Web-Based Lifestyle Interventions for Prostate Cancer Survivors: Qualitative Study

Background Exercise and a healthy diet can improve the quality of life and prognosis of prostate cancer survivors, but there have been limited studies on the feasibility of web-based lifestyle interventions in this population. Objective This study aims to develop a data-driven grounded theory of web-based engagement by prostate cancer survivors based on their experience in the Community of Wellness, a 12-week randomized clinical trial designed to support healthy diet and exercise habits. Methods TrueNTH’s Community of Wellness was a four-arm pilot study of men with prostate cancer (N=202) who received progressive levels of behavioral support (level 1: website; level 2: website with individualized diet and exercise recommendations; level 3: website with individualized diet and exercise recommendations, Fitbit, and text messages; and level 4: website with individualized diet and exercise recommendations, Fitbit and text messages, and separate phone calls with an exercise trainer and a registered dietitian). The primary aim of the study is to determine the feasibility and estimate the effects on behaviors (results reported in a separate paper). Following the 12-week intervention, we invited participants to participate in 4 focus groups, one for each intervention level. In this report, we used grounded theory analyses including open, axial, and selective coding to generate codes and themes from the focus group transcripts. Categories were refined across levels using embodied categorization and constant comparative methods. Results In total, 20 men with prostate cancer participated in the focus groups: 5, 4, 5, and 6 men in levels 1, 2, 3, and 4, respectively. Participants converged on 5 common factors influencing engagement with the intervention: environment (home environment, competing priorities, and other lifestyle programs), motivation (accountability and discordance experienced within the health care system), preparedness (technology literacy, health literacy, trust, and readiness to change), program design (communication, materials, and customization), and program support (education, ally, and community). Each of these factors influenced the survivors’ long-term impressions and habits. We proposed a grounded theory associating these constructs to describe the components contributing to the intuitiveness of a web-based lifestyle intervention. Conclusions These analyses suggest that web-based lifestyle interventions are more intuitive when we optimize participants’ technology and health literacy; tailor interface design, content, and feedback; and leverage key motivators (ie, health care providers, family members, web-based coach) and environmental factors (ie, familiarity with other lifestyle programs). Together, these grounded theory–based efforts may improve engagement with web-based interventions designed to support prostate cancer survivorship.


Introduction
Prostate cancer is the most common cancer among men in the United States, with more than 190,000 new diagnoses expected in 2020 [1]. The median age at diagnosis is 66 years, and 82% of men are aged 65 years or older [2]. Many men live for decades after their diagnosis and may benefit from adopting healthy dietary and exercise practices to combat prostate cancer symptoms and treatment-related side effects [3][4][5][6][7][8] in addition to improving their overall health.
Diet and exercise are associated with lower risk of prostate cancer progression [9], prostate cancer-specific mortality [10][11][12][13], and treatment-related side effects [14][15][16][17][18]. Specifically, cruciferous vegetables, vegetable fat, fish, and cooked tomatoes [19] have been associated with lower risk of prostate cancer progression and/or mortality, whereas whole milk and poultry with skin have been associated with increased risk of prostate cancer progression and/or mortality [19][20][21][22][23][24][25][26][27]. Physical activity has also been consistently associated with significant reductions in mortality [26], symptoms, and treatment-related side effects. The 2018 American College of Sports Medicine roundtable recommendations for cancer survivors include 30 min of moderate aerobic training 3 or more times a week for at least 8 to 12 weeks; resistance training alone or the addition of resistance exercise to an aerobic regimen may also improve symptoms [28]. The Exercise and Sports Science Australia recommends that the specifics of the multimodal exercise prescription and total weekly dosage be determined by the patient's needs or goals but similarly supports that cancer survivors should avoid inactivity [29]. Unfortunately, many prostate cancer survivors fail to meet physical activity or nutrition recommendations.
Web-based interventions have the potential as scalable modalities to deliver lifestyle interventions in prostate cancer survivors [30]. Previous studies have demonstrated the benefits of web-based interventions in supporting behavior change related to diet, exercise, and smoking cessation for noncancer populations [31][32][33][34][35]. However, there remains to be a lack of data on the specific types and quantities of intervention components needed to change behavior. Thus, we developed a trial [36] to assess the feasibility and acceptability of a web-based intervention for men with prostate cancer. The study focused on the diet and exercise factors mentioned earlier, with particular attention to whether progressive levels of support would lead to increasingly higher levels of behavioral change and improvements in other outcomes such as symptom reduction and quality of life. Our primary feasibility, acceptability, and behavior change results are presented elsewhere.
Given the success of recent web-based interventions, we were also interested in the insights underlying the behaviors of prostate cancer survivors navigating a web-based platform. The attitudes, motivations, and perspectives of cancer survivors engaging in a web-based lifestyle intervention are complex and require further study. A prior qualitative study exploring lifestyle change in prostate, colon, and breast cancer survivors after participation in web modules designed to promote physical activity and healthy eating examined barriers to behavior change (knowledge, motivation, and individual reactions to cancer diagnosis) using a thematic analysis approach [37]. However, qualitative evidence to specifically inform intervention design is lacking. To our knowledge, perceptions of web-based interventions for lifestyle change in prostate cancer survivors have not been investigated using a qualitative methodology; as such, a grounded theory qualitative investigation using data-driven analysis would be helpful to inform future web-based intervention design. In this report, we aim to explore the insights of prostate cancer survivors who engaged with a web-based lifestyle intervention and to provide grounded theory-based recommendations to guide future intervention design.

Design
We conducted a four-arm study called TrueNTH's Community of Wellness (NCT03406013) of men with prostate cancer (N=202) who were randomized to receive progressive levels of behavioral support. The details of the design of the pilot study have been previously published [36], and select screenshots from the website are presented in Figure 1. Men in level 1 had access to prostate cancer-specific diet and exercise resources through a static, informational website. Men in level 2 had access to the website and received individualized diet and exercise recommendations based on a self-report survey completed at the start of the study. Men in level 3 had access to the website and received individualized diet and exercise recommendations and also received a Fitbit device and text messages. Men in level 4 had access to the website and received individualized diet and exercise recommendations, received a Fitbit device and text messages, and were offered a 30-min phone call with an exercise trainer and a 30-min phone call with a registered dietitian. Of note, the Community of Wellness is one of many TrueNTH programs funded by the Movember Foundation, and some men participated in multiple TrueNTH programs concurrently. Reporting in this study is consistent with the consolidated criteria for reporting qualitative research [38].

Focus Groups
Men who completed the pilot study and consented to being contacted were invited via email to participate in a focus group. Briefly, for the primary pilot trial, men were recruited through hospital cancer registry databases at the University of California San Francisco, the Oregon Health and Sciences University, and the University of Colorado Denver; at the Cancer of the Prostate Strategic Urologic Research Endeavor registry of men with prostate cancer; and in clinics at the abovementioned institutions.
Each participant consented to participate in both the pilot study and focus group. In total, 48 men were willing to participate in a focus group; of these, 20 men could attend at the scheduled times ( Figure 2). We conducted 4 focus groups, one for each intervention level. As participants in the trial could reside throughout the United States, focus groups were conducted via Zoom, a secure, interactive audioconference platform. In the interest of confidentiality, we disabled video calling; however, we used screen sharing so that participants could comment on various aspects of the website. The focus groups were led by a female researcher (RG). She is an assistant professor in epidemiology with 10 years of experience researching urologic cancers, who previously worked in market research and web usability where she gained experience in qualitative research methods. RG first interacted with the participants when scheduling and conducting the focus group. Interviews were semistructured using interview guides (available in Multimedia Appendix 1) tailored to each group's intervention level (eg, individuals randomized to level 1 were asked about the website only, etc). Participants were prompted to answer hypothetically if they did not use or recall certain aspects of the program during the study period. Focus groups were recorded and transcribed. Quotations were edited for clarity, and field notes were made after focus groups. Focus groups took place between May and June 2019; the median time from the end of the study to the focus group was 7 months (Table 1) 19 (95) 6 (100) 5 (100) 3 (75) 5 (100) White 1 (5) 0 (0) 0 (0) 1 (25) 0 (0) Other 7 (6-12) 6 (6-12) 7 (6-12) 10 (7-14) 7 (7)(8) Months from intervention end date to focus groups, median (IQR)

Grounded Theory Analyses
We used a grounded theory approach [39,40]. Coding was completed manually by one investigator (EW) and reviewed with 4 other investigators (SK, JB, RG, and EV); axial codes were managed in Microsoft Excel.
We conducted open, axial, and selective coding ( Figure 3). Open, line-by-line coding generated data-driven codes that were refined into 15 axial codes. Ultimately, through embodied categorization [41] and constant comparative methods (to address the multiple levels) [42], we consolidated the data under 7 selective codes (categories). From these categories, a grounded theory surrounding prostate cancer survivors' use of web-based lifestyle interventions emerged. The codes and their relationships to one another were intermittently discussed and finalized among EW, RG, JB, EV, and SK.

Results
In total, 10% (20/200) men (of pilot study participants) with prostate cancer participated in the focus groups; 5, 4, 5, and 6 men in levels 1, 2, 3, and 4, respectively. The characteristics of the focus group participants are presented in Table 1 We identified 5 categories influencing intervention engagement: (1) environment (home environment, competing priorities, and other lifestyle programs), (2) motivation (accountability and discordance), (3) preparedness (technology literacy, health literacy, trust, and readiness to change), (4) program design (communication, materials, and customization), and (5) program support (education, ally, and community; Figure 3). We also identified the long-term effects of the interventions (impressions and habits). Each code represents an actionable component contributing to the overall intuitiveness and seamlessness of this web intervention, as demonstrated by participant quotes below.

Environment
Participants discussed the environmental factors influencing their participation and impressions of the program.

Home Environment
Participants' family members and geographic locations played roles in their perceptions and usage of the web-based program:

Motivation
Participants discussed factors influencing their motivation to participate in the program.

Accountability
Participants described or alluded to a sense of accountability:

Preparedness
In addition to environment and motivation, participants' unique skill sets and backgrounds influenced their ability to engage with the program.

Health Literacy
Participants demonstrated varying levels of health literacy (ability to communicate an understanding about prostate cancer and/or the purpose of the study), which affected their interest and engagement with the program:

Tech Literacy
Using a web-based intervention requires some baseline comfort using technology-the participants greatly varied in their preferences, which affected their engagement with the program:

Program Design
Participants reflected on the various components of the program and suggested improvements.

Communication
Comments about how participants hoped communications would be used and how they might be improved:

Customization (Flexibility)
Participants of all levels commented on their desire for increased customization and flexibility-many participants mentioned that their engagement in various aspects of the program would have changed if messaging delivery or content was customized:

Program Support
Participants communicated their expectations of various types of support from the web-based program.

Education
Participants from all levels provided suggestions on how to improve the educational component of the intervention:

Impressions and Habits
Ideally, lifestyle interventions help participants develop lifelong habits. In this quotation, one participant offers his thoughts on the long-term impacts of this intervention. This quotation and others reflect the participants' impressions of the program; these impressions add to their collective experiences with technology and health:

Intuitive Interventions
Each code generated in this study represents a unique mechanism for designing a more intuitive, web-based lifestyle intervention for prostate cancer survivors. By addressing the environment, we may transform factors that already exist in participants' lives as obstacles to reinforcing factors for improved engagement with the web-based program. By addressing participants' motivation, we may improve our ability to tailor web content and web-based communications. Understanding participants' preconceived attitudes based on past encounters with the health care system will allow us to actively address concerns and improve program adherence. We may influence preparedness when we assess and consider each participant's unique level of health and technological literacy, readiness to change, and trust and bolster these whenever possible through program content. Program design and program support are the most easily affected; we can increase intuitiveness through tailored communication, materials, and feedback, providing quality educational content, serving as allies, and generating community.
Noting the ways in which certain codes presented in the different intervention levels helped contextualize feedback. For example, participants in level 1, who received only web access to educational content, requested more communication, whereas participants in levels 2 to 4, who received increasing levels of behavioral support, provided details on ways in which the multiple forms of communication they received might be tailored. Participants in levels 3 and 4 received more types of behavior support and were also more likely to request more instructions or reminders orienting them to the program, as their interventions had more components. Conversely, some codes were commonly expressed across groups, such as competing priorities, readiness for change, flexibility, education, and community.
The relationships among these codes ( Figure 4 [41]) represent iterative, actionable pathways by which designers may increase program intuitiveness for prostate cancer survivors engaging in web-based interventions often via multiple mechanisms at once. For example, we might influence motivation (accountability and discordance) by improving program design in the following ways: (1) using Health on the Net [43] transparency and quality principles (quality, confidentiality, neutrality, transparency, community, and visibility) for certification, (2) communicating with clinical providers about participants' involvement in the program, (3) remaining sensitive to participants' guilt with failures to modify behaviors, and (4) leveraging participants' familiarity with existing lifestyle programs to optimize engagement. These and other grounded theory-based solutions ( Table 2) may result in a more accessible and integrated intervention for prostate cancer survivors.

Implications
Men with prostate cancer find themselves in an era of seemingly limitless access to medical information via the web. Technological advances impact their daily lives, and as technology and health care delivery are increasingly intertwined, their ability to maintain health may inevitably be influenced by their willingness to engage with technological interfaces [44,45]. We learned that prostate cancer survivors within this study were sensitive to discrepancies related to clinical evidence and practice. They developed heuristics for navigating copious information, they described an interest in transparent sources, and they voiced a desire for continuity and ongoing care. They discussed the emotional impact of their participation within the health care system; these cumulative experiences (including newer experiences with technology-based care) underlie all experiences with health-promoting interventions.

Qualitative Methodology
We used the grounded theory methodology because no comprehensive theory of web interventions for behavior change in prostate cancer survivors has been developed before. This methodologic approach is a strength because data-driven open coding is most equipped to interrogate the inherent assumptions held by study participants and researchers alike [40]. Another strength of the study was the interpretation of data across groups receiving progressive levels of lifestyle interventions.

Comparison With Prior Work
An intuitive, web-based interface is not a novel concept. In 1993, Nielsen [46] coined the term usability engineering, where the usability of a system is defined by (1) learnability, (2) efficiency, (3) memorability, (4) low error rate, and (5) satisfaction. Usability heavily overlaps with intuitiveness, although we believe intuitiveness emphasizes tailoring and program responsiveness, shifting the burden of anticipation on program designers rather than program users. The interest in temporal and user tailoring beyond usability is also illustrated by the growing literature on just in time adaptive interventions, which are designed to adapt according to changes in an individual's contexts over time. These interventions provide the most appropriate and timely support to their users (usually enabled by mobile and sensing technologies); their applications in health promotion are of particular interest [47].
Our findings suggest that intuitiveness will likely depend on both the context and the intended user. This qualitative study elucidates some of the key areas that can be optimized for intuitive use of an internet-based lifestyle intervention among well-educated, White prostate cancer survivors. Although we used a grounded theory approach and generated data-driven codes, many of the resulting codes and their relationships to one another (Figure 4) are corroborated by existing theories in public health, as described below.
The environment code (applied in instances where participants mention environmental factors impacting their program engagement) is corroborated by the idea of a multilevel intervention based on the social ecological model. The social ecological model by Bronfenbrenner and Morris suggests that the individual is enveloped and influenced by interpersonal, organizational, community, and public policy networks [48]. Readiness to change is supported by the transtheoretical model stages of change (with the stages of precontemplation, contemplation, preparation, action, maintenance, and termination) [49]. The idea that self-efficacy and agency influence how accountability is achieved (social cognitive theory) is highly consistent with motivation (accountability and discordance) [50]. Finally, the health belief model [51], which differentiates between behaviors in health and illness, is especially interesting when applied to lifestyle interventions in prostate cancer survivors. Prostate cancer survivors are in a unique position of having a chronic illness but also being in a position to engage in preventative health behaviors to deter recurrence or disease progression. The various components of the health belief model (perceived benefits vs perceived threat, self-efficacy, and cues to action) are impacted by the large majority of codes in our grounded theory model.

Limitations
Limitations of the study include the small subgroup sample size and lack of a theoretical sampling process parallel to the analyses. Overall, 10% (20/200) of eligible men were both interested and available to participate in the focus groups at the scheduled times. Although the smaller sample size is acceptable as our objectives were to explore themes using a grounded theory approach, this introduces a possible selection bias. In addition, not all participants fully participated in the web intervention as indicated, and the focus groups took place a median of 7 months after the interventions. Some men participated in multiple TrueNTH programs or were involved in other clinical trials. Although the longer follow-up period and competing priorities contributed valuable, realistic insight into the participants' lasting impressions and their habit formation, participants may not have recalled all the details of the intervention. In addition, although this was a multi-institutional study, the participants' experiences may primarily reflect viewpoints of educated, White men in the West and Mountain regions of the United States, where there may be disproportionately greater exposure to technology and overall better physical activity rates [52]. The lack of theoretical sampling and smaller subgroup sample size limits our ability to confidently comment on data saturation. In response to these limitations, we had a low threshold to include open codes in grounded theory, even if they were introduced by just 1 or 2 participants (ie, preparedness: trust, impressions, and habits); data-driven codes were also more likely to be elevated to axial or selective code status if the concepts they represented were supported by previous well-supported theories in public health. This qualitative study does not provide insight into which level of intervention performed best for this group of end users; however, it does provide researchers with important insights into the challenges of creating web-based approaches to support survivorship care that is both high tech and accessible. Further quantitative studies are needed to confirm the validity and directionality of these associations. Further work is needed to explore how our proposed theory applies to men with different sociodemographic characteristics.

Conclusions
Our study demonstrates that a web-based lifestyle intervention for men with prostate cancer can become intuitive and encourage adherence. These include addressing technological and health literacy, motivation, and environmental factors. In addition, flexible and transparent web design, integration of key stakeholders (ie, providers, family members), and effective coaching may improve the usability and intuitiveness of a web-based intervention to support prostate cancer survivorship. Men with prostate cancer tend to be older, have comorbidities, and balance multiple priorities; this may limit their ability to engage with a web-based lifestyle platform. A web intervention's potential to affect long-term change will depend on the intuitiveness of its components, allowing integration within an individual's daily life (eg, clinical support, familial involvement, preparedness for program participation). This grounded theory-based analysis may help guide future web intervention designs for cancer survivors. The convergence of our findings with well-established theories in public health suggests that certain aspects of our theory are broadly applicable to lifestyle intervention design, although this will require further study.