Inflammatory bowel disease (IBD), an umbrella term for chronic gastrointestinal diseases, including Crohn disease (CD) and ulcerative colitis (UC), affects approximately 0.3% of the population in North America and Oceania, 0.2% of the population in Europe, and rapidly rising proportions of the population in many other countries around the world [1,2]. The rising incidence has resulted in increased health care use and longer outpatient waiting lists [3]. One way to alleviate pressure on the health care system is to reduce waiting lists and health care use through patient self-management, that is, a patient’s “ability to manage the symptoms, treatment, physical and psychosocial consequences, and lifestyle changes inherent in living with a chronic condition” [4]. Being a chronic condition, IBD entails lifelong self-management; even in remission, IBD is known to have a severe impact on mental health, with recent evidence suggesting that approximately 30% of people with IBD report symptoms of anxiety or depression [5-7]. Self-management, including the patient’s ability to monitor their condition, is not only effective in relieving health care but is also well known to effectively maintain or even increase quality of life [4,8]. Self-management strategies are usually related to diet and the avoidance of perceived trigger foods [9-11], seeking support, stress reduction, experimentation, and lifestyle adaption [11,12].

In recent decades, information and communication technology (ICT) has played an ever-increasing role in the self-management of chronic conditions [13,14]. People living with IBD have been found to be generally willing to use ICT solutions for the self-management of their condition [15]. ICT solutions used for IBD self-management include websites [16]; smartphone apps [17,18]; wearable sensors [15,19]; and, lately, domotics such as “smart” toilet seats [20,21], which may soon arrive in the consumer market. Compared with standard outpatient-led care, ICT solutions (also known as eHealth technologies) have been shown to lead to improvements in relapse duration, disease activity, short-term medication adherence, quality of life and stress reduction, IBD knowledge, health care costs, and clinic visits [22].

Unfortunately, the efficacy of a majority of current ICT solutions for IBD self-management is marred by a lack of professional medical involvement and the limited use of evidence and international consensus guidelines [23]. Another important detrimental factor is a lack of patient input to inform the development of ICT solutions. Patient input is important because research shows that patients with IBD desire more information and actionable strategies with regard to managing their disease, shared decision-making, and reducing the burden of treatment and symptom control [24,25]. Furthermore, although using ICT for IBD self-management improves participants’ feelings of empowerment [25], the views, needs, and desires of users, as well as the barriers they face, have as yet largely been disregarded [15,21,26]. One recent study sheds light on patient self-management strategies and needs for ICT intervention content [27], but much remains unclear about prerequisites, desires, and barriers to using ICT for the self-management of IBD, such as individual differences in, and determinants for, the use of ICT; the impact of self-management strategies and needs on the use of ICT; current ICT use; and perceived use cases.

This study addresses the lack of patient perspectives on using ICT for the self-management of IBD in the literature by researching self-management strategies and needs, current ICT use, perceived ICT use cases, and the perceived efficacy of patients with IBD, and individual differences in ICT use in self-management. The study further elucidates perceived use cases, concerns, and barriers for a novel method to support IBD self-management by presenting participants with a description of a smart toilet seat currently under development at OnePlanet Research Center in Wageningen, Netherlands. Using low-fidelity prototypes such as descriptions or paper examples enables people to reflect on their potential use cases and barriers for ICT for IBD self-management [28,29]. The literature [21,30] shows that people differ in their use cases for smart technology, ranging from documenting aspects of their condition and using technology as a tool for diagnosis to signaling exacerbations and prescribing a course of action. Furthermore, people differ in terms of concerns and barriers, such as the transparency and visibility of technology use [31], privacy, and data agency (ie, who controls the dissemination of the data) [21]. What determines these differences, how they are interrelated, and how they relate to self-management strategies as yet remain unknown. This study contributes to addressing this gap in our knowledge.

To investigate the perspectives of people with IBD on using technology for the self-management of their condition and to identify the use cases and barriers for people with IBD regarding a smart toilet seat for IBD self-management, we performed a mixed methods study. We used a 3-part web-based questionnaire with both Likert-scale questions and questions with open-text fields. The first part of the questionnaire contained questions on demographics, health status, and disease self-management; the second part covered the use of ICT for disease self-management; and the third part focused on potential use cases and barriers for a specific ICT example, that is, a smart toilet seat. Responses to the Likert-scale questions were analyzed using descriptive statistics, correlations, and binomial general linear modeling, and these served as criteria for stratification in the qualitative analyses of the open-text questions. The answers to the open-text questions were analyzed and coded using summative content analysis and thematic qualitative analysis [32,33]. We then combined the quantitative and qualitative analyses through epistemic network analysis (ENA) [34-37], which enabled us to evaluate differences in the responses as well as their patterns across the stratification groups.

The study participants were members of the patients’ association Crohn & Colitis NL (CCNL [38]). CCNL has >10,000 members and sends out a monthly email newsletter. In this newsletter, we published a brief invitation to participate in the questionnaire, which was sent out to 2274 members, of whom 1633 (71.81%) identified as female, 634 (27.88%) as male, and 6 (0.26%) as other or unknown. The average age of the participants was 48.77 (SD 15.72) years. Of the 2273 members, 568 (24.98%) had UC, 707 (31.09%) had CD, and 999 (43.93%) had not disclosed their condition when subscribing to the CCNL newsletter. The inclusion criteria for participation were as follows: having IBD (CD, UC, or as-yet-undiagnosed IBD), being aged ≥18 years, and having access to the internet and email. There were no further inclusion or exclusion criteria.

Sample size considerations were driven by the qualitative analyses. A priori sample size calculations in qualitative research are subject to conceptual debate and practical uncertainty [39]. Saturation (ie, the moment when adding more data does not lead to new insights) is often seen as a criterion for whether to include more participants once analysis has started. Although saturation as a concept is contested [39], it is still the most practical measure of information maturity. As a rule of thumb, 20 to 40 participants are usually considered sufficient to achieve saturation [40]. Another measure is information power [41]: the density, depth, and specificity of the information required affects the required sample size. In this study, the aim was relatively broad, the combination of participants was less specific for the research question, there was little theory to inform the analysis, the questionnaire methodology allowed for little dialogue, and we planned to perform cross-case analyses. Therefore, we could expect the need to include more participants for a mature data set than rule-of-thumb approaches would suggest. For the ENA approach, no sample size considerations exist as yet, but a rule-of-thumb consideration is that there should be more participants than codes in each network [33-37], and it is advisable to achieve thematic saturation within each network. In this study, the maximum number of potential networks to compare would be when comparing age groups (5 groups, resulting in 5 networks). Therefore, we aimed to include at least 200 participants—5 times as many as rules of thumb would advise—to make ENA comparisons between groups possible.

After clicking the link in the newsletter, participants were taken to the questionnaire platform. Before entering the questionnaire, participants were presented with a consent form and asked to provide informed consent. They could exit the questionnaire at any time for any reason without any consequences. Participant data were only stored when they clicked the Send Results button at the end of the questionnaire. However, participants had the option to skip questions; any unanswered questions were recorded as missing data. Filling out the questionnaire took approximately 20 minutes. The questionnaire remained accessible for 3 weeks after participants received the invitation.

This study was deemed exempt from ethics approval under Dutch law by the ethics review board of the Máxima Medical Centre, Veldhoven (N23.030). Participants provided informed consent before starting the questionnaire by reading and agreeing to a statement outlining the goal and contents of the questionnaire. They were informed that all data would be stored anonymously, that answers would not be analyzed on an individual level, and that they had the right to skip any question. The statement also explained that filling out the questionnaire meant providing consent for the use of their questionnaire data for analysis and in scientific and professional publications. Finally, the statement advised them that they could quit participating at any time by closing the questionnaire and refraining from submitting their responses; no data were stored before the questionnaire was completed. Neither CCNL nor the researchers had access to any names or physical or email addresses of participants. After questionnaire completion, data that could potentially pose a privacy risk, such as age and occupation details, were pseudonymized by converting quantitative data into categorical data. No compensation was offered to participants for taking part in the study.

In research aimed at understanding participants’ experiences, there is a potential risk of researcher bias [42]. To enhance the integrity and credibility of qualitative research, researchers must assess how intersubjective components affect data collection and analysis. An instrument for this examination is reflexivity, which refers to researchers’ explicit, self-aware appraisal of their own roles [42,43].

The host institute of this study, OnePlanet Research Center, is a relatively new (it was founded 5 years ago) research institute focused on potential innovations in health and sustainability using sensor technology and artificial intelligence. One of its research programs is dedicated to gut health, with the smart toilet seat described in this study being an important part. The end goal of the program is the development of an integrated suite of sensors that informs a personal digital twin model designed to signal, measure, and prevent a range of health conditions.

SH is the principal behavioral scientist at OnePlanet Research Center, where he leads human factors research. His work focuses on the acceptability, usability, and efficacy of technological innovations designed to support healthy living. DT is a senior biomedical researcher at OnePlanet Research Center and is responsible for research projects within the smart toilet seat project. EW is the principal investigator of the smart bathroom program at OnePlanet Research Center; she leads all scientific and developmental activities for the smart toilet seat and other innovations. MD is the head of clinical research at the department of gastroenterology, Radboud University Medical Centre, Nijmegen. She specializes in treating IBD.

Participants filled out a 3-part questionnaire containing 31 questions on demographics; health status and disease self-management; the use of ICT for disease self-management; and potential use cases and barriers for a specific ICT example, that is, a smart toilet seat (refer to Textbox 1). The full questionnaire, translated into English, is available in Multimedia Appendix 1.

Data analysis consisted of 3 steps: quantitative analysis of the results of the questions used for stratification, summative content analysis and thematic qualitative analysis of the open-text question results, and ENA of the qualitative data using the quantitative results as stratification factors.

Quantitative analyses were performed using RStudio for Mac (version 2023.03.0+386; Posit Software, PBC) [45]. The questionnaire data were first imported into RStudio. Where necessary, data were converted into factorial format and labeled. For numerical data, we calculated means, SDs, minimum and maximum values, and medians. For all data, we printed bar plot histograms for visual information. We then calculated the heterogenous correlation matrix, consisting of Pearson product-moment correlations between numeric variables, polyserial correlations between numeric and ordinal variables, and polychoric correlations between ordinal variables. Next, we qualified the correlations by testing binomial general linear models for the 2 main research questions: what predicts the use of ICT for self-management? What predicts willingness to use the smart toilet seat innovation? Finally, all quantitative data were gathered in a data frame for use as stratification factors in the ENA step, using the source_from_df function from the rock package for R [46].

All source files were cleaned using the clean_sources function from the rock package. For each participant, answers to open-text questions were then grouped together per questionnaire part: (1) condition characteristics (ie, symptoms and comorbidities) and self-management (ie, predictors of exacerbations, perceived causes, self-management strategies, and self-management needs), (2) ICT use for IBD self-management (ie, which hardware and software participants currently use, perceived support for self-management through ICT, reasons not to use ICT, and desired uses of ICT), and (3) the smart toilet seat (ie, perceived use cases of the toilet, perceived barriers, data agency, who can receive data, and concerns).

All data were then manually segmented into utterances, and given a unique utterance identifier through the prepend_ids_to_sources function from the rock package. One researcher performed a summative content analysis [47] on the questions on symptoms, comorbidities, predictors of exacerbations, perceived causes, self-management strategies, self-management needs, and current technology use. Summative content analysis is well suited for counting and comparing keywords within content [47]. The researcher then performed a thematic analysis [32,33] on the results from the remaining open-text questions—perceived support for self-management through ICT, reasons not to use ICT, and desired uses of ICT for self-management from questionnaire part 2; perceived use cases, perceived barriers, data agency, who can receive data, and concerns from questionnaire part 3; and the open-text comment section—by coding and recoding all utterances using the i.rock web-based coding tool, a tool for creating human- and machine-readable coded qualitative data and then identifying themes within the codes. Thematic analysis is well suited to identify and interpret patterns of meaning within open qualitative data [33]. All coded utterances were then gathered in a data frame to use as analysis units in the ENA step, using the parse_sources function from the rock package.

To look for meaningful patterns and differentiations between subgroups in the qualitative data, we performed an ENA using the ENA web tool (version 1.7.0) [48,49]. ENA is a technique for modeling the structure of connections in qualitative data. It assumes that (1) it is possible to systematically identify a set of meaningful features in the data (codes), (2) the data have local structure (conversations), and (3) an important feature of the data is the way that codes are connected to one another within conversations [34-37]. ENA models the connections among codes by quantifying the co-occurrence of codes within a given moving stanza window in the conversations (in our analysis, this moving stanza window was set to 6 utterances), producing a weighted network graph in which nodes correspond to the codes, and edges reflect the relative frequency of co-occurrence (or connection) between 2 codes. The positions of the network graph nodes are fixed, and those positions are determined by an optimization routine that minimizes the difference between the plotted points and their corresponding network centroids. Because of this coregistration of network graphs and projected space, the positions of the network graph nodes—and the connections they define—can be used to interpret the dimensions of the projected space and explain the positions of plotted points in the space. Furthermore, the size of the nodes indicates the importance of the code for the network. We created graphs by selecting all utterances from each unique participant as conversations; the codes associated with one of the themes derived from the thematic qualitative analysis as codes; and the different values of the stratification factors derived from the quantitative analysis, plus the unique participants, as units.

Of the 2274 potential respondents, 724 (31.84%) participants with IBD filled out the questionnaire (ie, approximately 0.8%-1% of all people with IBD in the Netherlands) [50]. Participant demographic information (sex, age, and education; questions 1-3) is presented in Table 1. The results of the questionnaire are discussed in the subsections that follow; the results of part 1 (condition characteristics and self-management) are available in Table 1 (participant characteristics; questions 1-3), Table 2 (condition characteristics; questions 5, 6, and 10), Table 3 (differences in symptoms between participants with CD and those with UC; questions 8 and 9), and Table 4 (self-management skills, needs, and strategies; questions 11-17). The results of part 2 (ICT use for IBD self-management) are available in Table 5 (hardware and software use; questions 18-20). The results from the thematic analyses and ENAs on perceived use cases for ICT (questions 21-23) and the smart toilet seat (questions 24-31) are presented in detail in the following subsections. As the ENA may be unfamiliar to many readers, and it results in a large number of outcomes that may be difficult to interpret, we have added bullet points summarizing the main findings in each ENA subsection. An overview of all main themes, subthemes, codes, and code frequencies is available in Multimedia Appendix 2.

The coding process took place in batches of results from 25 participants and resulted in 165 codes. In the first batch, we used 137 unique codes; the second batch yielded another 10 unique codes, the third batch another 14 codes, the fourth batch another 7 codes, and the fifth batch another 9 codes. After these 5 batches, new codes never exceeded a threshold value of 2.5%, indicating full code saturation after 125 participants. Furthermore, all cells in the ENA analyses exceeded the rule-of-thumb threshold of 40 participants.

In this study, we aimed to investigate the perspectives of people with IBD regarding the use of ICT for self-management of their condition. To do so, we constructed a questionnaire and analyzed the responses provided by 724 participants. Through this analysis, we sought to identify the use cases and barriers perceived by individuals with IBD in adopting ICT-based solutions for managing their health condition. Among the participants in the study, more than half (405/724, 55.9%) were already using various forms of ICT for IBD self-management. The primary factor influencing their current use of ICT was their affinity for interacting with technology. The questionnaire also shed light on self-management needs and skills among persons with IBD. A large majority (never: 86/283, 30.6%; sometimes: 135/283, 48%) could often not predict exacerbations, and although many participants (203/283, 71.7%) thought they knew what caused exacerbations, most (203/283, 71.7%) did not know how to influence them. Participants indicated needs for social support (105/283, 37.1%), information support (81/283, 28.6%), and resilience and optimism (64/283, 22.6%); their current strategies were sleep and relaxation (65/283, 22.9%), avoiding trigger foods and maintaining a healthy diet (51/283, 18%), stress avoidance (36/283, 12.7%), and changes in medication (28/283, 9.9%).

Through ENA, distinct differences emerged between individuals who were using ICT and those who were not using ICT, particularly regarding their perceived use cases. Participants who were using ICT viewed it as a means to document their condition, monitor physical activity, and establish connections with health care professionals to share their insights. By contrast, the individuals not using ICT regarded it as a tool to document their condition, aiming to support professional diagnosis.

Upon considering the perceived cases and concerns related to using a smart toilet seat as an example of ICT for IBD self-management, the sample population can be categorized into 4 distinct groups. First, those who were not using ICT for the self-management of their condition and had no interest in using the smart toilet seat (N-N), accounted for 13.1% (70/535) of the respondents who provided a response to the question regarding their interest in the technology. Second, those who were not using ICT but expressed a desire to use the smart toilet seat (N-Y) constituted 26.4% (141/535) of the sample. Third, individuals who were using ICT for IBD self-management but were not interested in using the smart toilet seat (Y-N) represented 7.9% (42/535) of the participants. Finally, the majority of the respondents (282/535, 52.7%) were individuals who were using ICT for IBD self-management and expressed an interest in using the smart toilet seat (Y-Y).

These groups exhibited distinct variations in perceived use cases and concerns associated with the adoption of a smart toilet seat for IBD self-management. The N-N and Y-N groups lacked clear use cases for the smart toilet seat. The N-Y group members perceived the smart toilet seat as a means to identify detrimental changes in their condition, offering directions for action, and documenting information for health care professionals’ diagnoses. By contrast, although the Y-Y group members also viewed the smart toilet seat as a means to signal negative changes in their condition, they stood apart from the N-Y group in that they stated their intention to document these changes for their own self-management and decision-making.

Similarly, these groups expressed different concerns regarding the use of the smart toilet seat. The N-N group members expressed worries about the potential impact on daily life with their condition, including constant reminders of their health status. In addition, practical concerns, such as privacy and the accuracy of the measurements, were also prominent for this group. The Y-N group mainly raised concerns related to data agency and privacy issues. The N-Y group emphasized privacy concerns and practical aspects such as the potential hassle and associated costs. Finally, the Y-Y group members, again, shared concerns about privacy and data agency, along with the apprehension of others using the smart toilet seat, potentially compromising their personal information and health data.

This research provides new insights about patient perspectives on using ICT for IBD self-management, which are currently not available in the literature. Some of the results confirm earlier work with more general participant groups [21], in which signaling and documenting also appear as important use cases, with privacy and data agency emerging as major concerns. However, this study also offers valuable insights into effectively addressing the needs of individuals who exhibit an inclination toward technology interaction but remain hesitant to embrace proposed interventions. Providing assurance regarding privacy and data agency becomes essential in catering to this particular group.

The research shows that there is also a group of people who have low affinity for technology and would therefore not want to use an ICT-based intervention. They believed that such an intervention would be too much of a hassle and that it was not for them. These opinions need to be respected. However, a substantial number of participants in this group (N-N) talked about wanting to avoid constant reminders of their condition and overmedicalization of their condition. This may be a sign of detrimental coping by hiding or denying their condition [42,51], which may impact IBD coping and self-management negatively [42,43]. It is imperative to conduct further research in this area to gain a deeper understanding of the implications of these coping strategies on IBD coping and self-management. By exploring these aspects more comprehensively, we can develop tailored approaches and interventions that better align with the needs and preferences of individuals with low technology affinity while promoting effective disease management and self-care.

When it comes to other results from this study, we see that there is a participant bias in that more female individuals than male individuals participated. Although the participant rate for female and male individuals (60% and 40%, respectively) reflects sex differences in the occurrence of IBD in the Dutch population [52,53], the occurrence of CD and UC in the sample in this paper (approximately 55% and 45%, respectively) differs from population data (approximately 40% and 60%, respectively) [52,53]. This may be caused by the aforementioned participation bias, a recurring pattern in questionnaire research [54,55]. Similarly, the literature [54,55] on questionnaire research indicates that it is relatively common for older participants to be overrepresented compared to population numbers. However, in this study, we think this is not problematic because there were sufficient male participants (233/724, 32.2%) to achieve saturation and stratification, as well as adequate representation of people aged <31 years (76) and those aged between 31 and 40 years (109). Furthermore, highly educated participants were overrepresented in the sample. Notably, those with only primary school education, who can be expected to have limited health literacy, were underrepresented. Given the fact that this group often has limited language skills [52] and is therefore less likely to participate in web-based questionnaires, further research is necessary to learn about their needs and barriers.

Finally, when it comes to use cases for ICT for IBD in general, 3 major themes were constructed: learning about the condition, dealing with the condition, and connecting with others. The smart toilet seat only covers the first 2 themes; the third theme is difficult to integrate in this solution because toilet-related topics are often awkward to talk about or even taboo [21]. To fulfil people’s needs around connecting with others, a stand-alone solution would be preferable, especially considering the concerns that people have about privacy and data agency.

In conclusion, this study provides valuable insights into the perspectives of individuals with IBD on using ICT for self-management. The positive reception of the participants to the smart toilet seat concept suggests the potential for innovative ICT solutions in empowering patients to actively participate in their care. To facilitate wider adoption, it will be critical to address privacy concerns, ensure data security, and establish reliable ICT integration. Furthermore, the results of our research show that one-size-fits-all solutions will likely fail to reach many potential users of technological innovations for disease self-management because they do not connect to their perceived use cases or do not address their concerns. Use cases differ from simply signaling when something is amiss and providing an action perspective to informing personal scientific experiments. Patients’ concerns differ from fear of overmedicalization of their condition and constant reminders of the illness to concerns about cost, privacy, and data agency. This paper not only shows that people differ but also provides evidence that there are possibilities for profiling large groups of potential users with similar needs and barriers. The results of this study and methodologically similar future studies can inform implementation and dissemination strategies for technological innovations. Future research and development should not only focus on refining and validating the technologies to enhance their effectiveness and user acceptance, with extra focus on patients with limited health literacy or low affinity with technology interaction, but also strive to shed greater light on who the user is, what their needs and requirements are, and how their lived experience of their condition informs their use of technology for self-management.