Can discussions with patients and the public clarify missing data mechanisms for digital outcome measures?

Can discussions with patients and the public clarify missing data mechanisms for digital outcome measures? | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Can discussions with patients and the public clarify missing data mechanisms for digital outcome measures? Mia S. Tackney, Amber Steele, Marie-Louise Zeissler, Sofía S. Villar, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9293900/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Background Analysis of clinical trials with missing data requires statistical assumptions. When novel digital outcome measures are used, it is particularly important to understand trial participants’ experiences with the device. This can help illuminate the reasons for and specific patterns of missing data, thereby informing plausible missing data assumptions. Methods A patient and public involvement and engagement (PPIE) activity bridged statistical concepts on missing data with lived experiences of individuals who had participated in studies with digital outcome measures. Ten PPIE contributors attended three meetings, which were aimed to (i) introduce key statistical concepts and set the scene, (ii) discuss experiences of using digital devices in research studies, and (iii) consolidate learning and reflect on implications. In addition, contributors provided input via pre- and post-meeting surveys. Results Reasons for missing data in digital outcome measures were highly context dependent and varied according to the device, study population and environmental/cultural context. Identified reasons included operational aspects of the device, unresolved technology issues, and practicalities in daily life such as weather or season affecting comfort and the need to remove devices for religious, exercise or hygiene activities. Previous experience of using digital devices and receiving feedback from devices influenced levels of engagement. Although contributors did not report disengaging with devices in ways directly related to the outcome being measured, informative missingness was considered as important. Conclusions The PPIE activity was a feasible and valuable approach to exploring patient and public perspectives on the reasons for and patterns of missing data in digital outcome measures. Lived experience may shed light on reasons for missing data that may be overlooked by researchers; for example, religious activities affecting device removal was a key learning point. Incorporating PPIE discussions within ongoing trials may help inform statistical analyses and improve trial conduct. For future PPIE activities, purposeful recruitment and providing different modes of engagement, such as anonymous surveys, are recommended. missing data patient and public involvement and engagement digital outcome measures trial conduct Figures Figure 1 Figure 2 PLAIN ENGLISH SUMMARY Why Data from Digital Devices Goes “Missing” in Clinical Trials When researchers run medical studies, they sometimes use wearable devices to collect health information. For example, smartwatches are watches that can track your heart rate and physical activity, and continuous glucose monitors measure blood sugar levels. Sometimes, these devices stop collecting data. When this happens, gaps appear in the results. Why does this happen? Ten patient and public contributors shared their real-world experiences of using digital devices in research. They helped researchers understand why data might be missing. Some common reasons included: Technical problems: The device or app may stop working. Maintenance: Device might need charging, or some parts of the device need to be replaced, which can cause short gaps in data. Discomfort or personal reasons: People may remove devices if they feel uncomfortable, for example in hot weather, or during religious activities. The reasons listed above are mainly practical, such as technical problems or needing to replace part of the device. However, sometimes, missing data happens for a different reason that is linked to a person’s health. If someone’s symptoms get worse, they may be less likely to wear or use the device. This is especially important because missing data is directly related to what the study is trying to measure. Understanding these reasons can help researchers improve how studies are run, for example by giving clearer instructions and better preparing participants. It can also help statisticians carry out statistical analysis (using mathematical tools to make sense of the data) in the right way. For future patient and public involvement activities, it may help to recruit contributors who are participating in a specific clinical study to focus on a specific context, and to offer different ways to give feedback, such as anonymous surveys. Background Missing data assumptions in Clinical Trials with Digital Outcomes Statistical methods rely on specific assumptions, which act as fixed conditions under which the methods can best achieve their intended theoretical and inferential properties. In clinical trials, missing data is a pervasive issue, arising from technical difficulties, participant withdrawal, or missed assessments. A common assumption in the analysis of trials is that data are Missing at Random (MAR). This implies that the probability of data being missing relates to observed information (such as baseline characteristics or earlier measurements) rather than the unobserved value itself. If the assumption is incorrect, statistical estimates may be biased and standard errors may be invalid, potentially leading to incorrect conclusions about the treatment effect (Carpenter & Smuk, 2021 ). For example, analyses assuming that data are MAR may underestimate treatment effect if participants with poor outcomes are more likely to withdraw from the study. Consequently, sensitivity analyses are recommended, which assess whether results remain robust under alternative missing data assumptions (European Medicines Agency, 2017 ). Collecting digital outcome measures via digital health technologies introduces unique challenges regarding missing data. Step count data from accelerometers, glucose readings from continuous glucose monitors (CGMs), and patient-reported outcomes from smartphone applications may each have specific patterns of missing data and reasons for missing data (Di et al., 2022 ; Tackney, Carpenter, et al., 2024). In these settings, missing data often arises from technology-related issues, such as internet or Bluetooth connectivity failures, battery depletion, or device malfunction. Further, because these devices are typically operated by participants in their daily lives, the responsibility for data collection shifts towards the individual. Participant experience, therefore, becomes central to understanding missingness e.g., the removal of wearables due to discomfort or privacy concerns. Understanding these patterns of potential data loss is essential not only for evaluating device usability as part of the validation of digital devices (Bakker, Barge, Cobb, Cota, et al., 2024 ; Keogh et al., 2023 ), but also for identifying plausible missing data assumptions in the statistical analyses of the resulting data. The Need for Patient and Public Perspectives There is a clear need to understand trial participants’ experiences to identify sensible missing data assumptions for digital outcome measures. Two strands of previous work suggest this as a promising direction. Firstly, a priority-setting exercise on numerical aspects of clinical trials(Goulao et al., 2021 ) highlighted that Patient and Public Involvement and Engagement (PPIE) contributors felt well placed to challenge whether statistical modelling assumptions reflected real-world participant experiences. However, while contributors felt that explicit discussion of these assumptions increased research transparency, technical jargon often obscured them. This underscores the need for more effective communication and visibility of methodological assumptions. Secondly, tools have been developed to elicit expert and patient and public opinions on missing data. For example, Mason et al. (2017) asked doctors and nurses to quantify how outcomes for individuals with missing data might differ from those with observed data. These expert opinions then informed sensitivity analyses for the Missing at Random (MAR) assumption. Greenwood et al. conducted an elicitation exercise on missing data for a historical trial with patient and public partners who were not participants in the trial (Greenwood, Morris, O’Malley, Aucott, & Goulao, 2026 ; Greenwood, Morris, O’Malley, Aucott, Fagbemi, et al., 2026). There is a paucity of experience in exploring concepts of missing data with patient and public representatives who have lived experience of being in clinical studies. Challenges in PPIE for Statistical Methodology It is widely recognised that PPIE, where the public contributes to research design, conduct and delivery, can lead to improved outcomes and delivery of research (National Institute for Health Research, 2013 ). However, there are distinct challenges to meaningful PPIE in statistical methodology research. Statistical methods can appear abstract compared to patients’ lived experiences, so deliberate effort is required to bridge these conceptualisations and perspectives (Abell et al., 2023 ). Furthermore, the technical nature of the research requires establishing a shared foundation of knowledge for productive discussion. The appropriate level of information to share is essential, aligning with the Support and Learning principle of the UK Standards for Public Involvement (Crowe et al., 2020 ). Bridging the Gap To address these challenges, a growing body of practical expertise has emerged. This includes a community of practice, training workshops for statistical methodologists (Smith et al., 2025 ) successful co-production of animation videos (Worboys et al., 2023 ), glossaries of key terms (Booth et al., 2025 ) and toolkits to explain statistical concepts such as estimands (Cro et al., 2023 , 2025 ). While most developments focus on the communication and dissemination stages of the research cycle, less attention has been paid to embedding PPIE across the full statistical research lifecycle, such as in developing novel methods, evaluating methodological assumptions, or application of statistical methods (Broomfield et al., 2024 ). Involvement during the design stages is critical within the Support and Learning and Working Together principles of the UK Standards for Public Involvement and require specific planning for meaningful involvement. This article describes a PPIE activity designed to integrate patient and public perspectives into discussions on missing data for digital outcome measures in clinical trials. We describe how this PPIE activity was carried out, demonstrating a viable pathway for PPIE in this context and its value to methodological development. Our results offer recommendations for the conduct and analysis of trials using digital devices, as well as in future PPIE activities aimed at understanding informative missingness. Methods Patient and Public Involvement in Design The design of this activity involved Patient and Public Involvement and Engagement (PPIE) contributors at every stage. During the initial funding proposal, a discussion group with three members of the Cambridge University Hospitals (CUH) PPIE panel helped refine research aims and establish ongoing contributor involvement. Subsequently, eleven volunteers from the CUH PPIE panel reviewed the application’s Plain English Summary and PPIE plan. Their feedback focused on research relevance, text accessibility, and feasibility of the proposed activities. From these early engagements, three contributors with a specific interest in the field joined the Community Involvement Group for Digital Outcomes (CIG-DO). To further identify research priorities for digital outcome measures, we convened a multi-stakeholder event. This brought together 32 attendees, including clinicians, statisticians, computer scientists, implementation scientists, ethicists and health economists. Six PPIE contributors participated: four attended in person (two members of CIG-DO) to represent the patient and public perspective, while two contributed remotely via video and written comments of their lived experience digital research devices. The methodology and insights from this event have been previously published (Tackney, Steele, et al., 2024). Key Findings from Preliminary Engagement Feedback from these initial activities, alongside a foundation meeting with the CIG-DO, highlighted three specific areas requiring attention in subsequent phases: • Need to develop shared understanding of statistical methodology While PPIE contributors offered perceptive and practical solutions for preventing missing data (such as participant reminders), discussing statistical methodology required more dedicated time. Conveying the abstract concept that missing data must be addressed through statistical analysis, rather than prevention alone, remained a challenge that aligned with the Support and Learning standard. • Need for resources to support learning Some PPIE contributors found the technical jargon of the multi-stakeholder event overwhelming. They recommended providing a glossary of key terms, as well as a Plain English Summary for all speaker presentations, and a dedicated space for PPIE contributors to hold independent discussions (Tackney et al., 2024). Establishing a culture of learning with designated resources is vital to the Support and Learning principle. Need for clarity in expectations for PPIE contributors. During the multi-stakeholder event, some contributors felt uncertain about their specific role in discussions. This underscores the Working Together principle, and the need to jointly shape and define the purpose of the activity to ensure a shared understanding of roles expectations and responsibilities. Recruiting additional members to the CIG-DO To recruit additional members to the CIG-DO group, MST posted the opportunity to join the group on the CUH PPIE newsletter. Seven individuals expressed interest and had an initial discussion with MST and AS to discuss the research topic, the level of involvement that would be expected, and whether there were any specific needs. Objectives and Structure of Activity Between September – December 2025, each contributor attended three online sessions led by MST and supported by a facilitator (AS, with MLZ as an alternate). To accommodate the availability of the 10-member group and ensure that everyone had sufficient time to contribute, sessions utilised a blend of small group discussions and one-on-one meetings. PPIE contributors were compensated for their time spent in online meetings and providing feedback, as well as for remote working expenses, in accordance with NIHR payment guidelines (National Institute for Health and Care Research, 2024). There were two key objectives of the activity: 1. Exploratory objective: understand how missing data might arise in trials with digital outcome measures, drawing from the experiences who have lived experience of participation in research studies that used digital devices. 2. Focused objective: understand whether there could be informative missingness (when missingness relates to the outcome itself) for digital outcome measures. Session 1: Introduction to digital outcome measures Prior to the first session, the Plain English Summary of the funding proposal, and the meeting slide deck were shared with participants. A pre-session survey gathered PPIE contributors’ expectations for the activity. The session began with introductions from all attendees, followed by an overview of key statistical concepts, including digital outcome measures and missing data. Visualisations of minute-by-minute accelerometer step-count data, collected over seven days, were presented (e.g., Fig. 1). The group discussed potential reasons for missing data and the resulting implications for data analysis. To foster a culture of shared learning, technical terms used during meetings were deposited in a “Too Technical Term (TTT) Jar”, which MST described as being like a “swear jar.” PPIE contributors were encouraged to point out whenever a TTT (Too Technical Term) was used in discussions, and definitions were deposited in a shared document. In addition to this shared document containing specific terms related to digital outcome measures, PPIE contributors were given a link to a glossary of general statistical terms and video on statistical methodology (Booth et al., 2025; Worboys et al., 2023). To establish a shared understanding of the project, the group discussed the goals, expectations, and potential impact of the PPIE activities. MST emphasised that the meetings would not focus on one health condition or treatment (which may be typical for clinical PPIE activities), but rather around how participant experiences can inform statisticians’ understanding of how missing data may arise. Through the visualisations of missing data, MST explained how, as a statistician, one may not always have contextual information for why data may be missing, and that patient and public partners’ lived experiences could be informative. The group also discussed the potential impact of the PPIE activities; for example, the insights gained could improve the design, conduct and analysis of ongoing studies with digital outcome measures, such as the Edmond J. Safra Accelerating Clinical Trials for Parkinson’s Disease (EJS ACT-PD). After the first session, MST asked the group to watch Sophie Greenwood’s introductory video on missing data, which was co-produced with PPIE contributors from her PhD work (https://www.youtube.com/watch?v=DLBVXCru8cI) (Greenwood, 2025). A post-meeting survey invited reflections on the video. Furthermore, contributors with lived experience of using digital devices in research studies were invited to prepare a short talk explaining the context of the trial, the device used, and any reasons for missing data. Session 2: Listening to PPIE contributors’ experiences PPIE contributors with trial experience gave short talks on their experiences, guided by questions prepared by MST and AS. After each talk, MST asked additional questions to highlight specific statistical considerations. Contributors without lived experience participated by asking questions and sharing whether the experiences described aligned with their own expectations. To address the project’s second objective, MST reintroduced the concept of “informative missingness” (which was added to the “Too Technical Term” jar) and provided examples of how it occurs with digital outcome measures. MST asked the group to reflect on whether informative missingness is likely within each of the previously described contexts. Session 3: Feedback and summary meeting MST summarised experiences shared in Session 2 and identified key themes through discussions with AS, SSV and MLZ, and experts on devices and missing data. During the feedback meeting, MST shared these themes and described how the contributors’ experiences aligned with published findings on missing data patterns, e.g. analyses of CGM monitor data described by Kettermann et al. (2025), as shown in Fig. 2. Finally, MST invited PPIE contributors to share ideas on how future research and PPIE activities could better capture and account for informative missingness. Table 1 summarises the approaches used in this activity to address challenges identified in previous PPIE activities. Challenge identified in previous PPIE activities UK standard for Public Involvement Solutions implemented in this activity Need to develop shared understanding of statistical methodology Support and Learning • Initial discussion: MST and AS clarified that learning about statistical concepts would be part of the activity. • Session 1: statistical concepts were illustrated via visualisations of data from digital devices and linking to participant experiences. • Session 2: PPIE contributors shared personal experiences of using digital devices in research studies. Causes of missing data and patterns were discussed. • Session 3: MST explained how themes identified across PPIE contributors’ experiences provided insight on missing data considerations. Need for resources to support learning Support and Learning Resources were shared to support learning: • Before Session 1: Plain English Summary and slide deck • After Session 1: Sophie Greenwood’s missing data video and an existing glossary of general statistical terms. Technical terms were defined and deposited in a ‘Too Technical Term (TTT) jar,” which acted as a glossary of terms specific to digital outcome measures. Need for clarity in expectations for PPIE contributors Working Together • A pre-meeting survey before the first session asked PPIE contributors about their expectations of how they would like to be involved. • Session 1: PPIE contributors’ thoughts on expectations in the survey were discussed. Further, MST and AS’s previous experiences and challenges with doing PPIE in statistical methodology were openly discussed. Table 1: Challenges identified through previous PPIE activities for meaningful involvement in statistical methodology, and solutions that were implemented in this activity. Results As this was not a formal qualitative research study but was conducted within Patient and Public Involvement and Engagement (PPIE) remit, the following sections describe the themes and reflections discussed rather than formal research data. PPIE contributor backgrounds Of the ten individuals in the CIG-DO, the majority had previous experience of using digital devices for data collection in research studies or clinical trials. This included wearing Continuous Glucose Monitors (CGM) for a diabetes study (two individuals), wearing a wrist-worn device (e.g., FitBit or smartwatch) and completing digital assessments, such as electronic Patient-Reported Outcomes (ePROs) and voice recordings in research studies related to depression/anxiety, pain, and psychosis (four individuals). Some participants also brought professional research experience, further informing their interests in digital devices in health research. Session 1: Setting the Scene In the pre-session survey, several contributors emphasised that meetings should focus on topics that would matter to “ordinary patients”, and that technical or abstract concepts must always be placed in an understandable real-world context. In the first session, contributors found the visualisations of step count data interesting, noting that it provided a better understanding of the researcher’s perspective. There was discussion on the uncertainty of the accuracy of digital devices. In the post-meeting survey, PPIE contributors noted that they found the missing data video(Greenwood, 2025 ) to be accessible, clear and creative. One contributor suggested that it should be made available in other languages. Session 2: Key points from people’s experiences Reasons for missing data Technological issues Several PPIE contributors experienced major technological issues that could not be resolved, even when they asked research staff for help, such as an app that would systematically crash or a voice recording feature that did not work. These technological malfunctions would not only lead to missing data for specific outcomes but could also cause participant drop-out due to frustration. Additionally, contributors noted that instructions are often not fully adapted to digital platforms. For example, failing to state that an app response is time-sensitive can lead to unintended missing data. Practical issues with operating/setting up devices Practicalities of device maintenance can lead to missing data. Wearable devices which are not waterproof need to be removed for showering and swimming. CGM monitors require replacement of discs every 15 days; if a participant misses a reminder to replace it in time, there could be missing data. Once replaced, it may take up to an hour for the CGM monitor to sync up. Because removing a sensor effectively destroys it and participants have a limited supply, contributors noted that lost-to-follow up patterns are more likely than the intermittent missing data often seen with smartwatches, which participants can easily remove and put back on. Practical issues in daily use Daily comfort can impact missing data. One contributor reported removing their smartwatch during the hot weather due to discomfort, while another mentioned difficulty sleeping while wearing multiple CGMs. However, the latter noted that his previous experience with CGMs helped him to persevere without device removal, suggesting that device familiarity is a key mitigator of missing data. Cultural and religious factors also play a role. One PPIE contributor explained that, as a Muslim, she removes the smartwatch for ablution before prayer and occasionally forgets to put it back on. She also noted that participants are unlikely to wear digital devices during pilgrimages. Feedback from devices PPIE contributors discussed how device feedback can both mitigate and exacerbate missing data. One contributor found FitBit notifications for reaching 10,000 steps encouraging, which increased her engagement. On the other hand, a healthy participant using a CGM monitor for the first time in a trial experienced anxiety when observing changes in his projected glucose levels. Another PPIE contributor with prior CGM experience noted he was already accustomed to observing changes in his glucose levels, highlighting how previous experience with the device can play an important role in how participants interact with technologies. Informative missingness While the PPIE contributors generally followed protocols despite challenges, they identified informative missingness as a critical consideration. For example, in the context of a trial for depression, an increase in symptoms can be all-consuming making trial-related tasks feel like a greater cognitive burden. In such cases, the missing data is directly related to the condition being studied. Session 3: Consolidating Learnings Context matters When, how and why missing data occur may depend on the device, how the use of the device was communicated in the trial, the disease condition being studied, previous experience of the participants with the device, as well as environmental and cultural factors. The importance of the specific context in understanding missing data patterns was highlighted. There is potential for different reasons for missing data to occur together. For example, missing data due to practical challenges around operating devices, which may often be considered to occur at random, may disproportionately affect participants who are more unwell or more affected by their disease, pointing to the potential for informative missingness. Technical difficulties Recurring technical issues reinforced the need for rigorous usability testing before a study begins (Bakker, Barge, Cobb, & Cota, 2024 ). MST reflected a need to use more neutral language when discussing missing data, for example by discussing “when data collection might stop”, rather than “when participants remove devices”. Design of Future PPIE to focus on Informative missingness PPIE contributors recognised the challenge of gathering information about informative missingness. Two key challenges and potential solutions were discussed: selection bias and stigma. Firstly, in a PPIE context, the self-selected individuals who wish to take part in PPIE discussions are likely to be individuals who are interested in research and may be more likely to adhere to protocol. A potential solution is to target recruitment of PPIE contributors within a specific study, possibly using early readouts of digital devices, to identify individuals who have missing data. A second challenge is that there is a stigma associated with not adhering to protocol. To encourage honesty, contributors recommended clearly explaining the value proposition, emphasising how this information has the potential to improve trial analysis and interpretation of the results. They also suggested using anonymous surveys or encouraging open dialogue with staff. Finally, withdrawal forms in trials could explicitly ask if the decision to leave the study is related to their health condition. Results and recommendations are summarised in Table 2 . Post-session Feedback PPIE contributors noted that they enjoyed being able to share their experiences, as well as learning about others’ experiences in similar studies. Some individuals noted that the sessions helped them to imagine what happens to their data after the study and gave them a better understanding of trends in digital research. The full GRIPP2 short-form checklist is available in Table 3 in the Appendix . Table 2 Reasons for missing data, reported by PPIE contributors, and recommendations for trial conduct, trial analyses, and future PPIE activities that were discussed Reason for Missing Data reported by PPIE Contributors, and example quotation Recommendations for Trial Conduct and Analyses Recommendations for future PPIE activities Technical issues and with devices “The monitoring discs that I wear on my arm only last 15 days … they slowly stop working. If you don’t notice the reminder on your phone, ..., or your device has stopped working, there’s a window whereby that data is not gathered.” • Usability testing of digital devices before recruiting participants into a research study is crucial. • Provide clear instructions which have been adapted for the digital outcome. • Provide tech support throughout the trial. Given the potential for technological issues to be the underlying reason for missing data, keep language for PPIE discussions about missing data neutral, e.g. “when data collection might stop” rather than “when participants remove devices”. Practicalities of operating/setting up devices, and practical issues in daily life, e.g. discomfort due to weather, showering, religious activities. “We have to offer prayers 5 times a day, in which we need to do ablution (clean our face, hands, rinse our mouth etc) and if the device is not waterproof, then you need to remove it to do this activity, and most of the time I forgot to wear it [afterwards]." • Encourage ongoing feedback from trial participants during the study to understand factors which can lead to missing data. • Consider the specific context of the device and population when defining missing data, especially if it is via thresholds for wear time. • Consider whether to collect information on variables related to weartime for inclusion in imputation models to strengthen the MAR assumption, such as season/weather or regular activities that affect wear time such as religious or exercise activities (e.g. swimming). Conduct ongoing PPIE discussions during the study to understand what factors may impact missing data for the specific context. Aspects related to the outcome itself, such as increased disease severity leading to inability to engage with the study. These were felt to be important, but were not part of PPIE contributors’ personal experiences. “When you’re depressed, things seem so much more burdensome than they do when you’re not feeling depressed. The condition which is being treated could be a barrier to participation [for self-reported outcomes in an app].” • Collect reasons for missing digital data as part of the trial e.g. via a trial withdrawal form, through speaking to study staff or via feedback questionnaires or portals. • Make value proposition clear to participants: explain why it is important that reasons for missing data are known, and how this can ultimately help research. • Given that informative missingness is an important consideration, plan for and conduct sensitivity analyses to the MAR assumption. • Since PPIE contributors may typically be more likely to adhere to protocol, recruit individuals who may be more likely to have missing data. For example, early readouts from digital devices could identify individuals to contact. • Provide different modalities for giving feedback on PPIE activities; e.g. via anonymous surveys as well as ensuring it forms part of group discussions. Discussion Statistical analysis for digital outcome measures with missing data In clinical trials with digital outcome measures, it is essential to understand missing data mechanisms to identify appropriate statistical assumptions and align statistical analyses accordingly. A key recommendation from this activity is that Patient and Public Involvement and Engagement (PPIE) should be conducted for the purpose of understanding missing data and inform statistical considerations for digital trials. The PPIE activity provided insight into missing data patterns emerging from device-related operational factors. For instance, data loss occurs when CGM sensors require replacement during a study. Discussions also highlighted that weather and season may correlate with continuous weartime of smartwatches. Therefore, these factors may serve as important auxiliary variables in imputation models to strengthen the Missing at Random (MAR) assumption, particularly for physical activity outcomes, which are known to be seasonally variable (Garriga et al., 2022 ; Ridgers et al., 2015 ; Rothman et al., 2025 ). Previous imputation approaches for accelerometer data have linked trial data with weather data from the Met Office (Tackney et al., 2021 ), incorporating weather-related auxiliary variables that were predictive of both missingness and accelerometer outcomes. A key learning point was the potential for religious activities to contribute to missing data, for example, during pilgrimages or because of the need to remove specific devices for ablution. Religious groups may be underrepresented in trials that use digital health technologies, making it important to understand potential barriers to their participation (Digital Medicine Society, n.d.). Although Islamic perspectives of digital health technology have been described (Ali et al., 2025 ; Gamon, 2023 ; Sa’aid et al., n.d.), the impact of religious practice on trial-specific device use remains an under-explored research gap. It is important to explore all possible reasons for removing devices, with a need to ensure that this information captures perspectives from all backgrounds, to ensure that important factors such as religion or religiosity are not missed. Such factors could then potentially be used as auxiliary variables in imputation models in the analysis of trials. Given that PPIE contributors expected the meetings to be understandable to “ordinary patients”, discussions were focused around when, how and why missing data may occur, and whether data might be missing in a completely random way, whether missingness was influenced by specific factors (e.g. season) or whether there may be informative missingness. There was no discussion of which statistical methods were appropriate to use for which set of assumptions; this was felt by the researchers to be beyond the reasonable expectations communicated. Missing data discussions as a springboard to improved conduct PPIE sessions highlighted areas where improved study conduct could directly reduce the quantity of missing. These include resolving technical malfunctions before study initiation and improving communication regarding time-sensitive interface inputs. Conducting these discussions during or after a run-in period may be particularly effective for identifying and resolving practical or technological problems at an early stage. A key learning point was that discussions framed around statistical methodology in a PPIE context have the potential to improve study conduct. While researchers may need to steer the conversation back to statistical concepts, any feedback regarding conduct should be shared with relevant collaborators. Cross-disciplinary communication ensures that these insights maximize the overall value of the study. Recommendations for Future PPIE To obtain further insights on informative missingness, two important challenges and some solutions were discussed: Purposeful recruitment and reducing stigma. Firstly, as PPIE contributors who volunteer to participate in discussion are likely to be individuals who adhere to research protocol, recruiting individuals in a more purposeful way is needed. For example, if PPIE individuals are recruited within an ongoing study, early readouts from the digital devices to identify individuals who have missing data, who may be best placed to provide insight into reasons for missingness. Secondly, PPIE contributors recognised that speaking about not adhering to protocol can have a stigma, and it may be difficult for participants to openly admit to this. They recommended making the value proposition of discussing these aspects very clear, and also recommended allowing these contributions to be provided in different formats, including discussions with research staff or via anonymous surveys. Conclusion Engaging patient and public representatives in the discussion of missing data for digital outcome measures was shown to be a viable pathway to identifying potential reasons and patterns for missing data. The discussions brought to light the potential for operational aspects of the device to lead to data loss, as well as specific activities such as ablution before prayer to lead to device removal, which may be overlooked by researchers. These ongoing PPIE discussions can provide valuable recommendations for both statistical analyses and trial conduct. Further work is required to deepen the understanding of participant experiences on informative missingness. Abbreviations CGM: Continuous Glucose Monitor CIG-DO: Community Involvement Group for Digital Outcomes CUH: Cambridge University Hospitals ePROs: electronic Patient-Reported Outcomes MAR: Missing at Random PPIE: Patient and Public Involvement and Engagement Declarations Human ethics and consent to participate Not applicable. The Public and Patient Involvement and Engagement (PPIE) contributors involved in this work acted as members of the research team (Experts by Experience) and not as research participants. Therefore, formal consent to participate as a research subject, and ethical approval, were not required. Consent for Publication Not applicable Availability of data and materials Not applicable Competing Interests SSV is on the advisory board for PhaseV (unrelated to this work). Funding MST, Advanced Fellow, NIHR305417, is funded by the National Institute of Health and Care Research for this research project. The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. MLZ is funded by the Edmond J. Safra Foundation as part of the Edmond J. Safra Accelerating Clinical Trials in Parkinson’s Disease (EJS ACT-PD) Initiative. Authors Contributions MST conceived the research idea, led the investigation, and led the writing of the manuscript. AS facilitated PPIE meetings and provided guidance on conducting PPIE activities and expertise on digital health technologies. MLZ facilitated one PPIE meeting, provided guidance on conducting PPIE activities and the structure of the manuscript, and provided expertise on trial conduct. SSV provided expertise in design, analysis and conduct of clinical trials. SM, JD and FY provided expertise by experience in using digital health technologies in clinical studies, as well as expertise in PPIE. All authors reviewed and edited the manuscript. Acknowledgements Authors would like to thank Sophie Greenwood, James Carpenter, Alison Yarnall, Ríona McArdle and Jack Lumsdon for helpful discussions. Authors would like to acknowledge and thank all ten individuals in the CIG-DO for their invaluable contribution as experts by experience in the design and participation in the PPIE activities: Melissa Cox Jeremy Dearling Curie Freeborn Lorraine Hazlehurst Ian Hudson Carrol Lamouline Sarah Markham Yazan Mehyar Mike Willis Farheen Yameen References Abell L, Maher F, Begum S, Booth S, Broomfield J, Lee S, Smith E, Stannard R, Teece L, Vounzoulaki E, Worboys H, Gray LJ. Incorporation of patient and public involvement in statistical methodology research: a survey assessing current practices and attitudes of researchers. Res Involv Engagem. 2023;9(1). https://doi.org/10.1186/s40900-023-00507-5 . Ali SM, Saiyed MM, McAvoy A, Mackin R, Jay C, van der Veer SN. Determinants of mistrust in digital health research and approaches to address them among Muslim ethnic minorities living in the United Kingdom: a qualitative study. Int J Equity Health. 2025;24(1). https://doi.org/10.1186/s12939-025-02583-3 . Bakker JP, Barge R, Cobb B, Cota C. (2024). V3+: An extension to the V3 framework to ensure user-centricity and scalability of sensor-based digital health technologies . icity-and-scalability-of-sensor-based-digital-health-technologies/ Bakker JP, Barge R, Cobb B, Cota C, Guo CC, Hartog B, Horowicz-Mehler N, Izmailova ES, McClenahan S, Motola S, Patel S, Paun O, Schoone M, Sezgin E, Switzer T, Tandon A, van den Brink W, Vairavan S, Vandendriessche B, Goldsack JC. (2024). V3+: An extension to the V3 framework to ensure user-centricity and scalability of sensor-based digital health technologies . https://datacc.dimesociety.org/resources/v3-an-extension-to-the-v3-framework-to-ensure-user-centricity-and-scalability-of-sensor-based-digital-health-technologies/ Booth S, Wells M, Nevill C, Teece L, Czyznikowska B, Grewal-Santini G, Mancini M, Yameen F, Freeman SC. Plugging the gap’: development of a plain language glossary for statistical methodology research. Res Involv Engagem. 2025;11(1). https://doi.org/10.1186/s40900-025-00782-4 . Broomfield J, Hill M, Chandler F, Crowther MJ, Godfrey J, Guglieri M, Hastie J, Larkindale J, Mumby-Croft J, Reuben E, Woodcock F, Abrams KR. Developing a Natural History Model for Duchenne Muscular Dystrophy. PharmacoEconomics - Open. 2024;8(1):79–89. https://doi.org/10.1007/s41669-023-00450-x . Carpenter JR, Smuk M. Missing data: A statistical framework for practice. Biom J. 2021;63(5):915–47. https://doi.org/10.1002/bimj.202000196 . Cro S, Kahan BC, Patel A, Henley A, Joanna J, Hellyer P, Kumar M, Rahman Y, Goulão B. Starting a conversation about estimands with public partners involved in clinical trials: a co-developed tool. Trials. 2023;24(1). https://doi.org/10.1186/s13063-023-07469-9 . Cro S, Van Vogt E, Totton N, Lee E, Hellyer CJ, Kumar P, Rahman M, Y., Henley A. Supporting public involvement in defining estimands: a practical tool accessibly explaining the five key attributes of an estimand. Trials. 2025;26(1). https://doi.org/10.1186/s13063-025-08941-4 . Crowe S, Adebajo A, Esmael H, Denegri S, Martin A, McAlister B, Moore B, Quinn M, Rennard U, Simpson J, Wray P, Yeeles P. (2020). ‘All hands-on deck’, working together to develop UK standards for public involvement in research. In Research Involvement and Engagement (Vol. 6, Number 1). BioMed Central Ltd. https://doi.org/10.1186/s40900-020-00229-y Di J, Demanuele C, Kettermann A, Karahanoglu FI, Cappelleri JC, Potter A, Bury D, Cedarbaum JM, Byrom B. (2022). Considerations to address missing data when deriving clinical trial endpoints from digital health technologies. Contemporary Clinical Trials , 113 . https://doi.org/10.1016/j.cct.2021.106661 Digital Medicine Society. (n.d.). The Framework for Inclusive Development. DATAcc by DiMe. Retrieved 30. December 2025, from https://datacc.dimesociety.org/resources/the-framework-for-inclusive-development/ European Medicines Agency. ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials - Step 2b. Ema. 2017;44(August):1–23. . Gamon AMI. Ethics of Digital Health in Islamic Perspective. J Sci Technol. 2023;28(1):1–11. https://doi.org/10.20428/jst.v28i1.1993 . Garriga A, Sempere-Rubio N, Molina-Prados MJ, Faubel R. (2022). Impact of seasonality on physical activity: A systematic review. In International Journal of Environmental Research and Public Health (Vol. 19, Number 1). MDPI. https://doi.org/10.3390/ijerph19010002 Goulao B, Bruhn H, Campbell M, Ramsay C, Gillies K. Patient and public involvement in numerical aspects of trials (PoINT): exploring patient and public partners experiences and identifying stakeholder priorities. Trials. 2021;22(1). https://doi.org/10.1186/s13063-021-05451-x . Greenwood S. (2025). What do we mean by ‘missing data’ in clinical trials? A story of patient involvement . https://www.youtube.com/watch?v=DLBVXCru8cI Greenwood S, Morris TP, O’Malley L, Aucott L, Fagbemi O, Caie R, Goulao B. (2026). Can we elicit patient knowledge about missing data in trials? A co-design study to inform a patient-informed sensitivity analysis of missing data in trials using expert elicitation. Greenwood S, Morris TP, O’Malley L, Aucott L, Goulao B. (2026). Eliciting and incorporating patient’s opinions about missing data in randomised controlled trials . https://abdn.primo.exlibrisgroup.com/discovery/fulldisplay?context=L&vid=44ABE_INST:44ABE_VU1⟨=en&docid=alma9918660731505941 Keogh A, Ardle RM, Diaconu MG, Ammour N, Arnera V, Balzani F, Brittain G, Buckley E, Buttery S, Cantu A, Corriol-Rohou S, Delgado-Ortiz L, Duysens J, Forman-Hardy T, Gur-Arieh T, Hamerlijnck D, Linnell J, Leocani L, McQuillan T, Consortium MD. Mobilizing Patient and Public Involvement in the Development of Real-World Digital Technology Solutions: Tutorial. J Med Internet Res. 2023;25. https://doi.org/10.2196/44206 . Kettermann A, Dandi G, Clark J, Kim Y, Song J, Pucino F, Almario N, E. Navigating the Future: Considerations for use of Continuous Glucose Monitoring in Diabetes Trials. Therapeutic Innov Regul Sci. 2025. https://doi.org/10.1007/s43441-025-00880-1 . National Institute for Health and Care Research. (2024). Payment Guidance for Researchers and Professionals . National Institute for Health Research. (2013). Patient and Public Involvement in Health and Social Care Research: A handbook for researchers . Ridgers N, Salmon J, Timperio A. Too hot to move? Objectively assessed seasonal changes in Australian children’s physical activity. Int J Behav Nutr Phys Act. 2015;12:77. Rothman AMK, Middleton J, Zafar H, De Bie EMDD, Newman J, Varian F, Taylor J, Hitchcock F, Ashraf C, Patel J, Thompson AAR, Toshner M. (2025). Remote Actigraphy Is Seasonally Variable with Implications for Clinical Monitoring. MedRxiv , 2025.05.22.25328110. https://doi.org/10.1101/2025.05.22.25328110 Sa’aid HB, Azman A, Ramli SI, Kasim NH, Maon SS, Hasan HA, Zakaria AB. (n.d.). Islamic Bioethics in Health Technology Assessment: A Review of Key Ethical Frameworks . https://doi.org/10.47772/IJRISS Smith A, Worboys H, Begum S, Bennett D, Broomfield J, Cro S, Evans-Hill L, Greenwood J, Henley A, Mancini M, Royle KL, Saul H, Sergeant J, Stewart D, Tyrer F, Wason J, Yau C, Gray LJ. (2025). Incorporation of Patient and Public Involvement in Statistical Methodology Research: Summary of Workshop Proceedings. In Statistics in Medicine (Vol. 44, Numbers 15–17). John Wiley and Sons Ltd. https://doi.org/10.1002/sim.70159 Tackney MS, Carpenter JR, Villar SS. Unleashing the full potential of digital outcome measures in clinical trials: eight questions that need attention. BMC Med. 2024;22(1):413. https://doi.org/10.1186/s12916-024-03590-x . Tackney MS, Cook DG, Stahl D, Ismail K, Williamson E, Carpenter J. A framework for handling missing accelerometer outcome data in trials. Trials. 2021;22(1):1–18. https://doi.org/10.1186/s13063-021-05284-8 . Tackney MS, Steele A, Newman J, Fritzsche M-C, Lucivero F, Khadjesari Z, Lynch J, Abbott RA, Barber VS, Carpenter JR, Copsey B, Davies EH, Dixon WG, Fox L, González J, Griffiths J, Hinchliffe CHL, Kolanko MA, McGagh D, Villar SS. Digital endpoints in clinical trials: emerging themes from a multi-stakeholder Knowledge Exchange event. Trials. 2024;25(1):521. https://doi.org/10.1186/s13063-024-08356-7 . Worboys HM, Broomfield J, Smith A, Stannard R, Tyrer F, Vounzoulaki E, Czyznikowska B, Grewal-Santini G, Greenwood J, Gray LJ. Incorporation of patient and public involvement in statistical methodology research: development of an animation. Res Involv Engagem. 2023;9(1). https://doi.org/10.1186/s40900-023-00513-7 . Additional Declarations Competing interest reported. SSV is on the advisory board for PhaseV (unrelated to this work). Supplementary Files Appendix.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 10 Apr, 2026 Editor invited by journal 09 Apr, 2026 Editor assigned by journal 09 Apr, 2026 Submission checks completed at journal 09 Apr, 2026 First submitted to journal 01 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9293900","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":626653361,"identity":"8654eb3e-3dba-4641-ba6b-73658e7979cb","order_by":0,"name":"Mia S. Tackney","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtElEQVRIiWNgGAWjYHAC5sc/KuCcBKJ0sBkznCFRC4M0YxspWnT7zx8wLpxnJ6/bwPzwA2NbGmEtZjeSGR7P3JZsuO0Am7EEY1sOMVqYGQx4tx1g3HaAwYyBsa2CCC3nDzNI8M45YL/tAPs3IrUcSGaQ5m04kLjtAA/IFqIclmxmOONYcvK2wzzFEgnniPH++YOPH3yosbPddrx944cPZcmEtSAAMwORETkKRsEoGAWjgDAAAK0ROIVidwPtAAAAAElFTkSuQmCC","orcid":"","institution":"University of Cambridge","correspondingAuthor":true,"prefix":"","firstName":"Mia","middleName":"S.","lastName":"Tackney","suffix":""},{"id":626653362,"identity":"7629191b-d1e4-46ef-a9ae-cba4980cc105","order_by":1,"name":"Amber Steele","email":"","orcid":"","institution":"Cambridge University Hospitals NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Amber","middleName":"","lastName":"Steele","suffix":""},{"id":626653363,"identity":"3b96b751-c276-49ce-95e3-ef7b0114a2e4","order_by":2,"name":"Marie-Louise Zeissler","email":"","orcid":"","institution":"Newcastle University","correspondingAuthor":false,"prefix":"","firstName":"Marie-Louise","middleName":"","lastName":"Zeissler","suffix":""},{"id":626653364,"identity":"9b03e1ff-95da-4035-a0b8-b8b8b2b6ce44","order_by":3,"name":"Sofía S. Villar","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Sofía","middleName":"S.","lastName":"Villar","suffix":""},{"id":626653365,"identity":"a9fde7ea-aabc-4262-b232-d133ea149fc7","order_by":4,"name":"Jeremy Dearling","email":"","orcid":"","institution":"Expert by Experience","correspondingAuthor":false,"prefix":"","firstName":"Jeremy","middleName":"","lastName":"Dearling","suffix":""},{"id":626653366,"identity":"0ced3d1f-f67d-4150-962c-b498cb4cc013","order_by":5,"name":"Sarah Markham","email":"","orcid":"","institution":"King's College London","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Markham","suffix":""},{"id":626653370,"identity":"0780187d-35a8-4311-a642-9147afb95604","order_by":6,"name":"Farheen Yameen","email":"","orcid":"","institution":"Expert by Experience","correspondingAuthor":false,"prefix":"","firstName":"Farheen","middleName":"","lastName":"Yameen","suffix":""},{"id":626653375,"identity":"ffe7f4d1-8fb8-4709-bb07-5fb7cdb4541a","order_by":7,"name":"Community Involvement Group for Digital Outcomes (CIG-DO)","email":"","orcid":"","institution":"Expert by Experience","correspondingAuthor":false,"prefix":"","firstName":"Community","middleName":"Involvement Group for Digital Outcomes","lastName":"(CIG-DO)","suffix":""}],"badges":[],"createdAt":"2026-04-01 15:38:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9293900/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9293900/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107515257,"identity":"27ce18cd-5f2a-4f02-b783-fc5ac251c2cf","added_by":"auto","created_at":"2026-04-22 08:28:13","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":210113,"visible":true,"origin":"","legend":"\u003cp\u003eExample of seven days of accelerometer data from a trial participant. Here, intervals where individuals removed the device are indicated in red. There are areas of ambiguity; for example, on Wednesday, it is possible that the individual went to sleep earlier than other days, or it could be that they removed the device early but were still active in the evening. Figure reproduced from Tackney et al (2023) under the terms of the Creative Commons Attribution 4.0 License (CC BY 4.0).\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9293900/v1/7c80a97a37bdd62df8fd3ee8.png"},{"id":107515170,"identity":"db0e5e25-d71f-40a0-84c1-adb35aa71038","added_by":"auto","created_at":"2026-04-22 08:28:09","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":321021,"visible":true,"origin":"","legend":"\u003cp\u003eVisualization of missing data from CGM monitors. Heatmap displays the proportion of missing values during the baseline CGM observation period of a trial for 50 randomly selected patients. Each row represents one patient during the observation period, and each rectangle represents one day of observation. The vertical dashed line indicates the first 14 days of the observation period. Colors indicate the proportion of missing data for one day; yellow for minimal or no missing data, red for a high proportion of missing data, and grey for non-wear days. Rectangles that are outlined in black indicate days when a patient changed their sensor. Kettermann, A., Dandi, G., Clark, J. et al. Navigating the Future: Considerations for use of Continuous Glucose Monitoring in Diabetes Trials. \u003ca href=\"https://link.springer.com/journal/43441\"\u003eTherapeutic Innovation \u0026amp; Regulatory Science\u003c/a\u003e (2025). Reproduced with permission from SNCSC. This material is not part of the governing OA license but has been reproduced with permission.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9293900/v1/0bb498f619841606c0334d6f.png"},{"id":107515443,"identity":"1b8ee252-1211-45fa-917a-330b4a6d8a64","added_by":"auto","created_at":"2026-04-22 08:28:53","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1064107,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9293900/v1/1c9d3691-141f-44be-90f1-e44063acffd8.pdf"},{"id":107514850,"identity":"51ecc206-2ce5-453b-b10c-1dfec94ae0e2","added_by":"auto","created_at":"2026-04-22 08:27:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":16196,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-9293900/v1/f2e0ad8b034d3e2c436c96d6.docx"}],"financialInterests":"Competing interest reported. SSV is on the advisory board for PhaseV (unrelated to this work).","formattedTitle":"Can discussions with patients and the public clarify missing data mechanisms for digital outcome measures?","fulltext":[{"header":"PLAIN ENGLISH SUMMARY ","content":"\u003cp\u003e\u003cstrong\u003eWhy Data from Digital Devices Goes \u0026ldquo;Missing\u0026rdquo; in Clinical Trials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhen researchers run medical studies, they sometimes use wearable devices to collect health information. For example, smartwatches are watches that can track your heart rate and physical activity, and continuous glucose monitors measure blood sugar levels. Sometimes, these devices stop collecting data. When this happens, gaps appear in the results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWhy does this happen?\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTen patient and public contributors shared their real-world experiences of using digital devices in research. They helped researchers understand why data might be missing. Some common reasons included:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eTechnical problems:\u003c/strong\u003e The device or app may stop working.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eMaintenance:\u003c/strong\u003e Device might need charging, or some parts of the device need to be replaced, which can cause short gaps in data.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eDiscomfort or personal reasons:\u003c/strong\u003e People may remove devices if they feel uncomfortable, for example in hot weather, or during religious activities.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe reasons listed above are mainly practical, such as technical problems or needing to replace part of the device. However, sometimes, missing data happens for a different reason that is linked to a person\u0026rsquo;s health. If someone\u0026rsquo;s symptoms get worse, they may be less likely to wear or use the device. This is especially important because missing data is directly related to what the study is trying to measure.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUnderstanding these reasons can help researchers improve how studies are run, for example by giving clearer instructions\u0026nbsp;and better preparing participants.\u0026nbsp;It can also help statisticians carry out statistical analysis (using mathematical tools to make sense of the data) in the right way.\u003c/p\u003e\n\u003cp\u003eFor future patient and public involvement activities, it may help to recruit contributors who are participating in a specific clinical study to focus on a specific context, and to offer different ways to give feedback, such as anonymous surveys.\u003c/p\u003e\n"},{"header":"Background","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003eMissing data assumptions in Clinical Trials with Digital Outcomes\u003c/h2\u003e \u003cp\u003eStatistical methods rely on specific assumptions, which act as fixed conditions under which the methods can best achieve their intended theoretical and inferential properties. In clinical trials, missing data is a pervasive issue, arising from technical difficulties, participant withdrawal, or missed assessments. A common assumption in the analysis of trials is that data are Missing at Random (MAR). This implies that the probability of data being missing relates to observed information (such as baseline characteristics or earlier measurements) rather than the unobserved value itself.\u003c/p\u003e \u003cp\u003eIf the assumption is incorrect, statistical estimates may be biased and standard errors may be invalid, potentially leading to incorrect conclusions about the treatment effect (Carpenter \u0026amp; Smuk, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). For example, analyses assuming that data are MAR may underestimate treatment effect if participants with poor outcomes are more likely to withdraw from the study. Consequently, sensitivity analyses are recommended, which assess whether results remain robust under alternative missing data assumptions (European Medicines Agency, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eCollecting digital outcome measures via digital health technologies introduces unique challenges regarding missing data. Step count data from accelerometers, glucose readings from continuous glucose monitors (CGMs), and patient-reported outcomes from smartphone applications may each have specific patterns of missing data and reasons for missing data (Di et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Tackney, Carpenter, et al., 2024). In these settings, missing data often arises from technology-related issues, such as internet or Bluetooth connectivity failures, battery depletion, or device malfunction. Further, because these devices are typically operated by participants in their daily lives, the responsibility for data collection shifts towards the individual. Participant experience, therefore, becomes central to understanding missingness e.g., the removal of wearables due to discomfort or privacy concerns. Understanding these patterns of potential data loss is essential not only for evaluating device usability as part of the validation of digital devices (Bakker, Barge, Cobb, Cota, et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Keogh et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), but also for identifying plausible missing data assumptions in the statistical analyses of the resulting data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eThe Need for Patient and Public Perspectives\u003c/h2\u003e \u003cp\u003eThere is a clear need to understand trial participants\u0026rsquo; experiences to identify sensible missing data assumptions for digital outcome measures. Two strands of previous work suggest this as a promising direction. Firstly, a priority-setting exercise on numerical aspects of clinical trials(Goulao et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) highlighted that Patient and Public Involvement and Engagement (PPIE) contributors felt well placed to challenge whether statistical modelling assumptions reflected real-world participant experiences. However, while contributors felt that explicit discussion of these assumptions increased research transparency, technical jargon often obscured them. This underscores the need for more effective communication and visibility of methodological assumptions. Secondly, tools have been developed to elicit expert and patient and public opinions on missing data. For example, Mason et al. (2017) asked doctors and nurses to quantify how outcomes for individuals with missing data might differ from those with observed data. These expert opinions then informed sensitivity analyses for the Missing at Random (MAR) assumption. Greenwood et al. conducted an elicitation exercise on missing data for a historical trial with patient and public partners who were not participants in the trial (Greenwood, Morris, O\u0026rsquo;Malley, Aucott, \u0026amp; Goulao, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2026\u003c/span\u003e; Greenwood, Morris, O\u0026rsquo;Malley, Aucott, Fagbemi, et al., 2026). There is a paucity of experience in exploring concepts of missing data with patient and public representatives who have lived experience of being in clinical studies.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eChallenges in PPIE for Statistical Methodology\u003c/h3\u003e\n\u003cp\u003eIt is widely recognised that PPIE, where the public contributes to research design, conduct and delivery, can lead to improved outcomes and delivery of research (National Institute for Health Research, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). However, there are distinct challenges to meaningful PPIE in statistical methodology research. Statistical methods can appear abstract compared to patients\u0026rsquo; lived experiences, so deliberate effort is required to bridge these conceptualisations and perspectives (Abell et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Furthermore, the technical nature of the research requires establishing a shared foundation of knowledge for productive discussion. The appropriate level of information to share is essential, aligning with the \u003cem\u003eSupport and Learning\u003c/em\u003e principle of the UK Standards for Public Involvement (Crowe et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eBridging the Gap\u003c/h3\u003e\n\u003cp\u003eTo address these challenges, a growing body of practical expertise has emerged. This includes a community of practice, training workshops for statistical methodologists (Smith et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) successful co-production of animation videos (Worboys et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), glossaries of key terms (Booth et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and toolkits to explain statistical concepts such as estimands (Cro et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWhile most developments focus on the communication and dissemination stages of the research cycle, less attention has been paid to embedding PPIE across the full statistical research lifecycle, such as in developing novel methods, evaluating methodological assumptions, or application of statistical methods (Broomfield et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Involvement during the design stages is critical within the \u003cem\u003eSupport and Learning\u003c/em\u003e and \u003cem\u003eWorking Together\u003c/em\u003e principles of the UK Standards for Public Involvement and require specific planning for meaningful involvement.\u003c/p\u003e \u003cp\u003eThis article describes a PPIE activity designed to integrate patient and public perspectives into discussions on missing data for digital outcome measures in clinical trials. We describe how this PPIE activity was carried out, demonstrating a viable pathway for PPIE in this context and its value to methodological development. Our results offer recommendations for the conduct and analysis of trials using digital devices, as well as in future PPIE activities aimed at understanding informative missingness.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec7\"\u003e\n \u003ch2\u003ePatient and Public Involvement in Design\u003c/h2\u003e\n \u003cp\u003eThe design of this activity involved Patient and Public Involvement and Engagement (PPIE) contributors at every stage. During the initial funding proposal, a discussion group with three members of the Cambridge University Hospitals (CUH) PPIE panel helped refine research aims and establish ongoing contributor involvement. Subsequently, eleven volunteers from the CUH PPIE panel reviewed the application’s Plain English Summary and PPIE plan. Their feedback focused on research relevance, text accessibility, and feasibility of the proposed activities. From these early engagements, three contributors with a specific interest in the field joined the Community Involvement Group for Digital Outcomes (CIG-DO).\u003c/p\u003e\n \u003cp\u003eTo further identify research priorities for digital outcome measures, we convened a multi-stakeholder event. This brought together 32 attendees, including clinicians, statisticians, computer scientists, implementation scientists, ethicists and health economists. Six PPIE contributors participated: four attended in person (two members of CIG-DO) to represent the patient and public perspective, while two contributed remotely via video and written comments of their lived experience digital research devices. The methodology and insights from this event have been previously published (Tackney, Steele, et al., 2024).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\"\u003e\n \u003ch2\u003eKey Findings from Preliminary Engagement\u003c/h2\u003e\n \u003cp\u003eFeedback from these initial activities, alongside a foundation meeting with the CIG-DO, highlighted three specific areas requiring attention in subsequent phases:\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e• Need to develop shared understanding of statistical methodology\u003c/h3\u003e\n\u003cp\u003eWhile PPIE contributors offered perceptive and practical solutions for preventing missing data (such as participant reminders), discussing statistical methodology required more dedicated time. Conveying the abstract concept that missing data must be addressed through statistical analysis, rather than prevention alone, remained a challenge that aligned with the \u003cem\u003eSupport and Learning\u003c/em\u003e standard.\u003c/p\u003e\n\u003ch3\u003e• Need for resources to support learning\u003c/h3\u003e\n\u003cp\u003eSome PPIE contributors found the technical jargon of the multi-stakeholder event overwhelming. They recommended providing a glossary of key terms, as well as a Plain English Summary for all speaker presentations, and a dedicated space for PPIE contributors to hold independent discussions (Tackney et al., 2024). Establishing a culture of learning with designated resources is vital to the \u003cem\u003eSupport and Learning\u003c/em\u003e principle.\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cstrong\u003eNeed for clarity in expectations for PPIE contributors.\u003c/strong\u003e\u003c/p\u003e\n \u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eDuring the multi-stakeholder event, some contributors felt uncertain about their specific role in discussions. This underscores the\u0026nbsp;\u003cem\u003eWorking Together\u003c/em\u003e principle, and the need to jointly shape and define the purpose of the activity to ensure a shared understanding of roles expectations and responsibilities.\u003cbr\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cdiv id=\"Sec11\"\u003e\n \u003ch2\u003eRecruiting additional members to the CIG-DO\u003c/h2\u003e\n \u003cp\u003eTo recruit additional members to the CIG-DO group, MST posted the opportunity to join the group on the CUH PPIE newsletter. Seven individuals expressed interest and had an initial discussion with MST and AS to discuss the research topic, the level of involvement that would be expected, and whether there were any specific needs.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\"\u003e\n \u003ch2\u003eObjectives and Structure of Activity\u003c/h2\u003e\n \u003cp\u003eBetween September – December 2025, each contributor attended three online sessions led by MST and supported by a facilitator (AS, with MLZ as an alternate). To accommodate the availability of the 10-member group and ensure that everyone had sufficient time to contribute, sessions utilised a blend of small group discussions and one-on-one meetings. PPIE contributors were compensated for their time spent in online meetings and providing feedback, as well as for remote working expenses, in accordance with NIHR payment guidelines (National Institute for Health and Care Research, 2024).\u003c/p\u003e\n \u003cp\u003eThere were two key objectives of the activity:\u003c/p\u003e\n \u003cp\u003e1. Exploratory objective: understand how missing data might arise in trials with digital outcome measures, drawing from the experiences who have lived experience of participation in research studies that used digital devices.\u003c/p\u003e\n \u003cp\u003e2. Focused objective: understand whether there could be informative missingness (when missingness relates to the outcome itself) for digital outcome measures.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\"\u003e\n \u003ch2\u003eSession 1: Introduction to digital outcome measures\u003c/h2\u003e\n \u003cp\u003ePrior to the first session, the Plain English Summary of the funding proposal, and the meeting slide deck were shared with participants. A pre-session survey gathered PPIE contributors’ expectations for the activity.\u003c/p\u003e\n \u003cp\u003eThe session began with introductions from all attendees, followed by an overview of key statistical concepts, including digital outcome measures and missing data. Visualisations of minute-by-minute accelerometer step-count data, collected over seven days, were presented (e.g., Fig.\u0026nbsp;1). The group discussed potential reasons for missing data and the resulting implications for data analysis.\u003c/p\u003e\n \u003cp\u003eTo foster a culture of shared learning, technical terms used during meetings were deposited in a “Too Technical Term (TTT) Jar”, which MST described as being like a “swear jar.” PPIE contributors were encouraged to point out whenever a TTT (Too Technical Term) was used in discussions, and definitions were deposited in a shared document. In addition to this shared document containing specific terms related to digital outcome measures, PPIE contributors were given a link to a glossary of general statistical terms and video on statistical methodology (Booth et al., 2025; Worboys et al., 2023).\u003c/p\u003e\n \u003cp\u003eTo establish a shared understanding of the project, the group discussed the goals, expectations, and potential impact of the PPIE activities. MST emphasised that the meetings would not focus on one health condition or treatment (which may be typical for clinical PPIE activities), but rather around how participant experiences can inform statisticians’ understanding of how missing data may arise. Through the visualisations of missing data, MST explained how, as a statistician, one may not always have contextual information for why data may be missing, and that patient and public partners’ lived experiences could be informative. The group also discussed the potential impact of the PPIE activities; for example, the insights gained could improve the design, conduct and analysis of ongoing studies with digital outcome measures, such as the Edmond J. Safra Accelerating Clinical Trials for Parkinson’s Disease (EJS ACT-PD).\u003c/p\u003e\n \u003cp\u003eAfter the first session, MST asked the group to watch Sophie Greenwood’s introductory video on missing data, which was co-produced with PPIE contributors from her PhD work (https://www.youtube.com/watch?v=DLBVXCru8cI) (Greenwood, 2025). A post-meeting survey invited reflections on the video. Furthermore, contributors with lived experience of using digital devices in research studies were invited to prepare a short talk explaining the context of the trial, the device used, and any reasons for missing data.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\"\u003e\n \u003ch2\u003eSession 2: Listening to PPIE contributors’ experiences\u003c/h2\u003e\n \u003cp\u003ePPIE contributors with trial experience gave short talks on their experiences, guided by questions prepared by MST and AS. After each talk, MST asked additional questions to highlight specific statistical considerations. Contributors without lived experience participated by asking questions and sharing whether the experiences described aligned with their own expectations.\u003c/p\u003e\n \u003cp\u003eTo address the project’s second objective, MST reintroduced the concept of “informative missingness” (which was added to the “Too Technical Term” jar) and provided examples of how it occurs with digital outcome measures. MST asked the group to reflect on whether informative missingness is likely within each of the previously described contexts.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\"\u003e\n \u003ch2\u003eSession 3: Feedback and summary meeting\u003c/h2\u003e\n \u003cp\u003eMST summarised experiences shared in Session 2 and identified key themes through discussions with AS, SSV and MLZ, and experts on devices and missing data. During the feedback meeting, MST shared these themes and described how the contributors’ experiences aligned with published findings on missing data patterns, e.g. analyses of CGM monitor data described by Kettermann et al. (2025), as shown in Fig.\u0026nbsp;2.\u003c/p\u003e\n \u003cp\u003eFinally, MST invited PPIE contributors to share ideas on how future research and PPIE activities could better capture and account for informative missingness.\u003c/p\u003e\n \u003cdiv\u003e \u0026nbsp;\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv\u003eTable 1\u003c/div\u003e\n \u003cdiv\u003e\n \u003cp\u003esummarises the approaches used in this activity to address challenges identified in previous PPIE activities.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"3\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eChallenge identified in previous PPIE activities\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eUK standard for Public Involvement\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eSolutions implemented in this activity\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eNeed to develop shared understanding of statistical methodology\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eSupport and Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003e• Initial discussion: MST and AS clarified that learning about statistical concepts would be part of the activity.\u003c/p\u003e\n \u003cp\u003e• Session 1: statistical concepts were illustrated via visualisations of data from digital devices and linking to participant experiences.\u003c/p\u003e\n \u003cp\u003e• Session 2: PPIE contributors shared personal experiences of using digital devices in research studies. Causes of missing data and patterns were discussed.\u003c/p\u003e\n \u003cp\u003e• Session 3: MST explained how themes identified across PPIE contributors’ experiences provided insight on missing data considerations.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eNeed for resources to support learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eSupport and Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eResources were shared to support learning:\u003c/p\u003e\n \u003cp\u003e• Before Session 1: Plain English Summary and slide deck\u003c/p\u003e\n \u003cp\u003e• After Session 1: Sophie Greenwood’s missing data video and an existing glossary of general statistical terms.\u003c/p\u003e\n \u003cp\u003eTechnical terms were defined and deposited in a ‘Too Technical Term (TTT) jar,” which acted as a glossary of terms specific to digital outcome measures.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eNeed for clarity in expectations for PPIE contributors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eWorking Together\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003e• A pre-meeting survey before the first session asked PPIE contributors about their expectations of how they would like to be involved.\u003c/p\u003e\n \u003cp\u003e• Session 1: PPIE contributors’ thoughts on expectations in the survey were discussed. Further, MST and AS’s previous experiences and challenges with doing PPIE in statistical methodology were openly discussed.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eTable\u0026nbsp;1: \u003cem\u003eChallenges identified through previous PPIE activities for meaningful involvement in statistical methodology, and solutions that were implemented in this activity.\u003c/em\u003e\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eAs this was not a formal qualitative research study but was conducted within Patient and Public Involvement and Engagement (PPIE) remit, the following sections describe the themes and reflections discussed rather than formal research data.\u003c/p\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePPIE contributor backgrounds\u003c/h2\u003e \u003cp\u003eOf the ten individuals in the CIG-DO, the majority had previous experience of using digital devices for data collection in research studies or clinical trials. This included wearing Continuous Glucose Monitors (CGM) for a diabetes study (two individuals), wearing a wrist-worn device (e.g., FitBit or smartwatch) and completing digital assessments, such as electronic Patient-Reported Outcomes (ePROs) and voice recordings in research studies related to depression/anxiety, pain, and psychosis (four individuals). Some participants also brought professional research experience, further informing their interests in digital devices in health research.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eSession 1: Setting the Scene\u003c/h2\u003e \u003cp\u003eIn the pre-session survey, several contributors emphasised that meetings should focus on topics that would matter to \u0026ldquo;ordinary patients\u0026rdquo;, and that technical or abstract concepts must always be placed in an understandable real-world context.\u003c/p\u003e \u003cp\u003eIn the first session, contributors found the visualisations of step count data interesting, noting that it provided a better understanding of the researcher\u0026rsquo;s perspective. There was discussion on the uncertainty of the accuracy of digital devices. In the post-meeting survey, PPIE contributors noted that they found the missing data video(Greenwood, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) to be accessible, clear and creative. One contributor suggested that it should be made available in other languages.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eSession 2: Key points from people\u0026rsquo;s experiences\u003c/h2\u003e \u003cdiv id=\"Sec20\" class=\"Section3\"\u003e \u003ch2\u003eReasons for missing data\u003c/h2\u003e \u003cdiv id=\"Sec21\" class=\"Section4\"\u003e \u003ch2\u003eTechnological issues\u003c/h2\u003e \u003cp\u003eSeveral PPIE contributors experienced major technological issues that could not be resolved, even when they asked research staff for help, such as an app that would systematically crash or a voice recording feature that did not work. These technological malfunctions would not only lead to missing data for specific outcomes but could also cause participant drop-out due to frustration. Additionally, contributors noted that instructions are often not fully adapted to digital platforms. For example, failing to state that an app response is time-sensitive can lead to unintended missing data.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003ePractical issues with operating/setting up devices\u003c/h2\u003e \u003cp\u003ePracticalities of device maintenance can lead to missing data. Wearable devices which are not waterproof need to be removed for showering and swimming. CGM monitors require replacement of discs every 15 days; if a participant misses a reminder to replace it in time, there could be missing data. Once replaced, it may take up to an hour for the CGM monitor to sync up. Because removing a sensor effectively destroys it and participants have a limited supply, contributors noted that lost-to-follow up patterns are more likely than the intermittent missing data often seen with smartwatches, which participants can easily remove and put back on.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003ePractical issues in daily use\u003c/h2\u003e \u003cp\u003eDaily comfort can impact missing data. One contributor reported removing their smartwatch during the hot weather due to discomfort, while another mentioned difficulty sleeping while wearing multiple CGMs. However, the latter noted that his previous experience with CGMs helped him to persevere without device removal, suggesting that device familiarity is a key mitigator of missing data.\u003c/p\u003e \u003cp\u003eCultural and religious factors also play a role. One PPIE contributor explained that, as a Muslim, she removes the smartwatch for ablution before prayer and occasionally forgets to put it back on. She also noted that participants are unlikely to wear digital devices during pilgrimages.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eFeedback from devices\u003c/h2\u003e \u003cp\u003ePPIE contributors discussed how device feedback can both mitigate and exacerbate missing data. One contributor found FitBit notifications for reaching 10,000 steps encouraging, which increased her engagement. On the other hand, a healthy participant using a CGM monitor for the first time in a trial experienced anxiety when observing changes in his projected glucose levels. Another PPIE contributor with prior CGM experience noted he was already accustomed to observing changes in his glucose levels, highlighting how previous experience with the device can play an important role in how participants interact with technologies.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eInformative missingness\u003c/h2\u003e \u003cp\u003eWhile the PPIE contributors generally followed protocols despite challenges, they identified informative missingness as a critical consideration. For example, in the context of a trial for depression, an increase in symptoms can be all-consuming making trial-related tasks feel like a greater cognitive burden. In such cases, the missing data is directly related to the condition being studied.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eSession 3: Consolidating Learnings\u003c/h2\u003e \u003cdiv id=\"Sec27\" class=\"Section4\"\u003e \u003ch2\u003eContext matters\u003c/h2\u003e \u003cp\u003eWhen, how and why missing data occur may depend on the device, how the use of the device was communicated in the trial, the disease condition being studied, previous experience of the participants with the device, as well as environmental and cultural factors. The importance of the specific context in understanding missing data patterns was highlighted. There is potential for different reasons for missing data to occur together. For example, missing data due to practical challenges around operating devices, which may often be considered to occur at random, may disproportionately affect participants who are more unwell or more affected by their disease, pointing to the potential for informative missingness.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eTechnical difficulties\u003c/h2\u003e \u003cp\u003eRecurring technical issues reinforced the need for rigorous usability testing before a study begins (Bakker, Barge, Cobb, \u0026amp; Cota, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). MST reflected a need to use more neutral language when discussing missing data, for example by discussing \u0026ldquo;when data collection might stop\u0026rdquo;, rather than \u0026ldquo;when participants remove devices\u0026rdquo;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eDesign of Future PPIE to focus on Informative missingness\u003c/h2\u003e \u003cp\u003ePPIE contributors recognised the challenge of gathering information about informative missingness. Two key challenges and potential solutions were discussed: selection bias and stigma.\u003c/p\u003e \u003cp\u003eFirstly, in a PPIE context, the self-selected individuals who wish to take part in PPIE discussions are likely to be individuals who are interested in research and may be more likely to adhere to protocol. A potential solution is to target recruitment of PPIE contributors within a specific study, possibly using early readouts of digital devices, to identify individuals who have missing data.\u003c/p\u003e \u003cp\u003eA second challenge is that there is a stigma associated with not adhering to protocol. To encourage honesty, contributors recommended clearly explaining the value proposition, emphasising how this information has the potential to improve trial analysis and interpretation of the results. They also suggested using anonymous surveys or encouraging open dialogue with staff. Finally, withdrawal forms in trials could explicitly ask if the decision to leave the study is related to their health condition.\u003c/p\u003e \u003cp\u003eResults and recommendations are summarised in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePost-session Feedback\u003c/h3\u003e\n\u003cp\u003ePPIE contributors noted that they enjoyed being able to share their experiences, as well as learning about others\u0026rsquo; experiences in similar studies. Some individuals noted that the sessions helped them to imagine what happens to their data after the study and gave them a better understanding of trends in digital research.\u003c/p\u003e \u003cp\u003eThe full GRIPP2 short-form checklist is available in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e in the \u003cspan refid=\"Sec36\" class=\"InternalRef\"\u003eAppendix\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eReasons for missing data, reported by PPIE contributors, and recommendations for trial conduct, trial analyses, and future PPIE activities that were discussed\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eReason for Missing Data reported by PPIE Contributors, and example quotation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRecommendations for Trial Conduct and Analyses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRecommendations for future PPIE activities\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTechnical issues and with devices \u003c/p\u003e \u003cp\u003e\u003cem\u003e\u0026ldquo;The monitoring discs that I wear on my arm only last 15 days \u0026hellip; they slowly stop working. If you don\u0026rsquo;t notice the reminder on your phone, ..., or your device has stopped working, there\u0026rsquo;s a window whereby that data is not gathered.\u0026rdquo;\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026bull; Usability testing of digital devices before recruiting participants into a research study is crucial.\u003c/p\u003e \u003cp\u003e\u0026bull; Provide clear instructions which have been adapted for the digital outcome.\u003c/p\u003e \u003cp\u003e\u0026bull; Provide tech support throughout the trial.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGiven the potential for technological issues to be the underlying reason for missing data, keep language for PPIE discussions about missing data neutral, e.g. \u0026ldquo;when data collection might stop\u0026rdquo; rather than \u0026ldquo;when participants remove devices\u0026rdquo;.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePracticalities of operating/setting up devices, and practical issues in daily life, e.g. discomfort due to weather, showering, religious activities. \u003c/p\u003e \u003cp\u003e\u003cem\u003e\u0026ldquo;We have to offer prayers 5 times a day, in which we need to do ablution (clean our\u0026nbsp;face, hands, rinse our mouth\u0026nbsp;etc) and if the device is not waterproof, then you need to remove it to do this activity, and most of the time I forgot to wear it [afterwards].\"\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026bull; Encourage ongoing feedback from trial participants during the study to understand factors which can lead to missing data.\u003c/p\u003e \u003cp\u003e\u0026bull; Consider the specific context of the device and population when defining missing data, especially if it is via thresholds for wear time.\u003c/p\u003e \u003cp\u003e\u0026bull; Consider whether to collect information on variables related to weartime for inclusion in imputation models to strengthen the MAR assumption, such as season/weather or regular activities that affect wear time such as religious or exercise activities (e.g. swimming).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eConduct ongoing PPIE discussions during the study to understand what factors may impact missing data for the specific context.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAspects related to the outcome itself, such as increased disease severity leading to inability to engage with the study. These were felt to be important, but were not part of PPIE contributors\u0026rsquo; personal experiences.\u003c/p\u003e \u003cp\u003e\u003cem\u003e\u0026ldquo;When you\u0026rsquo;re depressed, things seem so much more burdensome than they do when you\u0026rsquo;re not feeling depressed. The condition which is being treated could be a barrier to participation [for self-reported outcomes in an app].\u0026rdquo;\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026bull; Collect reasons for missing digital data as part of the trial e.g. via a trial withdrawal form, through speaking to study staff or via feedback questionnaires or portals.\u003c/p\u003e \u003cp\u003e\u0026bull; Make value proposition clear to participants: explain why it is important that reasons for missing data are known, and how this can ultimately help research.\u003c/p\u003e \u003cp\u003e\u0026bull; Given that informative missingness is an important consideration, plan for and conduct sensitivity analyses to the MAR assumption.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026bull; Since PPIE contributors may typically be more likely to adhere to protocol, recruit individuals who may be more likely to have missing data. For example, early readouts from digital devices could identify individuals to contact.\u003c/p\u003e \u003cp\u003e\u0026bull; Provide different modalities for giving feedback on PPIE activities; e.g. via anonymous surveys as well as ensuring it forms part of group discussions.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e "},{"header":"Discussion","content":"\u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis for digital outcome measures with missing data\u003c/h2\u003e \u003cp\u003eIn clinical trials with digital outcome measures, it is essential to understand missing data mechanisms to identify appropriate statistical assumptions and align statistical analyses accordingly. A key recommendation from this activity is that Patient and Public Involvement and Engagement (PPIE) should be conducted for the purpose of understanding missing data and inform statistical considerations for digital trials.\u003c/p\u003e \u003cp\u003eThe PPIE activity provided insight into missing data patterns emerging from device-related operational factors. For instance, data loss occurs when CGM sensors require replacement during a study. Discussions also highlighted that weather and season may correlate with continuous weartime of smartwatches. Therefore, these factors may serve as important auxiliary variables in imputation models to strengthen the Missing at Random (MAR) assumption, particularly for physical activity outcomes, which are known to be seasonally variable (Garriga et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ridgers et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Rothman et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Previous imputation approaches for accelerometer data have linked trial data with weather data from the Met Office (Tackney et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), incorporating weather-related auxiliary variables that were predictive of both missingness and accelerometer outcomes.\u003c/p\u003e \u003cp\u003eA key learning point was the potential for religious activities to contribute to missing data, for example, during pilgrimages or because of the need to remove specific devices for ablution. Religious groups may be underrepresented in trials that use digital health technologies, making it important to understand potential barriers to their participation (Digital Medicine Society, n.d.). Although Islamic perspectives of digital health technology have been described (Ali et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Gamon, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Sa\u0026rsquo;aid et al., n.d.), the impact of religious practice on trial-specific device use remains an under-explored research gap. It is important to explore all possible reasons for removing devices, with a need to ensure that this information captures perspectives from all backgrounds, to ensure that important factors such as religion or religiosity are not missed. Such factors could then potentially be used as auxiliary variables in imputation models in the analysis of trials.\u003c/p\u003e \u003cp\u003eGiven that PPIE contributors expected the meetings to be understandable to \u0026ldquo;ordinary patients\u0026rdquo;, discussions were focused around when, how and why missing data may occur, and whether data might be missing in a completely random way, whether missingness was influenced by specific factors (e.g. season) or whether there may be informative missingness. There was no discussion of which statistical methods were appropriate to use for which set of assumptions; this was felt by the researchers to be beyond the reasonable expectations communicated.\u003c/p\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003eMissing data discussions as a springboard to improved conduct\u003c/h2\u003e \u003cp\u003ePPIE sessions highlighted areas where improved study conduct could directly reduce the quantity of missing. These include resolving technical malfunctions before study initiation and improving communication regarding time-sensitive interface inputs. Conducting these discussions during or after a run-in period may be particularly effective for identifying and resolving practical or technological problems at an early stage.\u003c/p\u003e \u003cp\u003eA key learning point was that discussions framed around statistical methodology in a PPIE context have the potential to improve study conduct. While researchers may need to steer the conversation back to statistical concepts, any feedback regarding conduct should be shared with relevant collaborators. Cross-disciplinary communication ensures that these insights maximize the overall value of the study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section3\"\u003e \u003ch2\u003eRecommendations for Future PPIE\u003c/h2\u003e \u003cp\u003eTo obtain further insights on informative missingness, two important challenges and some solutions were discussed: Purposeful recruitment and reducing stigma.\u003c/p\u003e \u003cp\u003eFirstly, as PPIE contributors who volunteer to participate in discussion are likely to be individuals who adhere to research protocol, recruiting individuals in a more purposeful way is needed. For example, if PPIE individuals are recruited within an ongoing study, early readouts from the digital devices to identify individuals who have missing data, who may be best placed to provide insight into reasons for missingness.\u003c/p\u003e \u003cp\u003eSecondly, PPIE contributors recognised that speaking about not adhering to protocol can have a stigma, and it may be difficult for participants to openly admit to this. They recommended making the value proposition of discussing these aspects very clear, and also recommended allowing these contributions to be provided in different formats, including discussions with research staff or via anonymous surveys.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eEngaging patient and public representatives in the discussion of missing data for digital outcome measures was shown to be a viable pathway to identifying potential reasons and patterns for missing data. The discussions brought to light the potential for operational aspects of the device to lead to data loss, as well as specific activities such as ablution before prayer to lead to device removal, which may be overlooked by researchers. These ongoing PPIE discussions can provide valuable recommendations for both statistical analyses and trial conduct. Further work is required to deepen the understanding of participant experiences on informative missingness.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCGM: Continuous Glucose Monitor\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;CIG-DO: Community Involvement Group for Digital Outcomes\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;CUH: Cambridge University Hospitals\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;ePROs: electronic Patient-Reported Outcomes\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;MAR: Missing at Random\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;PPIE: Patient and Public Involvement and Engagement\u0026nbsp;\u003c/p\u003e\n"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eHuman ethics and consent to participate\u003cbr\u003e\u003c/strong\u003eNot applicable. The Public and Patient Involvement and Engagement (PPIE) contributors involved in this work acted as members of the research team (Experts by Experience) and not as research participants. Therefore, formal consent to participate as a research subject, and ethical approval, were not required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication\u003c/strong\u003e\u0026nbsp;\u003cbr\u003e\u0026nbsp;Not applicable\u003cbr\u003e\u0026nbsp;\u003cbr\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003cbr\u003e\u0026nbsp;\u003c/strong\u003eNot applicable\u0026nbsp;\u003cbr\u003e\u0026nbsp;\u003cbr\u003e\u003cstrong\u003eCompeting Interests\u003cbr\u003e\u0026nbsp;\u003c/strong\u003eSSV is on the advisory board for PhaseV (unrelated to this work).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;MST, Advanced Fellow, NIHR305417, is funded by the National Institute of Health and Care Research for this research project. The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. MLZ is funded by the Edmond J. Safra Foundation as part of the Edmond J. Safra Accelerating Clinical Trials in Parkinson’s Disease (EJS ACT-PD) Initiative.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors Contributions\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;MST conceived the research idea, led the investigation, and led the writing of the manuscript. AS facilitated PPIE meetings and provided guidance on conducting PPIE activities and expertise on digital health technologies. MLZ facilitated one PPIE meeting, provided guidance on conducting PPIE activities and the structure of the manuscript, and provided expertise on trial conduct. SSV provided expertise in design, analysis and conduct of clinical trials. SM, JD and FY provided expertise by experience in using digital health technologies in clinical studies, as well as expertise in PPIE. All authors reviewed and edited the manuscript.\u0026nbsp;\u003cbr\u003e\u0026nbsp;\u003cbr\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u0026nbsp;\u003cbr\u003e\u0026nbsp;Authors would like to thank Sophie Greenwood, James Carpenter, Alison Yarnall, Ríona McArdle and Jack Lumsdon for helpful discussions. Authors would like to acknowledge and thank all ten individuals in the CIG-DO for their invaluable contribution as experts by experience in the design and participation in the PPIE activities:\u0026nbsp;\u003cbr\u003e\u0026nbsp;\u003cbr\u003e\u0026nbsp;Melissa Cox\u003cbr\u003e\u0026nbsp;Jeremy Dearling\u003cbr\u003e\u0026nbsp;Curie Freeborn\u0026nbsp;\u003cbr\u003e\u0026nbsp;Lorraine Hazlehurst\u003cbr\u003e\u0026nbsp;Ian Hudson\u003cbr\u003e\u0026nbsp;Carrol Lamouline\u0026nbsp;\u003cbr\u003e\u0026nbsp;Sarah Markham\u0026nbsp;\u003cbr\u003e\u0026nbsp;Yazan Mehyar\u0026nbsp;\u003cbr\u003e\u0026nbsp;Mike Willis\u0026nbsp;\u003cbr\u003e\u0026nbsp;Farheen Yameen\u003cbr\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbell L, Maher F, Begum S, Booth S, Broomfield J, Lee S, Smith E, Stannard R, Teece L, Vounzoulaki E, Worboys H, Gray LJ. Incorporation of patient and public involvement in statistical methodology research: a survey assessing current practices and attitudes of researchers. Res Involv Engagem. 2023;9(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40900-023-00507-5\u003c/span\u003e\u003cspan address=\"10.1186/s40900-023-00507-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAli SM, Saiyed MM, McAvoy A, Mackin R, Jay C, van der Veer SN. Determinants of mistrust in digital health research and approaches to address them among Muslim ethnic minorities living in the United Kingdom: a qualitative study. Int J Equity Health. 2025;24(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12939-025-02583-3\u003c/span\u003e\u003cspan address=\"10.1186/s12939-025-02583-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBakker JP, Barge R, Cobb B, Cota C. (2024). \u003cem\u003eV3+: An extension to the V3 framework to ensure user-centricity and scalability of sensor-based digital health technologies\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eicity-and-scalability-of-sensor-based-digital-health-technologies/\u003c/span\u003e\u003cspan address=\"http://icity-and-scalability-of-sensor-based-digital-health-technologies/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBakker JP, Barge R, Cobb B, Cota C, Guo CC, Hartog B, Horowicz-Mehler N, Izmailova ES, McClenahan S, Motola S, Patel S, Paun O, Schoone M, Sezgin E, Switzer T, Tandon A, van den Brink W, Vairavan S, Vandendriessche B, Goldsack JC. (2024). \u003cem\u003eV3+: An extension to the V3 framework to ensure user-centricity and scalability of sensor-based digital health technologies\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://datacc.dimesociety.org/resources/v3-an-extension-to-the-v3-framework-to-ensure-user-centricity-and-scalability-of-sensor-based-digital-health-technologies/\u003c/span\u003e\u003cspan address=\"https://datacc.dimesociety.org/resources/v3-an-extension-to-the-v3-framework-to-ensure-user-centricity-and-scalability-of-sensor-based-digital-health-technologies/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBooth S, Wells M, Nevill C, Teece L, Czyznikowska B, Grewal-Santini G, Mancini M, Yameen F, Freeman SC. Plugging the gap\u0026rsquo;: development of a plain language glossary for statistical methodology research. Res Involv Engagem. 2025;11(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40900-025-00782-4\u003c/span\u003e\u003cspan address=\"10.1186/s40900-025-00782-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBroomfield J, Hill M, Chandler F, Crowther MJ, Godfrey J, Guglieri M, Hastie J, Larkindale J, Mumby-Croft J, Reuben E, Woodcock F, Abrams KR. Developing a Natural History Model for Duchenne Muscular Dystrophy. PharmacoEconomics - Open. 2024;8(1):79\u0026ndash;89. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s41669-023-00450-x\u003c/span\u003e\u003cspan address=\"10.1007/s41669-023-00450-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarpenter JR, Smuk M. Missing data: A statistical framework for practice. Biom J. 2021;63(5):915\u0026ndash;47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/bimj.202000196\u003c/span\u003e\u003cspan address=\"10.1002/bimj.202000196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCro S, Kahan BC, Patel A, Henley A, Joanna J, Hellyer P, Kumar M, Rahman Y, Goul\u0026atilde;o B. Starting a conversation about estimands with public partners involved in clinical trials: a co-developed tool. Trials. 2023;24(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-023-07469-9\u003c/span\u003e\u003cspan address=\"10.1186/s13063-023-07469-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCro S, Van Vogt E, Totton N, Lee E, Hellyer CJ, Kumar P, Rahman M, Y., Henley A. Supporting public involvement in defining estimands: a practical tool accessibly explaining the five key attributes of an estimand. Trials. 2025;26(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-025-08941-4\u003c/span\u003e\u003cspan address=\"10.1186/s13063-025-08941-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrowe S, Adebajo A, Esmael H, Denegri S, Martin A, McAlister B, Moore B, Quinn M, Rennard U, Simpson J, Wray P, Yeeles P. (2020). \u0026lsquo;All hands-on deck\u0026rsquo;, working together to develop UK standards for public involvement in research. In \u003cem\u003eResearch Involvement and Engagement\u003c/em\u003e (Vol. 6, Number 1). BioMed Central Ltd. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40900-020-00229-y\u003c/span\u003e\u003cspan address=\"10.1186/s40900-020-00229-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi J, Demanuele C, Kettermann A, Karahanoglu FI, Cappelleri JC, Potter A, Bury D, Cedarbaum JM, Byrom B. (2022). Considerations to address missing data when deriving clinical trial endpoints from digital health technologies. \u003cem\u003eContemporary Clinical Trials\u003c/em\u003e, \u003cem\u003e113\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cct.2021.106661\u003c/span\u003e\u003cspan address=\"10.1016/j.cct.2021.106661\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDigital Medicine Society. (n.d.). The Framework for Inclusive Development. DATAcc by DiMe. Retrieved 30. December 2025, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://datacc.dimesociety.org/resources/the-framework-for-inclusive-development/\u003c/span\u003e\u003cspan address=\"https://datacc.dimesociety.org/resources/the-framework-for-inclusive-development/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEuropean Medicines Agency. ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials - Step 2b. Ema. 2017;44(August):1\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003c/span\u003e\u003cspan address=\"http://www.ema.europa.eu/contact\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGamon AMI. Ethics of Digital Health in Islamic Perspective. J Sci Technol. 2023;28(1):1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.20428/jst.v28i1.1993\u003c/span\u003e\u003cspan address=\"10.20428/jst.v28i1.1993\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarriga A, Sempere-Rubio N, Molina-Prados MJ, Faubel R. (2022). Impact of seasonality on physical activity: A systematic review. In \u003cem\u003eInternational Journal of Environmental Research and Public Health\u003c/em\u003e (Vol. 19, Number 1). MDPI. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijerph19010002\u003c/span\u003e\u003cspan address=\"10.3390/ijerph19010002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoulao B, Bruhn H, Campbell M, Ramsay C, Gillies K. Patient and public involvement in numerical aspects of trials (PoINT): exploring patient and public partners experiences and identifying stakeholder priorities. Trials. 2021;22(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-021-05451-x\u003c/span\u003e\u003cspan address=\"10.1186/s13063-021-05451-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenwood S. (2025). \u003cem\u003eWhat do we mean by \u0026lsquo;missing data\u0026rsquo; in clinical trials? A story of patient involvement\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.youtube.com/watch?v=DLBVXCru8cI\u003c/span\u003e\u003cspan address=\"https://www.youtube.com/watch?v=DLBVXCru8cI\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenwood S, Morris TP, O\u0026rsquo;Malley L, Aucott L, Fagbemi O, Caie R, Goulao B. (2026). Can we elicit patient knowledge about missing data in trials? A co-design study to inform a patient-informed sensitivity analysis of missing data in trials using expert elicitation.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenwood S, Morris TP, O\u0026rsquo;Malley L, Aucott L, Goulao B. (2026). \u003cem\u003eEliciting and incorporating patient\u0026rsquo;s opinions about missing data in randomised controlled trials\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://abdn.primo.exlibrisgroup.com/discovery/fulldisplay?context=L\u0026amp;vid=44ABE_INST:44ABE_VU1⟨=en\u0026amp;docid=alma9918660731505941\u003c/span\u003e\u003cspan address=\"https://abdn.primo.exlibrisgroup.com/discovery/fulldisplay?context=L\u0026amp;vid=44ABE_INST:44ABE_VU1⟨=en\u0026amp;docid=alma9918660731505941\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeogh A, Ardle RM, Diaconu MG, Ammour N, Arnera V, Balzani F, Brittain G, Buckley E, Buttery S, Cantu A, Corriol-Rohou S, Delgado-Ortiz L, Duysens J, Forman-Hardy T, Gur-Arieh T, Hamerlijnck D, Linnell J, Leocani L, McQuillan T, Consortium MD. Mobilizing Patient and Public Involvement in the Development of Real-World Digital Technology Solutions: Tutorial. J Med Internet Res. 2023;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2196/44206\u003c/span\u003e\u003cspan address=\"10.2196/44206\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKettermann A, Dandi G, Clark J, Kim Y, Song J, Pucino F, Almario N, E. Navigating the Future: Considerations for use of Continuous Glucose Monitoring in Diabetes Trials. Therapeutic Innov Regul Sci. 2025. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s43441-025-00880-1\u003c/span\u003e\u003cspan address=\"10.1007/s43441-025-00880-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Institute for Health and Care Research. (2024). \u003cem\u003ePayment Guidance for Researchers and Professionals\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003c/span\u003e\u003cspan address=\"http://www.nihr.ac.uk/about-us/who-we-are/policies-and-guidelines/payment-guidance-researchers-and-professionals\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Institute for Health Research. (2013). \u003cem\u003ePatient and Public Involvement in Health and Social Care Research: A handbook for researchers\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRidgers N, Salmon J, Timperio A. Too hot to move? Objectively assessed seasonal changes in Australian children\u0026rsquo;s physical activity. Int J Behav Nutr Phys Act. 2015;12:77.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRothman AMK, Middleton J, Zafar H, De Bie EMDD, Newman J, Varian F, Taylor J, Hitchcock F, Ashraf C, Patel J, Thompson AAR, Toshner M. (2025). Remote Actigraphy Is Seasonally Variable with Implications for Clinical Monitoring. \u003cem\u003eMedRxiv\u003c/em\u003e, 2025.05.22.25328110. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2025.05.22.25328110\u003c/span\u003e\u003cspan address=\"10.1101/2025.05.22.25328110\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSa\u0026rsquo;aid HB, Azman A, Ramli SI, Kasim NH, Maon SS, Hasan HA, Zakaria AB. (n.d.). \u003cem\u003eIslamic Bioethics in Health Technology Assessment: A Review of Key Ethical Frameworks\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.47772/IJRISS\u003c/span\u003e\u003cspan address=\"10.47772/IJRISS\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith A, Worboys H, Begum S, Bennett D, Broomfield J, Cro S, Evans-Hill L, Greenwood J, Henley A, Mancini M, Royle KL, Saul H, Sergeant J, Stewart D, Tyrer F, Wason J, Yau C, Gray LJ. (2025). Incorporation of Patient and Public Involvement in Statistical Methodology Research: Summary of Workshop Proceedings. In \u003cem\u003eStatistics in Medicine\u003c/em\u003e (Vol. 44, Numbers 15\u0026ndash;17). John Wiley and Sons Ltd. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/sim.70159\u003c/span\u003e\u003cspan address=\"10.1002/sim.70159\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTackney MS, Carpenter JR, Villar SS. Unleashing the full potential of digital outcome measures in clinical trials: eight questions that need attention. BMC Med. 2024;22(1):413. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12916-024-03590-x\u003c/span\u003e\u003cspan address=\"10.1186/s12916-024-03590-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTackney MS, Cook DG, Stahl D, Ismail K, Williamson E, Carpenter J. A framework for handling missing accelerometer outcome data in trials. Trials. 2021;22(1):1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-021-05284-8\u003c/span\u003e\u003cspan address=\"10.1186/s13063-021-05284-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTackney MS, Steele A, Newman J, Fritzsche M-C, Lucivero F, Khadjesari Z, Lynch J, Abbott RA, Barber VS, Carpenter JR, Copsey B, Davies EH, Dixon WG, Fox L, Gonz\u0026aacute;lez J, Griffiths J, Hinchliffe CHL, Kolanko MA, McGagh D, Villar SS. Digital endpoints in clinical trials: emerging themes from a multi-stakeholder Knowledge Exchange event. Trials. 2024;25(1):521. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-024-08356-7\u003c/span\u003e\u003cspan address=\"10.1186/s13063-024-08356-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWorboys HM, Broomfield J, Smith A, Stannard R, Tyrer F, Vounzoulaki E, Czyznikowska B, Grewal-Santini G, Greenwood J, Gray LJ. Incorporation of patient and public involvement in statistical methodology research: development of an animation. Res Involv Engagem. 2023;9(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40900-023-00513-7\u003c/span\u003e\u003cspan address=\"10.1186/s40900-023-00513-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-research-methodology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmrm","sideBox":"Learn more about [BMC Medical Research Methodology](http://bmcmedresmethodol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmrm/default.aspx","title":"BMC Medical Research Methodology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"missing data, patient and public involvement and engagement, digital outcome measures, trial conduct","lastPublishedDoi":"10.21203/rs.3.rs-9293900/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9293900/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eAnalysis of clinical trials with missing data requires statistical assumptions. When novel digital outcome measures are used, it is particularly important to understand trial participants\u0026rsquo; experiences with the device. This can help illuminate the reasons for and specific patterns of missing data, thereby informing plausible missing data assumptions.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA patient and public involvement and engagement (PPIE) activity bridged statistical concepts on missing data with lived experiences of individuals who had participated in studies with digital outcome measures. Ten PPIE contributors attended three meetings, which were aimed to (i) introduce key statistical concepts and set the scene, (ii) discuss experiences of using digital devices in research studies, and (iii) consolidate learning and reflect on implications. In addition, contributors provided input via pre- and post-meeting surveys.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eReasons for missing data in digital outcome measures were highly context dependent and varied according to the device, study population and environmental/cultural context. Identified reasons included operational aspects of the device, unresolved technology issues, and practicalities in daily life such as weather or season affecting comfort and the need to remove devices for religious, exercise or hygiene activities. Previous experience of using digital devices and receiving feedback from devices influenced levels of engagement. Although contributors did not report disengaging with devices in ways directly related to the outcome being measured, informative missingness was considered as important.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThe PPIE activity was a feasible and valuable approach to exploring patient and public perspectives on the reasons for and patterns of missing data in digital outcome measures. Lived experience may shed light on reasons for missing data that may be overlooked by researchers; for example, religious activities affecting device removal was a key learning point. Incorporating PPIE discussions within ongoing trials may help inform statistical analyses and improve trial conduct. For future PPIE activities, purposeful recruitment and providing different modes of engagement, such as anonymous surveys, are recommended.\u003c/p\u003e","manuscriptTitle":"Can discussions with patients and the public clarify missing data mechanisms for digital outcome measures?","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-22 08:26:07","doi":"10.21203/rs.3.rs-9293900/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-04-11T01:26:44+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-09T10:38:34+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-09T07:00:34+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-09T07:00:04+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Research Methodology","date":"2026-04-01T15:22:02+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-research-methodology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmrm","sideBox":"Learn more about [BMC Medical Research Methodology](http://bmcmedresmethodol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmrm/default.aspx","title":"BMC Medical Research Methodology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4adf7511-f16a-41a9-8811-70c2d688f769","owner":[],"postedDate":"April 22nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-22T08:26:08+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-22 08:26:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9293900","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9293900","identity":"rs-9293900","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}