Review Article | Published:

Patient-reported outcomes in cancer care — hearing the patient voice at greater volume

Nature Reviews Clinical Oncology volume 14, pages 763772 (2017) | Download Citation

Abstract

Recording of patient-reported outcomes (PROs) enables direct measurement of the experiences of patients with cancer. In the past decade, the use of PROs has become a prominent topic in health-care innovation; this trend highlights the role of the patient experience as a key measure of health-care quality. Historically, PROs were used solely in the context of research studies, but a growing body of literature supports the feasibility of electronic collection of PROs, yielding reliable data that are sometimes of better quality than clinician-reported data. The incorporation of electronic PRO (ePRO) assessments into standard health-care settings seems to improve the quality of care delivered to patients with cancer. Such efforts, however, have not been widely adopted, owing to the difficulties of integrating PRO-data collection into clinical workflows and electronic medical-record systems. The collection of ePRO data is expected to enhance the quality of care received by patients with cancer; however, for this approach to become routine practice, uniquely trained people, and appropriate policies and analytical solutions need to be implemented. In this Review, we discuss considerations regarding measurements of PROs, implementation challenges, as well as evidence of outcome improvements associated with the use of PROs, focusing on the centrality of PROs as part of 'big-data' initiatives in learning health-care systems.

Key points

  • The collection of electronic patient-reported outcome (ePRO) data at the point of care is feasible and usually provides more reliable information on the patient's experience than that reported by clinicians

  • ePRO data are of research quality, and are an important element of learning health-care systems and 'big-data' initiatives

  • To maximize the utility of big data in learning health-care systems, they must include the 'patient's voice' via the incorporation of PROs into routine care

  • To date, big-data initiatives have not adequately included ePROs; this situation can be addressed through data standardization

  • The routine collection of ePRO data simultaneously improves the quality of care for patients and facilitates big-data initiatives

  • Evidence obtained over the past few years indicates that the routine collection and inclusion of ePRO data in patient care improves clinical outcomes

Main

The past decade has ushered in increased attention to patient-reported outcomes (PROs), which provide a direct measurement of patients' experiences, often via validated scales assessing patient-centred parameters that include symptom burden, mood, physical function, quality of life (QoL), or distress, among others. PROs are an attractive topic in light of contemporary trends in health-care innovation that emphasize the patient experience as a key measure of the quality of care; PROs are a critical element of person-centred, high-quality care for patients with cancer, as advocated by the US National Academy of Medicine1. Published definitions of PROs vary slightly. The FDA defines a PRO as “a measurement based on a report that comes directly from the patient about the status of a patient's health condition without amendment or interpretation of the patient's response by a clinician or anyone else” (Ref. 2). In light of the diversity of PRO measures, we favour a broad definition, recognizing that, in some situations, PROs are not even 'outcomes' (Ref. 3). A consensus-based white paper on this topic was released by the PRO Task Force of the National Patient-Centered Clinical Research Network (PCORnet) in 2015 (Ref. 4). The authors of this document recommend the FDA's definition of PRO, but also specifically recommend considering reports from a proxy (that is, a spouse, significant other, or formal or informal caregiver) “in circumstances where communication via proxy is the sole method of communication with patients” (Ref. 4).

PROs are just a component of the growing efforts to collect various types of data directly from or about patients, for example, via biosensors, home-based digital devices, such as scales and blood-pressure monitors, actigraphy measures of physical activity, nutritional diaries, or descriptive reports of patient well-being from caregivers. These patient-generated health data (PGHD) provide a more-complete view of the patient experience outside the environment of care, at a resolution beyond what can be gleaned directly by the clinician. Importantly, these data do not replace those gathered by clinicians directly or indirectly using medical tests; rather, they provide information about variables different to those that can be measured via any examination manoeuvre, laboratory test, or imaging modality. According to the PCORnet guidance, PGHD are “health-related data — including health history, symptoms, biometric data, treatment history, lifestyle choices, and other information — created, recorded, gathered, or inferred by or from patients or their designees (for example, care partners or those who assist them) to help address a health concern” (Ref. 5) (adopting the definition from the Office of the National Coordinator for Health Information Technology). Together, PROs and PGHD can be grouped into the general category of patient-centred outcomes4.

Historically, PROs were used solely as research tools. They were collected on paper as part of a research protocol, and were used largely as secondary or exploratory end points in clinical trials or observational studies. As such, these data were not typically made available to the clinical team, and had no influence on patient care. Over time, PROs have taken a prominent place in oncology research. PROs are currently used regularly as secondary end points in oncology drug trials, and sometimes even as primary end points (Box 1). Indeed, PROs are increasingly recognized as meaningful, reproducible data elements that are, in some cases, more accurate than those from direct assessments by clinicians6,7,8.

Box 1: Domains typically assessed in outcomes research in oncology

Some patient-reported outcome (PRO) instruments focus on very specific constructs, including anxiety, pain, or depression, while others assess overall health-related quality of life (QoL), and might include subdomains assessing more specific aspects thereof, such as physical, emotional, social, and functional well-being, as in the Functional Assessment of Cancer Therapy-General (FACT-G) instrument82. A tool that deserves to be mentioned is the European Organisation for Research and Treatment of Cancer (EORTC)-QLQC30 (Ref. 83), which incorporates nine multi-item scales: five functional scales, three symptom scales, and a global health and QoL scale. The PROMIS84 family of measures is also of great importance. The development of these measures was fostered by the US NIH for open-source use across clinical trials and other research projects, as a sort of 'common currency' of PRO data. Many PROMIS scales have been developed, some of which are mentioned below.

While a complete list of all PRO instruments is beyond the scope of this Review, herein we highlight some commonly used measures and tools that incorporate them.

• Emotional well-being: part of FACT-G82, EORTC-QLQC30 (Ref. 83) and the PROMIS tool for emotional distress85

• Functional well-being: part of FACT-G82

• Physical well-being: part of FACT-G82 and EORTC-QLQC30 (Ref. 83)

• Psychological distress: measured with the National Comprehensive Cancer Network Distress Thermometer52

• Satisfaction with care: measured with FAMCARE86

• Social well-being: part of FACT-G82 and EORTC-QLQC30 (Ref. 83)

• Spiritual well-being: measured with Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being (FACIT-Sp)87

• Cognitive function: part of EORTC-QLQC30 (Ref. 83)

• Role function: part of EORTC-QLQC30 (Ref. 83)

• Global health status and QoL: part of FACT-G82 and EORTC-QLQC30 (Ref. 83)

• Symptoms: measured with a variety of tools including the Edmonton Symptom Assessment System (ESAS)88, Memorial System Assessment Scale (MSAS)89, MD Anderson Symptom Inventory (MDASI)53, Brief Pain Inventory90, and the PROMIS tools for pain behaviour91, pain interference92, sexual function93, and sleep disturbance94

A prominent example of the increasing importance and influence of PROs is the COMFORT-I study of ruxolitinib, a novel targeted drug for the treatment of myelofibrosis9. In this randomized controlled trial, a PRO-based symptom score specifically designed for patients with this malignancy was one of three secondary end points, along with response duration and overall survival. In a separate analysis10 of overall QoL scores from this study, improvements were detected for patients in the ruxolitinib arm: 45.9% of patients reported a reduction of ≥50% in the total symptom score between baseline and week 24, compared with 5% of patients in the placebo arm. The reported improvements in spleen size, QoL, and related symptom scores were among the factors that led to the approval of ruxolitinib by the FDA in November 2011. PRO-based studies in patients with myelofibrosis also spawned different sophisticated analyses of symptom clusters among patients with other myeloproliferative neoplasms11,12. The approval of gemcitabine for locally advanced or metastatic pancreatic adenocarcinoma in 1996 is an important earlier example of the role of PRO-based end points in oncology drug development13. In this case, a combined clinical benefit variable, including patient-reported pain score, patient-reported use of analgesics, performance status, and weight change, was used to assess improvements in patients' experiences. The definition of clinical benefit required a sustained improvement (lasting ≥4 weeks) in at least one of these parameters, without reported worsening in the other parameters. According to this composite measure, 24% of patients receiving gemcitabine derived a benefit from therapy, compared with 5% of those receiving 5-fluorouracil (P = 0.002); the 12-month overall survival was also improved with gemcitabine (18% versus 2%)14.

Similarly, several landmark randomized controlled trials of specialist palliative-care interventions in patients with advanced-stage cancer have used PRO-based end points to assess clinical benefit15,16,17,18,19 (Table 1). The demonstrated benefits of these interventions in symptom management, QoL, and mood state have prompted revolutionary changes in clinical practice. Concordantly, early administration of palliative care along with anticancer treatments is currently a recommended standard practice for patients with advanced-stage cancer and/or a substantial symptom burden20. In one study21, PROs were used to assess outcomes for caregivers, demonstrating that early administration of palliative care had a beneficial effect for other members of the family unit.

Table 1: Randomized trials of early palliative care in patients with cancer

The increasing use of PRO measures in clinical trials has also been fostered by the creation of cooperative research groups. The European Organisation for Research and Treatment of Cancer (EORTC), for example, has had a QoL working group in place since 1980. This working group developed a scale to evaluate changes in QoL, the EORTC-QLQC30, which has since been validated extensively as an outcome measure for use in clinical trials22. Indeed, the EORTC hosts a periodic conference on this topic, and has fostered a very active working group in this area, resulting in guidelines for the development of outcome measures and in the creation of computerized adaptive testing methods for administering the EORTC-QLQC30 (Ref. 22). In the USA, similar efforts by the Alliance for Clinical Trials in Oncology cooperative group have led to the development of an electronically collected PRO (ePRO) toxicity reporting system for use in clinical trials23. Traditionally, toxicity reporting in clinical trials has been done by clinicians using the Common Terminology Criteria for Adverse Events (CTCAE) schema; the new PRO-CTCAE enables patients to self-report toxicities and adverse events at least as reliably as clinician-based reporting24. These initiatives illustrate that the use of PROs in studies involving patients with cancer has changed research practices over the past few years.

These prominent uses of PROs reflect the increasing importance of evaluating the patient experience quantitatively, although the use of PROs has been limited mainly to the realm of research and PROs are not universally perceived as a useful tool in the routine management of individual patients with cancer25. This situation is changing: a growing number of efforts to integrate PROs into routine clinical care processes are currently in place. This paradigm shift is changing the landscape of PROs in oncology, particularly after the publication in 2017 of the results of a clinical trial in which an association was observed between improvements in clinical outcomes and the use of PROs26. Indeed, PGHD, and patient-centred outcomes in general, are an increasing area of interest in cancer care27. Herein, we focus on PROs, but many of the points we discuss also apply to those other categories. In particular, we summarize the current use of PROs in the care of patients with cancer. Our aim is to highlight the remarkable progress made in PRO-related research in the past decades, and to demonstrate the central role of PROs as a tool to enhance the patient's voice in cancer care in the 21st century.

Advantages of the use of PROs

Electronic collection of PRO data is feasible and reliable. Several published studies have demonstrated that ePRO-data collection is feasible in the clinic28,29,30,31,32. In studies using ePROs conducted at the Duke Breast Cancer Clinic (Durham, North Carolina, USA), most patients (75%) felt that the ePRO system helped them remember to ask their clinicians about troublesome symptoms, and 88% said they would recommend this system to other patients29. Interestingly, investigators also reported increased patient satisfaction over time, as patients see how reporting of symptoms can influence the care that they receive. Similar results have been obtained in other experiences with ePRO-data collection. For example, investigators at the Memorial Sloan Kettering Cancer Center (New York, New York, USA) created a web-based symptom-reporting tool to collect data from patients receiving chemotherapy while they were at home between hospital visits33. The rate of monthly compliance with symptom reporting was 83%, and fell only in the month before death (to 35%). The results of an initial pilot feasibility study revealed that the vast majority of patients (96%) found the system useful and would recommend it to other patients28. Furthermore, the results of several studies suggest that ePRO-data collection is feasible even among patients undergoing very intensive therapies, such as stem-cell transplantation. For example, in a study conducted at the University of North Carolina, 100% of patients undergoing stem-cell transplantation completed PRO assessments, with 94% of them doing so electronically instead of on paper34. Similarly, many patients have begun reporting their experiences via online registry systems, many of which use surveys containing various PRO measures. By 'donating' their data to the advocacy groups that administer these registries, patients can amplify their voice and increase their visibility, which is especially important for patients with rare malignancies, via the power of crowd-sourced data collection. One prominent example is the Cancer Experience Registry, an initiative of the Cancer Support Community35, in which data from >10,000 patients and caregivers have been collected to date. Taken together, these experiences show that ePRO-data collection does not need to be burdensome. In fact, published reports suggest that patients value the opportunity to tell their story using these tools.

Evidence also indicates that ePRO-data collected as part of routine care can be of research-standard quality. In comparison with paper-based PRO-data collection, the 'gold standard' used in research studies, ePRO data were found to be equivalent and similarly complete36,37,38,39,40,41,42. For example, in a study comparing paper-collected PROs with ePRO-data36, the electronically collected responses validly reflected those provided on paper for most of the scales evaluated. The results of a large meta-analysis43 also indicate that electronic data collection methods are valid, and equivalent to paper-based data collection. In addition, data from other studies show that the same symptoms or QoL-related features tend to be under-recognized or under-reported by clinicians in comparison with patients5,44. Thus, the standard approach of many research studies to the use of QoL indicators is to rely on patient-reported data, rather than on clinician-reported data. Concerns about the feasibility of ePRO-data collection in specific populations (such as older patients or those with visual impairment) are commonly expressed, but on the basis of our experience with ePROs over the past decade, the use of a stylus and/or larger fonts generally mitigates these concerns — such that very few patients opt for the use of backup paper forms, and most have little or no difficulty completing the assessments. Indeed, the findings of a meta-analysis of 72 studies45 support prior reports indicating that paper-based and electronically based data collection methods are, at least, equivalent. Other published evidence supports these observations45,46, with most ePRO studies demonstrating high levels of feasibility and usability, high satisfaction levels, and high reliability compared to paper-based data collection, even in populations including older patients and/or patients with low levels of health literacy45,46. These issues need to be further studied, however, across different cultural contexts, disease types, and societies with different degrees of technology penetrance.

Another important aspect of ePRO-data collection is the monitoring of adverse events in clinical trials. Historically, clinicians have reported toxicity-related outcomes in anticancer drug trials; however, the results from a study in which toxicity-related data were directly reported by patients indicate that this approach improves the accuracy of reporting in some domains7. For instance, patients generally reported symptoms earlier, and more often, than clinicians7. In addition, patients' symptom reports correlated better with overall QoL scores than did clinicians' symptom reports7. Evidence from some studies suggests that nurses report the most common treatment-related toxicities more accurately than clinicians, but patient self-reporting is the most reliable modality47. Thus, the US National Cancer Institute promoted the development of the PRO-CTCAE23, a PRO-focused version of the CTCAE48 reporting system. The aim of using the PRO-CTCAE is to encourage patient-reporting of toxicity data, in an ePRO environment, rather than relying on only clinician-based reporting23,49,50. Reliable and valid toxicity data are generated using PRO-CTCAE, which is currently being applied in a variety of clinical trials and explored by the FDA.

PRO-assessment tools do not need to be burdensome; many of these instruments are short and easy to complete. For example, the National Comprehensive Cancer Network (NCCN) Distress Thermometer51 is a 0–10 item ordinal scale, with an accompanying 'problem list' of 39 questions with a yes or no answer. This tool is used in clinical practice to assess the presence of unmet health-care needs in individual patients, and to identify appropriate specialty services for them, such as financial counselling or family therapy. Patients can complete this questionnaire in a few minutes at their point of care — a procedure that compares favourably with the use of more-extensive mood-assessment tools, such as the Hospital Anxiety and Depression Scale52. At the Duke Cancer Institute clinics, the NCCN Distress Thermometer is completed by patients in the exam room, while they are waiting to be seen. The MD Anderson Symptom Inventory53 is another commonly used symptom-screening tool that includes a small number of questions and is easily completed by patients in a short period of time. Other available and commonly used PRO tools in oncology have been reviewed in detail elsewhere54.

Importantly, the feasibility of ePRO-data collection depends greatly on the approach used. When such data are not actively used to guide care, patients are less interested in completing ePROs, which they perceive might not improve communication with their cancer-care team55. Accordingly, we have found that the mechanism of integration of PRO-data collection in health-care-related processes (such as clinical workflows) directly influences data completeness and quality36. Thus, clinicians aiming to obtain reliable and complete data for research and quality-monitoring purposes must meaningfully engage patients in the process. Some clinical features, such as the provision of longitudinal care, improve the completeness of PRO-data collection. Patient-related features, such as having a supportive spouse or partner who attends clinic visits with the patient, also increase the completeness of data collection56. In both situations (supportive clinical environment and/or favourable patient-related features), this improvement is probably related to the patients being more willing to complete ePRO assessments when they can see how doing so affects the care that they receive. In addition, ePROs have proved to be useful for clinicians at the point of care during patient visits57. For example, by facilitating more-thorough and reliable collection of symptom data from patients than could be done verbally in the limited time of a face-to-face clinical encounter, ePRO data enable clinicians to better recognize unmet symptom needs and trends over time, thus facilitating tailored interventions. The optimal integration of PRO-data collection into routine health care is an area of ongoing investigation, however, and should be considered an important priority in PRO research.

In order to ensure that PRO-data collection occurs at the point of care, the technology used needs to be ubiquitous and well-integrated with the routine processes of health-care delivery. This integration can be influenced substantially by health-care policy, as illustrated by an interesting initiative conducted in 2015–2016. In early 2015, the US Center for Medicare and Medicaid Innovation launched their Oncology Care Model (OCM) programme58. In the early stages, this plan included PRO-based quality measures, which inherently required wide-scale PRO-data collection. Electronic health-record vendors and other companies soon thereafter pledged to develop software to support the participation of oncology practices in the OCM. Software vendors initially devoted substantial technology resources to designing and building ePRO systems that would comply with the OCM conditions. In 2016, however, the OCM requirement for PRO-based quality measures was dropped. Immediately, technology vendors abandoned PRO-focused development activities in order to concentrate their technology-development resources on other OCM requirements. This experience demonstrates that technology vendors develop solutions in response to external requirements. Thus, the promotion of widespread routine PRO-data collection across health-care systems could be facilitated by an external demand (for example, reimbursement requirements). New US health-care-policy initiatives, such as the 21st Century Cures Act, have reinstated some external demands for PRO-data collection; great expectations surround this new initiative. Signed into law in December 2016, this Act, among many actions, demanded the creation of the Oncology Center for Excellence at the FDA, with an emphasis on patient-focused drug development and the “measurement of the patient experience”59, as is achieved via ePRO-data collection.

ePRO-data collection leads to changes in patient care. The patient's voice is inadequately accounted for in current health-care systems60,61. Thus, the notion of collecting data directly from patients has great face value; after all, patients are 'the experts on their own experiences'. The results of studies with patient-reported versions of oncology toxicity reporting schemes show that clinicians do not reliably agree with the feelings of patients, and that they consistently underestimate the incidence and severity of symptoms8. Moreover, another study showed that the degree of symptom burden documented in medical records is an underestimate of that reported by the patients themselves62. Similarly, patients often report symptoms that clinicians would not expect, or do not routinely ask about. For example, in a study of patients with breast or gastrointestinal cancer63,64, ePRO assessments revealed that sexual concerns were prevalent in these populations (reported by 53% of patients with breast cancer and 57% of patients with gastrointestinal cancer). When patients are asked directly about their experiences, changes can be made in the care they receive, resulting in a quality improvement.

Similarly to other types of data, ePRO data can be collected, tracked, subjected to trend analysis, and meaningfully interpreted. For example, in some well-validated and commonly used QoL scales, such as the Functional Assessment of Cancer Therapy-General (FACT-G) scale, thresholds for clinical interpretation have been established65. These thresholds, sometimes called minimally important differences, are also considered in the subscales of FACT-G for the assessment of the clinical relevance of different outcomes (for example, changes in a patient's emotional well-being). Other commonly used symptom-assessment scales are already used in routine clinical practice, either on paper or electronically, to facilitate standardized symptom-assessment and longitudinal monitoring. One such scale, the Edmonton Symptom Assessment System (ESAS), is a well-validated symptom-distress screening tool that is widely used in various oncology and palliative-care settings as part of routine care to improve the quality of care delivery66,67,68. The ESAS is a straightforward symptom-screening instrument that consists of 12 questions, rated on an ordinal scale of 0–10, with results that can be easily used by clinicians to better understand their patient's experience of illness, from the standpoint of symptom burden.

An example that demonstrates the influence of ePRO-data collection in improving patients' experiences is a study carried out at Memorial Sloan Kettering Cancer Center, with results published in 2016 (Ref. 69). In this trial69, investigators randomly assigned patients with advanced-stage cancer treated with chemotherapy to either use an electronic-based symptom-reporting tool or to receive standard supportive care. More patients in the ePRO-intervention arm than in the standard-care arm had improved QoL and fewer had deterioration of QoL. In addition, patients in the intervention arm were less likely than those in the standard-care arm to require care in the emergency room. Of note, the duration of chemotherapy was longer for patients in the intervention arm than for those in the standard-care arm69. The authors hypothesized that the intervention facilitated earlier recognition of and increased attention on symptoms, thereby enabling patients to remain on treatment for longer than patients in the standard-care arm, who might have stopped treatment sooner owing to less-effective management of toxicities. Importantly, extended follow-up data from this trial26 show a median overall survival improvement of approximately 5 months for patients in the intervention arm compared with those in the standard-care arm — a greater survival benefit than that conferred by most of the drugs newly approved for the treatment of advanced-stage solid tumours in 2016 (Ref. 70). This survival improvement was observed despite the fact that participating clinicians were not given any explicit instructions regarding how they should respond to the reported information; indeed, they were given only printouts of reported symptoms at clinic visits, and nurses received electronic alerts when patients reported severe or worsening symptoms. Thus, a relatively straightforward intervention, solely requiring the collection of ePRO data from patients receiving therapy and the delivery of such data to clinicians, resulted in the improvement of several important patient-centred outcomes26,69. In another study with results published in 2017 (Ref. 71) similar improvements in symptom severity and duration were achieved using a telephone-based symptom-reporting system, coupled with nurse practitioner follow-up visits to address persistent or severe symptoms. This randomized trial involved 358 patients beginning chemotherapy at four oncology practices across the USA, who called an automated monitoring system every day to report the severity of 11 symptoms71. Only those patients in the intervention arm (n = 180) received automated self-management coaching, as well as telephone follow-up support from nurse practitioners for poorly controlled symptoms71. The intervention resulted in a 43% reduction in symptom severity for patients in the intervention arm compared with those in the usual-care arm (P <0.001)71.

The collection of ePRO data extends the clinician's ability to detect and address issues that are important for the patient. Few clinicians have the time to conduct an 80-point systems assessment questionnaire, such as the Patient Care Monitor instrument30, at each visit, but patients can complete such a questionnaire electronically in <10 minutes in the waiting room29. The resulting data can be fed directly, and in real-time, back to the treating clinician at the point of care without disrupting the clinical workflow, thereby improving the clinician's ability to detect important issues. In a study of such a system, researchers found that many patients with advanced-stage solid tumours reported problems with sexual function (57% of those with gastrointestinal cancer, and 53% of those with breast cancer)72. Notably, before ePRO-data collection was undertaken, the likelihood of clinicians asking patients about sexual issues, or of patients bringing up such a sensitive subject was low. This situation led to a local quality-improvement intervention that helped to address sexual dysfunction, in the clinic, in a way that would not have been possible without the insights gleaned from the use of ePROs. Moreover, the evidence suggests that patients report sensitive symptoms more honestly when using electronic data-collection tools than they do when using paper-based methods73.

Patient-reported data can also be prognostic of downstream outcomes, such as survival. In their most simplified form, PRO data can be used as a substitute for the prognostic data that have historically been collected by the clinician; an example is performance status. In cancer care, this prognostic variable has been known to be critical for decades, but only in the past few years has longitudinal patient-reported performance status been shown to correlate with important outcomes, including survival74. Other variables, such as patient-reported pain scores, can be used to follow response to therapy over time, and have sufficient validity to drive regulatory decision-making for new therapeutic agents, as exemplified in the aforementioned approval of gemcitabine for the treatment of pancreatic cancer13. Another example is the approval of 223Ra for the treatment of painful bone metastases in patients with prostate cancer on the basis of improvements in pain scores75,76. In addition, a consistent relationship exists between baseline patient QoL and survival, as demonstrated in an analysis of pooled data from 30 randomized trials supported by the EORTC, in which investigators demonstrated that the addition of QoL-related data to their prognostic model improved the accuracy of overall survival predictions by an additional 6% in comparison with the use of demographic and clinical factors alone77.

For ePROs to be useful at the point of care, however, they must be structured in a way that is conducive to their use by busy clinical staff. This approach requires the use of visually intuitive interfaces (for example, 'dashboards') that facilitate the easy recognition of patterns and trends in a hectic environment, and with short appointment times. With the right strategy, ePROs can augment the armamentarium and reach of the busy clinician, rather than complicate it. Unfortunately, systems that support this strategy remain scarce25. Importantly, PROs enhance, but do not replace, the assessments made by clinicians.

PROs, technology, 'big data', and learning health-care systems. Technological improvements have driven the increasing use of ePROs in oncology. When paper-based data collection was the only modality available, and the analysis of PROs required the use of complex scoring schemes that had to be calculated manually or using statistical software, PROs did not have a meaningful role in clinical care. Currently, however, tablet computers, faster processers, wireless networks, and touch screens facilitate the collection of patient-reported data at the point of care, and the immediate availability of these results in actionable formats for treating clinicians. As structured numerals, PRO-related data are similar to currency-related data: they are rapidly analysable, easy to manipulate and compute, and readily linked to other types of data. For the first time, ePROs are positioned to meaningfully inform patient care on an individual, real-time basis.

Concurrently with the increasing attention on PROs, international discussions about 'big data' and learning health systems (LHSs) have been ongoing78,79. The US Institute of Medicine (now named the National Academy of Medicine) describes an LHS as a health system that relies on an iterative innovation process “designed to generate and apply the best evidence for the collaborative health-care choices of each patient and provider; to drive the process of discovery as a natural outgrowth of patient care; and to ensure innovation, quality, safety, and value in health care” (Ref. 80). LHSs are only possible, however, if they include the collection and curation of high-quality data, amid data interoperability. PROs are an important, yet often overlooked part of these initiatives, lending richness and context to more-traditional data elements in health care, such as laboratory values, vital signs, radiographs, or biopsy findings. Thus, patient-reported data are a 'bridge' between objective data from electronic health records (EHRs) and the subjective experience of individual patients. This rich, contextual information dramatically enhances the utility of a dataset, and makes comparative-effectiveness research more meaningful3. For example, if large datasets increasingly included PRO data, comparative studies of QoL across different treatment regimens with similar clinical efficacy against the same tumour type would be feasible. Currently, such studies are rarely conducted because they require head-to-head comparison of two drugs from different companies, with the possibility of associating one agent with unfavourable outcomes, or are simply too difficult to perform because of the number of patients required and the relative rarity of certain types of cancer. To maximize the clinical value and impact of big-data innovations, these initiatives must include ePRO data81.

To accomplish the merger of traditional EHR data with PRO measurements or other PGHD, interoperable data elements must be used. Interoperable refers to the use of validated scales applied consistently across clinical practices, health systems, and EHR platforms, along with homogeneous data standardization. Unfortunately, little agreement about these issues and insufficient collaboration regarding data interoperability had been reached at the time of preparation of this manuscript. No standard guidelines exist regarding the timing and frequency of data collection, the scales required, and the optimal approach to the incorporation of PROs into clinical workflows. In the near future, progress and agreement in these areas should be encouraged.

The collection of ePRO data improves the quality of care delivered to the patient, as well as the utility of the data 'donated' by the patient in an LHS. Each interest serves the other, but the results of the experiences with ePROs conducted to date demonstrate that patients must be able to perceive that the data they contribute can affect the care they receive. The collection of large volumes of patient-reported data is not, in itself, helpful to each individual patient who reports their experience. Only when PRO data are collected in structured, 'trackable', and 'trendable' ways, and when they are fed back into the management of both individual patients and larger patient populations, will they truly become meaningful in cancer care, both in the present and in the future.

Barriers to implementation

Despite the evidence indicating that ePRO-data collection enables better estimation of patient concerns, potentially improves QoL, has prognostic value, and generates reliable data that can be incorporated into downstream analyses, ePROs are not currently used routinely in patient care. The prevalent barriers that prevent the integration of ePROs in routine patient management are largely related to two closely intertwined challenges: logistical problems and technological gaps. Regarding logistics, many competing priorities are present in a clinic for patients with cancer, such as lengthy and emotionally charged patient discussions, the interpretation of genomic tests, diagnostic dilemmas, or the creation of EHRs. The addition of a new ePRO-related stream, even if valuable, must be compelling and efficient enough to be accommodated into the current requirements without taking up the time of the clinical staff. Furthermore, clinicians must believe that PRO-data collection is more practical, valid, and reliable than the current system based on recording feedback given orally and clinician observations; technological and workflow solutions are required, and must be seamlessly adopted for such efforts to succeed. Regarding technology, the existing EHR systems have not yet integrated ePRO tools in a sufficiently robust manner to enable their easy adoption. A few vanguard practices have begun to incorporate ePRO-data collection into the routine management of patients, but most health-care facilities have not yet done so. Hence, despite the overwhelming evidence of the many benefits of ePRO-data collection described herein, the replacement of the current status quo remains a challenge. Without appropriate health policy and reimbursement measures, the transition will remain slow.

Conclusions

The collection of ePRO data can potentially improve the quality of the care administered to patients with cancer, but great effort is needed for PROs to become routinely used in clinical practice. Their implementation requires specific training of clinical staff and appropriate policies, and the use of analytical solutions designed for this purpose. In a few instances, PROs are used in routine clinical practice, but we are not aware of many centres in which this approach is common31. The most frequently reported examples tend to come from larger academically oriented practices (Box 2).

Box 2: Use of patient-reported outcomes in routine clinical practice: the example of Cancer Care Ontario

The Cancer Care Ontario (Canada) strategy95, in which patient-reported outcome (PRO) data are collected using the Edmonton Symptom Assessment System (ESAS) from all patients seen as part of routine care, is a vanguard example of the use of PROs in routine clinical practice. The ESAS is a validated tool for symptom screening and monitoring that provides useful information to clinicians about the severity of nine common symptoms occurring in patients with cancer88,96. At Cancer Care Ontario, this assessment tool is used together with symptom-management guides, and even some algorithms95, to provide actionable decision-support in the clinical setting. The proportion of patients who received screening through this scheme has steadily increased since its implementation in 2009, and clinician satisfaction with the system is high97. This system is used in Cancer Care Ontario quality-monitoring and quality-improvement schemes, to which other PROs, such as patient-reported performance-status measurements, have subsequently been added95. The challenges reported with the implementation of this system include inadequate documentation of the changes in the care of patients who reported high scores, and difficulties in tracking information and trends longitudinally. Reported keys to success include dedicated leadership (both administrative and clinical), dedicated infrastructure support, use of electronic PROs, and clinician engagement.

Clinicians seeking to implement an ePRO system in their local clinics should pay attention to several specific issues. First, ePROs must improve patient care, not hinder it — data collection must be integrated into existing clinical and documentation workflows, and into the electronic record, through mechanisms that ensure that health-care processes are enhanced, but not slowed down, by PROs. For example, a symptom-screening tool can save time for clinicians if the data can be fed into clinical documentation as a review of systems assessment, ensuring that time can be dedicated to addressing specific and relevant problems through face-to-face discussions with the patient, rather than presenting the patient with long lists of questions about issues they might actually not be having. Second, ePRO data must be actionable; that is, such data must be available at the point of care, when patients are being seen, and clear instructions must be available about what can or should be done with the results. This approach first requires the selection of an appropriate instrument to collect PROs. A common complaint about some PRO-related measures is that they are not clearly linked with an intervention that could improve them. One example is the use of specific assessments, such as symptom-assessment tools (for example, the ESAS), or disease-related modules developed for patients with a specific problem, which are more likely to result in actionable measures than the use of global QoL or well-being assessments. A corollary is that education and training might be required to ensure that clinicians are equipped to know how to respond to abnormal ePROs, and are aware of the resources available to act on the results. Indeed, the collection of ePRO data is not sufficient: an action needs to be taken on the basis of the results obtained in order to improve patient care. Third, encouraging clinicians to 'buy in' to the ePRO concept is critical. Clinicians might initially think of ePROs as another administrative burden or documentation challenge that will only slow down and increase the difficulty of performing routine procedures in the clinical setting. Thus, ensuring that clinicians are involved and engaged in the development and implementation of PRO measures is essential, and can increase their awareness that ePROs can streamline and improve care if thoughtfully integrated into the local context. Clinician engagement might require different approaches in distinct contexts; a one-size-fits-all approach to ePROs in cancer care does not yet exist.

The use of PRO data can change the care provided to patients with cancer, improve their QoL, and incorporate the patient's voice into the design of big-data initiatives. Efforts to increase the routine use of PROs in the care of patients with cancer will be warmly welcomed.

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    , , & (eds) Delivering high-quality cancer care: charting a new course for a system in crisis (National Academies Press, 2013).

  2. 2.

    US Food and Drug Administration. Guidance for industry — patient-reported outcome measures: use in medical product development to support labeling claims. Health Qual. Life Outcomes 4, 79 (2006).

  3. 3.

    , , & Assessing the value of patient-generated data to comparative effectiveness research. Health Aff. 33, 1220–1228 (2014).

  4. 4.

    et al. Reaching consensus on patient-centered definitions: a report from the Patient-Reported Outcomes PCORnet Task Force. NIH Collaboratory (2015).

  5. 5.

    , , & Patient-generated health data, white paper. HealthIT.gov (2012).

  6. 6.

    , , , & How accurate is clinician reporting of chemotherapy adverse effects? A comparison with patient-reported symptoms from the Quality-of-Life Questionnaire C30. J. Clin. Oncol. 22, 3485–3490 (2004).

  7. 7.

    et al. Adverse symptom event reporting by patients versus clinicians: relationships with clinical outcomes. J. Natl Cancer Inst. 101, 1624–1632 (2009).

  8. 8.

    et al. Reliability of adverse symptom event reporting by clinicians. Qual. Life Res. 21, 1159–1164 (2012).

  9. 9.

    et al. Effect of ruxolitinib therapy on myelofibrosis-related symptoms and other patient-reported outcomes in COMFORT-I: a randomized, double-blind, placebo-controlled trial. J. Clin. Oncol. 31, 1285–1292 (2013).

  10. 10.

    et al. A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N. Engl. J. Med. 366, 799–807 (2012).

  11. 11.

    et al. Distinct clustering of symptomatic burden among myeloproliferative neoplasm patients: retrospective assessment in 1470 patients. Blood 123, 3803–3810 (2014).

  12. 12.

    & Symptom burden in hematologic malignancies. Blood 123, 3686–3687 (2014).

  13. 13.

    et al. Patient-reported outcomes supporting anticancer product approvals. J. Clin. Oncol. 25, 5094–5099 (2007).

  14. 14.

    et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J. Clin. Oncol. 15, 2403–2413 (1997).

  15. 15.

    et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N. Engl. J. Med. 363, 733–742 (2010).

  16. 16.

    et al. Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: the Project ENABLE II randomized controlled trial. JAMA 302, 741–749 (2009).

  17. 17.

    et al. Early versus delayed initiation of concurrent palliative oncology care: patient outcomes in the ENABLE III randomized controlled trial. J. Clin. Oncol. 33, 1438–1445 (2015).

  18. 18.

    et al. Early palliative care for patients with advanced cancer: a cluster-randomised controlled trial. Lancet 383, 1721–1730 (2014).

  19. 19.

    et al. Emergency department-initiated palliative care in advanced cancer: a randomized clinical trial. JAMA Oncol. 2, 591–598 (2016).

  20. 20.

    et al. American Society of Clinical Oncology provisional clinical opinion: the integration of palliative care into standard oncology care. J. Clin. Oncol. 30, 880–887 (2012).

  21. 21.

    et al. Benefits of early versus delayed palliative care to informal family caregivers of patients with advanced cancer: outcomes from the ENABLE III randomized controlled trial. J. Clin. Oncol. 33, 1446–1452 (2015).

  22. 22.

    & Quality of life research within the EORTC — the EORTC QLQ-C30. Eur. J. Cancer 38 (Suppl. 4), 125–133 (2002).

  23. 23.

    et al. Development of the National Cancer Institute's patient-reported outcomes version of the common terminology criteria for adverse events (PRO-CTCAE). J. Natl. Cancer Inst. 106, dju244 (2014).

  24. 24.

    et al. Validity and reliability of the US National Cancer Institute's Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). JAMA Oncol. 1, 1051–1059 (2015).

  25. 25.

    Patient-reported outcomes — harnessing patients' voices to improve clinical care. N. Engl. J. Med. 376, 105–108 (2017).

  26. 26.

    et al. Overall survival results of a trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment. JAMA 318, 197–198 (2017).

  27. 27.

    & Outcomes research: integrating PROs into the clinic — overall survival benefit or not, it's worth the trouble. Nat. Rev. Clin. Oncol. 14, 529–530 (2017).

  28. 28.

    et al. Patient online self-reporting of toxicity symptoms during chemotherapy. J. Clin. Oncol. 23, 3552–3561 (2005).

  29. 29.

    et al. Feasibility and acceptability to patients of a longitudinal system for evaluating cancer-related symptoms and quality of life: pilot study of an e/Tablet data-collection system in academic oncology. J. Pain Symptom Manage. 37, 1027–1038 (2009).

  30. 30.

    et al. Validation of the Patient Care Monitor (Version 2.0): a review of system assessment instrument for cancer patients. J. Pain Symptom Manage. 40, 545–558 (2010).

  31. 31.

    , & Electronic patient-reported outcome systems in oncology clinical practice. CA Cancer J. Clin. 62, 337–347 (2012).

  32. 32.

    et al. PatientViewpoint: a website for patient-reported outcomes assessment. Qual. Life Res. 18, 793–800 (2009).

  33. 33.

    et al. Feasibility of long-term patient self-reporting of toxicities from home via the Internet during routine chemotherapy. J. Clin. Oncol. 31, 2580–2585 (2013).

  34. 34.

    et al. Feasibility of frequent patient-reported outcome surveillance in patients undergoing hematopoietic cell transplantation. Biol. Blood Marrow Transplant. 19, 450–459 (2013).

  35. 35.

    Cancer Support Community. Cancer Experience Registry. Cancer Support Community (2017).

  36. 36.

    et al. Improving health care efficiency and quality using tablet personal computers to collect research-quality, patient-reported data. Health Serv. Res. 43, 1975–1991 (2008).

  37. 37.

    et al. Automated collection of quality-of-life data: a comparison of paper and computer touch-screen questionnaires. J. Clin. Oncol. 17, 998–1007 (1999).

  38. 38.

    , , & Assessing functional status of elderly adults via microcomputer. Percept. Mot. Skills 82, 689–690 (1996).

  39. 39.

    , , & Computerized assessment of common mental disorders in primary care: effect on clinical outcome. Fam. Pract. 13, 120–126 (1996).

  40. 40.

    , , , & Electronic quality of life questionnaires: a comparison of pen-based electronic questionnaires with conventional paper in a gastrointestinal study. Qual. Life Res. 4, 21–26 (1995).

  41. 41.

    , & Evaluation of a computer interview system for use with neuro-otology patients. Clin. Otolaryngol. Allied Sci. 14, 3–9 (1989).

  42. 42.

    , , , & The Cancer Care Monitor: psychometric content evaluation and pilot testing of a computer administered system for symptom screening and quality of life in adult cancer patients. J. Pain Symptom Manage. 26, 1077–1092 (2003).

  43. 43.

    , & Equivalence of electronic and paper-and-pencil administration of patient-reported outcome measures: a meta-analytic review. Value Health 11, 322–333 (2008).

  44. 44.

    et al. Symptomatic toxicities experienced during anticancer treatment: agreement between patient and physician reporting in three randomized trials. J. Clin. Oncol. 33, 910–915 (2015).

  45. 45.

    et al. Equivalence of electronic and paper administration of patient-reported outcome measures: a systematic review and meta-analysis of studies conducted between 2007 and 2013. Health Qual. Life Outcomes 13, 167 (2015).

  46. 46.

    et al. Review of electronic patient-reported outcomes systems used in cancer clinical care. J. Oncol. Pract. 10, e215–e222 (2014).

  47. 47.

    et al. Clinician versus nurse symptom reporting using the National Cancer Institute-Common Terminology Criteria for Adverse Events during chemotherapy: results of a comparison based on patient's self-reported questionnaire. Ann. Oncol. 20, 1929–1935 (2009).

  48. 48.

    et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin. Radiat. Oncol. 13, 176–181 (2003).

  49. 49.

    et al. The association between clinician-based common terminology criteria for adverse events (CTCAE) and patient-reported outcomes (PRO): a systematic review. Support Care Cancer 24, 3669–3676 (2016).

  50. 50.

    , & Methods for implementing and reporting patient-reported outcome (PRO) measures of symptomatic adverse events in cancer clinical trials. Clin. Ther. 38, 821–830 (2016).

  51. 51.

    & Cancer distress screening. Needs, models, and methods. J. Psychosom. Res. 55, 403–409 (2003).

  52. 52.

    et al. Screening for psychologic distress in ambulatory cancer patients. Cancer 103, 1494–1502 (2005).

  53. 53.

    et al. Assessing symptom distress in cancer patients: the M. D. Anderson Symptom Inventory. Cancer 89, 1634–1646 (2000).

  54. 54.

    et al. Recommendations for incorporating patient-reported outcomes into clinical comparative effectiveness research in adult oncology. J. Clin. Oncol. 30, 4249–4255 (2012).

  55. 55.

    et al. Long-term toxicity monitoring via electronic patient-reported outcomes in patients receiving chemotherapy. J. Clin. Oncol. 25, 5374–5380 (2007).

  56. 56.

    et al. Integration of electronic patient-reported outcomes into routine cancer care: an analysis of factors affecting data completeness. JCO Clin. Cancer Informat. (2017).

  57. 57.

    , , & Patient-reported outcomes in the evaluation of toxicity of anticancer treatments. Nat. Rev. Clin. Oncol. 13, 319–325 (2016).

  58. 58.

    Center for Medicare & Medicaid Innovation. Oncology Care Model. CMS.gov (2017).

  59. 59.

    US Food and Drug Administration. Patient-focused drug development. US Food and Drug Administration (2017).

  60. 60.

    The missing voice of patients in drug-safety reporting. N. Engl. J. Med. 362, 865–869 (2010).

  61. 61.

    & Quality of life in higher resolution: the next generation of comparative effectiveness research in malignant hematology. Haematologica 98, 823–824 (2013).

  62. 62.

    , , & Agreement between patient-reported symptoms and their documentation in the medical record. Am. J. Manag. Care 14, 530–539 (2008).

  63. 63.

    et al. Patient experiences with communication about sex during and after treatment for cancer. Psychooncology 21, 594–601 (2012).

  64. 64.

    , , , & Sexual concerns in cancer patients: a comparison of GI and breast cancer patients. Support Care Cancer 18, 1179–1189 (2010).

  65. 65.

    , , , & General population and cancer patient norms for the Functional Assessment of Cancer Therapy-General (FACT-G). Eval. Health Prof. 28, 192–211 (2005).

  66. 66.

    et al. Cancer Care Ontario's experience with implementation of routine physical and psychological symptom distress screening. Psychooncology 21, 357–364 (2012).

  67. 67.

    , & Use of a modified ESAS in cancer patients: a pilot study of patient and staff experiences. Int. J. Palliat. Nurs. 13, 610–616 (2007).

  68. 68.

    , , , & A'Hern, R. The use of the Edmonton Symptom Assessment Scale (ESAS) within a palliative care unit in the UK. Palliat. Med. 12, 75–82 (1998).

  69. 69.

    et al. Symptom monitoring with patient-reported outcomes during routine cancer treatment: a randomized controlled trial. J. Clin. Oncol. 34, 557–565 (2016).

  70. 70.

    & Approvals in 2016: questioning the clinical benefit of anticancer therapies. Nat. Rev. Clin. Oncol. 14, 135–136 (2017).

  71. 71.

    et al. Automated home monitoring and management of patient-reported symptoms during chemotherapy: results of the symptom care at home RCT. Cancer Med. 6, 537–546 (2017).

  72. 72.

    , & Sexual concerns in lung cancer patients: an examination of predictors and moderating effects of age and gender. Support Care Cancer 19, 161–165 (2011).

  73. 73.

    et al. Use of tablet personal computers for sensitive patient-reported information. J. Support Oncol. 7, 91–97 (2009).

  74. 74.

    , , , & Longitudinal patient-reported performance status assessment in the cancer clinic is feasible and prognostic. J. Oncol. Pract. 7, 374–381 (2011).

  75. 75.

    et al. Alpha emitter radium-223 and survival in metastatic prostate cancer. N. Engl. J. Med. 369, 213–223 (2013).

  76. 76.

    et al. A randomized, dose-response, multicenter phase II study of radium-223 chloride for the palliation of painful bone metastases in patients with castration-resistant prostate cancer. Eur. J. Cancer 48, 678–686 (2012).

  77. 77.

    et al. Baseline quality of life as a prognostic indicator of survival: a meta-analysis of individual patient data from EORTC clinical trials. Lancet Oncol. 10, 865–871 (2009).

  78. 78.

    et al. Rapid-learning system for cancer care. J. Clin. Oncol. 28, 4268–4274 (2010).

  79. 79.

    A rapid-learning health system. Health Aff. 26, w107–w118 (2007).

  80. 80.

    , & (eds) The learning healthcare system: workshop summary (National Academies Press, 2007).

  81. 81.

    et al. Electronic patient-reported data capture as a foundation of rapid learning cancer care. Med. Care 48, S32–S38 (2010).

  82. 82.

    et al. The functional assessment of cancer therapy scale: development and validation of the general measure. J. Clin. Oncol. 11, 570–579 (1993).

  83. 83.

    et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J. Natl Cancer Inst. 85, 365–376 (1993).

  84. 84.

    , , , & The use of PROMIS and assessment center to deliver patient-reported outcome measures in clinical research. J. Appl. Meas. 11, 304–314 (2010).

  85. 85.

    et al. Item banks for measuring emotional distress from the Patient-Reported Outcomes Measurement Information System (PROMIS®): depression, anxiety, and anger. Assess 18, 263–283 (2011).

  86. 86.

    Validity and reliability testing of the FAMCARE Scale: measuring family satisfaction with advanced cancer care. Soc. Sci. Med. 36, 693–701 (1993).

  87. 87.

    , , , & Measuring spiritual well-being in people with cancer: the functional assessment of chronic illness therapy — Spiritual Well-being Scale (FACIT-Sp). Ann. Behav. Med. 24, 49–58 (2002).

  88. 88.

    , , , & The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J. Palliat. Care 7, 6–9 (1991).

  89. 89.

    et al. The Memorial Symptom Assessment Scale: an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur. J. Cancer 30A, 1326–1336 (1994).

  90. 90.

    & Pain assessment: global use of the Brief Pain Inventory. Ann. Acad. Med. Singapore 23, 129–138 (1994).

  91. 91.

    et al. Development and psychometric analysis of the PROMIS pain behavior item bank. Pain 146, 158–169 (2009).

  92. 92.

    et al. Development of a PROMIS item bank to measure pain interference. Pain 150, 173–182 (2010).

  93. 93.

    et al. Initial report of the cancer Patient-Reported Outcomes Measurement Information System (PROMIS) sexual function committee: review of sexual function measures and domains used in oncology. Cancer 115, 1142–1153 (2009).

  94. 94.

    et al. Development and validation of patient-reported outcome measures for sleep disturbance and sleep-related impairments. Sleep 33, 781–792 (2010).

  95. 95.

    et al. Population-based standardized symptom screening: Cancer Care Ontario's Edmonton Symptom Assessment System and performance status initiatives. J. Oncol. Pract. 10, 212–214 (2014).

  96. 96.

    , & The Edmonton Symptom Assessment System: a 15-year retrospective review of validation studies (1991—2006). Palliat. Med. 22, 111–122 (2008).

  97. 97.

    et al. Cancer care professionals' attitudes toward systematic standardized symptom assessment and the Edmonton Symptom Assessment System after large-scale population-based implementation in Ontario, Canada. J. Pain Symptom Manage. 51, 662–672 (2016).

  98. 98.

    et al. Effect of inpatient palliative care on quality of life 2 weeks after hematopoietic stem cell transplantation: a randomized clinical trial. JAMA 316, 2094–2103 (2016).

  99. 99.

    et al. Effect of inpatient palliative care during hematopoietic stem-cell transplant on psychological distress 6 months after transplant: results of a randomized clinical trial. J. Clin. Oncol. (2017).

  100. 100.

    et al. Effects of early integrated palliative care in patients with lung and gi cancer: a randomized clinical trial. J. Clin. Oncol. 35, 834–841 (2017).

  101. 101.

    , & Early palliative care for patients with hematologic malignancies: is it really so difficult to achieve? Curr. Hematol. Malig. Rep. 12, 300–308 (2017).

Download references

Acknowledgements

The work of T.W.LB. is supported by an American Cancer Society Mentored Research Scholar Grant (MRSG-15-185-01-PCSM) and a Cambia Health Foundation Sojourns Scholars Award.

Author information

Affiliations

  1. Duke University School of Medicine, Box 2715, Durham, North Carolina 27710, USA.

    • Thomas W. LeBlanc
    •  & Amy P. Abernethy
  2. Duke Cancer Institute, 2424 Erwin Road, Suite 602, Durham, North Carolina 27705, USA.

    • Thomas W. LeBlanc
  3. Flatiron Health, 200 5th Avenue, New York, New York 10010, USA.

    • Amy P. Abernethy

Authors

  1. Search for Thomas W. LeBlanc in:

  2. Search for Amy P. Abernethy in:

Contributions

Both authors researched data for the article, discussed the article contents, and wrote, reviewed, and edited the manuscript before submission.

Competing interests

T.W.LB. previously consulted for Flatiron Health, New York, New York, USA, and has served on an advisory board of the Cancer Support Community, Washington, District of Columbia, USA. A.P.A. is an employee of Flatiron Health.

Corresponding author

Correspondence to Thomas W. LeBlanc.

About this article

Publication history

Published

DOI

https://doi.org/10.1038/nrclinonc.2017.153

Rights and permissions

To obtain permission to re-use content from this article visit RightsLink.