Digital Health for Optimal Supportive Care in Oncology: Benefits, Limits, and Future Perspectives

Digital solutions with ePROs and self-management can be incorporated in supportive care in oncology practice and provide benefits to: patients, e.g., reduced symptom burden and distress, increased symptom reporting, improved overall survival; healthcare professionals, with targeted patient management; payors, potentially with reduced supportive care-related costs and hospitalizations.

The International Agency for Research on Cancer estimated that in 2018, there were 18.1 million new cancer cases worldwide and 9.6 million cancer-related deaths [1]. A global surveillance report suggests a trend toward increased survival [2], with some cancers progressing to chronicity. However, the total burden of new cancer cases is increasing, and new therapies are generally more costly [3]. Additionally, more drugs are available in oral formulations for home administration, with reduced face-to-face surveillance by healthcare professionals (HCPs). Novel approaches for optimal patient management that allow containment of healthcare costs are urgently needed [4].

The new approaches should focus on patient-centered care with integration of tumor-directed treatment and patient-directed supportive and palliative care throughout the disease journey [5, 6]. The goals of management are to achieve improvements in not only overall survival (OS) but also patient-reported outcomes (PROs) such as quality of life (QOL) [7], fewer emergency department visits, and self-reported improvements in symptoms [7, 8].

The intensive development over recent years of therapies with novel mechanisms of action, including molecular-targeted therapies, immuno-oncology therapies, and precision radiation oncology, has transformed the oncology treatment landscape [9, 10]. These advances have increased the complexity of treatment (combination of therapies) and required modifications in the patient pathway (oral treatment intake at home versus hospitalization) to ensure quality care. The real-world toxicity profile of novel agents may not always correlate with that observed in clinical trials and may result in unanticipated toxicities [11, 12]. Increased availability of oral therapies for home administration results in less healthcare supervision during treatment, whereas the prolonged use of such treatments as long-term maintenance may be associated with the emergence of new toxicities [13]. Therefore, careful monitoring of adverse events (AEs) during self-administration of treatments at home is becoming essential to facilitate prompt intervention to reduce their severity and duration.

Patients must therefore manage symptoms and treatment-related side effects without direct medical supervision; home administration of anticancer treatments also increases the chance of nonadherence and administration errors by patients [14]. With immunotherapeutic treatments, the timely identification of toxicities is crucial since many symptoms may improve with prompt intervention [15]. Additionally, a potential shortage in oncology services and workforce linked to the increasing cancer incidence and complexity of cancer treatments [16] has highlighted the need for new strategies to ensure that all patients receive optimal treatment and care throughout the continuum of disease.

Advances in digital communications and medical technologies have led to the digitalization of healthcare [17]. Increased access and uptake of such technologies among physicians and patients yields large amounts of potentially usable data, which, in the context of electronic health records (EHRs), forms an important part of physicians’ decision-making. Self-reported data is extensively used in healthcare. Patient-level data provide real-world medical information, with opportunities for improved clinical decision-making, patient empowerment, improved health outcomes, and cost reductions [18, 19, 20]. However, patient confidentiality and compliance with local and global data privacy regulations need to be ensured.

Digitalized healthcare comprises eHealth, telemedicine, telemonitoring, and digital therapeutics (Fig. 1).

The terms digital health, telehealth, and eHealth are interchangeable and are defined as the provision of healthcare services supported by telecommunications or digital technology to improve or support healthcare services. eHealth solutions can be part of each step of the healthcare process (i.e., prevention, diagnosis, decision-making, treatment/intervention, and follow-up).

Telemedicine represents medical services provided remotely to patients by HCPs using telecommunications platforms. Healthcare activities, such as patient evaluation, diagnosis, or treatment, are performed by HCPs without the need for inpatient consultation, although the legal status of such consultations varies according to jurisdiction [21].

Telemonitoring is the use of digital technology to frequently or continuously monitor patients’ vital signs or any other symptoms. The information is assessed remotely by HCPs to inform the patient and caregivers about the actions needed for appropriate symptom management and treatment advice.

Digital therapeutics embed algorithms based on medical guidelines and best practices, which transform collected data into actionable insights, with the objective to bring value to evidence-based clinical outcomes (from clinical studies or real-world evidence). They may be used alone or in conjunction with drugs and medicinal products, medical devices, or other therapies, to enhance and support medical treatment. According to the risk level of the embedded algorithms, the digital therapeutics may be classified as medical devices. Depending on the regulatory status, they may be used on prescription only (prescription digital therapeutics).

A further technology of relevance to the broad concept of digitalized healthcare is artificial intelligence with capabilities of machine learning, which may be defined as the use of computer algorithms to make successful predictions about future events based on past experiences [22].

From a health outcomes perspective, digital health can be grouped into solutions connected to sensors or not and that capture ePROs to allow patient monitoring only or those that allow patient monitoring and symptom management by HCPs, covering remote areas, or symptom management by the patients themselves with or without real-time decision support for self-management. Patients receive individualized guidance, from a simple recommendation to call their HCP, to a suggestion to begin a specific treatment intake.

The Multinational Association of Supportive Care in Cancer defines supportive care in cancer as “the prevention and management of the adverse effects of cancer and its treatment. This includes management of physical and psychological symptoms and side effects across the continuum of the cancer experience from diagnosis through treatment to post-treatment care. Enhancing rehabilitation, secondary cancer prevention, survivorship, and end-of-life care are integral to supportive care.” (About MASCC. mascc.org/about-mascc. Accessed January 11, 2019). Whereas there has been significant progress in anticancer treatment, improvements for optimal supportive care are still needed at all stages of the cancer treatment pathway [5]. Currently, supportive care interventions’ assessment of patient QOL and medical outcomes remains limited, and QOL endpoints are insufficiently reported for clinical trials of novel therapies [23].

A number of evidence-based supportive care guidelines have been developed, but their implementation in routine clinical practice is suboptimal and the opportunity to improve control of symptoms is often forfeited [24]. This highlights the need for more optimal use of guidelines, for personalized and patient-centered care that is delivered in a timely manner.

Digital solutions present an opportunity to address certain unmet needs in prevention or management of adverse events in patients with cancer including (1) increased communication between patients, providers, and their communities [18]; (2) education of patients and caregivers; (3) integration of standard clinical assessments with PROs measured during routine clinical practice; (4) help of patients in monitoring their respective conditions [18]; (5) improved patient empowerment and self-management; and (6) improved evidence from clinical trials on the basis of PRO endpoints in studies evaluating anticancer treatments and prospective evaluations of supportive care interventions and real-world efficiency of care for cancer patients.

The objectives of the present review are to evaluate the state of digital health solutions in oncology supportive care allowing collection of ePRO and focused on symptom management and to identify benefits and limitations.

Guidance of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement was followed in the conduct of this study (Fig. 2).

The MEDLINE Public Library of Medicine (PubMed) database and the Cochrane Library were explored from December 1, 2008, to November 30, 2018, for relevant studies using the following search terms: (1) MEDLINE, “cancer or oncology” AND “telehealth or eHealth” AND “symptom management” or “symptom monitoring”; (2) Cochrane Library (title abstract keywords), “cancer or oncology” AND “telehealth or eHealth” AND “symptom”. Clinicaltrials.gov search was performed using the following search strategy: “cancer or oncology” (condition or disease) AND “telehealth or eHealth” (other terms) AND “symptom” (outcomes measures).

Screening of publication was done by 2 researchers on titles and abstracts and then full-text to ensure eligibility to the following criteria.

Adult cancer patients, all randomized controlled trials (RCTs) or feasibility and pilot studies that evaluated the effectiveness of telehealth or eHealth solutions in supportive cancer care were eligible for inclusion in English language.

Studies involving pediatric patients and those evaluating solutions at the palliative phase were excluded.

For results retrieved from clinicaltrials.gov, not completed studies or studies without published results were excluded.

Retrieved studies were reviewed, and those evaluating solutions at palliative latest phase of cancer were removed from the analysis.

Outcomes of interest were as follows for each digital solution identified: description of the digital solution including PRO for supportive care in oncology, with remote monitoring, with/without patient automated symptoms self- management, its benefits, limitations, drivers of and barriers to adoption; unmet needs; PRO data including QOL outcomes; AE incidence, severity, and management; emergency room (ER) admissions and hospitalizations; health resource utilizations; and survival outcomes including OS.

Search results were critically analyzed by the authors for relevance to the focus of this review. Two researchers extracted the data. The authors analyzed systematically according to outcomes of interests detailed above the study results to critically discuss the impact on outcomes of the various digital solutions.

A total of 206 articles have been identified through databases searches in Medline, Cochrane, and Clinicaltrials.gov. Twenty-four (24) additional records were provided from other sources (manual search, cross-references). We excluded narrative reviews (23), publications which titles and abstracts were about pediatric population or focused on palliative care phase of cancer (60), and other records (27) (not completed results in clinical trials, conference abstracts, not in English language, cross references to full-text articles).

Regarding the 120 selected articles, another 54 full-text articles were excluded because of absence of study results, duplicates, or design (exclusion when not a RCT nor a feasibility study).

Finally, 66 full-text articles and associated clinical trials are included in this review.

The review results outlining the status of clinical evidence regarding digital health solutions that collect ePRO for supportive care in oncology are summarized in Table 1 [7, 8, 25-86]. These 38 digital solutions can be classified into 2 main categories: the first, 17 digital solutions based on PRO collection only, and the second, 21 digital solutions providing also self-management. The key findings are summarized according to outcome.

Clinical evidence for digital health solutions evaluated in feasibility or randomized controlled studies are also summarized in Table 1 [7, 8, 25-86].

From the patient perspective, some of the key factors identified for the uptake of the digital tools included the following:

1Ease of use [30, 38, 44, 51, 53, 55, 64, 80];

2Reassurance [28, 30, 41, 48, 49, 55, 59, 70];

3High usability and usefulness [37, 42, 44, 58, 62, 74];

4Improved communication with HCPs [27, 29, 30, 53, 58] www.owise.uk;

5Correct generation of system alerts and fast response to alerts [28, 70, 80];

6Patient empowerment [29, 30, 69]; and

7The convenience of real-time reporting of symptoms [28, 37];

One study evaluating the extent of patient use of a Web-based intervention reported that reduction of cancer symptom distress was a key driver of uptake, with use of the intervention resulting in a significant reduction in distress score [47].

Conversely, some of the barriers for adoption encountered by patients were as follows:

1Problems with technology or connectivity [48, 49, 69, 80];

2Limited usefulness [29, 30, 71];

3Lack of clarity of the language used [29, 30]; and

4Generation of false alerts [69].

Whereas higher education level, current employment, and low levels of social support have been associated with uptake, lower education level and non-working status may be barriers to accessing interventions [47, 84].

Fewer studies have assessed the feasibility of digital solutions from the HCP perspective. The most important reasons for adoption reported by HCPs were the usability and usefulness of the tool [26, 38, 52, 58], and the most commonly reported barrier was problems with technology or connectivity [31, 75].

Interestingly, while some tools were perceived as a burden due to increased workload [28], others did not impact the working time of HCPs [69, 85].

Most studies presented ePRO data, including symptom distress and burden, pain, depression, and adherence.

A meta-analysis of 9 studies reported a statistically significant benefit for digital interventions in patients with cancer-related fatigue, with moderate benefits also observed for QOL and depression [45].

Several studies showed a significant reduction compared with usual care in symptom-related distress on the basis of measures that included Short-Form (SF)-36, Memorial Symptom Assessment Scale (MSAS), Symptom Distress Scale-15 (SDS-15), and Functional Assessment of Cancer Therapy-Head & Neck Scale (FACT-HN) [36, 40, 45, 47, 50, 78, 85]. Symptom benefit was observed in conjunction with automated home or Web-based symptom self-management systems.

Studies also reported a reduction in depression [73, 85], symptom severity [33, 53], pain [43, 56, 73, 77], and need for symptom management support [40].

An RCT enrolling 766 patients with solid tumors receiving outpatient chemotherapy demonstrated that self-reporting of 12 common cancer-related symptoms led to significant improvement in QOL, as measured by the EuroQol EQ-5D Index [8].

Two studies used the European Organization for Research and Treatment of Cancer Qualify of Life Questionnaire Core 30 (EORTC-QLQ-C30) for QOL assessment [43, 52]. One of these used the EORTC-QLQ-C30 and the Hospital Anxiety and Depression Scale (HADS) as an intervention, with a larger proportion of patients who reported these measures to their oncologists showing clinically meaningful improvements in QOL compared with a control group, despite no detectable changes in patient management [52].

An RCT evaluating the impact of an internet-based exercise intervention reported significant improvement in EORTC-QLQ-C30 scores for global health status, physical, role, and cognitive functioning, together with improvements in pain severity on the Brief Pain Inventory compared with control [43].

In another study of a Web-based intervention, the addition of self-care instructions and communication coaching to Electronic Self-report Assessment–Cancer (ESRA-C) of symptoms and QOL resulted in significant increase in reporting fatigue, pain, and physical function issues. However, differences between groups in symptom distress reported by patient did not reach significance [46].

Finally, a report found benefit for patient QOL, including increased symptom identification and management, and improved functional status following electronic collection of Therapy-Related Symptom Checklist for Adults (TRSC) [81].

A prospective study compared survival in patients with lung cancer who were assigned to weekly symptom self-reporting via a Web application intervention for early detection of relapse with a retrospective group of control patients [60]. Median OS was improved for the patients assigned to the intervention compared with the historical control arm.

Survival outcomes were also reported in 2 RCTs. A single-center trial reported that integration of ePROs into the routine care of patients with metastatic cancer led to increased survival compared with usual care [7]. At a median follow-up of 7 years, median OS was 31.2 months (95% CI, 24.5–39.6) in the group that provided self-report of 12 common symptoms, with severe or worsening symptoms triggering an email alert and follow-up care by a nurse practitioner with escalation as needed. In comparison, median OS in the group assigned to usual care was 26.0 months (95% CI, 22.1–30.9; difference, 5 months; P = 0.03). In patients with advanced lung cancer, a multicenter study reported that intervention involving a Web-based follow-up algorithm to assess weekly patient symptom self-reports compared with routine follow-up resulted in median OS of 19.0 (95% CI, 12.5-noncalculable) and 12.0 months (95% CI, 8.6–16.4), respectively (P = 0.001) [61]. In addition, the performance status at first relapse was 0 to 1 for 76% of patients in the intervention arm compared with 33% in the control arm (2-sided P < 0.001); anticancer treatment was considered to be optimal in 72% and 33%, respectively (2-sided P < 0.001). In the final OS analysis for this study, median OS was 22.5 months in the intervention group and 14.9 months in the control group (hazard ratio, 0.59 [95% CI, 0.37–0.96]; P = 0.03) [87].

The effect of digital solutions on the number of ER visits, hospital days, or utilization of healthcare resources is not commonly evaluated in clinical studies. Some solutions, involved in patient monitoring providing or not providing feedback for self-management, have been associated with a reduction in ER visits, unplanned hospitalizations, and hospital days [8, 35, 73]. Additionally, use of a telehealth system for rehabilitation of patients with head and neck cancer following chemo-/radiotherapy resulted in fewer and shorter appointments until discharge compared with usual care and was accompanied by a significant cost-reduction for patients, specifically in travel costs [51]. On the contrary, one study using a Web-based intervention that included review by a nurse practitioner found no differences compared with control with respect to healthcare resource use, including oncology-related appointments, number of physician visits, or medical tests [75]. The effect of digital solutions on overall healthcare costs needs further assessment [8, 35, 73, 75].

Benefits and limitations of introducing a patient-management solution in oncology, according to stakeholders of digital solutions in the healthcare system, are summarized in Table 2 and illustrated in Figure 3. These benefits and limitations were identified in the selected publications and from the authors experience and opinion. Lots of benefits have been identified of important impact on all stakeholders (patients, physicians, caregivers, nurses, healthcare system, pharmaceutical company), with limitations related to technical dealing, regulatory constraints, costs, and changes in practices.

Although the clinical benefits of remote patient monitoring have been demonstrated in clinical trials [7, 62], achieving optimal supportive care requires strategies that go beyond ePRO apps/systems. Such benefits are not obtained solely through the assessment of outcomes of interest but also through appropriate management in response to assessments. Even if benefits have been confirmed in the setting of RCTs, there is a need to continue to evaluate ePRO efficacy and efficiency in real-world conditions, with ongoing assurances of data security and privacy, to provide relevant information for optimal self-management.

Several factors need to be considered for a high-quality symptom self-management system. Guidance from the treating physician is critical. Electronic self-reported assessment tools for cancer-related symptoms and QOL can increase communication between patients and HCPs and promote discussion that is focused on symptoms and QOL. Digital tools that give advice to patients on the reporting of symptoms to HCPs have been shown to increase symptom reports by patients during visits. However, these have not been shown to impact practitioner responses, indicating that guideline adherence and commitment by the medical team is also needed. The collection of information regarding related clinical symptoms and the medication received requires integration with electronic real-time monitoring of symptoms into oncologists’ routine clinical practice. When real-time monitoring is used, beneficial outcomes in terms of symptom management have been identified [88], with the potential for further optimization when structured patient education or practitioner-/nurse-led symptom counseling is in place. Optimization of digital tools requires integration with the patients’ EHRs, thereby allowing continuity in the flow of patient-related data and the healthcare support systems.

Digital health solutions need to be integrated into the patient pathway and in healthcare team practices for optimal supportive care in oncology in line with appropriate guidelines. How this integration is implemented is debatable, with consideration given as to whether the digital tool is merged into current healthcare systems in a gradual or disruptive manner. The European Society for Medical Oncology (ESMO) has developed a Magnitude of Clinical Benefit Scale (ESMO-MCBS) to assess the extent of the clinical benefit from new and effective anticancer therapies measuring improvement in survival, disease-free survival, response, grade 3–4 toxicities, and QOL measures [89]. MCBS-based assessment of the digital tools as part of anticancer therapies and the use of MCBS for the development of clinical guidelines would ease this integration.

There are challenges in the development of a digital solution for supportive care of cancer patients. Setting up and conducting clinical trials for the evaluation of digital tools is a long process, especially because digital solutions need to be quickly available for evaluation in real-world settings. The principal difficulties are in developing and implementing a solution to fit the needs of all or most patients, while achieving the necessary patient compliance to change with the new digital tool and integrate it into care and maintaining enough adaptability for its use in different regulatory systems and healthcare centers. Implementation may be associated with challenges in staff having to deal with new technologies, accepting and adapting to changes, and the potential for reorganization of multidisciplinary teams/treatment centers. Maintenance of the device may also introduce complexity since device utility is dependent on updates in accordance with relevant guidelines, as well as drug safety information, approval of new drugs, and the use of different drugs from the same class. Oncologic therapy is by its nature complex, with sequential phases, and device utility will need to reflect the use of different antitumor regimens, including radiotherapy and radio-chemotherapy, and combination of drugs. Uptake of the technology may be dependent on oncologist perceptions of patients’ willingness to adopt new technologies, as well as the actual willingness of patient subgroups, particularly elderly patients, to embrace digital solutions. Finally, digital solutions should be perceived as facilitators of in-person communication between patient and practitioner.

This review offers elements for scoping digital solution based on feasibility studies on limited level of evidence or still limited numbers of patients evaluated on RCT.

Several clinical studies have already demonstrated reliability, feasibility, and clinical value (various symptoms, QOL, and OS) with efficacy of ePRO collection through digital solutions. The ideal digital solution in the setting of supportive care in oncology would present with the following characteristics (Fig. 4): it would be user-friendly, intuitive, and engaging to meet the immediate needs of the end-users; it would also be efficient at processing and delivering relevant information to provide supportive care as its principal aim. In thinking about its place in the supportive care setting, the ideal digital solution is not intended as a replacement for the practitioner; rather, its intended value would be in providing additional information that is appropriate to the care of the patient and the specific issues associated with their disease in real time. This information would be sufficiently detailed but not overcomplicated and presented in a language the patient understands in order to be accessible by the patient for effective symptom self-management [90]. The digital solution would maintain existing expectations regarding patient confidentiality and data privacy [91], cybersecurity, compliance with regulatory requirements, and being updated according to the most recent evidence-based practice. It would be operational throughout the entire course of the disease and for all anticancer treatments. Its built-in flexibility would enable adaptation of the digital tool to all territories, institutions, and centers and to all different care needs according to whether treatment is delivered in the community or at a regional center, such that it also serves patients who live in remote areas. It would be customizable to adapt to the needs of the individual patient. It would have a seamless connection with HCPs’ systems. Integration with patients’ EHR would allow for rapid follow-up and intervention as appropriate by HCPs in response to system alerts triggered by patient reports of clinically relevant events. It would have a high level of acceptance both by HCPs and patients, allowing its complete adoption and full integration in the patient pathway and in routine clinical practice. For digital solutions with proven clinical and cost benefits, reimbursement policies would be in place to ensure availability for implementation through defined market access programs. Finally, the ideal digital solution would not only provide the means for patient self-management of anticancer treatment-related symptoms but would also provide psychosocial support and improve QOL. Although a single system would not be able to address all needs – treatment adherence, symptom management, alignment with guidelines, medication reminders, medical and nutritional information, resources for social support, and coping strategies – it is important that digital tools find common ground, with solutions offered to address key challenges in the setting of supportive care in cancer.

Editorial and medical writing assistance was provided by Iratxe Abarrategui, PhD, CMPP, Aptitude Health, The Hague, The Netherlands, and Howard Christian, PhD, Mediscinz Communications Limited. The authors are responsible for all content and editorial decisions for this manuscript.

This work was supported by funding from Voluntis S.A., Suresnes, France.

The authors have a medical expertise contract with Voluntis. However, they do not receive any specific remuneration in consideration of this publication.

Aapro M, Bossi P, Dasari A, Fallowfield L, Gascón P, Geller M, Jordan K, Kim J, Martin K, Porzig S. Digital health for optimal supportive care in oncology: benefits, limits, and future perspectives. Support Care Cancer. 2020;28(10):4589–4612. (DOI: 10.1007/s00520–020–05539–1). ^© 2020 The Author(s). (Reprint), licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/deed.de).

Reprinted from Supportive Care in Cancer 2020;28:4589–4612.