数字技术越来越多地用于健康研究,以从更广泛的人群中收集现实世界的数据。新一轮的数字健康研究主要依靠数字技术来进行完全远程的研究。远程数字健康研究有望带来比传统、在人的研究。然而,据报道,这些研究通常会受到参与者流失的影响,其来源在很大程度上仍未得到充分研究。
为未来远程数字健康研究规划做出贡献,我们提出了研究登记和完成的概念框架和假设。该框架引入了影响远程数字健康研究结果的3个参与标准:(1)参与者的动机和激励或推动,(2)参与者任务复杂性,(3)科学要求。这项研究的目标是从以人为本的角度为远程数字健康研究的计划和实施提供信息。
我们进行了范围审查,以收集有关参与远程数字健康研究的信息,关注影响参与者注册和保留的方法论方面。在PubMed上进行了全面搜索,CINAHL,和WebofScience数据库,我们的研究还包括来自引文搜索的其他来源。我们包括完全远程进行的数字健康研究,包括招聘期间至少一项框架标准的信息,研究的入职或保留阶段,并包括研究登记或完成结果。进行定性分析以综合纳入研究的结果。
我们报告了37项纳入研究的定性结果,这些研究根据目标样本量计算揭示了已实现的中位参与者入学人数的高值,128%(IQR100%-234%),和研究完成中位数,48%(IQR35%-76%)。对于向具有外在动机的参与者提供激励或推动的研究,观察到研究完成的中位数增加(62%,IQR43%-78%)。在没有激励或推动的情况下,降低参与者的任务复杂性并不能提高研究入学率中位数(103%,IQR102%-370%)或完成(43%,IQR22%-60%)在观察性研究中,与提供更多激励或推动的介入研究相比(研究完成率中位数为55%,IQR38%-79%)。此外,在评估的远程数字健康研究中,完成指标存在不一致,其中只有大约一半的研究(14/27,52%)基于整个研究期间的参与者保留.
很少有研究以一致的方式报道参与因素和研究结果。这可能限制了我们研究的证据基础。由于观察到在数字健康研究中对具有数字素养技能的参与者的偏好,我们的评估也可能受到出版偏见或不具代表性的研究样本的影响。然而,我们发现,未来的远程数字健康研究计划可以受益于针对特定的参与者档案,提供激励和推动,并降低研究复杂性以改善研究结果。
Digital technologies are increasingly used in health research to collect real-world data from wider populations. A new wave of digital health studies relies primarily on digital technologies to conduct research entirely remotely. Remote digital health studies hold promise to significant cost and time advantages over traditional, in-person studies. However, such studies have been reported to typically suffer from participant attrition, the sources for which are still largely understudied.
To contribute to future remote digital health study planning, we present a conceptual framework and hypotheses for study enrollment and completion. The framework introduces 3 participation criteria that impact remote digital health study outcomes: (1) participant motivation profile and incentives or nudges, (2) participant task complexity, and (3) scientific requirements. The goal of this study is to inform the planning and implementation of remote digital health studies from a person-centered perspective.
We conducted a scoping review to collect information on participation in remote digital health studies, focusing on methodological aspects that impact participant enrollment and retention. Comprehensive searches were conducted on the PubMed, CINAHL, and Web of Science databases, and additional sources were included in our study from citation searching. We included digital health studies that were fully conducted remotely, included information on at least one of the framework criteria during recruitment, onboarding or retention phases of the studies, and included study enrollment or completion outcomes. Qualitative analyses were performed to synthesize the findings from the included studies.
We report qualitative findings from 37 included studies that reveal high values of achieved median participant enrollment based on target sample size calculations, 128% (IQR 100%-234%), and median study completion, 48% (IQR 35%-76%). Increased median study completion is observed for studies that provided incentives or nudges to extrinsically motivated participants (62%, IQR 43%-78%). Reducing task complexity for participants in the absence of incentives or nudges did not improve median study enrollment (103%, IQR 102%-370%) or completion (43%, IQR 22%-60%) in observational studies, in comparison to interventional studies that provided more incentives or nudges (median study completion rate of 55%, IQR 38%-79%). Furthermore, there were inconsistencies in measures of completion across the assessed remote digital health studies, where only around half of the studies with completion measures (14/27, 52%) were based on participant retention throughout the study period.
Few studies reported on participatory factors and study outcomes in a consistent manner, which may have limited the evidence base for our study. Our assessment may also have suffered from publication bias or unrepresentative study samples due to an observed preference for participants with digital literacy skills in digital health studies. Nevertheless, we find that future remote digital health study planning can benefit from targeting specific participant profiles, providing incentives and nudges, and reducing study complexity to improve study outcomes.