As little is known about the use of patient portals for the delivery of palliative and end-of-life care, a scoping review of the academic literature was conducted to identify sources describing currently available patient portal PCEOL tools. Our methods were aligned with methodological framework for scoping reviews by Arksey and O’Malley [26], which includes the following stages: identifying the research questions; identifying relevant studies; study selection; charting the data; and collating, summarizing, and reporting the results. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) extension for Scoping Review checklist guided our reporting (Multimedia Appendix 1) [27]. Our scoping review goals and procedures were registered with the Open Science Framework on March 26, 2020 (DOI 10.17605/OSF.IO/N34JZ).

To identify PCEOL patient portal tools described in the academic literature, we searched three databases (Ovid MEDLINE, CINAHL, and Web of Science) for peer-reviewed sources, including both qualitative and quantitative studies. The search terms included patient portal in combination with PCEOL terms such as palliative care, hospice, end-of-life, terminal illness, cancer, advance directives, symptom management, ACP, grief, and caregiver support. Our specific search strategy with exclusion terms is presented in Multimedia Appendix 2. The PCEOL search terms have previously been used to identify digital health solutions directed toward palliative care [28,29]. No time limits restricted the search, and only sources in English were included. We used Covidence, a web-based systematic review management program to remove duplicates, track citations, and screen references [30]. Our search was conducted in March 2020.

Sources were included in the scoping review if they (1) described a developed feature, resource, tool, or intervention that focused on at least one domain of PCEOL as defined by the National Coalition for Hospice and Palliative Care [31]; (2) targeted adults with serious illness (a health condition with a high risk of mortality impacting daily functioning or quality of life) [32] or their family and caregivers; and (3) were offered via a patient portal. The domains of palliative care include the following:

To determine whether sources met the inclusion criteria, 2 coders (MPI and CV) independently reviewed the academic sources by title and abstract.

Full texts of sources included by title and abstract were downloaded and further assessed for eligibility by 2 coders (MPI and CV). We determined whether the portal tool was targeted toward patients, caregivers, or both by reviewing the full text of the included sources for explicit mention of the intended users of the portal tool. The full text was also reviewed for a thorough description of the tool’s purpose and available features, apps, and content. Sources that presented a protocol or planned portal feature, did not provide a detailed description of the resources or content (eg, abstract and review articles), did not focus on a PCEOL domain, targeted a pediatric or non–seriously ill population, examined the user characteristics and perspectives of general portal use, were not published in English, described a tool that was not tethered to a patient portal (eg, free-standing website or mobile app), or for which the full text was not retrievable, were excluded from the review. Pediatric populations were excluded because of their different care needs, access to technology, and technological skills.

The coders regularly reviewed the sources and discussed questions and possible disputes. Both coders agreed on the features included in the review, with a third reviewer (JDP) reconciling disputes. A summary of the search and screening processes is shown in Figure 1.

The full text of each of the included sources was reviewed, and the elements and features of the described PCEOL patient portal tool were abstracted by 2 reviewers (MPI and CV) into a Microsoft Excel spreadsheet. To ensure consistency, the reviewers double-extracted five articles before dividing the remaining articles. The abstracted data included the following components: portal name, health system user, developer (research team or organization), target illness and/or palliative care element, intended audience (patient, patient’s caregiver, or both), and a summary of the features. When described, patient and caregiver perspectives about the tool, including measurements and qualitative reports of acceptability or usability and user satisfaction, were abstracted.

We used an iterative and mixed inductive-deductive approach to summarize the charted information and results from the included papers to address our aims [26]. The charted information from each paper was reviewed extensively to identify similarities and differences in patient portal features, PCEOL domains, study population or targeted audience, and acceptability and usability findings (when available). The first, second, and senior authors (MPI, CV, and JDP) reviewed and agreed upon the thematic groupings of the results.

The initial database search identified 796 unique sources that we screened by title and abstract (Figure 1). At this stage, we excluded 89.4% (712/796) of sources and reviewed 10.6% (84/796) of full-text sources. We excluded an additional 65 sources after full-text review because they presented a protocol or planned portal feature (7/65, 11%), examined user characteristics and perspectives of general portal use (6/65, 9%), described a tool that was not tethered to a patient portal (17/65, 26%), did not focus on a specific PCEOL domain (6/65, 9%), or was targeted at a pediatric or non–seriously ill population (8/65, 12%). At this stage, we also excluded sources that did not provide detailed descriptions of the tool’s intended audience, content, and/or features (12/65, 18%) and sources for which the full text was not retrieved through the 3 separate university libraries we had access to (9/65, 13%). Using these methods, we identified 18 sources eligible for inclusion. During the full-text review of a selected article [33], an additional article that met the inclusion criteria [34] was identified in the text, bringing the total number of included sources to 19.

We abstracted data from 19 sources published between the years 2013 and 2020 that describe 12 unique PCEOL patient portal tools (Table 1). The identified tools were developed by researchers and clinicians with the specific intent of enhancing PCEOL care. The identified PCEOL patient portal tools address the following domains: ethical and legal (n=5), physical (n=5), and psychological aspects of care (n=2). No tools address spiritual or bereavement support or hospice care education. Tools addressing the physical and psychological aspects of palliative care focus on quality of life and psychosocial support tools and are designed for cancer (n=6) and chronic kidney disease (n=1) patients. Patients are the intended audience for most PCEOL patient portal tools (n=9). Most tools were developed in an academic setting (n=11) and in the United States (n=9).

Portal use measures were described for nearly all (n=10) patient portal tools in at least one study. Table 1 illustrates these findings in detail. Portal use was most often measured by the mean number of log-ins and duration in minutes [40,41,47], number of registered users [45], features accessed [38,40,41,47], or completed activities such as completing an AD or other ACP-related documentation [22,24,33-36] or inputting a daily goal [38].

Overall, an overview of appointments and medical records were the most frequently accessed features for portals [38,40,41,47]. The Patient-Centered Toolkit assessed that education on patients’ problem or condition was the least visited feature; however, this could be explained by the study’s in-patient setting where the patients and caregivers had more frequent access to providers [38,39].

Studies for four of the portals specifically described methods of promoting portal use, either by reminders via electronic messages [35,42,48] or in-person [39]. The impact of these promotions of use is mixed; two of the studies did not collect data on portal use to correspond to their promotion efforts. After sending electronic messages prompting patients to complete their e-PRO assessment before their scheduled appointment, Wagner et al [42] found that 80% of recipients read the message, while 33% of them completed the electronic patient-reported outcomes (e-PRO) assessment.

Usability, acceptability, and user satisfaction were formally evaluated for nine of the identified PCEOL patient portal tools (Table 2). Only two studies have specifically assessed the user perspectives of older adults [37,49]. In general, usability and acceptability rankings were high, and users were satisfied with the tools. Users valued tools that allowed the electronic collection of patient-reported outcomes and symptoms. Aside from their perceived convenience, these tools allow users to gain a sense of control over their health, improve symptom management, and access novel resources [40,41,44,47,48].

Users viewed these tools as a means of improving communication. In particular, the integration of tools with electronic medical records was valued. In ACP tools, users appreciated that the advance directive forms were easily retrievable and viewable for their providers and family [23,24,37]. The tools addressing quality of life and psychosocial support were also perceived to improve communication with the care team [39,44,47,48]. There was evidence of increased care plan concordance between patients [39], and users noted that the tools better prepared them for their clinical visits [44,47,48]. Users of the CaS-PET felt that the tool helped them determine what questions to ask during their medical appointments.

Patient portals may offer accessible avenues for keeping individuals with serious illness and their caregivers connected to PCEOL resources. Although the use of patient portals among patients with serious life-limiting illness is of growing interest, previous studies suggest that patients with serious illness and their caregivers are potentially extensive users of patient portal tools that are commonly available, particularly to access care team information, review laboratory results, and obtain medical records [50,51]. Interest in using patient portals among this population may be high and potentially an excellent venue for providing PCEOL-specific information and resources. This scoping review of 19 articles described available PCEOL features via patient portals tethered to electronic health records. Overall, the review highlighted some of the ways patient portals are offering PCEOL resources, as well as opportunities for future work to leverage patient portals to increase access and awareness of PCEOL care.

The results identified 12 unique patient portal resources supporting the following PCEOL domains: ethical or legal (in the case of this review, exclusively ACP education; n=5), physical (including symptom monitoring, goals of care, and patient-reported outcomes; n=5), and psychological support (n=2). Most resources were targeted toward patient audiences (n=9), with the remainder being targeted toward caregivers, or both patients and caregivers. Only two resources focused specifically on the user characteristics of older adults [37,49], which indicates a clear opportunity to improve our understanding of use and usability in this high-need patient population.

Evidence demonstrates that patients living with serious illnesses and their family members and caregivers can benefit from palliative care [52]. However, there are nearly 1 million patients who are admitted to the hospital annually who could benefit from palliative care and yet do not receive this specialized care [53]. In addition, people living in rural communities have less access to palliative care services than their urban counterparts [54]. Although we recognize that palliative care is highly personalized, patient portals offer the opportunity to leverage technology to promote access and education regarding PCEOL issues. Incorporating PCEOL resources into patient portal features may serve to introduce or supplement specialty palliative care services or bridge the gap where patients have no access to palliative care.

Furthermore, the role of digital health solutions, including patient portal resources, is more important than ever in the wake of the COVID-19 pandemic. Health systems across the United States, and the globe, are forced to incorporate digital care as part of their regular repertoire of available patient care as safety precautions and social distancing have become the norm [55,56]. Individuals who are appropriate for palliative care are highly vulnerable and must be especially considered when addressing how to bring quality virtual health care to individuals during a pandemic. These patients have an increased risk of contracting the virus and experiencing related complications (eg, hospitalization or death) because of their age and health status. They have also been uniquely impacted by the pandemic in that they still require continuous disease management and may experience disproportionate exacerbations of psychosocial needs [57]. Finally, people are dying or experiencing bereavement in isolation at higher rates in the wake of the pandemic [58,59], emphasizing the need for creative palliative care solutions, including better access to PCEOL information and resources for individuals with serious illness and their loved ones.

Nearly all the portals included in this review focused on quality of life, symptom management, or ACP education or completion of advance directives. ACP may be better leveraged through patient portals than other resource types given its more limited and specific goal than some of the more complex and nuanced needs and components inherent in PCEOL care. Similarly, PROMIS tools and other electronic mechanisms for patient-reported outcomes are well suited for patient portals.

Although the legal and physical aspects are essential pieces of quality and holistic PCEOL services, equally important are the psychological and spiritual aspects, as well as end-of-life education around hospice. Only two of the portals reviewed focused on the psychological needs of patients with serious illness [47,49]. None of the portals included in the review offered features addressing grief and bereavement or information about hospice or comfort care. Despite the evidence supporting the benefits of hospice care [60-62], lack of information and understanding remains one of the most common barriers to hospice use [63-65]. Furthermore, bereaved individuals and families often benefit from educational resources related to the grief process [66]. Future research and development on portal tools for patients with serious illness and their family members should emphasize psychological and spiritual PCEOL domains and consider tools and educational resources regarding grief, bereavement, and hospice.

It is important to emphasize the special consideration of older adult patients with serious illness and patient portal tool development, availability, and adoption among PCEOL populations. In total, 80% of older adults in the United States have one or more chronic conditions, many of which can lead to serious illness [9]. As the older adult population continues to grow, and subsequently the number of older adults experiencing serious illness increases, the focus of patient portals’ development and research must continue to include the perspectives and needs of patients aged ≥65 years. Despite the currently low adoption rate of portals by older patients and caregivers, research demonstrates that there is an interest in such tools among this group [37,67], an interest that might well be addressed by improving the design and availability of solutions tailored to and especially suited to the needs of older patients [68]. Future research on PCEOL patient portal development, use, and acceptability should explicitly incorporate older adult needs and feedback.

To the best of our knowledge, this scoping review is the first to identify and describe patient portals that offer PCEOL features. However, this is not without limitations. First, the review only included results from academic literature and may be missing commercially available portals that were developed by industry without concurrent publication in the peer-reviewed literature. In addition, we did not evaluate the full text of articles in which we were unable to retrieve access, potentially overlooking additional PCEOL patient portal resources. Although some of the articles described the efficacy on clinically meaningful outcomes, we did not address this information as it was beyond the scope of our scoping view. Future research should focus on the efficacy of such tools. Finally, this review had a small sample size of articles (n=19) covering 12 different patient portals that are predominantly US based, potentially limiting the results to a Westernized PCEOL perspective.

The scoping review highlighted an important gap in PCEOL resources offered via patient portals linked to electronic health records, as well as opportunities for leveraging patient portals as a means of offering education and support for patients with serious illness. The results suggest that there are ongoing efforts to offer various PCEOL supports through patient portals, particularly with a focus on ACP. However, our review resulted in a small number of portals that met our criteria, of which included only a few of the expansive elements of PCEOL care. Future research and development of patient portals would benefit from offering comprehensive PCEOL features to increase access and education for patients experiencing serious illness.