- formed upon stroke symptom onset (which may be shaped by one’s attitude towards performing the behavior); iii) self-efficacy for stroke symptom recognition and calling 911; and iv) subjective norms about calling 911. The questionnaire included multiple choice questions with most having the option for the respondent to provide their own answer if none of the options was considered appropriate. Response options were based on data from previous surveys and focus groups assessing reasons for delay after stroke symptom onset conducted in similar communities (Hsia et al., 2011b; Skolarus et al., 2013). Several questions asked respondents to rate their level of agreement with a written statement on a Likert scale. There was one open-ended question (see supplementary file for questionnaire.) After creating the questionnaire, we performed cogni- tive interviewing on a convenience sample of stroke sur- vivors to ensure that they understood the questions as intended. Cognitive interviewing is a technique used to study how material is understood and processed by the in- tended target audience. We used the cognitive interview- ing technique of verbal probing, asking participants questions to probe their understanding of the survey ques- tions (Willis, 2005). The convenience sample consisted of stroke survivors attending a session of a stroke survivor support group. Based on the cognitive interviewing, slight modifications were made to the wording of questions to improve comprehensibility. The hospital stroke coordina- tor distributed the final questionnaire to stroke patients, who self-administered the questionnaire. Data analysis Given the types of questions asked on the question- naire, we used a mixed-methods approach to analyze sur- [page 70] [Qualitative Research in Medicine & Healthcare 2022; 6:10639] Article Figure 1. Theory of planned behavior. No n- co mm er cia l u se on ly vey responses. Quantitative methods were used to de- scribe the proportion of respondents who selected differ- ent answer choices and the proportion who provided the same free-text responses. The number of questionnaire re- spondents (N=19) was too small to permit inferential sta- tistics on the questionnaire data. After identifying the most frequent responses, we used a qualitative approach by using the theory of planned behavior to interpret the re- sponses. Specifically, we categorized responses into the beliefs that comprise the theory of planned behavior. We then used survey responses to inform development of the intervention by linking MoAs to behavior change tech- niques (Carey, et al., 2019; Figure 2). The following sec- tion presents a description of the survey participants, organization of survey responses into the theory of planned behavior, response proportions, and linking of mechanisms of action to behavior change techniques. Results We surveyed 19 stroke patients between December 2020 and April 2021. Participants had a mean (± standard deviation) age of 57 ± 8 years old. Of the participants, 58% were men, 63% had at least some college education, 63% self-reported their race as Black, 37% self-reported their race as White, and 16% had a self-reported past his- tory of stroke. Percentages indicate the proportion of sur- vey respondents who provided the response. Theory of planned behavior: Behavioral beliefs The most frequent beliefs about stroke symptoms that were identified as reasons for waiting to come to the hos- pital included that the symptoms would go away (73.7%) and that the symptoms were not serious (31.6%). The ma- jority of stroke survivors (57.9%) understood that they would get an ambulance by calling 911. Even though many stroke survivors (57.9%) thought that they would have to pay for the ambulance, very few stroke survivors (15.8%) indicated that the thought of having to pay for an ambulance caused them to delay in calling 911. Theory of planned behavior: Normative beliefs The majority of participants (89.5%) held normative beliefs about calling 911, with the opinion that most of their friends or family would call 911 if they had symp- toms of stroke. Many stroke survivors (21.1%) thought that their friends and family would not be able to recog- nize stroke symptoms or were unsure (31.6%) if their friends and family would recognize stroke symptoms. The majority of stroke survivors (63.2%) called a family mem- ber when their stroke symptoms started, and this was the first action after symptom onset in 31.6% of stroke sur- vivors. Many stroke survivors (47.4%) eventually pre- sented to the hospital because family told them to do so. Theory of planned behavior: Control beliefs Some participants (31.6%) identified not wanting to miss work as a reason for waiting to call 911. Most par- ticipants (63.2%) stated that, considering what they knew prior to coming to the hospital, they would know what to do if someone was having a stroke. The majority of par- ticipants (63.2%) also stated that, considering what they knew prior to coming to the hospital, they would have been able to tell a 911 operator why they thought someone was having a stroke. How survey results informed development of intervention materials: Linking MoAs to BCTs We adapted the intervention pamphlet based on the questionnaire responses. Adaptation of the video was not needed to address the survey responses. In modifying the intervention, we directly addressed behavioral beliefs that were identified by stroke sur- vivors as reasons for waiting to come to the hospital— including that the symptoms would go away and that the symptoms were not serious—and, therefore, might in- fluence their attitude towards calling 911. We addressed stroke survivors’ attitude towards the behavior of calling 911 by including information about health conse- quences. We modified the intervention pamphlet to in- clude the sentence Don’t wait for the signs to go away, STROKE IS SERIOUS!. We also added the sentence Faster treatment means improved health! (Table 1). Since stroke survivors expressed knowledge that calling 911 would result in dispatch of an ambulance, we did not reinforce this fact. Instead, we emphasized the mes- sage to call 911 and get an ambulance as soon as stroke symptoms start. While stroke survivors held normative beliefs that their friends or family would call 911 upon stroke symp- tom onset, many thought that their friends or family would not or might not be able to recognize the signs of [Qualitative Research in Medicine & Healthcare 2022; 6:10639] [page 71] Article Figure 2. Linking theory to mechanisms of action to behavior change techniques. No n- co mm er cia l u se on ly stroke. The pamphlet already addressed the lack of knowl- edge or skills by providing written and pictorial descrip- tions of stroke signs and how to check for those signs. We added instruction on when to call 911 by stating Notice just one sign? REACT- Call 911!. Since many stroke sur- vivors called family when their stroke symptoms started, we provided verbal persuasion about the stroke survivor’s capability of calling 911 by stating Time to call 911! Make it the FIRST thing you do! (Table 1). Participants expressed self-efficacy in recognizing stroke symptoms and calling 911. A control belief identi- fied by some stroke survivors included not wanting to miss work which has the potential to influence their atti- tude toward calling 911. We therefore emphasized the salience of the consequence of not calling 911 by includ- ing the reason why rapid stroke treatment is important, specifically The longer a stroke goes without treatment, more of the brain dies! We also emphasized throughout the intervention pamphlet that stroke is an emergency and that stroke is treatable (Table 1). Cognitive interviews and intervention refinement After having adapted the intervention pamphlet, we interviewed patients presenting to the emergency depart- ment of the local hospital to determine whether they un- derstood the content of the intervention pamphlet and video as we intended. Materials and Methods Setting The emergency department of the same local hospital in which questionnaires were administered. Participants Eligible emergency department patients were 18 years of age or older, English speaking, likely to be discharged home as determined by the emergency department triage system, and lacked severe pain or any other condition which would distract from participation. Data collection The research coordinator screened the emergency de- partment triage list for patients who met the eligibility cri- teria described above and enrolled participants after obtaining informed consent. The research coordinator per- formed cognitive interviews in the emergency department at the patient’s bedside. Participants were asked to read the intervention pam- phlet in sections. While participants were reading the pam- phlet, the research coordinator noted any challenging words or sentence structures. After each section, the research co- ordinator asked questions to assess the meaning that the participant had obtained from reading the section. Partici- pants then viewed the intervention video in sections. After each section of the video, the research coordinator asked questions to assess the participant’s understanding of the video. (See supplementary file for cognitive interview guides for the initial versions of the pamphlet and video.) Interviews were audio recorded and transcribed. Results We enrolled 40 participants between July 2021 and November 2021. Enrolled participants had a mean (± standard deviation) age of 47 ± 16. Of the participants, 40% were men, 57.5% had at least some college educa- tion, 37.5% self-reported their race as Black, 50% self-re- ported their race as White, 12.5% self-reported their race as Other, and 10% had a self-reported history of stroke. Cognitive interviews and intervention adaptation After reviewing the interview transcripts, we identified sections of the pamphlet that were misunderstood by many participants. We modified the wording of those sections. The intervention video was edited to remove repetitive messages. We confirmed the acceptability of the modified [page 72] [Qualitative Research in Medicine & Healthcare 2022; 6:10639] Article Table 1. Link between the mechanism of action to the behavior change technique. MoA-BCT links (Carey et al., 2019) Stroke Intervention MoA BCT MoA BCT Attitude towards the behavior Information about health Symptoms will go away; Don’t wait for the signs to go away, consequences symptoms are not serious STROKE IS SERIOUS. Faster treatment means improved health! Knowledge/skills Instruction on how to perform Friends and family would not or Notice just one sign? REACT- Call 911! the behavior might not be able to recognize stroke symptoms. Beliefs about capabilities Verbal persuasion about First action after symptom onset Time to call 911! Make it the FIRST capability was calling family. thing you do! Attitude towards the behavior Salience of consequences Do not want to miss work. The longer a stroke goes without treatment, more of the brain dies! BCT, Behavior Change Technique; MoA, Mechanism of Action. No n- co mm er cia l u se on ly pamphlet and video via an additional round of cognitive in- terviews of 10 new emergency department patients who were also identified by screening the emergency depart- ment triage list. We found that participants understood these modified intervention materials as intended. Intervention materials and possible intervention outcomes The intent of the interviews was to determine partici- pants’ understanding of the intervention pamphlet and video, but we also identified possible mechanisms by which the intervention materials might produce the in- tended outcome of increasing stroke symptom recognition and intent to call 911. Evoking personal experiences with stroke Although we did not ask participants about their prior knowledge or experience with stroke, several participants recounted stories about their personal experience with fam- ily members or friends who had had a stroke. Through these stories, participants affirmed their knowledge of stroke symptoms, the impact of stroke, and evaluated their past experiences with stroke symptom recognition and response in light of what was described in the intervention material. For example, when one participant was asked their under- standing of the stroke signs presented in the pamphlet, the participant stated “Like I said, I have my mother and my grandmother who have it. I know all the signs.” Another participant stated, “If you…like if I saw that the mouth or the face was drooping, I’d be on the phone with 911 in- stantly…because that’s to me the first sign of what I saw my cousin go through.” When asked to explain what was meant by the phrase disability caused by stroke can leave a person unable to do everyday, normal activities, one par- ticipant drew on their personal observation of stroke to de- scribe the impact of stroke and stated, “…normally it changes up a person’s living in, like you said, daily activi- ties, because I know a lot of people have strokes, they lose their jobs, and they can’t return to work because they’re high risk.” When one participant was asked their under- standing of the pamphlet’s explanation of how to check for stroke signs, the participant evaluated their past experience with evaluating and responding to stroke signs: Because Mom was just like laying on the couch. I would have never thought…or did I pick her arm up? Now I know if I’m around anybody I think’s maybe having one, I’ll know to raise both their arms. And, and I did get the ambulance. Some of the family didn’t think they got there fast enough, of course…. But I think they got there in a reason- able time. The intervention materials caused the participants to re- flect on their own knowledge and experiences with stroke. Drawing upon a sense of community The intervention video contained scenes of local land- marks, and participants were asked what thoughts came to mind when seeing them. Some participants linked the local landmarks to the relevance of the intervention for their community. One participant stated, “[it means] that it’s [stroke is] happening closer than we think…showing you that it’s right here in our community.” One participant similarly stated, “When they see that, they think, this is in my neighborhood, or this is in our area.” Another par- ticipant said, “They really trying to get Flint people to care about, um, strokes.” The local landmarks evoked a sense of comfort and community, as one participant remarked, “Ah, I feel a little more comfortable just seeing it from Flint. It’s Flint people.” Another participant stated, “Oh, home. I’m a Flintstone,” and a different participant stated, “Um, that it’s kind of like our hometown.” Eliciting emotional responses Although not specifically asked, some participants ex- pressed their emotional response to the intervention con- tent. When asked their understanding of the phrase Stroke is common, some participants voiced fear in response to this fact. According to one participant, “It means that a lot of people have strokes, and it is…it is very…it is very common, yes. Actually, my-my significant other had a mild stroke years ago. It’s frightening.” Another partici- pant stated, “That’s kind of scary, too. I don’t think it should be common, but that’s life….” Other participants expressed fear when reading the sentence on the interven- tion pamphlet that stroke causes part of the brain to die. Discussion In the sections above, we described how qualitative data informed development of our stroke education inter- vention. We found that grounding questionnaire items in a behavior change theory facilitated adaptation of our in- tervention. Our questionnaire revealed attitudes that may be as- sociated with a delay in calling 911 upon stroke symptom recognition. We addressed these attitudes in the interven- tion pamphlet by adding information about the health con- sequences of stroke and about the consequences of not calling 911. Survey participants identified possible short- comings in knowledge, skills, and capability associated with calling 911 for stroke. We therefore added instruction on when to call 911 and persuasion about their capability of calling 911. We showed how a behavioral change theory—the the- ory of planned behavior—was useful in developing our stroke education intervention. Using the theory of planned behavior, we were able to identify behavioral, normative, and control beliefs that influenced the behavior of delay in [Qualitative Research in Medicine & Healthcare 2022; 6:10639] [page 73] Article No n- co mm er cia l u se on ly going to the hospital upon stroke symptom onset. We adapted our intervention to address those beliefs by identi- fying the links between MoAs, or theoretical constructs, and active components of the intervention, or BCTs. Link- ing MoAs to BCTs in the intervention adaptation or design phase can thus contribute to more effective behavioral in- terventions (Carey et al., 2019; Michie et al., 2018). Through cognitive interviews that were intended to as- sess participants’ understanding of the intervention mate- rials, we learned that the intervention materials caused some participants to reflect on their personal experience with family members or friends who had a stroke. These participants constructed their own narrative around recog- nition of stroke signs or response to stroke signs. The use of narratives in health interventions has been shown to positively change attitudes, beliefs, intentions, and behav- iors in Black American populations (Ballard et al, 2021; Houston, et al., 2011), which is a finding that has great relevance given the target population of our intervention. Through our qualitative methodology, we also found that the visual depiction of community landmarks in our intervention video elicited a sense of community belong- ing—i.e., connectedness between an individual and his/her community (Hystad & Carpiano, 2012)—in some of our participants. A dose-response relationship has been reported between a person’s perceived degree of commu- nity belonging and health behavior change, such that a stronger sense of community belonging is associated with increased likelihood of undertaking health behavior change (Hystad & Carpiano, 2012). The sense of commu- nity belonging evoked by including landmarks in our in- tervention video might, therefore, be a mechanism through which the intervention can effect positive behav- ior change. Conclusions In this manuscript, we have described how our quali- tative data informed development of our behavioral inter- vention. We showed how grounding data collection instruments in behavior change theory facilitates interpre- tation of the results and identification of MoAs that can be linked to BCTs to be included in the intervention. We also showed how qualitative data collected for interven- tion development can provide insight into how the inter- vention might produce its intended effect. Our hope is that greater reflection and more detailed reporting of how qualitative data are used in intervention development will facilitate the development of effective behavioral inter- ventions. References Aparicio, H. J., Carr, B. G., & Kasner, S. E., Kallan, M. J., Al- bright, K. C., Kleindorfer, D. O., & Mullen, M. T. (2105). Racial disparities in intravenous recombinant tissue plas- minogen activator use persist at primary stroke centers. Journal of the American Heart Association, 4(10), e001877. Ajzen, I. (1991). The theory of planned behavior. Organiza- tional Behavior and Human Decision Processes, 50(2), 179- 211. Ballard, A. M,, Davis, A., Hoffner, & C. A. (2021). The impact of health narratives on persuasion in African American women: A systematic review and meta-analysis. Health Commununication, 36(5),560-71. Boehme, A. K., Siegler, J. E., Mullen, M. T., Albright, K. C., Lyerly, M. J., Monlezun, D.J., Jones, E.M., Tanner, R., Gon- zales, N. R., Beasley, T. M., Grotta, J. C., Savitz, S. I., & Martin-Schild, S. (2014). Racial and gender differences in stroke severity, outcomes, and treatment in patients with acute ischemic stroke. Journal of Stroke and Cerebrovascu- lar Disease, 23(4), e255-61. Boutron., I, Altman, D. G., Moher, D., Schulz, K. F., & Ravaud, P. (2017) CONSORT statement for randomized trials of non- pharmacologic treatments: A 2017 update and a consort ex- tension for nonpharmacologic trial abstracts. Annals of Internal Medicine, 167(1), 40-7. Carey, R. N., Connell, L. E., Johnston, M., Rothman, A. J., de Bruin, M., Kely, M. P., & Michie, S. (2019). Behavior change techniques and their mechanisms of action: A syn- thesis of links described in published intervention literature. Annals of Behavioral Medicine, 53(8), 693-707. Colligan, E. M., Pines, J. M., Colantuoni, E., & Wolff, J. L. (2017). Factors associated with frequent emergency depart- ment use in the medicare population. Medical Care Re- search and Review, 74(3), 311-27. Craig, P., Dieppe, P., Macintyre, S., Michies, S., Nazareth, I., & Petticrew, M. (2008). Developing and evaluating complex interventions: The new medical research council guidance. British Medical Journal, 337, a1655. Duggleby, W., Peacock, S., Ploeg, J., Swindle, J., Kaewwilai, L., & HeunJung, L. (2020). Qualitative research and its im- portance in adapting interventions. Qualitative Health Re- sarch, 30(10), 1605-13. Faysel, M. A., Singer, J., Cummings, C., Stefanov, D. G., & Levine, S. R. (2019). Disparities in the use of intravenous t- Pa among ischemic stroke patients: Population-based recent temporal trends. Journal of Stroke and Cerebrovascular Diease, 28(5): 1243-51. Fjeldsoe, B. S., Miller, Y. D., O'Brien, J. L., & Marshall, A. L.(2012). Iterative development of mobilemums: A physical activity intervention for women with young children. Inter- national Journal of Behavioral Nutrition and Physical Ac- tivity, 9:151. Hsia, A. W., Edwards, D. F., Morgenstern, L.B., Wing, J. J., Brown, N. C., Coles, R., Loftin, S., Wein, A., Koslosky, S. S., Fatima, S., Sánchez, B. N., Fokar, A., Gibbons, M. C., Shara, N., Jayam-Trouth, A., & Kidwell, C.S. (2011a). Racial disparities in tissue plasminogen activator treatment rate for stroke: A population-based study. Stroke, 42(8): 2217-21. Hsia, A. W., Castle, A., Wing, J, J., Edwards, D.F., Brown, N,C,, Higgins, T.M., Wallace, J. L., Koslosky, S. S., Gibbons, M. C., Sánchez, B. N., Fokar, A., Shara, N., Morgenstern, L. B., & Kidwell, C.S. (2011b). Understanding reasons for delay in seeking acute stroke care in an underserved urban popu- lation. Stroke, 42(6), 1697-701. Houston, T. K., Allison, J. J., Sussman, M., Horn, W., Holt, C. L., Trobaugh, J., Salas, M., Pisu, M., Cuffee, Y. L., Larkin, [page 74] [Qualitative Research in Medicine & Healthcare 2022; 6:10639] Article No n- co mm er cia l u se on ly D., Person, S. D., Barton, B., Kiefe, C. I., & Hullett, S. (2011). Culturally appropriate storytelling to improve blood pressure: A randomized trial. Annals of Internal Medicine, 154(2), 77-84. Hystad, P., & Carpiano, R. M. (2012). Sense of community-be- longing and health-behaviour change in Canada. Journal of Epidemiology and Community Health, 66(3), 277-83. Johnston, S. C., Fung, L. H., & Gillum, L. A. (2001). Utilization of intravenous tissue-type plasminogen activator for is- chemic stroke at academic medical centers: The influence of ethnicity. Stroke, 32(5), 1061-8. Katan, M., & Luft, A. (2018). Global burden of stroke. Seminars in Neurolology, 38(2), 208-1. Lacy, C. R., Suh, D. C., Bueno, M., & Kostis, J. B. (2001) Delay in presentation and evaluation for acute stroke: Stroke time registry for outcomes knowledge and epidemiology (S.T.R.O.K.E.). Stroke, 32(1), 63-9. Lewin, S., Gelenton, C., & Oxman, A. D. (2009). Use of quali- tative methods alongside randomised controlled trials of complex healthcare interventions: Methodological study. British Medical Journal, 339, b3496. Lichtman, J. H., Watanabe, E., Allen, N. B., Jones, S. B., Dostal, J., & Goldstein, L. B. (2009). Hospital arrival time and in- travenous t-pa use in us academic medical centers, 2001- 2004. Stroke, 40, 3845-50. Michie, S., Carey, R. N., Johnston, M., Rothman, A. J. de Bruin, M., Kelly, M. P., & Connell, L. E. (2018). From theory-in- spired to theory-based interventions: A protocol for devel- oping and testing a methodology for linking behaviour change techniques to theoretical mechanisms of action. An- nals of Behavioral Medicine, 52(6), 501-12. Nasr, D. M., Brinjikji, W., Cloft, H. J., & Rabinstein, A. A. (2013). Racial and ethnic disparities in the use of intra- venous recombinant tissue plasminogen activator and out- comes for acute ischemic stroke. Journal of Stroke and Cerebrovascular Disease, 22(2), 154-60. National Institute of Neurological Disorders and Stroke, rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. (1995). The New England Journal of Med- icine, 333(24), 1581-7. O'Cathain, A., Croot, L., Duncan, E., Rousseau, N., Sworn, K., Turner, K. M., Yardley, L., & Hoddinott, P. (2019). Guidance on how to develop complex interventions to improve health and healthcare. British Medical Journal, 9, e029954. Robles, M. C., Springer, M. V., Corches, C.L., Burke, J. F., Lin, C. C., Oliver, A., & Skolarus, L.E. (2020). Stroke ready very brief intervention improves immediate postintervention stroke preparedness. Circulation: Cardiovascular Quality and Outcomes, 13(12), e006643. Sandelowski, M., & Leeman, J. (2012). Writing usable qualita- tive health research findings. Qualitative Health Research, 22(10), 1404-13. Siegler, J.E., Boehme, A. K., Albright, K. C., & Martin-Schild, S. (2013). Ethnic disparities trump other risk factors in de- termining delay to emergency department arrival in acute ischemic stroke. Ethnicity & Disease, 23(1), 29-34. Skolarus, L. E., Murphy, J. B., Zimmerman M. A,, Bailey, S., Fowlkes, S., Brown, D. L, Lisabeth, L.D., Greenberg, E., & Morgenstern, L. B. (2013). Individual and community de- terminants of calling 911 for stroke among african americans in an urban community. Circulation: Cardiovascular Qual- ity and Outcomes, 6(3), 278-83. Springer, M. V., Labovitz, D. L., & Hochheiser, E. C. (2017). Race-ethnic disparities in hospital arrival time after ischemic stroke. Ethnicity & Disease, 27(2), 125-32. U.S. Census Bureau. (2021). U.S. Census bureau quickfacts. Available from: https://www.census.gov/quickfacts/ flintci- tymichigan. Wei, H. G., & Camargo, C. A., Jr. (2000) Patient education in the emergency department. Academic Emergency Medicine, 7(6), 710-7. Wigginton, B., Thomson, Z. O., Sandler, C. X., & Reeves, M. M. (2020). Reflexive intervention development: Using qual- itative research to inform the development of an intervention for women with metastatic breast cancer. Qualitative Health Research, 30(5), 666-78. Williams, O., Teresi, J., Eimicke, J. P., Abel-Bey, A., Has- sankhani, M., Valdez, L., Gomez Chan, L., Kong, J., Ramirez, M., Ravenell, J., Ogedegbe, G., & Noble, J. M. (2019). Effect of stroke education pamphlets vs a 12-minute culturally tailored stroke film on stroke preparedness among black and hispanic churchgoers: A cluster randomized clin- ical trial. JAMA Neurolology, 76(10), 1211-8. Williams, O., & Noble, J. M. (2008). “Hip-hop” stroke: A stroke educational program for elementary school children living in a high-risk community. Stroke, 39(10), 2809-16. Willis, G. B. (2005). Cognitive interviewing. SAGE Publica- tions, Inc. Yardley, L., Morrison, L., Bradbury, K., & Muller, I. (2015). The person-based approach to intervention development: Appli- cation to digital health-related behavior change interven- tions. Journal of Medical Internet Research, 17(1): e30. [Qualitative Research in Medicine & Healthcare 2022; 6:10639] [page 75] Article No n- co mm er cia l u se on ly https://www.census.gov/quickfacts/flintcitymichigan https://www.census.gov/quickfacts/flintcitymichigan