Over the past few decades, the management of patients with cancer has considerably changed. Owing to earlier detection of cancer and improved treatment strategies, the number of cancer survivors has increased [1]. For this reason, patients with cancer currently require long-term follow-up care, similar to patients with chronic diseases. It is expected that the increased need of follow-up care, in combination with an overall increase of cancer patients due to an aging society, will intensify the use of health care services and increase health care costs [2]. Current health care systems cannot accommodate these increased demands and are deemed unsustainable [3,4]. An additional problem that requires a reevaluation of oncological health care services is the multidisciplinary, and sometimes multi-institutional, aspect of current care. This refers to the fact that cancer patients are usually cared for by multiple health care providers (HCPs), and that, as a result of centralization of complex care to high-volume centers [5], patients may receive parts of their treatment in multiple hospitals [6-9]. These aspects result in the fragmentation of cancer care [10].

The challenges facing cancer care are recognized by the World Health Organization [11] and the Health Program of the European Union [12], and have prompted them to think of different health care structures that enable the delivery of affordable, coordinated, patient-centered oncological care. A solution in restructuring health care for oncological patients could be found in telemedicine or electronic health (eHealth). Moreover, the COVID-19 pandemic, which led to the sudden forced implementation of telemedicine, has demonstrated the potential of telemedicine and eHealth to many patients and doctors [13]. These technologies therefore seem to be indispensable for sustainable care.

For these reasons, the time seems right to deploy eHealth much more widely in oncological care. An updated, accepted definition of eHealth is lacking, but eHealth can be perceived as an umbrella term for all digital communication and information technologies that aid in health care or health care services [14,15]. Over the past decades, multiple eHealth interventions have been developed to support oncological care [16,17]. The eHealth-based intervention of interest in this review is a digital care platform. We define a digital care platform as an eHealth-based tool that aims to increase coordinated and patient-centered care. A digital care platform incorporates several different features, which separately have been the subject of study elsewhere [18-21]. First, digital care platforms may provide patients with information specific to their situation. Second, they can provide an overview of the patients’ personal health records, including appointments, medical results, and correspondence. Third, digital care platforms may also offer direct, secure messaging with HCPs (eg, electronic consultation [e-Consult]). We consider these three characteristics as key features of a digital care platform. Some additional features may include the registration of patient-reported outcomes (PROMs), a patient forum, the possibility to exchange patient-related information between different health care institutes, and enable communication between HCPs and general practitioners.

Studies on the use of eHealth-based tools and their effect on care processes have been previously described for patients with chronic diseases [22,23]. A recent systematic review by Tighe et al [23] indicated that digital platform–like interventions such as self-management tools have a positive effect on physical activity and disease-related quality of life. A systematic review by Kooij et al [22] indicated that digital interventions had positive effects on patient confidence and HCP satisfaction, but that firm conclusions on its clinical effects could not be drawn. Studies on the use of eHealth in oncological care have reported favorable results, as it has been associated with improved patient-provider communication [24]; improved coping with cancer-related symptoms such as fatigue, depression, anxiety, and physical activity [16,25-28]; and improved medication adherence and higher patient satisfaction [20,27]. Some studies have also reported favorable effects on quality of life [29,30]. However, these studies investigated the effect of only one feature of a digital care platform, for instance the registration of PROMs. Previous reviews did not study the digital care platform as a central part of the oncological health care service or the effect of digital care platforms on enhancement of information, continuity of care, or patient- and HCP-reported experiences. The effect on patient self-efficacy has only been studied for one feature of a digital care platform, namely self-management programs [25]. The outcome parameters of interest (ie, enhancing availability of information, self-efficacy, continuity of care, and patient-reported and HCP-reported experiences) can also be placed in the Quality of Care framework provided by the Institute of Medicine [31]. The six domains of quality of care are safe, effective, patient-centered, timely, efficient, and equitable care. Enhancing availability of information is a measure of effective and equitable care, self-efficacy is a measure of patient-centered care, patient-reported and HCP-reported experiences are measures of safe and patient-centered care, and continuity of care is a measure of efficient care and arguably also timely care.

Additionally, we did not identify any existing systematic reviews that focused on the barriers and facilitators that exist for the successful implementation of a digital care platform in oncological care. This is important to assess so as to adequately determine the feasibility of a digital care platform as part of routine oncological care.

The aim of this systematic review was two-fold: (1) to collect the best available evidence of the effect of a digital care platform on quality of care for oncological patients by focusing on enhancement of available information, self-efficacy, continuity of care (including communication), and patient-reported and HCP-reported experiences; and (2) to analyze the currently reported barriers and facilitators for implementation of a digital platform in oncological health care.

This systematic review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42020199282) [32] and was carried out according to the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) guidelines [33].

The PubMed (Medline), Embase, CINAHL, and Cochrane Library databases were searched for the period from January 1, 2000 until May 27, 2020. Search terms used included “neoplasms” and affiliated terms combined with “patient portal” or “digital care” or “eHealth.” The aim of this search query was to provide search results on literature involving both the effect of a digital care platform on quality of care for oncological patients as well as barriers and facilitators for implementation. The exact search query is shown in Textbox 1.

Studies of interest included randomized controlled trials (RCTs), prospective studies, and retrospective cohort studies. In addition, qualitative studies in which questionnaires were distributed or stakeholders were interviewed to investigate barriers and facilitators were also included. All studies were required to generate empirical data. Studies not written in English were also screened on the condition that they presented an English abstract. A translator could translate the full text if the abstract seemed to be eligible, thus avoiding a language bias. Systematic and narrative reviews, conference abstracts, and single case reports were excluded.

For our first objective, to collect the best available evidence of the effect of a digital care platform on quality of care, studies were required to involve oncological patients ≥18 years old. These clinical studies had to assess the effectiveness of a digital care platform. Although we provided our definition of a digital care platform in the Introduction, considering the exploratory stage of research of digital platforms in oncological care, we suspected that there would be a rather limited number of studies investigating such an extensive digital care platform. For this reason, for inclusion in our systematic review, a digital care platform was required to have at least two of the following key features: (1) provide general information concerning the disease as education (eg, symptoms, treatment, follow-up, prognosis); (2) provide patient-specific information concerning their medical file, such as planned appointments, treatments, and lab results; and (3) enable patients to communicate with their physician or specialized nurse via chatting, e-Consult, or email. Figure 1 depicts these and other features of a digital care platform.

Studies that did not involve a digital care platform with at least two of the three above-mentioned features and studies that provided insufficient or vague details concerning the digital intervention were excluded from the analysis to assess our primary objective.

For the second objective, identification of barriers and facilitators for implementation of a digital care platform in oncological care, studies were not required to involve a digital care platform that was already implemented. They could comprise studies that actually implemented a digital care platform and subsequently described the barriers and facilitators for implementation, but they could also comprise studies that identified barriers and facilitators based on a hypothetical discussion with stakeholders. In the latter case, the digital care platform in question was not yet implemented or developed.

Two separate reviewers (JH and LV) screened the search output on titles and abstracts using Rayyan software [34]. During this screening process, the reviewers were blinded to each other’s decisions. Studies with a contradictory judgment were discussed. In case of a remaining discrepancy, senior reviewers (RH and MS) were asked to arbitrate. The full texts of the selected studies were screened for eligibility. Reference lists of studies that were included after full-text screening were checked for additional eligible studies (snowball method). In cases in which multiple eligible studies reported on the same dataset (≥50% overlap of sample size), we excluded the article with the shortest follow-up.

Two authors (JH and LV) extracted data using a shared template. These data included: (1) author, year of publication; (2) country; (3) study design; (4) patient population, sample size; (5) platform that was the subject of the study; 6) features of the platform; and (7) the outcome parameters, including enhancement of available information, self-efficacy, continuity of care, and patient- and HCP-reported experiences. If possible, data were pooled; otherwise, synthesis of the results was performed qualitatively.

Barriers and facilitators for implementation of digital care platforms were categorized according to the framework of Grol and Wensing [35]. This framework categorizes barriers and facilitators at six different levels: (1) innovation, which involves advantages in practice, feasibility, credibility, accessibility, and attractiveness; (2) individual professional, which concerns the awareness, knowledge, attitude, motivation to change, and behavioral routines of the involved professionals; (3) patient, which involves knowledge, skills, attitude, and compliance of the patients; (4) social context, which concerns opinion of colleagues, culture of the network, collaboration, and leadership; (5) organizational context, which includes the organization of the care processes, staff, capacities, resources, and structures; and (6) economic and political contexts, which involve regulations, financial arrangements, and policies. The frequency of the barriers and facilitators mentioned, and the quality of the studies were used to prioritize barriers and facilitators.

The Mixed Methods Appraisal Tool (MMAT) version 2018 was used to critically appraise the included studies [36]. The MMAT is designed to help reviewers appraise the quality of empirical studies with different methods. These include quantitative RCTs, quantitative nonrandomized trials, quantitative descriptive studies, as well as qualitative and mixed methods studies. Each category includes five different quality parameters, all requiring to be assessed by answering “yes,” “no,” or “cannot tell” (maximum total score=5). Two authors (JH and LV) independently appraised all studies. In case of disagreement, RH and MS were asked to arbitrate. Studies with a total score of 1-2 were considered to be of low quality, a study with a score of 3 was considered to be of moderate quality, and studies with total scores of 4-5 were considered high-quality studies.

The initial search generated 6789 articles. After removal of duplicates and initial screening of titles and abstracts, 52 studies remained for full-text assessment. Of these 52 studies, 17 articles met our eligibility criteria and were included in this systematic review. Figure 2 depicts the PRISMA flowchart for the study screening process.

Quality assessment of the included studies is shown in Table 3. Of the clinical studies, four studies were of low quality [37,41,43,44], three of moderate quality [38,39,42], and one study was assessed as a high-quality study [40]. Regarding the barrier and facilitator studies, three studies were of low quality [49,51,53], one of moderate quality [48], and five were of high quality [45-47,50,52].

A digital care platform for oncological patients has the potential to improve quality of care through the improved availability of information and positive effect on self-efficacy. Although continuity of care was not studied as a primary outcome in the identified studies, based on focus group interviews with HCPs, a digital care platform potentially improves continuity of care by optimizing the exchange of patient information across institutes. Patient-reported experiences such as satisfaction with the platforms were considerably positive. Our barrier and facilitator analysis indicated that the majority of barriers exist at the professional level. This included a concern of increased workload for HCPs and release of unattended medical information to patients. The majority of facilitators were identified at the patient and innovation levels. The patient’s ability to become more informed, empowered, and involved in their care was identified as a prominent facilitator. Another relevant facilitator, at the organization level, is the digital care platform’s potential to improve information exchange between HCPs across different institutes. This is relevant as this may improve continuity of care and diminish fragmentation of oncological care. Among the clinical studies, the majority were of low to intermediate quality. Regarding the barrier and facilitator studies, the majority were of high quality.

The positive effects of a digital care platform described in this review are consistent with prior research studying isolated features of digital care platforms [16,29,54]. A digital care platform seems to be a helpful medium in providing patients with general disease-related and personal information. This is a welcoming result as previous studies have described that a considerable number of patients are dissatisfied with information provision [55,56]. Moreover, 40%-80% of all medical information provided by the HCP during a consultation is forgotten or remembered incorrectly [57]. Therefore, the need of patients to read or reread medical information that is relevant for their situation is an important facilitator for a digital care platform. In this systematic review, we identified studies that showed positive effects on self-efficacy, but no convincing clinically significant effects. In another systematic review, which studied the effectiveness of eHealth-based self-management tools, significant yet small effects on self-efficacy were described [25]. The authors explain that the tool was able to increase self-efficacy by enabling patients to enhance participation in their care trajectory. A more recent randomized trial refuted this finding [58]. In this RCT, the effect of an eHealth-based self-management tool (Oncokompas) was investigated in 600 cancer patients. At 3 and 6 months, patients were assessed on self-efficacy with the General Self-Efficacy scale. No significant effects on self-efficacy were observed (mean difference 0.5, 95% CI –0.4 to 1.4, P=.31). The authors partially attributed the lack of effect to the included patient population, which mostly consisted of patients diagnosed with cancer 2 years prior to the study. The authors suggested that this population is perhaps less in need of a self-management tool to increase self-efficacy as they have more experience and “know-how” compared with their newly diagnosed peers. The hypothesis that the effect of an eHealth-based tool may vary considerably depending on the patient population is important to consider for assessing the effectiveness of a digital care platform. In our systematic review, we included studies with patients in different stages of their disease trajectory and with different cancer types, and all effects regarding self-efficacy were positive but not significant. On the one hand, this aspect can be perceived as a limitation because the heterogeneity of patients may underestimate the effect of a digital care platform. On the other hand, it is a strength as we now have an overview of the effect of a digital care platform on self-efficacy of a broad cancer patient population, thereby increasing the generalizability of our results.

Similarity with prior work concerns the limited number of studies investigating combined features of a digital care platform. During the screening process of this review, we encountered numerous studies that only investigated one feature of the platform. This finding is similar to the findings of a review by Kruse et al [20], which concluded that many studies assessed the effect of one feature of a digital care platform and that a full platform was rarely studied in clinical practice. To adequately ascribe outcomes as an effect of the intervention, it is reasonable to first study an isolated feature of a digital care platform. However, it is absolutely vital to also study the combined features in a digital care platform for two key reasons: (1) because this is likely to be the eHealth intervention that is implemented in practice, and (2) because the effect of multiple features may not be equivalent to the sum of effects of a single feature.

The fact that we assessed the effects of digital care platforms that had at least two key features can therefore be considered a strength of this review. In addition, we assessed the effect of a digital care platform on quality of care parameters such as continuity of care, and described the patient- and HCP-related experiences. Other reviews did not, and instead only assessed the effect of a digital care platform with one feature or chose to study different endpoints such as fatigue, physical activity, depression, quality of life, and self-management abilities [20,25,29,30,54,59]. An additional strength is that we analyzed barriers and facilitators for implementation. This allows for a good understanding of what is needed for successful implementation. A final strength is that the search query was quite elaborate without exclusion of non-English articles, thereby minimizing the possibility of missing valuable studies. For these reasons, this systematic review provides a comprehensive overview of the best available evidence of the effect of a digital care platform on quality of care for oncological patients.

However, this systematic review also has some limitations. First, the included clinical studies were mostly early-stage, single-arm prospective studies where feasibility and acceptability were investigated. Although we did include three studies with a more advanced research methodology (ie, RCTs), these studies were all appraised as studies with low to intermediate quality. This makes interpretation of these RCTs quite challenging. One RCT reported a significant difference in patient satisfaction scores favoring a type of digital care platform [37]; however, the minimal difference (3.48 vs 3.28 on a 1-5 scale) puts into question the clinical relevance of this finding. A second limitation of this review concerns the heterogeneity of study designs and of the digital care platforms. Despite our predefined criteria of what a digital care platform should include, they were still quite different from each other with respect to the specific web format, options, and intended use, among other aspects. This heterogeneity precluded a pooled analysis with quantitatively measured outcome parameters. A third limitation concerns the study population and its representativeness. Three-quarters of the clinical studies included breast cancer patients. This study population mainly consists of well-educated females [60]. Indeed, 72%-73% of the breast cancer population in the studies included in this review had a college or university degree. It is plausible that a high education is an important factor for successful use of a digital care platform. For this reason, results of this review should be interpreted with caution and cannot simply be generalized to other cancer patients. A final limitation concerns the barrier and facilitator analysis. Studies included in this review almost exclusively focused on the barriers and facilitators mentioned by the end users of the platform (ie, the patients and HCPs). Evidently, their perspectives are crucial, but barriers and facilitators proposed by other stakeholders should also be investigated, such as stakeholders with an organizational, economic, and political background (eg, health insurers). This latter group can provide insight into what is required to realize the structural financing of a digital care platform.

Although there is a growing body of literature that describes the positive effects of digital care platforms, high-quality studies describing the effectiveness of these platforms integrated in oncological care are currently lacking. In addition, most studies implemented a digital care platform in one target patient population such as breast cancer patients. More evidence is required concerning the desirability and use of a platform in patients with other types of cancers and education levels. More specifically, it is important to investigate whether patients with a lower level of education can benefit from digital care platforms to the same extent as their well-educated peers to avoid widened health disparities. The same applies to the patient population with limited internet access or internet skills. Studies in this review were performed in Australia, North America, and western Europe. The usefulness of digital care platforms in countries with low internet access is likely very different.

Concerning barriers and facilitators, future studies should aim to further elucidate barriers and facilitators at the organizational level and the economic and political levels. Ultimately, this is required for a digital care platform to become a successful eHealth-based tool in the improvement of quality of care for patients living with cancer.

Digital care platforms have a favorable effect on availability of information and enhancement of self-efficacy. Additionally, they could potentially serve as a valuable medium to improve continuity of care by optimizing communication between patients and HCPs and among HCPs. The vast majority of patients are positive about a digital care platform and its ability to meet their needs in improving the availability of information and patient involvement. Although these results are favorable, they were mostly generated by early-stage, nonrandomized studies with a specific patient population. To fully understand whether a digital care platform is able to increase quality of care by supporting the delivery of coordinated, patient-centered oncological care, more advanced studies such as RCTs are required, as well as studies investigating the barriers and facilitators at the economic and political levels.