This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Digital innovations continue to shape health and health care. As technology socially integrates into daily living, the lives of health care consumers are transformed into a key source of health information, commonly referred to as patient-generated health data (PGHD). With chronic disease prevalence signaling the need for a refocus on primary prevention, electronic PGHD might be essential in strengthening proactive and person-centered health care.
This study aimed to review and synthesize the existing literature on the utilization and implications of electronic PGHD for primary disease prevention and health promotion purposes.
Guided by a well-accepted methodological framework for scoping studies, we screened MEDLINE, CINAHL, PsycINFO, Scopus, Web of Science, EMBASE, and IEEE Digital Library. We hand-searched 5 electronic journals and 4 gray literature sources, additionally conducted Web searches, reviewed relevant Web pages, manually screened reference lists, and consulted authors. Screening was based on predefined eligibility criteria. Data extraction and synthesis were guided by an adapted PGHD-flow framework. Beyond initial quantitative synthesis, we reported narratively, following an iterative thematic approach. Raw data were coded, thematically clustered, and mapped, allowing for the identification of patterns.
Of 183 eligible studies, targeting knowledge and self-awareness, behavior change, healthy environments, and remote monitoring, most literature (125/183, 68.3%) addressed weight reduction, either through physical activity or nutrition, applying a range of electronic tools from socially integrated to full medical devices. Participants generated their data actively (100/183, 54.6%), in combination with passive sensor-based trackers (63/183, 34.4%) or entirely passively (20/183, 10.9%). The proportions of active and passive data generation varied strongly across prevention areas. Most studies (172/183, 93.9%) combined electronic PGHD with reflective, process guiding, motivational and educational elements, highlighting the role of PGHD in multicomponent digital prevention approaches. Most of these interventions (110/183, 60.1%) were fully automatized, underlining broader trends toward low-resource and efficiency-driven care. Only a fraction (47/183, 25.6%) of studies provided indications on the impact of PGHD on prevention-relevant outcomes, suggesting overall positive trends, especially on vitals (eg, blood pressure) and body composition measures (eg, body mass index). In contrast, the impact of PGHD on health equity remained largely unexplored. Finally, our analysis identified a list of barriers and facilitators clustered around data collection and use, technical and design considerations, ethics, user characteristics, and intervention context and content, aiming to guide future PGHD research.
The large, heterogeneous volume of the PGHD literature underlines the topic’s emerging nature. Utilizing electronic PGHD to prevent diseases and promote health is a complex matter owing to mostly being integrated within automatized and multicomponent interventions. This underlines trends toward stronger digitalization and weaker provider involvement. A PGHD use that is sensitive to identified barriers, facilitators, consumer roles, and equity considerations is needed to ensure effectiveness.
The emergence of digital health innovations is expected to continue shaping the organization and delivery of health services [
A landmark whitepaper by the US Office of the National Coordinator for Health Information Technology defines PGHD as health-related information created by patients or their designees outside traditional health care contexts [
As the prevalence of chronic diseases continues to rise, many health care systems face unprecedented challenges that deem it necessary to refocus on prevention [
Our overarching objective targets the synthesis of the literature on the overall utilization of electronic PGHD for primary disease prevention and health promotion purposes. Specific objectives include (1) providing an overview of applied PGHD types and tools, as well as their aims, purposes, and contexts, (2) exploring health care consumer, provider, and technology responsibilities, as well as potential interactions among them, and (3) synthesizing broader implications of electronic PGHD on health outcomes and equity.
Our methodology was guided by Arksey and O’Malley’s framework for scoping studies and Levac, Colquhoun, and O’Brien’s conceptual extensions [
Our research question was formed by an iterative process by getting acquainted with the literature, identifying existing evidence gaps, as well as by regular exchange and expert consultation. Our question consisted of the 3 previously mentioned study objectives across the underlying dimensions of (1) data generation and collection, (2) sharing or communication, (3) interpretation, and (4) utilization. We narrowed the definition of electronic PGHD to data generated by consumer-facing means, excluding information that was collected through standardized, provider-driven methods, such as predefined questionnaires [
With the support of a specialized librarian and preliminary literature review, we developed an extensive and purposively sensitive search strategy, applied to 7 electronic databases that included MEDLINE, Cumulative Index to Nursing and Allied Health Literature, PsycInfo, Scopus, Web of Science, EMBASE, and Institute of Electrical and Electronics Engineers Digital Library. The searches were conducted on February 1, 2018. We additionally hand-searched 5 key electronic journals and 4 gray literature sources, complemented by Web searches, using the first 10 page results of 3 engines and thorough screenings of 6 relevant Web pages. Our last research steps consisted of (1) the manual reference list screening of all eligible studies and (2) author consultations, requesting input on potentially missed or unpublished work. A more detailed description of our study identification strategy is provided in previously published protocol [
The full search strategy and search terms are provided in
A total of 2 members of the research team (VN and PL) independently conducted a screening of the titles and abstracts, as well as full text screening against a set of predefined eligibility criteria (
Addresses electronic patient-generated health data (PGHD), as defined by this review, and additionally does the following:
Includes at least one sentence on the electronic PGHD tool or type.
Includes at least one sentence on how these are used or created.
Addresses PGHD that are available in a digital format at the point of utilization for intended health-related purposes, irrespective of the generation process.
Has a main focus on primary prevention and health promotion and falls within one the following domains:
Preventing initial occurrence of disease in healthy or high-risk individuals.
Mitigating risk in healthy or high-risk individuals.
Promoting existing health.
Describes, explores, and analyzes some form of health care consumer and provider involvement, where
Addresses an adult population.
Is a primary study published in English or German between January 1, 2003, and January 31, 2018.
Data extraction was conducted by 2 reviewers (VN and PL), guided by a predefined, flexible data extraction form, to capture the review’s objectives and corresponding research questions. The final form was refined and validated through consultation and expert feedback. Impact data were broadly extracted in terms of significance and direction. Equity data were extracted according to Cochrane Equity Group recommendations [
The whole process, including data charting (Step 4), was guided by an adapted PGHD flow framework, provided in
An external PGHD expert was consulted twice during the conceptualization stage who provided content-related feedback. A total of 3 stakeholders, a provider-partner (physician) and 2 consumer-partners, were consulted during the final manuscript preparation stages to ensure that our interpretations were relevant and understandable.
The deduplicated database search resulted in 8556 citations, which were screened by titles and abstracts. Full-text appraisal was deemed eligible for 305 studies, of which 199 did not fulfill our inclusion criteria. In total, the electronic database searches yielded 106 included studies. Interrater agreement reached 84% (42/50) (k=0.411) for a sample of 50 studies during title and abstract screening and 93% (14/15) (k=0.636) for a sample of 15 studies during full-text review. Complementary searches, including hand searching and searching gray literature sources, led to the inclusion of 30 studies, whereas reference list screenings and author consultations led to 47 additional studies, resulting in a total of 183 inclusions. A list of all excluded references at full-text screening, including justifications, is provided in
Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart. PGHD: patient-generated health data.
Most eligible studies were published as scientific journal articles (n=162), followed by sections of conference proceeding collections (n=13) and published theses (n=8). With an average of 22 studies per year, most were published from 2011 onward, whereas the number of publications averaged to around 6 studies a year between 2003 and 2010. More than half of the studies were conducted in North America (n=107), with 105 from the United States and 2 from Canada. European research followed with 38 studies, most of which were conducted in the United Kingdom. The remaining were conducted in Australia and New Zealand (n=18), Asia (n=13), and the Middle East (n=1). A total of 6 studies had an international scope. Randomized controlled trials constituted most of the applied methodologies (n=93), followed by quantitative nonrandomized approaches (n=47), mixed-method designs (n=30), and purely qualitative methodologies (n=13). The majority aimed to demonstrate effectiveness and efficacy (n=99), followed by mixed and purely exploratory aims (n=52), whereas less than a quarter explored feasibility and acceptability of interventions (n=32). The duration of identified studies ranged from a single examination to up to 2 years. More detailed information on study characteristics, including percentages, is provided in
The most commonly addressed prevention area was weight management, primarily in the form of physical activity and nutrition, which consists of 68.3% (125/183) of the identified literature. This is followed by 12.0% (22/183) of studies with a broader focus on health and well-being. These studies did not exclusively focus on one prevention area and included combinations of chronic and infectious disease, as well as mental health. About 7.7% (14/183) of the literature addressed cardiometabolic health, whereas 7.1% (13/183) focused on substance use prevention, targeting tobacco and alcohol. Healthy aging, such as prevention of falls, cognitive decline, and bone health, was the subject of 6 studies (6/183, 3.3%), followed by 2 studies on breastfeeding (2/183, 1.1%) and 1 study on skin cancer prevention (1/183, 0.5%).
We continued our analysis by synthesizing information on the aims of generating and sharing PGHD for primary disease prevention and health promotion purposes. We identified that enabling health consumers to generate their own health information aims at (1) promoting healthy behavior (142/183, 77.5%), (2) increasing health knowledge and self-awareness (120/183, 65.5%), (3) enabling healthy environments (60/183, 32.7%), and (4) enhancing remote monitoring (20/183, 10.9%). Most studies (134/183, 73.2%) targeted 2 or more of those aims. A similar pattern was observed within prevention areas, with health behavior change and knowledge gain being the most commonly addressed aims. This was not the case for substance use prevention, where enabling healthy environments outweighed knowledge gain. Not all studies adhered strictly to those aims, with 21.3% (39/183) deviating from purely preventive purposes and additionally using PGHD as outcome measures, for example, to quantify the effects of interventions and for secondary analyses.
Description and examples of patient-generated health data aims.
Aim | Description | Example from the literature |
Increase health knowledge and self-awareness | Increase in knowledge and cognition about one’s health, well-being, and behavior, with no particular focus on how to translate this knowledge into action and concrete behavior | Participants record dietary intake and receive weekly feedback with summaries on their fruit, vegetable, and junk food intake [ |
Promote healthy behavior | Help translate one’s knowledge into action, behavior change, and skill development, targeting health improvement and maintenance | Participants record dietary intake and receive nutritional feedback and additional individual dietary targets, recipes, and a meal plan for achieving those [ |
Enable healthy environments | Enable environments and contexts that facilitate health and well-being | Participants record physical activity in a digital partnership with family members or friends, creating an environment of healthy social pressure and support [ |
Enhance remote monitoring | Enable the remote monitoring of individual health and well-being parameters, by health care and wellness providers | Participants record blood pressure, blood glucose, weight, and body fat at home and sent data electronically to medical professional who monitors and provides personalized physical activity plans [ |
Successful prevention undoubtedly requires a clear definition of health care consumer responsibilities. Our analysis identified 3 broad consumer roles. The first consisted of passive PGHD generation (20/183, 10.9%), in which consumers used sensor-based devices to automatically collect and transmit information. Such an approach was predominantly applied in physical activity, weight loss, and overall health and well-being, capturing data that did not require manual entries, such as step counts, heart rate, and sleep quality. The more common second and third roles consisted of fully (100/183, 54.6%) or partially (63/183, 34.4%) active consumers, requiring occasional to regular actions. Active consumer involvement is key for capturing data that are not easily captured automatically, such as consumed meals and the quantity of smoked cigarettes. The term
Patient-generated health data–related consumer roles and examples.
Consumer roles | Examples from the literature | |
|
||
Fully active data generation | Take picture of meal and optionally add descriptions, visit website to add further contextual information [ |
|
Partially active data generation | Manually record stress levels and automatically capture data by wearing heart monitor [ |
|
Passive data generation | Carry mobile phone an physical activity monitor that generates PGHDa automatically [ |
|
|
||
Low-intensity data sharing | PGHD automatically stored on mobile phone based database and automatically transmitted in an encrypted manner [ |
|
High-intensity data sharing | Share data manually from monitors to website (directly or via a docking station) [ |
aPGHD: patient-generated health data.
Data generation is often followed by data sharing to third parties or across devices and storage locations. Information on data sharing was provided by about 73.2% (134/183) of the literature. We defined high-intensity sharing as any transmission of PGHD that requires concrete consumer action. High-intensity sharing was applied in 91 studies (91/183, 49.7%). Half of those (39/91, 43%) indicated more demanding actions requiring the active transfer of PGHD to external devices (eg, external computer) or storage locations (eg, server and website). In contrast, low-intensity sharing describes the automatic transmission of PGHD, which was applied in 43 (43/183, 23.5%) studies. We did not identify any difference of distribution between higher or lower sharing intensity across most prevention areas, except for cardiometabolic health and weight loss (
On the basis of the initial framework by Shapiro et al, we developed an extended and more comprehensive conceptual framework [
Our enriched framework shows that the 3 identified consumer roles are linked to different PGHD tools, ultimately creating different clusters of data. Although we identified interventions stopping at that level (single-component), the majority entailed additional intervention components (multicomponent), with and without human provider involvement. Thus, prevention and health promotion impact can be achieved at 3 levels, as the lower arrows indicate. Across the different elements, from the consumer to the provider, 5 common areas of barriers and facilitators can inhibit or promote the effective use of electronic PGHD. The framework additionally visualizes the link between identified PGHD aims and additional intervention components, as well as the involvement of health care providers. This framework fulfills the function of providing a simplified process overview, ultimately fostering a better understanding of PGHD utilization across prevention areas. All framework components are detailed throughout the results section.
Distribution overview of key themes across prevention areas.
Prevention areas | Weight control, physical activity, nutrition (n=125), n (%) | Overall health and well-being (n=21), n (%) | Cardio- metabolic health (n=14), n (%) | Substance use (smoking and alcohol; n=14), n (%) | Healthy aging (n=6), n (%) | Breastfeeding (n=2), n (%) | Skin cancer (n=1), n (%) | |
|
||||||||
|
Active data generation | 65 (52.0) | 11 (52) | 4 (29) | 14 (100) | 3 (50) | 2 (100) | 1 (100) |
|
Partially active data generation | 46 (36.8) | 7 (33) | 9 (64) | 0 (0) | 1 (17) | 0 (0) | 0 (0) |
|
Passive data generation | 14 (11.2) | 3 (15) | 1 (7) | 0 (0) | 2 (33) | 0 (0) | 0 (0) |
|
||||||||
|
High-intensity data sharing | 43 (34.4) | 9 (43) | 8 (58) | 4 (29) | 1 (17) | 0 (0) | 1 (100) |
|
Low-intensity data sharing | 51 (40.8) | 8 (38) | 3 (21) | 4 (29) | 2 (33) | 0 (0) | 0 (0) |
|
Unclear or not described | 31 (24.8) | 4 (19) | 3 (21) | 6 (42) | 3 (50) | 2 (100) | 0 (0) |
|
||||||||
|
Support and motivate PGHDb | 2 (1.6) | 3 (14) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
|
Review and analyze PGHD | 23 (18.4) | 1 (5) | 5 (36) | 3 (21) | 1 (17) | 1 (50) | 0 (0) |
|
Support and Motivate PGHD and review and analyze PGHD combined | 12 (9.6) | 0 (0) | 4 (29) | 1 (7) | 0 (0) | 0 (0) | 0 (0) |
|
Non-PGHD-related involvement | 14 (11.2) | 2 (10) | 1 (7) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
|
No involvement at all | 74 (59.2) | 15 (71) | 4 (28) | 10 (72) | 5 (83) | 1 (50) | 1 (100) |
|
||||||||
|
Nonhealth-related (eg, phone) | 104 (83.2) | 18 (86) | 12 (86) | 14 (100) | 5 (83) | 2 (100) | 1 (100) |
|
Health-related (eg, pedometer) | 65 (52.0) | 12 (87) | 10 (71) | 0 (0) | 2 (33) | 0 (0) | 0 (0) |
|
Medical (eg, glucometer) | 2 (1.6) | 3 (14) | 5 (36) | 5 (36) | 0 (0) | 0 (0) | 0 (0) |
|
||||||||
|
Reflective | 113 (90.4) | 16 (76) | 13 (93) | 11 (79) | 5 (83) | 2 (100) | 0 (0) |
|
Process guiding | 99 (79.2) | 9 (43) | 13 (93) | 12 (86) | 5 (83) | 1 (50) | 1 (100) |
|
Motivational | 88 (70.4) | 8 (38) | 7 (50) | 13 (93) | 4 (67) | 0 (0) | 1 (100) |
|
Educational | 84 (67.2) | 6 (29) | 10 (71) | 10 (71) | 2 (33) | 2 (100) | 0 (0) |
aConsumer roles are described in patient-generated health data–related consumer roles and examples table. Additional intervention components are defined in detail in descriptions of intervention components.
bPGHD: patient-generated health data.
cStudies were assigned to multiple hardware and additional intervention component categories, for which rows do not add up to 100%.
Enriched conceptual framework of electronic patient-generated health data (PGHD) flow and use for primary disease prevention and health promotion.
Prevention-targeted PGHD were generated through 3 broad types of hardware, often used in combination. The first included nonhealth-related products that are mostly well integrated into daily living (157/183, 85.8%), such as computers and mobile phones. The second entailed health-related devices that are less societally integrated (90/183, 49.2%), such as pedometers and heart rate monitors. The third included more specialized medical devices (15/183, 8.2%), such as glucometers and blood pressure monitors. Beyond manual and automatic PGHD collection, their most common functions included the provision of additional intervention components, such as goal setting and reminders, analysis and visualization of data, provision of feedback, sharing and storage of PGHD, and communication and interaction with third parties. Another key function was the provision of cues and visualizations, such as using color schemes, pictures, avatars, and other virtual elements to support the interpretation of PGHD.
The identified electronic PGHD were categorized in 4 broad types. Most studies (78/183, 42.6%) addressed textual or numerical data, requiring manual entry and an active consumer. This was followed by waves or signals (22/183, 12.0%) that did not require manual collection and audiovisual (video) (4/183, 2.2%) as well as photographic (2/183, 1.1%) PGHD, which again required an active user. Textual or numerical data were mostly used in weight control, substance use prevention, and healthy aging. Photographic data were applied in healthy eating, whereas audiovisual PGHD were commonly utilized in smoking and alcohol prevention. Waves and signals were primarily applied in weight control, well-being, cardiometabolic health, and healthy aging. Finally, almost half of the studies used 2 or more forms of digital PGHD (77/183, 42.1%), with the most common combinations being that of textual or numerical with waves or signals (58/77, 75%) and textual or numerical with photographic data (7/77, 9%). Textual or numerical with waves or signals was used across the spectrum of health domains from weight control to diabetes prevention and usually included initial sensor data that were then manually recorded by users. Textual or numerical with photographic data was applied in dietary interventions, where users took pictures of their meals and added descriptions.
In 172 (172/183, 93.9%) of the identified studies, PGHD were embedded in larger multicomponent preventive interventions. Our analysis identified 4 overarching components to which PGHD were combined with, categorized as (1) reflective, (2) process guiding, (3) motivational, and (4) educational. Their descriptions and examples are provided in
Reflective: All intervention components that are based on simple feedback of generated patient-generated health data (PGHD), with no additional educational information on how these are to be interpreted and applied. Examples include PGHD reports and summaries, as well as access to unstructured data.
Process guiding: All intervention components that aim to provide general support on the generation of PGHD, the use of technology, the compliance to intervention guidelines, and the response to problems arising from these processes. They include technical advice, instructions on when and how to collect and share, and problem-solving advice. Guidance in understanding and applying PGHD falls out this category’s scope (see point 4. Educational).
Motivational: All intervention components that are based on techniques that target the motivation of users to collect PGHD and apply those for healthy behavior changes. They include the provision of rewards and incentives, persuasion techniques, goal setting, reminders, motivational feedback, social support, as well as entertainment elements, such as gamification.
Educational: All intervention components that go beyond the simple feedback of generated PGHD (reflective), being attached to additional information that targets knowledge and skill enhancement, as well as knowledge testing. In contrast with process guiding, this category focuses on understanding and applying PGHD. They include the provision of newsletters, in-person counseling, remote coaching, educational podcasts, quizzes, and knowledge tests. Technical guidance on the generation and share of PGHD falls out of this category’s scope (see point 2. Process guiding).
The integration and utilization of electronic PGHD varied across additional intervention components. In combination with reflective components, PGHD were mostly used for self-referencing, such as the visualization of progress over time, enabling users to track individual health goals. In the context of motivational components, PGHD were repeatedly utilized to enable social comparison, such as contrasting data to normative or peer-generated values, often generating certain social pressure for healthier lifestyles. When integrated with educational components, PGHD were used to inform and provide individualized recommendations and counseling, aligned to the progress and capabilities of individual participants or participant subgroups. Finally, in combination with process guiding components, PGHD were key to identifying individual challenges, allowing for tailored and problem-focused support, while ensuring that adherence to intervention guidelines (eg, dietary or exercise plans) was monitored.
Less than half of the literature described the role of health care providers in the implementation (73/183, 39.9%) of interventions and only 30.6% (56/183) involved providers that had clearly designated PGHD-related responsibilities. The remaining proportion of the literature (110/183, 60.1%) addressed predominantly automatic programs. The proportion of studies without provider involvement was larger across all prevention areas, except for those on cardiometabolic health and breastfeeding promotion. Health care providers included an array of professionals, including physicians, nurses, dieticians, psychologists, health consultants, fitness experts, and trainers. Our thematic analysis identified 2 main clusters of PGHD-related provider roles. The first role (5/55, 9%) is that of a supporter, including the prompting, overseeing, and motivating of PGHD use, which was primarily found in weight control, nutrition and well-being interventions. The second role (34/55, 62%) is that of a reviewer, consisting of analyzing PGHD to inform counseling, personalize advise, conduct remote monitoring, and complement medical data, which was common in weight control, nutrition, cardiometabolic health, and substance use prevention. In 31% (17/55) of the studies, mainly in weight control, cardiometabolic health, and substance use prevention, providers held both roles simultaneously. In addition, we identified that provider-consumer interactions predominantly occurred remotely (36/73, 49%), either via the PGHD tool itself (eg, data collection website) or through other supporting channels (eg, email), both in a synchronous (eg, telephone) or asynchronous fashion (eg, forums). In-person interactions were less common (14/73, 19%) and more often combined with remote elements (17/73, 23%). One study (1/73, 1%) involved no direct interaction with consumers, whereas 5 (5/73, 7%) lacked clear interaction descriptions. Our findings additionally suggested that the involvement of health care providers was linked to the previously described PGHD aims, as one of those, namely the aim of enhanced remote monitoring, inevitably relies on data review by a provider.
Assessed prevention-relevant outcomes were broadly categorized into: (1) vitals and body composition measures (eg, body mass index, blood pressure, blood glucose, and heart rate), (2) behavioral change (eg, physical activity, eating habits, and lifestyle factors), and (3) knowledge change (eg, health literacy and awareness). About a quarter of the identified literature (47/183, 25.7%) provided indications on the potential impact of PGHD on preventive outcomes. These studies either had PGHD as a distinct or single component in one of their intervention arms (13/47, 28%), or as a part of multicomponent interventions (34/47, 72%), with sections that explored the associations between PGHD (eg, adherence to data collection) and outcomes. The majority explored implications on vitals and body composition-related outcomes (37/47, 79%). Most of those studies reported statistically significant beneficial trends (n=27), followed by nonsignificant effects (n=8) and mixed results (n=4). Outcomes in health behavior were less commonly addressed (15/47, 32%) and provided no clear tendencies, with an equal number of studies providing statistically significant beneficial (n=4) and nonsignificant (n=4) trends, as well as a relatively large proportion of unclear or mixed results (n=3). Health knowledge outcomes were the least commonly (2/47, 4%) addressed, with one study reporting nonsignificant associations between PGHD and health knowledge and one reporting mixed results. Most of these studies included active (27/47, 57%) and partially active consumers (8/47, 17%), whereas only one study entailed passive consumers (1/47, 2%). For the studies with active and partially active user engagement, a proportionally equal number of them reported statistically significant, mixed, and nonstatistically significant results. One study that included passive consumers did not provide enough information to be meaningfully compared.
A larger proportion of the literature (98/183, 53.6%) addressed interventions with multiple components and did not entail analyses on the relationship between PGHD components and prevention outcomes. Although their results could not be directly linked to PGHD, the overall picture suggested beneficial trends, with 23% (22/98) providing almost entirely positive results. Mixed results were indicated by 69% (68/98) of studies, almost all of which included at least one significantly positive outcome. A smaller proportion of interventions (8/96, 8%) did not identify beneficial effects at all. The remaining part of the literature (38/183, 20.8%) focused on feasibility and usability results instead, which is not reported in further detail here.
Considering equity as an important outcome for all health interventions, we extracted information linked to implications for subgroups that are commonly divided by health inequalities, as defined by the Cochrane Equity Group [
About 89.6% (164/183) of studies provided information on potential barriers and facilitators of electronic PGHD. Both barriers and facilitators were clustered around 5 recurring themes: (1) data collection and use, (2) technical and design considerations, (3) ethics, safety, and trust, (4) user characteristics, and (5) context and content. Data collection and use (127/164, 77.4%) addressed the levels of ease, difficulty, and burden of electronic PGHD generation, the adaptability of data collection to user needs, and associated resource demands (eg, time, costs). Technicalities and design (84/164, 51.2%) covered the functional maturity of PGHD technology, the facilitating role of mobile and interoperable devices, as well as the importance of dynamic, user-appealing, and simple designs. Ethics, safety, and trust (55/164, 33.5%) entailed barriers and facilitators around privacy, trustworthiness, credibility, and reliability. The category of user characteristics (72/164, 43.9%) highlighted consumer-related elements, such as digital literacy, knowledge, sociodemographic determinants, and overall attitudes toward PGHD technologies. Finally, the last category of content and context (148/164, 90.2%) included elements around contextual resources, such as PGHD support and interaction with providers. It additionally addressed the role of technology and intervention content, such as the combination of PGHD with other behavior change communication techniques.
Data collection and use (n=49):
Burdensome data collection
Inflexible data entry
Retrospective data entry: incentive to manipulate data
Unstructured data: information overload
Automatized recording: feeling of no control over data
Costly
Technicalities and design (n=39):
Immature or nonfunctional
Unappealing design
Nonuser-friendly functions
Ethics, safety, and trust (n=32):
Privacy and security concerns
Nontrustworthy patient-generated health data (PGHD) tools and data
Sociocultural resistances
Low-quality and unreliable PGHD
User characteristics (n=38):
Low digital literacy and no previous experience
Negative attitudes toward PGHD
Mismatch with daily life routines
Nonperceived usefulness
Sociodemographics (eg, young age, low education)
Content and context (n=44):
Missing data interpretation and general support
Missing in-person contact
Missing or too frequent reminders
Missing provider resources to evaluate PGHD
Missing (financial) incentives or rewards
Missing or insensitive feedback
Unrealistic goals
Nonengaging environment (eg, no social support)
Data collection and use (n=78):
Simple and low-effort data collection
Highly flexible data entry
Retrospective data entry: incentive to correct data
Time-efficient and intuitive data output
Automatized recording: noninterference with daily life
Free or low cost
Technicalities and design (n=45):
Technically mature and interoperable
User-engaging and appealing design
Dynamic design: interactive and modifiable
Mobile
Ethics, safety, and trust (n=23):
Credible patient-generated health data (PGHD) tools
Trustworthy, reliable, and complete PGHD
Processes that do not invade privacy
User characteristics (n=34):
Digital literacy and previous experience
Preexisting motivation and readiness to use PGHD
Self-efficacy
Perceived PGHD usefulness and relevance
Content and context (n=104):
Available guidance and support
Available human interaction
Sensitive reminders that do not disturb
Data visualizations and summaries
Motivating rewards and (financial) incentives
Immediate, sensitive, and motivating feedback
Realistic goal setting
Social support (eg, peer interactions)
Content and context personalization
Enabled data access, ownership, and control
Fun elements (eg, gamification)
Novel elements (eg, geofence triggered support)
Our review described a large and dynamically emerging volume of the literature on the use of electronic PGHD for primary disease prevention and health promotion purposes. Beyond quantity, the literature manifested large methodological and thematic heterogeneity, adding to the topic’s conceptual complexity. Our results enabled the development of an enriched conceptual framework (
The identified literature predominantly focused on weight control, through physical activity and nutrition, which was consistent with previous reviews that addressed digital health interventions across prevention areas [
Our thematic analysis identified certain recurring patterns of PGHD generation. We broadly classified consumer roles as passive, partially active, and fully active and identified that the proportions of these vary across prevention areas. Acknowledging that consumer roles are closely linked to PGHD types, we found that certain prevention areas are being dominated by 1 or 2 types of PGHD. On one hand, weight control, alcohol and smoking prevention, and overall health and well-being seemed to be mostly addressed by technologies that require the manual collection of textual or numerical data, while on the other hand, cardiometabolic disease prevention was primarily addressed by a combination of PGHD types that require a mix of active and passive data generation. In contrast, entirely passive data generation was only identified for weight control, overall health and well-being, cardiometabolic health, and healthy aging. Although not focused on prevention, a review by Vagesna et al [
Considering that the sophistication and reliability of PGHD technology varies across prevention areas, these patterns are expected. To be adequately targeted and well-informed, prevention often requires very specific consumer action and PGHD input. On one hand, for certain areas, such as addiction prevention or dietary intake, this input is entirely behavioral and not easily captured automatically. This includes the exact number of smoked cigarettes and consumed alcohol drinks, the type of consumed drinks, the percentage of alcohol content in each drink or the portions of consumed meals, all of which currently cannot be reliably or cost effectively collected by sensor-based devices, while on the other hand, sophisticated and highly commercialized fitness trackers are increasingly being improved to reliably capture certain activities and bodily functions, such as physical exercise and heart rate. Exercise-based weight loss, well-being promotion (eg, sleep quality), and healthy aging (eg, fall prevention) are prevention areas in which such devices can be applied to, which explains the prevalence of passive PGHD generation. In between the two extremes, there are prevention approaches that inherently require combinations of measures, such as in diabetes prevention (eg, dietary intake and physical activity), which in turn allow for a partially active and partially passive generation of PGHD.
Linking consumer roles to identified barriers and facilitators suggests some conflicting dynamics. Passive PGHD generation might be less burdensome, but may also lead to lower consumer engagement. Conversely, active generation involves more effort but may simultaneously trigger higher user motivation. In their review on wearable monitoring technology, Baig et al [
The relatively large proportion of studies that described automatic prevention systems, constituting 60.1% (110/183) of the identified literature, underlines a broader trend toward low-recourse and efficiency-driven care [
Most identified studies integrated electronic PGHD within multicomponent interventions, either complementing or facilitating other intervention components (eg, enabling self-reflection, facilitating social comparison, informing counseling, and directing guidance). A systematic review on the use of technology for weight reduction identified a similar trend, with 19 out of 27 studies combining PGHD with counseling feedback [
Summarizing the results of single-component interventions, exploratory analyses (eg, associations between PGHD and health outcomes), and overall effects of multicomponent interventions, the overall directions suggest a predominantly positive PGHD impact on prevention. These trends are expected, considering the existing evidence on the association between monitoring one’s own health and preventive outcomes [
Finally, the existence of conflicting barriers and facilitators highlighted the currently emerging nature and potential knowledge gaps of the topic. When do reminders become disturbances in one’s daily life and when are they the key to ensuring prevention adherence? Are automatic and simple data collection methods preferred by consumers because of being less burdensome, or do they counteract user engagement and motivation? Do financial rewards act purely as incentives to collect data and adhere to preventive guidelines, or could they become incentives for data falsification? Are PGHD tools that allow for retrospective data entries beneficial because of added flexibility, or do they add to the risk of data manipulation? Although these uncertainties may be indicators of an emerging topic that requires more research, they might also be the result of electronic health and prevention complexity. Neither digitalization nor prevention are static or fixed phenomena [
Despite our rigorous methodological approach, our study is subjected to some limitations. First, although PGHD may be key throughout the continuum of care, our review’s scope is restricted to primary disease prevention and health promotion. As such, our overall findings might not be applicable to the domains of treatment, disease management, and rehabilitation. This scope is narrower than defined in our protocol and has been chosen for practical and conceptual reasons. Retaining a broader scope would have led to an unmanageable volume of the literature and challenges in meaningfully synthesizing the results. Second, the chosen definition of electronic PGHD, which emphasized the aspect of patient’s control, led to the exclusion of standardized and more provider-driven approaches. Broadening the definition might provide a better understanding of PGHD-based prevention within health care contexts and the interaction of such data streams with health care provider infrastructures. Third, the variety and evolvement of definitions and terms to describe PGHD, as well as prevention, might have led to missing out a few terms and the associated literature. To compensate for that, we conducted thorough hand searches, reference list screenings, and author consultations. Finally, our scoping methodology and heterogeneous output did not allow for robust synthesis and comparison of effects.
The patterns we identified may support users, patients, and providers in understanding the complexity of utilizing electronic PGHD for prevention purposes. Beyond technical maturity, providers need to consider the wider implications that data collection might have on patients and consumers, such as its interference with daily living, personal beliefs, and digital literacy. Users and providers need to be sensitive to ethical and trust concerns, while ensuring that the PGHD environments are motivating and supportive enough to facilitate adherence and successful prevention.
Scoping reviews are often conducted to assess the feasibility of conducting a full systematic review [
Our review provides a comprehensive picture of the literature on electronic PGHD use for primary disease prevention and health promotion purposes, enabling a broader identification of processes and patterns. The high heterogeneity in the scope and content of identified studies underlines the topic’s emerging nature. This is reflected by the variety of identified PGHD-generating technologies, resulting in diverse data types and different consumer responsibilities. Utilizing electronic PGHD to prevent disease and promote health is a complex matter. In the literature, this complexity arises from electronic PGHD being mostly integrated into multicomponent and automatized interventions, limiting our ability to assess their individual preventive impact, and underlining an overall trend toward larger consumer responsibility. The broad set of identified barriers and facilitators, some being conflicting, highlights the need for a holistic understanding of such enabling factors, as well as for a stronger focus on ethical challenges, which is currently lacking.
Protocol deviations and justifications.
Search strategy and utilized keywords.
Adapted conceptual framework of electronic patient-generated health data flow.
List of studies excluded at full-text appraisal, with reasons.
List and characteristics of included studies.
Extracted study characteristics.
Participant characteristics.
List of included studies, grouped by prevention area.
Patient-generated health data tools and their functionalities.
patient-generated health data
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
The authors would like to thank Shannon Hubbs for acting as a third reviewer. The salary of VN was kindly provided by the Béatrice Ederer-Weber Foundation. The sponsor was not involved in any phase of this study. No further funding was received. The authors would like to thank Dr med Andreas Burkhard, Maria Giovanna Caruso, and Quentin Leonard Pannier for critically reviewing and providing feedback on the final manuscript. All authors provided submission approval for the final manuscript version.
VN contributed to the conceptualization and implementation of the study. VN wrote and edited the manuscript. MM and MAP contributed to the conceptualization of the study, supervised the entire process, and critically edited the manuscript. FE contributed to the design of the study, provided regular feedback, and critically edited the manuscript. PL contributed to the implementation of study and critically edited the manuscript. The main screening and review procedures were conducted by VN and PL.
None declared.