DESIGN FOR EHEALTH EQUITY The development and application of design knowledge for the participatory design of eHealth interventions for people with a low socioeconomic position JASPER FABER
Design for eHealth Equity The development and application of design knowledge for the participatory design of eHealth interventions for people with a low socioeconomic position.
Design for eHealth Equity The development and application of design knowledge for the participatory design of eHealth interventions for people with a low socioeconomic position. Dissertation for the purpose of obtaining the degree of doctor at Delft University of Technology by the authority of the Rector Magnificus, prof. dr. ir. T.H.J.J. van der Hagen, chair of the Board for Doctorates to be defended publicly on Wednesday 6 November 2024 at 12:30 o’clock by Jasper Stephan FABER Master of Science in Integrated Product Design, Delft University of Technology, the Netherlands born in Bloemendaal, the Netherlands
This dissertation has been approved by the promotors. Composition of the doctoral committee: Rector Magnificus chairperson Dr. V. T. Visch Delft University of Technology, promotor Dr. H. J. G. van den Berg-Emons Erasmus University Medical Center, promotor Dr. J. J. Kraal Delft University of Technology, copromotor Independent members: Prof. Dr. P. J. Stappers Delft University of Technology Prof. Dr. H. M. C. Kemps Eindhoven University of Technology Prof. Dr. C. Bolman Open University of the Netherlands Dr. S. M. Kelders University of Twente Dr. L. E. Oldenhof Erasmus School of Health Policy and Management Prof. Dr. R. H. M. Goossens Delft University of Technology, reserve member This work was conducted within the framework of the Medical Delta program, eHealth and self-management for a healthy society, with funding support from Capri Hartrevalidatie. Medical Delta is gratefully acknowledged for providing financial support for the printing costs of this thesis. Keywords: eHealth, behavior change, participatory design, low socioeconomic position, cardiac rehabilitation Printing and layout by: Ridderprint | www.ridderprint.nl Design: Yakeem Pfluger, Persoonlijk Proefschrift ISBN 978-94-6506-398-0 Copyright© 2024 by J. S. Faber
Contents Chapter 1: Introduction 9 1.1 Background 10 1.2 Problem statement and knowledge gaps 11 1.3 Aim of dissertation 13 1.4 Project background 14 1.5 Thesis outline 15 Part A: Knowledge inquiry 19 Chapter 2: Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 21 2.1 Introduction 23 2.2 Methods 25 2.3 Results 30 2.4 Discussion 39 2.5 Conclusion 45 Chapter 3: Participatory design for and with patients with low health literacy 49 3.1 Introduction 51 3.2 Methods 53 3.3 Results 56 3.4 Discussion 61 3.5 Conclusion 63 Part B: Development of the knowledge tool 65 Chapter 4: Guide development for eHealth interventions targeting people with a low socioeconomic position 67 4.1 Introduction 69 4.2 Methods 71 4.3 Results 77 4.4 Discussion 83 4.5 Conclusion 86
Part C: Application Cycle 89 Chapter 5: Application of the Inclusive eHealth Guide during the development of an eHealth intervention for and with cardiac patients with a low socioeconomic position 91 5.1 Introduction 93 5.2 Methods 95 5.3 Results 103 5.4 Discussion 105 5.5 Conclusion 108 Chapter 6: Feasibility and effects of an eHealth intervention to support patients with a low socioeconomic position during their waiting period preceding cardiac rehabilitation 111 6.1 Introduction 113 6.2 Methods 114 6.3 Results 120 6.4 Discussion 125 6.5 Conclusions 129 Chapter 7: General discussion 131 7.1 Part A – Knowledge inquiry: Attitudes and participatory design 132 7.2 Part B – Knowledge tool: The development of the Inclusive eHealth Guide 134 7.3 Part C – Application cycle: Applying the Inclusive Health Guide in cardiac rehabilitation. 135 7.4 Implications 136 7.5 Strengths and limitations 145 7.6 Future directions 147 7.7 General conclusion 151 References 154 Appendices 174 Summary 190 Samenvatting 194 Acknowledgments 198 About the author 202 List of publications 203
CHAPTER 1 Introduction
Chapter 1 10 1.1 Background Imagine a world in which you live nearly 30 years less in good health because of the socioeconomic circumstances in which you live. While this sounds like a dystopian scenario, it actually is a harsh reality. Currently, in the Netherlands, there is an average 7-year difference in lifespan between individuals with the highest and lowest levels of education (RIVM, 2017). This gap widens to 27 years when considering a person’s health span, which refers to the years of good health they can enjoy (RIVM, 2017). People with lower levels of education tend to develop non-communicable chronic diseases (NCDs, e.g., cardiovascular disease, diabetes, and obesity) at an earlier age compared to their more highly educated counterparts (Mackenbach et al., 2008; Mackenbach et al., 2019; Stringhini et al., 2017). Similar disparities are observed across varying income (Jarvandi et al., 2012; Kaplan et al., 1996) and occupation levels (Ravesteijn et al., 2013; Volkers et al., 2007). Together, individuals with lower education, income, and occupational levels are referred to as those with a low socioeconomic position (SEP) (Braveman et al., 2005; Havranek et al., 2015). The ‘health gap’ between socioeconomic classes displays one of the most concerning examples of inequality within our current society. Moreover, the higher prevalence of NCDs among people with a low SEP leads to prolonged healthcare needs, a challenge that extends to both the individual and society (Adler & Stewart, 2010; Drewnowski et al., 2014; Latulipe et al., 2015; Mackenbach et al., 2008; Shishehbor et al., 2006). A major reason for the higher prevalence of NCDs in groups with a low SEP is the greater prevalence of an unhealthy lifestyle compared to groups with a high SEP. Studies have shown that people with a low SEP are more likely to display lower levels of leisuretime physical activity (Beenackers et al., 2012; Gidlow et al., 2016), increased television viewing time (Clark et al., 2010; King et al., 2010), poorer diet (Darmon & Drewnowski, 2008), and more smoking behavior (Hiscock et al., 2012) compared to people with a high SEP. A multitude of interconnected factors, including stress, low literacy, poor living conditions, poor parenting, lack of social support, and low self-efficacy, contribute to this unfavorable health behavior (Marmot, 2005; Pampel et al., 2010). The complexity of these interconnected factors makes it challenging to address the underlying causes of an unhealthy lifestyle. Lifestyle interventions have shown promising outcomes in areas such as physical activity, diet, and quitting smoking in the general population. Interventions focusing on diet and exercise have led to significant changes in body weight and physical activity (Greaves et al., 2011). In addition, behavioral approaches have proven to generally reduce tobacco usage (Stead et al., 2016). The rise of eHealth technologies has further transformed the approach to lifestyle interventions in recent years. Through
Introduction 11 1 digital platforms, mobile applications, and wearable devices, eHealth interventions can potentially make health behavior more engaging and accessible, particularly when they are based on theoretical frameworks and behavior change techniques (Webb et al., 2010). While traditional interventions have been shown to improve health behavior, several studies also emphasize that eHealth interventions can improve physical activity, diet, and sedentary behavior (Schoeppe et al., 2016) and lead to smoking cessation (Taylor et al., 2017; Whittaker et al., 2016). While traditional and eHealth interventions hold substantial promise for improving health behavior in the general population, evidence suggests that their success could be more evident in low socioeconomic populations (Reiners et al., 2019; Yamin et al., 2011). 1.2 Problem statement and knowledge gaps Traditional lifestyle interventions have been largely unsuccessful in changing the behavior and improving the health of people with a low SEP (Bull et al., 2015; White et al., 2009). This can be attributed to several constraints related explicitly to low-SEP groups, such as stressful life situations (Marmot, 2005), accessibility issues (Coupe et al., 2018), inadequate social support (Moroshko et al., 2011), experienced stigma and distrust in healthcare (Armstrong et al., 2007), and low health literacy (Paasche-Orlow & Wolf, 2007). Another potential reason is that some individuals within the target group are less willing to engage in health-promoting behavior (Hardcastle et al., 2015). After all, it is worth questioning whether academics’ views on health genuinely resonate with the values, beliefs, and priorities of those we aim to help. We might be operating from a standpoint that equates health with longevity and quality of life. At the same time, some individuals in the target group might prefer living a fulfilling life, even if it means that it may be shorter or less “healthy” by our standards (Heutink et al., 2010; Wardle & Steptoe, 2003). This dissonance could lead to a lack of willingness from the target group to engage with “our” interventions. eHealth interventions possess inherent qualities that could mitigate some barriers regarding lifestyle interventions for low-SEP groups. eHealth platforms are often customizable, allowing for adaptations that better suit the problematic life situations frequently encountered by low-SEP groups. They can provide information in accessible multi-media formats, aligning with the needs of those with low (health) literacy levels (Michie et al., 2009). Moreover, the virtual nature of eHealth makes it more accessible than traditional interventions, as it can be accessed from any location with an internet connection, thereby partially bypassing accessibility issues (Hill & Powell, 2009). However, despite these advantages, current eHealth interventions seem to need to catch up to their potential in low-SEP populations. Several studies have indicated that eHealth interventions
Chapter 1 12 remain largely ineffective for people with a low SEP (Veinot et al., 2018). Multiple key factors must be in place for an eHealth intervention to be effective. First, accessibility is crucial; the intervention should reach its intended audience and be supported by the necessary technological infrastructure and device availability. Second, the target group must find the intervention acceptable, indicating willingness and ability to use it. Finally, adherence is essential; the target group should consistently engage with the intervention throughout its intended duration. Currently, eHealth interventions need to catch up in reaching and retaining adherence among individuals with a low SEP (Reiners et al., 2019; Yamin et al., 2011). Additional barriers that could account for this include inadequate digital (health) literacy (Cashen et al., 2004; Estacio et al., 2019), skeptical or less confident attitudes toward technology (Choi & Dinitto, 2013) and lack of resources (Cashen et al., 2004). Bottom-up, participatory approaches serve as a transformative lens to address the challenges, needs, skills, and preferences of the target group by actively involving them in the design process of eHealth interventions (van Gemert-Pijnen et al., 2011). Essentially, participatory approaches shift the paradigm from designing “for” to designing “with” the target group. This orientation is deeply rooted in human-centered design and design research, as it prioritizes the lived experiences, insights, and contextual nuances of the users (Sanders & Stappers, 2008; Spinuzzi, 2016). Given the complex interplay of factors contributing to low uptake and engagement of eHealth interventions in low-SEP groups, tailored participatory approaches are recommended above top-down, one-size-fits-all strategies (Braveman et al., 2005). Indeed, participatory approaches have shown success in ensuring the intervention is aligned with the specific challenges, skills, and needs of the target group and may facilitate the uptake of the developed interventions (Lee et al., 2022; Neuhauser, 2017). Therefore, integrating participatory design in developing eHealth lifestyle interventions for low-SEP groups could be a crucial strategy to reach equitable eHealth interventions. However, these approaches can be challenging when working with hard-to-reach groups, such as those with a low SEP. eHealth professionals (e.g., designers, developers, researchers, and care providers) often face practical challenges in reaching these groups for participatory design, including low health literacy, distrust toward the research team, cultural differences, and stigmatization (Bonevski et al., 2014; Stowell et al., 2018). These barriers can be time-consuming and challenging to overcome, especially under tight budgets and timelines. While the body of scientific knowledge on addressing the barriers related to the participatory design for and with groups with a low SEP is growing (Bonevski et al., 2014; Stowell et al., 2018), there has been limited effort to translate this knowledge into practical guides that specifically aid in the design of eHealth interventions for and
Introduction 13 1 with low-SEP groups. Practical guides are essential because they act as roadmaps for eHealth professionals, condensing complex research findings into actionable steps that can be easily implemented, particularly when resources are constrained (Graham et al., 2006). Indeed, there are existing practical guides for eHealth development (van GemertPijnen et al., 2011) and for participatory design more broadly (Sanders et al., 2010; Spinuzzi, 2016), but none of these resources focus on the complexities of engaging specifically with low-SEP groups in such a process. Some guides do target low-SEP groups specifically, yet they often limit their scope by focusing primarily on addressing literacy-related barriers. This involves addressing digital literacy by designing more userfriendly and understandable interfaces and addressing traditional literacy by ensuring that the information provided is easily comprehensible (Choi & Dinitto, 2013). However, successful behavior change goes beyond these surface-level factors; it also requires tackling the more fundamental elements that underpin motivation—such as contextual (e.g., accessibility, social support, and influences), psychological (e.g., self-efficacy, perceived barriers), and emotional (e.g., stressful life situations) factors. 1.3 Aim of dissertation The equity challenges currently experienced during the design of eHealth interventions result in these interventions often being designed as a one-size-fits-all solution, which unintentionally favors those with high health literacy, motivation, and willingness to engage with the intervention and access to technology and the internet. However, this approach leaves out those who may need eHealth interventions the most: people with a low SEP. This leads to the possibility that eHealth interventions are not helping to address health disparities but instead exacerbate them. Therefore, there is a pressing need for a comprehensive and practical tool that integrates the known barriers and facilitators regarding inclusive eHealth design to inform the design of eHealth interventions aligned with the needs of low-SEP groups. The main aim of this dissertation was to develop a practically applicable knowledge tool that aligns with the needs of professionals and helps to facilitate the designing of eHealth interventions tailored to people with a low SEP. To accomplish this aim, this thesis consists of several studies that relate to each other based on the knowledge-to-action (KTA) framework (Graham et al., 2006). This framework is often used in healthcare research to move knowledge into actionable strategies. It involves the three dynamically interacting concepts of knowledge inquiry, the development of a knowledge tool, and action cycles (Figure 1.1). To facilitate the main aim of developing the knowledge tool, we address two additional aims. First, we engaged in knowledge inquiry by addressing critical knowledge
Chapter 1 14 gaps about the attitudes of individuals with a low SEP toward health, healthcare, and eHealth and how participatory design can better engage these groups in research and design processes. Second, we engaged in an action cycle representing a design process of an eHealth intervention in the specific context of cardiac rehabilitation (CR). This involves adapting and applying the knowledge tool in a real-world setting to assess the applicability of the knowledge and evaluate the outcomes of the resulting intervention. These outcomes serve to refine and improve the knowledge tool’s applicability in specific settings, acting as an initial step in the iterative refinement needed to sustain knowledge use in future studies to develop eHealth interventions for low-SEP populations. Figure 1.1 Visualization of the thesis approach modified from KTA framework presented in Field et al. (2014). 1.4 Project background This project was a collaborative effort funded by the Medical Delta as part of the theme “eHealth and self-management for a healthy society” and Capri Cardiac Rehabilitation. It involved multiple institutions, including Delft University of Technology, Erasmus University Medical Center, Leiden University, Leiden University Medical Center, and Capri Cardiac Rehabilitation. The project’s primary objective was to develop and evaluate a knowledge tool designed to support professionals developing eHealth interventions together with
Introduction 15 1 and for low-SEP populations. Two PhD projects contributed to this goal. The project that was carried out at Leiden University and Leiden University Medical Center was performed by Isra Al-Dhahir. She adopted a broad, top-down approach and explored barriers and facilitators in eHealth design for low-SEP groups, primarily through a literature review and consultation with professionals. She later evaluated the acceptance of the knowledge tool’s content among professionals. The PhD project that constitutes this thesis was carried out at the Delft University of Technology, Erasmus University Medical Center, and Capri Cardiac Rehabilitation. The approach adopted a bottom-up perspective, in contrast to Isra Al-Dhahir’s approach, by concentrating on participatory design with the target group in specific contexts. Although each PhD candidate largely worked independently, developing the knowledge tool was a collaborative effort. It aimed to integrate the top-down (professionals) and bottom-up (target group) perspectives. 1.5 Thesis outline In line with the three aims presented in section 1.3, this thesis is divided into three parts that address knowledge inquiry, development of the knowledge tool, and the application cycle (Figure 1.2). Figure 1.2 Thesis outline schematic and chapter division. SEP: socioeconomic position; CR: cardiac rehabilitation.
Chapter 1 16 Part A: Knowledge inquiry: Attitudes and participatory design In the research described in Chapter 2, we investigated whether people with a low SEP are willing to improve their health through technology by exploring their attitudes toward health, healthcare, and eHealth through a community-based participatory research approach. Nine individual attitude profiles and two general attitudes regarding health, healthcare, and eHealth are described. In Chapter 3, we describe a participatory design process within a case study aimed to develop an eHealth intervention to improve medication adherence of asthma patients with low health literacy. We demonstrate the challenges of performing participatory design with hard-to-reach groups and propose three participatory design strategies that could facilitate such a participatory process. Part B: Development of the knowledge tool: The Inclusive eHealth Guide In Chapter 4 we describe the development of the knowledge tool: The Inclusive eHealth Guide. We describe how the research of part A and the work performed by Isra Al-Dhahir is synthesized into a practical guide to support the development of eHealth interventions for people with a low SEP. We describe how we developed the guide with professionals working with eHealth and people with a low SEP through participatory design to ensure the guide matches their practical needs. We identified 16 requirements the guide needed to comply with and developed the guide accordingly. The result is an open-ended website with recommendations, user portraits, practical knowledge, examples, and references. Part C: Application cycle: Applying the Inclusive eHealth Guide during the design of an eHealth intervention for CR patients with a low SEP. In Chapter 5 we demonstrate how the Inclusive eHealth Guide was applied in a participatory design process of an eHealth intervention for patients with a low SEP through a specific case study within the context of CR. This case study allowed us to explore specific use cases and challenges that provided insight into the application of the guide. CR provided a valuable setting for this study as it often focuses on lifestyle changes and involves a range of interventions that could be delivered effectively through eHealth. In Chapter 6, we evaluate the feasibility of the resulting intervention using a mixedmethod randomized controlled feasibility study. This chapter sheds light on the potential
Introduction 17 1 value of the guide while developing eHealth interventions tailored toward people with a low SEP. General Discussion Finally, in Chapter 7, we reflect on the findings from the three parts of this dissertation and their implications for the design of equitable eHealth. It also provides a discussion on the strengths and limitations and future directions.
PART A: KNOWLEDGE INQUIRY Attitudes and participatory design
This chapter is published as: Faber, J. S., Al-Dhahir, I., Reijnders, T., Chavannes, N. H., Evers, A. W. M., Kraal, J. J., van den Berg-Emons, H. J. G, & Visch, V. T (2021). Attitudes Toward Health, Healthcare, and eHealth of People with a Low Socioeconomic Status: A Community-Based Participatory Approach. Frontiers in Digital Health, 3. doi: 10.3389/ fdgth.2021.690182 All research data and code supporting the findings described in this chapter are available in 4TU. Centre for Research Data at 10.4121/7a2ca4e2-acca-4585-9d08-85d9f8139896 CHAPTER 2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position Building on the foundation in the introduction, this chapter delves into the first question of why eHealth interventions may be less successful for individuals with a low socioeconomic position (SEP). A limitation in existing research is its oversight of the target group’s perspectives. To address this, our study adopted a community-based participatory research approach, emphasizing the direct involvement of those with a low SEP. Through this, we aimed to understand the target group’s attitudes toward their health, healthcare, and eHealth. We present nine distinct profiles reflecting varied attitudes toward these areas, ultimately distilling them into two overarching attitudes: the “Optimistically Engaged” and “Doubtfully Disadvantaged”. Our findings suggest that the assumption of a uniform unwillingness among people with a low SEP to engage with healthy behavior and eHealth interventions may not be entirely accurate. Our research indicates that there is diversity in attitudes within the low-SEP group and that the majority exhibits a willingness to engage in health-promoting behaviors. This suggests that the issue could stem more from the design of interventions, which might not adequately address the diverse needs of the group, than an unwillingness to participate in eHealth and healthy behaviors.
Chapter 2 22 Abstract Background | Low socioeconomic position (SEP) is associated with a higher prevalence of unhealthy lifestyles compared to a high SEP. Health interventions that promote a healthy lifestyle, like eHealth solutions, face limited adoption in lowSEP groups. To improve the adoption of eHealth interventions, their alignment with the target group’s attitudes is crucial. Objective | This study investigated the attitudes of people with a low SEP toward health, healthcare, and eHealth. Methods | We adopted a mixed-method community-based participatory research approach with 23 members of a community center in a low-SEP neighborhood in the city of Rotterdam, the Netherlands. We conducted a first set of interviews and analyzed these using a grounded theory approach resulting in a group of themes. These basic themes’ representative value was validated and refined by an online questionnaire involving a different sample of 43 participants from multiple community centers in the same neighborhood. We executed three focus groups to validate and contextualize the results. Results | We identified two general attitudes based on nine profiles toward health, healthcare, and eHealth. The first general attitude, Optimistically Engaged, embodied approximately half our sample and involved light-heartedness toward health, loyalty toward healthcare, and eagerness to adopt eHealth. The second general attitude, Doubtfully Disadvantaged, represented roughly a quarter of our sample and was related to feeling encumbered toward health, feeling disadvantaged within healthcare, and hesitance toward eHealth adoption. Conclusions | The resulting attitudes strengthen the knowledge of the motivation and behavior of people with a low SEP regarding their health. Our results indicate that negative health attitudes are not as evident as often claimed. Nevertheless, intervention developers should still be mindful of differentiating life situations, motivations, healthcare needs, and eHealth expectations. Based on our findings, we recommend eHealth should fit into the person’s daily life, ensure personal communication, be perceived usable and useful, adapt its communication to literacy level and life situation, allow for meaningful self-monitoring and embody self-efficacy enhancing strategies.
2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 23 2.1 Introduction Low socioeconomic position (SEP) is associated with a higher prevalence of unhealthy lifestyles compared to a high SEP (Stringhini et al., 2010). Consequently, people with a low SEP are at increased risk of chronic diseases (e.g., cardiovascular disease, diabetes, and obesity) (Drewnowski et al., 2014; Mackenbach et al., 2008; Shishehbor et al., 2006). eHealth interventions such as monitoring devices, online communication platforms, and serious games have been proven effective in changing behavior and promoting a healthy lifestyle in various domains. However, these interventions are less successful in changing the behavior of people with a low SEP due to low reach, less adherence during the intervention or less effectiveness of the interventions (Bull et al., 2015; Busch & van der Lucht, 2012; Busch & Schrijvers, 2010; Michie et al., 2009; Reiners et al., 2019). A crucial factor in facilitating the adoption, and therefore success, of eHealth interventions, is the alignment with a person’s attitude toward using this technology (Garavand et al., 2016; Venkatesh et al., 2003). Moreover, successfully achieving a lifestyle change, a primary goal of such interventions, requires the person to have a positive attitude toward their health and health services (Ajzen, 1991). eHealth is designed to expect its intended users to have a positive and pro-active health attitude. However, considering the growth of current health inequalities, such interventions would have a bigger impact when they can support groups not sharing these attitudes. A multitude of studies point out that people with a low SEP have unfavorable attitudes toward their health, healthcare, and eHealth. For instance, Wardle and Steptoe (2003) found that health attitudes within the low-SEP groups are specifically characterized by a lower consciousness about health and less often thinking about the future. Other studies have identified more passive attitudes toward healthcare (Schröder et al., 2018) and less confident attitudes toward digital health interventions (Choi & Dinitto, 2013) within low-SEP groups. Nevertheless, there is insufficient evidence to inform researchers and designers about these attitudes. The complexity of studying health values within contrasting sociodemographic environments poses various emotional and ethical challenges such as perceived harms, feelings of stigmatization, and anxiety toward research and the research team (Birks et al., 2007; Bonevski et al., 2014; Stuber et al., 2020). As a result, hard-to-reach groups are minimally included in research efforts. Moreover, existing evidence is difficult to generalize toward other contexts. Measurements of attitudes are highly context-dependent and are expected to differ by country, setting, and time (Eagly & Chaiken, 2007). Financial wellbeing and accessibility of health sources, for example, will not have a profound impact within countries that have unemployment
Chapter 2 24 funds, state-funded healthcare, and relatively good public transportation. Consequently, we have a lack of evidence to support the research and design of eHealth interventions that align with the attitudes of people with a low SEP. The rise of eHealth in current healthcare systems opens up exciting new possibilities to improve healthcare quality and efficiency. However, with the increased use of technical innovations and digital systems come unintended, unpredictable, and adverse consequences for individuals. Due to the underrepresentation of these specific societal groups, interventions are minimally aligned toward their attitudes. Consequently, these interventions face the risk of not being adopted and therefore unintentionally contribute to rising health inequalities. Researchers and designers should carry the responsibility to harness the potential of eHealth to create benefit for all groups in society, not merely for those that are motivated to perform a healthy lifestyle (Viswanath & Kreuter, 2007). To engage the target group in the research process, an approach is needed that is comprehensive, culturally sensitive, and builds upon a relationship-based personal approach (Stuber et al., 2020). Community-based participatory research (CBPR), a socio-culturally sensitive approach, which creates a trustful and long-lasting relationship between researcher and participant, has been effectively applied in culturally contrasting contexts (Israel, 2013; Unertl et al., 2016). For example, Henderson et al. (2013) successfully implemented a CBPR approach to develop a tailored web-based diabetes self-management tool in a low-resource setting in the United States. Such an approach can engage hard-to-reach groups in the research process yet has not been applied in the context of attitudes in low-SEP groups. In addition, focusing on a community instead of a person’s individual characteristics is increasingly being recognized as a valuable approach. Studies that focus on these characteristics imply that these are the cause of poor health outcomes, which carries the risk of increasing stigma (Auerswald et al., 2017). It is becoming increasingly known that contextual community factors, such as the availability of healthy food, experiences of discrimination, and neighborhood poverty, also have a significant relation to poor health outcomes (Schüz, 2017; Winkleby & Cubbin, 2003). The resulting knowledge could improve the alignment of health services toward attitudes of low-SEP populations, thereby facilitating their adoption. Currently, eHealth interventions aimed at these populations have only been minimally tailored, for example, by simplifying text and including images and videos (Kock et al., 2019). However, there is currently limited evidence reporting how interventions could be tailored toward psychological characteristics, such as attitudes with regard to eHealth. Although some
2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 25 studies report on the relationship between attitudes and interventions (Bukman et al., 2014; Coupe et al., 2018), the resulting knowledge is difficult to apply in the design of interventions directly. Forms of practical knowledge, such as data-driven patientprofiles, have been used in the past to tailor content, context, and delivery of care toward individual preferences (Dekkers & Hertroijs, 2018). Yet, such a form of knowledge has not been developed for attitudes of people with a low SEP toward their health, healthcare, and eHealth in general. This study aims to achieve design-relevant knowledge about the attitudes of people with a low SEP toward their health, healthcare, and eHealth. To achieve this, we took a community-based participatory research approach to facilitate responsible engagement of the target group in the research process. The resulting knowledge can facilitate the design and alignment of health services toward the different attitudes of low-SEP populations. This will result narrowing current health disparities by developing interventions that are more acceptable, satisfactory, and user-friendly. 2.2 Methods Our methodology revolved around the principles of CBPR. CBPR is a partnership approach to research that equitably involves community members, organizational representatives, and researchers in all aspects of the research process (Israel, 2013). Our CBPR approach consisted of three separate phases (Figure 2.1) in which the outcomes of each phase were used in the next. Figure 2.1 Overview of project phases and corresponding methods, materials, and analysis products.
Chapter 2 26 2.2.1 Sampling and recruitment We initiated our collaboration with a community center located in a neighborhood in Rotterdam, the Netherlands. The neighborhood was selected based on its neighborhood SEP, a combined measure of neighborhood income, education, and occupation (CBS, 2019). The neighborhood in which the community center is situated has been one of the lowest scoring neighborhoods on livability; a combined measure of its social, physical and safety index (Gemeente Rotterdam, 2020). The area therefore is on the agenda as one of the focus-neighborhoods of the municipality of Rotterdam. Sixty-eight percent of the inhabitants have a migration background, compared to 52% in Rotterdam. In addition, 59% of the households have a low income compared to 52% in Rotterdam. Finally, 34% of the inhabitants have a low education, compared to 32% in Rotterdam (de Graaf, 2018). The participants were sampled based on their affiliation with the community center and their living area (neighborhood SEP). The community center situated in this neighborhood facilitates inhabitants that struggle with fundamental aspects of their life. They focus on poverty, occupation, living, social contacts, upbringing, and safety. We included participants living in the selected neighborhood with the following affiliations with the community center: (1) Visitors (Vi): Persons who visit the community center regularly and require support. (2) Volunteers (Vo): Unemployed persons who performed volunteering work in the community center in exchange for state funding. (3) Key persons (Kp): Social workers who have close relationships with the community members. In this study, Kp’s were not considered as part of the target group as they are employed at the community center and are in the role of providing support. However, since they interact with Vi’s and Vo’s on a daily basis, we included them to learn about attitudes within the community from the Kp’s perspective. In that light, we did not include Kp’s in the second phase of the study as we were solely interested in acquiring a deeper understanding of the attitudes we observed in the first phase. Finally, it should be noted that Vo’s could visit the community center as Vi’s as well. For this study, we considered persons a Vo when they had at least one regular weekly shift at the community center. In phase one, we sampled the participants conveniently and recruited them faceto-face at the community center. In the second phase, Vo’s and Vi’s were purposively sampled and recruited face-to-face. In phase three, we recruited participants for the questionnaire through an advertisement on the community center’s Facebook page and WhatsApp group and through Kp’s of various community centers within the same neighborhood. The participants for the focus groups were recruited through a question attached at the end of the digital questionnaire and by approaching them face-to-face at the community center. Because of the come-and-go nature of the community center,
2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 27 some participants frequently visiting the community center participated in each of the three phases, while others only participated in one. 2.2.2 Procedure and materials In phase one, we aimed to form a trustful research partnership with the community and narrow down the research scope by simultaneously exploring and identifying specific research directions. We initiated the partnership by attending community gatherings and organizing health-themed lunch events at the community center. Such immersive activities have been used and proven successful in creating a relationship in various other CBPR efforts (Israel, 2013). During these activities, we addressed the research scope by engaging in unstructured interviews with community members individually. Based on an initial literature review, a backlist of topics guided the interviews and helped to steer them toward our research questions (Wilson, 2014). We divided the topic questions into three overarching research themes: attitudes toward health, healthcare, and eHealth. For example, we explored the attitude toward health with questions such as “How important is it for you to live long?” Questions such as “What do you think of your doctor’s advice?” and “What do you think of a technology that could help you live healthier?” referred to the attitude toward healthcare and eHealth, respectively. Data was captured by taking quick field notes during the visits and elaborating on them into comprehensive reports directly afterward. In phase two, we investigated the specific directions resulting from the first phase more extensively through semi-structured interviews. In contrast to unstructured interviews, these interviews are more formal and intimate, which comes conjointly with emotional challenges when discussing sensitive and stigma-inducing topics (Renzetti & Lee, 1993). Therefore, the pre-established trusting relationship between participants and the researcher was an essential facilitator. The interviews (N = 10) were conducted at the community center in a separate room with the participants individually and took approximately thirty minutes. We developed the interview guide structuring the interview based on the research directions from the first phase. For example, we explored how the participants perceived their health with the question: “What do you have to do to become 100% healthy?” The data was collected by audio-recording and transcribing the interviews. We progressed to the subsequent phase when we achieved theoretical saturation. In the third phase, we validated and generalized the insights from phase two and discovered general attitudes through the data-driven profiles. Meanwhile, we had to consider the newly introduced COVID-19 regulations. Therefore, we developed a digital questionnaire which we distributed digitally to members of community centers. This questionnaire presented the resulting insights of the second phase and asked the
Chapter 2 28 participants to rate the extent to which they felt the insight reflected themselves. By distributing this online questionnaire, we reached a more extensive and diverse sample. In addition, we gathered quantitative data that we used to validate our preliminary results and develop data-driven attitude profiles. Questionnaires, frequently being long and textual, are at risk of being disengaged by their participants as they depend on reading comprehension. This risk holds especially true for participants with lower education attainment. The use of graphics in previous studies has successfully engaged low-literate participants with questionnaires (Maceviciute et al., 2019). Therefore, we synthesized our insights toward visual two-frame storyboards. We executed several pilot sessions to reduce the chance that participant understandings would not match the story’s original implication and adjusted any inaccuracies accordingly. A 6-point Likert scale accompanied the stories in the questionnaire. The stories were grouped under their representative category. Each group concluded with an open-ended question regarding the corresponding category. See Figure 2.2 for an example of the consciousness page in the questionnaire. In addition, we asked participants to report their age, gender, educational attainment, and neighborhood. The online questionnaire was designed and distributed using Qualtrics. Finally, we performed focus groups to validate and contextualize the profiles that resulted from the questionnaire. Each focus group meeting consisted of three to four participants, lasted for approximately one hour, and was audio-recorded. The focus groups took place in a large and ventilated room at the community center that allowed maintaining 1,5-meter distance between the participants according to the COVID-19 regulations. 2.2.3 Data analysis In phases one and two, we transcribed the audio recordings verbatim and analyzed them together with the field reports and qualitative questionnaire data using the software package ATLAS.ti. Throughout the qualitative analysis, we followed the grounded-theory approach outlined by Corbin and Strauss (1990), as it is specifically useful in discovering social processes focused on social change and improvement (de Boer, 2011). We continuously broke down the data and collected it under similar content in the form of concepts using open coding techniques. For example, we created the concept perceived barriers to refer to quotes where participants mentioned barriers that decreased their motivation to perform healthy behavior. Subsequently, we grouped related concepts toward overarching categories based on attitude theory constructs such as Beliefs, Feelings, Motivation, and Opportunity (Eagly & Chaiken, 2007; Fazio & Towles-Schwen, 1999). Two independent researchers (JF and IA) developed the concepts together to improve the reliability of the results.
2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 29 Figure 2.2 An example of the visual questionnaire distributed in phase 3. The storyboards represent the concepts found within the consciousness category.
Chapter 2 30 In phase three, we imported the Likert scores of the concepts and categories obtained from the questionnaire as variables into SPSS. We performed k-means cluster analyses on the concepts based on Euclidian distance for health, healthcare, and eHealth with SPSS. We determined the optimal number of clusters with the Elbow method using the factoextra and NbClust packages in R. We used an ANOVA to identify the concepts with significant (p < 0,05) contribution to the cluster segmentation. The concepts with an insignificant contribution were removed from further analysis. To validate the clusters, we performed an ANOVA with the category scores as independent and the clusters themselves as dependent variables. Using a post-hoc ANOVA, we defined the resulting clusters based on significant differences between mean scores of the concept variables. We created profiles by further clarifying and enriching these clusters by analyzing the qualitative data from the questionnaire and focus group discussions. This was done by extending on the existing categories and concepts and using the same grounded-theory approach as used in previous phases. Finally, we performed a principal component analysis (PCA) using the factoextra package in R to discover correlations between concepts from different profiles. 2.2.4 Ethics The study protocol was approved by the Human Research Ethics Committee of Delft University of Technology (approval numbers 953, 1064, and 1141). Through our relationship-based CBPR approach we aimed to limit the impact of emotional and ethical challenges such as perceived harm, feelings of stigmatization, and anxiety toward research and the research team. In the first phase, we briefed our participants orally about the nature of the study as a formal written consent in this first introduction phase would obstruct a trustful interaction. The participants provided their consent verbally to the researcher (JF). In phases 2 and 3, when the relationship was more solid, written informed consent was provided. 2.3 Results 2.3.1 Participants During the unstructured interviews in the first phase, we spoke with 16 different members of the community center. These members consisted of eight Vi’s, two Vo’s and six Kp’s. In the second phase, we interviewed five Vo’s and five Vi’s. In phase one and two, we did not collect demographic data. In the third phase, 45 participants responded to the questionnaire. From these latter responses, we excluded three participants not living in
2 Attitudes toward health, healthcare, and eHealth of people with a low socioeconomic position 31 our target neighborhood from analysis. The participants’ mean age in this final sample was 52 years (SD = 11.10), 21% was male and 79% was female. Most of this sample (67%) had a low to medium education, which was defined as not having attained a follow-up education. This is relatively high compared to 59% in the Netherlands. Ten participants participated in the focus groups: two Kp’s, five Vo’s and three Vi’s. 2.3.2 Phase 1 and 2 – Exploration and specification The unstructured interviews of phase one yielded 30 pages of field reports containing 85 coded segments. The semi-structured interviews of phase two yielded ten interview transcripts containing 359 coded segments. The grounded theory analysis resulted in 58 concepts within nine categories related to attitudes toward health, healthcare, and eHealth. Examples of the categories found are consciousness about health, motivation to perform healthy behavior and satisfaction toward healthcare. Examples of identified concepts are Interest in health, Perceived barriers, and loyalty toward healthcare provider. Table 2.1 presents an overview of the concepts and categories included in the third phase. We excluded categories conveying a limited number of concepts (N = 1) or not fitting the attitude theory constructs (N = 1). We selected the concepts to include (N = 29) in the third phase based on the number of associated coded segments and discussion by the two analysts. Table 2.1 Concepts (N = 29) under their categories (N = 9) resulting from grounded theory analysis including number (N) of associated codes, description, and exemplary quotes (translated). Concept N Description Quote Category: Health Beliefs [Being healthy is…] Working on health 30 When one frequently performs healthy behavior such as physical activity and maintaining a healthy diet. “I’m eating healthy, I only drink in the weekends […] I frequently do yoga […] Yes I think that I’m being healthy” (Vo3) Absence of complaints 12 The absence of complaints, symptoms, and disease. “There was a time when I was heavier. I struggled with shortness of breath and cholesterol, and I don’t know what else.” (Vi6) Participation 12 Being able to go out and participate in society. “The first thing you have to do is to get up early and just go somewhere […] Otherwise you will not have active contacts with people who provide a positive influence or create chances for you” (Vi3) Balance 10 Maintaining a balance between unhealthy and healthy behavior. “I have other things. I don’t drink for instance so that makes up for it quite a lot.” (Vo5) Life under control 10 When you have a roof above your head and no major financial or social struggles. “Unhealthy is when you don’t have a roof above your head, and you have to roam the streets.” (Vi5)
Chapter 2 32 Concept N Description Quote Category: Consciousness [about health is impacted by…] Complaints 19 The experience of health-related symptoms and complaints. “I haven’t visited the doctor in 30 years. My last painkiller I used when I was at high school” (Vo5) Incident 13 The consideration of a healthrelated incident in the past. “Yes, a significant impression. Before that [the incident] I was just flying blind.” (Vo3) Concern 11 The extent to which one is concerned about their health. “You can come up with all sorts of graphs, but I don’t, I just don’t want to worry about it. Maybe it is just very easy the way I live.” (Vi5) Interest 3 The level of interest one has in their health. “It doesn’t interest me […]. I just eat whatever I like” (Vi3) Category: Motivation [to perform healthy behavior is impacted by…] Future perspective 22 The consideration of its value toward future health. “How important is the future for you?” ”Well, I just hope to continue like this.” (Vo3) Perceived barriers 20 The amount of financial, social, and environmental barriers one perceives. “I have always had a one-sided diet. A lot of cheese for example. We didn’t have a fridge at work.” (Vi1) Feeling 6 The extent to it contributes to the subjective emotional state one experiences. “Do you think it’s important to do it [performing healthy behavior]?” ”Yes, it makes you feel better.” (Vo2) Enjoyment 5 The extent to which it impacts the level of joy in one’s life. “No, I don’t really consider it [being healthy] that much. You also would want to enjoy life” (Vo4) Category: Control [One perceives to have over health is impacted by] Support 24 The amount of support one receives on managing their health. “What facilitates you in doing it [healthy behavior]?” “To be honest, my friend. […] She supports me and shows me the ropes.” (Vo2) Self-efficacy 14 The level of capabilities one perceives to have to change health-related behavior “But you are not eager to quit, are you?” ”I am my boy, however, I’m not able to. If you have a pill for me that I take, and it makes me quit…” (Vi5) Chance 13 The belief that what happens regarding health is all based on chance and coincidence. “I’ll not reach the age of 110, I’m not that healthy. Although, it doesn’t say much actually because there are people who are 100 years old and they still smoke.” (Vi5) Fatalism 5 The belief that what happens regarding one’s health is subjugated to fate or destiny. “You can’t really do something about it [getting sick]. The only thing you can do is watch out [for accidents], that is the only thing.” (Vi3) Category: Healthcare experience [Is impacted by…] Communication 13 The quality of communication with the healthcare provider. “I would like them to take more time for people like me, who do not fully understand it. Sometimes I really feel like a foreigner.” (Vo2) Table 2.1 Continued
www.ridderprint.nlRkJQdWJsaXNoZXIy MTk4NDMw