Patient portals (also known as tethered personal health records) consist of an internet-based application that accesses the electronic health record of a health care organization and provides timely access to medical records, laboratory results, appointment bookings, repeat prescriptions, and secure messaging with health care professionals, among other content and functionality [1]. Patient portals aim to engage patients and carers in managing their care, which has been found to improve health outcomes, the quality of care, and patient safety [2]. Patient portals are well established in UK family practice, with electronic health records being commonplace in 96% of general practices for almost 3 decades [3]. In UK acute care hospitals, the use of handwritten inpatient records remains widespread [3], and as such, patient portals are less common. Global Digital Exemplar (GDE) trusts are internationally recognized providers of exceptional and efficient National Health Service (NHS) care via world-class digital technology and information and are committed to sharing best practices and supporting the widespread adoption of patient portals [4]. The future vision of the NHS is to create a single access point to acute care hospitals with integrated systems that share and exchange data securely with other health and care providers [5]. However, the integration of portals with the existing systems is currently a barrier to their adoption, in addition to clinical engagement, information governance, low patient awareness, and resources [6]. Furthermore, among the patients who currently access portals, engagement or meaningful use is often limited.

The greatest benefit to patients and the health service can be achieved by optimizing portal use among older people [7]. Older people (aged ≥65 years) are less likely to use a patient portal than younger people (86% of adopters are aged <65 years) [8], yet they are the highest users of the health service, with more than half (54%) of them experiencing multimorbidity [9]. Older people are more at risk of serious complications and hospital-acquired infections, and they may experience frailty and other mobility problems that hinder their access to health centers. Moreover, older age is the greatest risk factor for mortality from COVID-19 [10]. Barriers are exacerbated when older patients lack access to and experience of using technology, have lower levels of education, and have low health literacy [11-13]. However, older people have been found to express interest in using a patient portal independently or with a carer, irrespective of their health literacy level, previous portal use, or experience seeking health information over the internet [11]. The COVID-19 pandemic has accelerated the rate of adoption of digital technologies in health care settings by necessitating remote visits, communication, and monitoring, which are especially important for people managing long-term health conditions [14]. The need and demand for more flexible access to health services are unlikely to diminish.

There are numerous reviews of patient portal features, functionality, adoption, and implementation, with the vast majority focusing on family practice settings. A review examining portal use in multiple health care settings was published by Antonio et al [14]. This umbrella review explored the current state of evidence for patient portals, with a specific focus on portal technology. It identified several factors that influence portal adoption, including patient circumstances, interest, and satisfaction; portal usability; provider attitudes; and service use [14]. Another review across multiple health care settings reported a range of patient characteristics that impact portal use, such as age, ethnicity, education, health literacy, health status, and carer role [7], and factors that impact patient portal engagement, such as provider endorsement and portal usability. The authors argue that future research should aim to boost portal engagement among specific populations most likely to benefit from its use [7]. This review set out to scope the literature on older people and acute care settings to inform the broader aim of our research program—to develop an evidence-based implementation strategy for portal use and engagement among older people in acute care settings. Despite efforts toward the widespread uptake of and engagement with patient portals across acute NHS trusts, there is limited research into evidence-based strategies for successful engagement [15]. Engagement strategies, such as advertisement campaigns or training for patients, carers, and staff, must be tailored to the targeted population groups and the local context to be effective [16]. Tailoring can improve equity within the patient group. Identifying contextual factors that impact the routine implementation of patient portals in acute care hospitals is the first step toward developing an evidence-based implementation strategy for older people.

The aim of this scoping review was to identify and synthesize the literature on contextual factors that impact the implementation of patient portals in acute care hospitals and among older people.

Our primary interest was to improve the engagement with patient portals among older people in acute care hospitals. However, we needed a broad scope of the literature to capture learning from studies in multiple settings (family practice and acute care hospitals) and patient groups (general population and older people), given the potential generalizability of the findings across settings.

This scoping review followed the methodological framework described by Arksey and O’Malley [17] and the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) reporting guidelines [18]. The framework was selected to achieve our broad aim of summarizing what is known about our primary area of interest, to synthesize findings, and to highlight key gaps in the literature.

To build on the existing literature, we used a theoretical framework to provide a semantic structure to the synthesis of our findings. The NASSS (Non-adoption, Abandonment, Scale-up, Spread, and Sustainability) framework [19] provided the basis for summarizing the results. This framework was developed to analyze the varied outcomes of technological innovations in health and social care and to help inform the implementation of such technologies. The NASSS framework comprises 7 domains: (1) the condition or illness, (2) the technology, (3) the value proposition, (4) the adopter system (intended users), (5) the organization, (6) the wider system (especially regulatory, legal, and policy issues), and (7) a final cross-cutting domain that considers adaptation and embedding over time. Each of the 7 domains can be “simple” (ie, few components and predictable), “complicated” (ie, many components but still largely predictable), or “complex” (ie, many components interacting in a dynamic and unpredictable way). Crucially, NASSS surfaces factors that are often unacknowledged in technology implementation programs, helping to move beyond the identification of individual barriers and enablers in recognition of the dynamic interactions between the domains, for example, the relationship between the individual adopter and the organizational or wider system context.

A search strategy was developed in collaboration with an academic librarian. As a preliminary examination of the literature indicated that only a few reviews focused solely on acute care hospitals or older adults and because the specific health care setting was not always immediately clear, we decided not to include filters for population or health care setting in order not to exclude potentially relevant publications.

The full search strategy is shown in Textbox 1. Search terms related to patient portals and systematic reviews were used. The search strategy used a combination of medical subject headings and free-text words. Searches were restricted to 2010 to account for the pace of development in portal technology. Searches were conducted on June 16, 2020, and included the following databases: MEDLINE and Embase via the Ovid platform, CINAHL and PsycINFO via the EBSCO platform, and the Cochrane Library. Reference lists of the included reviews were screened for additional literature.

To generate sufficient breadth of coverage for the scoping review, inclusion criteria were defined to capture maximum learning with respect to the implementation of tethered patient portals among older people and in acute care hospitals. Specifically, 3 categories of reviews were eligible for consideration:

The inclusion criteria were as follows: reviews published since 2010 in English; reviews focused on the implementation of tethered patient portals (as defined in the Introduction section); reviews focused on patient portals for older adults (in settings that include acute care hospitals); reviews focused on patient portals for patients, health care professionals, managers, and budget holders in acute care hospitals; reviews of reviews of patient portals in settings that include acute care hospitals; reviews focused on contextual factors (ie, barriers and facilitators) that impact the implementation of patient portals; systematic reviews; scoping reviews; narrative reviews; qualitative meta-syntheses; meta-ethnographies; and reviews of reviews.

The exclusion criteria were as follows: reviews published before 2010; reviews not in English; reviews not focused on the implementation of tethered patient portals; reviews focused on the technical aspects of patient portals; reviews in family practice settings only; nonsystematic reviews; secondary analyses of the existing data sets; discussions of literature for theory building or critique; summaries of literature for information or commentary; editors’ discussions; letters; conference abstracts; and theses; and reviews whose full text was not available.

Search results were imported into EndNote reference management software (Clarivate Analytics), and duplicates were removed automatically and double checked manually. Two reviewers (JH and ZK) independently screened titles and abstracts. Any discrepancies in the articles identified for full-text screening were discussed, and consensus was reached. Full-text articles of potentially eligible reviews were assessed independently by 2 reviewers (JH and ZK) against the prespecified inclusion and exclusion criteria. Discrepancies were resolved through discussion. The reasons for exclusion were recorded and included in the PRISMA diagram (Figure 1).

In accordance with the Arksey and O’Malley framework for scoping reviews [17], a data charting form was compiled in Microsoft Excel (Microsoft Corp), which contained a row for each included study and columns to record general study information, namely authors, publication date, country of origin, review aim, health care setting, participants, definition of patient portal, theoretical framework, database searches, inclusion and exclusion criteria, data extraction method, quality assessment method, method of analysis or synthesis, and the number of included studies. The included studies were examined to determine the extent of study overlap between the included reviews. The data charting form was also used to extract data on the contextual factors (barriers and facilitators) that impacted the implementation of patient portals in accordance with the 7 domains of the NASSS framework (refer to the NASSS Framework section), together with the authors’ recommendations for future research. The form was piloted on 4 studies. A single reviewer (JH) read each study and extracted the study characteristics and data on contextual factors from the results and discussion sections. The discussion sections were included in the charting process, as they often provide additional material to enhance the interpretation of review findings. The data extraction for each of the 4 articles was cross-checked by another team member (TJB, ZK, JL, and FS) to verify whether data charting was performed in accordance with the 7 NASSS framework domains. Data charting was then completed by a single reviewer (JH) and independently verified by another member of the team (ZK).

A narrative approach was used to summarize the results of the scoping review. In line with the usual practice for scoping reviews [17], no attempt was made to assess the quality of the included reviews or the weight of the evidence with respect to the implementation of patient portals.

A total of 2065 references were identified (Figure 1). After the removal of duplicates (n=606, 29.35%), another 1340 (64.9%) records were excluded based on the title and abstract, leaving 119 (5.76%) full-text articles to be assessed for eligibility. Of the 119 full papers assessed, 10 (8.4%) met the criteria for inclusion in this scoping review. Scrutiny of the reference lists of the included reviews did not generate additional literature for inclusion. Multimedia Appendix 1 provides details on the characteristics of the included reviews [15,20-28].

The reviews were published between 2015 and 2020. Of the 10 included reviews, 4 (40%) originated from the United States [15,20-22], 2 (20%) each from the United Kingdom [23,24] and the Netherlands [25,26], and 1 (10%) each from Australia [27] and Iran [28]. Overall, 3 (30%) systematic reviews addressed patient portals in acute care hospitals [15,20,23], 5 (50%) systematic reviews of reviews examined patient portals in both acute care hospitals and other care settings [24-28], and 2 (20%) systematic reviews addressed the implementation of patient portals among older adults (in multiple settings, including acute care hospitals) [21,22].

Of the included reviews, 7 (70%) did not specify a particular framework for the analysis of the results [20-24,27,28], 1 (10%) used the System Engineering Initiative for Patient Safety model to categorize interventions [15], 1 (10%) used the Clinical Adoption Framework [25], and 1 (10%) applied the Problem Solving Cycle [26]. Reviews in which no framework was specified considered portal design, use, and usability [20]; input, process, and output factors [27]; content and capabilities [28]; impact on outcome measures [24]; barriers to and facilitators of adoption and user experience [21]; characteristics of older users, evaluation of outcome measures and results, and barriers to and facilitators of use [22]; and impact on trust and communication and consideration of ethical issues [23].

To establish the breadth of coverage of this scoping review, an indication of the extent of overlap of studies in the included reviews was determined. The included reviews contained 206 studies (156 primary studies, 75.8%, and 50 reviews, 24.3%), excluding the 109 references of Otte-Trojel et al [26], which, unfortunately, we were not able to obtain for scrutiny. Table 1 provides a summary of the extent of study overlap. Only 19 studies appeared in >1 included review, suggesting limited duplication in the scoping review.

Table 2 shows that the contextual factors influencing implementation identified in the included reviews tended to cluster in specific domains: (1) the condition, (2) the technology, and (3) the value proposition. Certain aspects within these domains received more coverage than others, such as sociocultural factors and comorbidities, the usability and functionality aspects of the technology, and the demand-side value. The included reviews that used a theoretical framework [15,26,28] pointed to a focus on a narrow range of components of patient portal adoption, usually people and use factors.

There are links among the different domains. For example, the severity of an individual’s illness can affect their ability to interact with portal technology, raising questions about expectations of the involvement of lay caregivers (domain 4: the adopter system), organizational policies on proxy access (domain 5: the organization), privacy and security features (domain 2: the technology), and information governance (domain 6: the wider context). Similarly, organizational leadership and support (domain 5: the organization) for the development and implementation of portals can ensure inclusive and iterative portal design (domain 1: the technology), addressing not only usability and functionality issues but also the perceived value (domain 3: the value proposition) of the technology, as well as concerns about the impact of portals on health professional roles and identities (domain 4: the adopter system) and workload and workflow (domain 5: the organization).

There are gaps in the literature pertinent to the consideration of the provision of patient portals among older people in acute care hospitals, including the lack of consideration of the diversity of older adults and their needs, the question of interoperability between systems (likely to be important where care involves multiple services), the involvement of lay caregivers and looking beyond short-term implementation to ways in which portal use can be sustained.

This scoping review provides an overview of the contextual factors that impact the implementation of patient portals through an exploration of the emerging literature on patient portal use and engagement in acute care hospitals and among older people. Patients with chronic disease or greater disease severity were found to engage more with portals; however, comorbidities associated with the aging process were identified as barriers to portal use (domain 1: the condition). Perceived benefits from the supply side supported the adoption of portals, such as the potential for lower hospital costs, as did benefits from the demand side, such as engagement in care or self-management (domain 3: value proposition). Training for patients and staff should address technology literacy, the use of portal features, capacity-related concerns (integration of portals into workflows), and perceived value among health care professionals (domain 2: the technology). Older patients may lack help or support to access a portal; however, expectations of the involvement of lay caregivers in the adoption of patient portals were not reported (domain 4: the adopter system). Organizational leadership facilitates portal adoption, such as working with developers on portal design, developing policies for user training, and integrating portals into clinical workflows (domain 5: the organization). The development of national and international information exchange (ie, interoperability) and other standards (eg, security) was as a facilitator of portal implementation within the wider context (domain 6: the wider context). The reviews did not report on the long-term adaptability or sustainability of patient portals or organizational resilience. There were concerns that most portals are implemented within a single organization and that implementation across organizational contexts or in fragmented care delivery contexts would be challenging. This is important for the care of older people (domain 7: embedding and sustaining).

The diversity of older adults and their patient portal needs are not well understood. Older patients are more likely to experience chronic disease or greater severity of disease, and patients with chronic illness and greater severity of disease were found to engage more with portals owing to the perceived benefits of self-management, empowerment, and enhanced patient care. However, comorbidities related to age, such as vision and hearing loss, decreased dexterity and mobility, and declining cognitive function impede portal use. Variability in portal use and engagement among older people will, in part, be linked to the reasons for variability in internet use. Low income is the largest impediment to internet use among older people, followed by being aged >75 years, living alone, mobility, and memory or concentration problems [29]. This scoping review found that lower socioeconomic status, increasing age, male sex, and non-White ethnicity were factors associated with low adoption. People of lower socioeconomic status, older people, and people with mobility and memory or concentration problems are regular users of acute care services, making it an ideal setting to address these inequalities in patient portal access and engagement. Training programs and other engagement activities must directly target these inequalities to prevent any unintended exacerbation of the gray digital divide caused by the introduction or widespread use of a patient portal.

An important gap in the literature identified by this review was the lack of consideration of the involvement of the wider care network or lay caregivers in the adoption of patient portals. Many impediments to internet use among older people are linked to the increased likelihood of receiving care, for example, older age, mobility, and memory or concentration problems [29]. Older people are found to value proxy access to patient portals [30], with motivators including help to manage care, in the event of an emergency and lack of technology experience [31]. However, older patients express concerns when portals contain access to stigmatized conditions and financial commitments [32]. A review of 20 US health systems found that half of them had proxy access functionality, although only a few allowed the patient to specify role-based privileges [33]. The provision of separate proxy access should be accompanied by the provision of more control for patients over the information they wish to share [34]. This review found that organizational policies such as universal access; security protocols, including proxy access; adherence audits; and data availability and timing will facilitate portal development and implementation. To further aid the organizational readiness for technology-supported change, wider contextual factors must be considered at the planning stage in the form of policy shifts and patient developer specifications regarding the facility for internal and external information exchange. There is a need to engage with new ways of managing and talking about people’s data, which may require a different skill set, that is, different stakeholders round the table.

None of the included studies addressed the procurement process for patient portals. In the United Kingdom, there are a handful of providers that offer patient portals to acute care hospitals, such as Epic and Cerner. Patient portals differ in the extent to which they provide an off-the-shelf product or a tailored product with features that can be switched on or off depending on organizational readiness and capacity to facilitate them, such as patient-clinician communication. This review found that decentralized architectures were more compatible with local needs and implemented more quickly but were associated with higher costs [28]. Furthermore, there is no information on the level of support provided by the technology suppliers for the use of their products. This is anticipated to be a major organizational level barrier to implementation, which needs to be promptly addressed to facilitate the scale-up of portal use in acute care hospitals across the United Kingdom. Portals that can integrate into preexisting systems or offer data sharing and exchange were identified as facilitators [20]. Interoperability of health and care systems and other community services, such as the police and social services, is placed high on the NHS agenda [5], but although organizations may desire data sharing among themselves, the loss of control over shared data may serve as a barrier to portal adoption and highlights the complexity of this approach.

This scoping review used a comprehensive set of search terms to identify literature from electronic databases and followed robust procedures for citation and full-text screening in duplicate. Study overlap is a recognized limitation of reviews of reviews, where the primary studies may be reported in >1 systematic review and hence findings are overemphasized. This review included 156 primary studies and 50 reviews. We found that 19 (9.2%) of these 206 studies appeared in >1 review, indicating that study overlap was minimal, although we did not track overlap in the reference lists of all the included reviews. The NASSS framework provides a semantic structure by which to explore multilevel contextual factors impacting the implementation of digital health interventions. NASSS has largely been used to predict and evaluate implementation programs, but more recently, the framework has been used to synthesize review findings [35,36].

The number of reviews that focused exclusively on acute care hospitals (3/10, 30%) and older people (2/10, 20%) was low, which led to a broader scope of the extensive literature, primarily conducted in family practice and other patient groups, to capture learning and potential generalizability of the findings across settings and patients. In broadening the scope of the review, there were similarities with the umbrella review conducted by Antonio et al [14], who used a similar search strategy and a knowledge translation tool to present their findings. Our review was designed and our searches were conducted before the publication of the review conducted by Antonio et al [14]. We believe these similarities reinforce our robust approach to reviewing and synthesizing the literature, particularly as our design aimed to scope and map the literature rather than to systematically review it. The key differences between the reviews include our focus on older people and acute care hospitals; our review design aimed to scope rather than systematically review and appraise the literature; and our application of the NASSS framework. The NASSS has been referenced in >70 JMIR published studies since its publication in 2017, enabling easy comparison with the wider literature.

Papers were selected in accordance with our definition of a patient portal, and we were guided by the authors’ description of a patient portal. Multimedia Appendix 1 shows the definition of patient portals in each of the included studies. The use of the NASSS framework is concerned with the complexity of the use of portals; therefore, all data were considered according to the framework to produce a “big picture” aggregation of what is known about the implementation of patient portals. The included reviews were limited to those published in English; however, we did not exclude reviews that included non-English studies.

There is substantial literature on the contextual factors impacting patient portal use, with approximately 200 studies identified by the included reviews. However, there are few studies that evaluated interventions to address disparities in the use of patient portals [15]. As highlighted in a systematic review on the implementation of complex interventions in family practice [37], implementation studies exploring contextual factors tend to focus on surveys and qualitative studies, which are valuable in providing individual stakeholder perspectives but need triangulation with other methods. Observation and document analysis should accompany interviews to capture a more complete picture of the contextual factors at play, in particular, the wider context. As with any study exploring or evaluating the determinants of implementing a complex intervention, the features and functionality of the portal should be described in detail using established guidance [38] to enable reflection on the transferability of the findings to other settings. Finally, where interviews are used to explore the determinants of implementation among patients and staff, portal use data could be used to prompt further examination of use and sustained use.

This review provides implications for portal adoption and implementation that can inform current practice. This review found that cost, interoperability, trialability, and adaptability were all facilitators of portal adoption. Among hospitals in England deciding which portal product to adopt, GDE trusts play an important role in sharing detailed journeys through a digital technology via GDE blueprints, which are intended to promote scale-up, spread, and sustainability. To maximize the impact of GDE blueprints, they must be reported in an honest and transparent manner, with details on the challenges as well as the benefits of portals’ adoption, engagement, and sustained use. Descriptions of portal implementation must clearly delineate implementation strategies, such as detailed information on training for staff and patients (ie, content, frequency, and format) and communication strategies for the technical, interpersonal, and workflow aspects of patient portals. NHS Digital has created a Personal Health Records adoption toolkit, which offers generic support to organizations looking to implement a patient portal [39]. Furthermore, GDE trusts act as buddy sites to support other trusts, known as “Fast Followers,” for example, by sharing software, IT teams, standard processes, and could possibly assist with the selection and implementation of a patient portal; this approach is a powerful knowledge mobilization strategy enabling successful models to be scaled-up across the NHS [4].

This scoping review found that contextual factors influencing patient portal implementation tended to cluster in specific domains: (1) the condition, (2) the technology, and (3) the value proposition. Certain aspects within these domains received more coverage than others, such as sociocultural factors and comorbidities, the usability and functionality aspects of the technology, and the demand-side value. There are gaps in the literature pertinent to the consideration of the provision of patient portals for older people in acute care hospitals, including the lack of consideration of the diversity of older adults and their needs, the question of interoperability between systems (likely to be important where care involves multiple services), the involvement of lay caregivers, and looking beyond short-term implementation to ways in which portal use can be sustained. There is substantial literature on the contextual factors impacting patient portal use. Future research should focus on evaluating strategies that address disparities in use and promote engagement with patient portals among older people in acute care settings.