Identification of studies

Treatment burden is a relatively recent concept, and it was considered unlikely at the outset that many studies would explicitly use the terms “treatment burden” or “burden of treatment” in the title, abstract, or index. To overcome this challenge, treatment burden theoretical models [4, 7, 13, 26], conceptual reviews [3, 6, 13, 18, 27], and measurement scales [16, 17, 28] were examined to identify examples of “work” or self-management behaviours that could be incorporated into database searches. A database search strategy (S1 Appendix) was devised in consultation with a senior medical librarian that encompassed terms relating to cancer, self-management behaviours, and qualitative research. Several iterations of the search strategy were tested, and search terms were refined to return studies relevant to the review question. Medline, CINAHL and EMBASE databases were searched from the year 2000 onwards. Cancer treatments have evolved over time, and it was important that patient experiences reflected modern cancer treatment pathways.

Inclusion and exclusion criteria are summarised in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Inclusion and exclusion criteria.

https://doi.org/10.1371/journal.pone.0286308.t001

Titles and abstracts were imported into Proquest Refworks (https://refworks.proquest.com/) and duplicates removed. Titles and abstracts were reviewed independently by two authors (RA, JC, AK, EY, and VV). Full texts were retrieved for all studies which were judged by at least one author to be potentially eligible for inclusion. Full texts were reviewed independently by at least two authors and any disagreements were resolved by discussion with a third author.

Data extraction and synthesis

A data extraction form was designed in Microsoft Word to capture descriptive data about the paper, the key themes, messages, and main findings described by the original study authors. Verbatim participant quotations about treatment burden were also extracted. Data extraction was performed independently by at least two authors, and three authors (RA, LD, and AK) met to agree the final data extraction form for each study. A unified document containing the data extracted from all studies was circulated to the whole review team.

Study characteristics were summarised using descriptive statistics. Thematic synthesis [29] was used to synthesise qualitative study data. All authors met to discuss key findings and concepts that were present across multiple studies. Codes (short descriptive terms) were agreed, which represented meaningful concepts present across the studies. The lead author coded the extracted study data in NVivo version 12, noting similarities, differences, and relationships between the codes. Analytical themes were generated, discussed, and refined within the whole review team. The studies were also examined for any similarities or differences in treatment burden between different cancer types. Analytical themes (similar to third order interpretations in meta-ethnography [30]) remained true to the findings of the original studies but went beyond the original research to generate additional insights into concepts of treatment burden after cancer.

Quality assessment

Study quality assessment was undertaken independently by two authors (JC and EY) using a modified CASP checklist [31]. Areas of uncertainty or disagreement were resolved by discussion with a third author (LD). Studies were not excluded based on quality assessment.

Ethical considerations

No participant quotations were used in this review from any study without an explicit statement of ethical approval, and all quotations used in this review were published in a peer reviewed journal and already in the public domain.

Characteristics of included studies

Descriptions of the 48 original qualitative research articles, their study population and scope are given in S2 Table. Studies included individuals with breast cancer (n = 19, 39.6%) [32–50], colorectal cancer (n = 13, 27.1%) [51–63], lung cancer (n = 5, 10.4%) [64–68], and prostate cancer (n = 2, 4.2%) [69, 70]. Nine studies (18.8%) included individuals with a mix of cancer types [71–79]. The studies were conducted in the United States (n = 13, 27.1%) [33–35, 38, 45, 46, 48, 49, 63, 67, 72, 73, 79], UK or Ireland (n = 11, 22.9%) [36, 41, 51–53, 56, 71, 75–78], Canada (n = 7, 14.6%) [39, 50, 54, 61, 65, 69, 70], Taiwan (n = 4, 8.3%) [42, 57, 58, 66], Sweden (n = 4, 8.3%) [32, 55, 60, 74], China (n = 2, 4.2%) [37, 68], and Germany (n = 2, 4.2%) [40, 59]. The remaining five studies were conducted in Australia, Turkey, Thailand, Denmark, and Indonesia (each n = 1) [43, 44, 47, 62, 64].

The studies included 1250 participants (median 16, range 3 to 178 participants), and collected qualitative data through interviews (n = 33, 68.8%) [36–42, 44, 46–50, 52, 53, 55–58, 60–62, 64–68, 72–75, 77, 78], focus groups (n = 6, 12.5%) [32, 33, 45, 51, 54, 59], free text written responses (n = 1, 2.1%) [63], or a combination of approaches such as interviewing and focus groups, integration of diary, drawing methods, and free text responses (n = 8, 16.7%) [34, 35, 43, 69–71, 76, 79].

Analytical themes

The analytical themes and their relationships are summarised in Fig 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Analytical themes and relationships between them.

https://doi.org/10.1371/journal.pone.0286308.g002

Theme one: The work of cancer and cancer management

Cancer management was compared to or described as “work” by participants. The wide-ranging work of cancer management is summarised in Table 2. The cancer experience was time-consuming and included cognitive work, practical work, and relational work. Participant quotations exemplifying the work of cancer management are presented in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. The cognitive, practical and relational work of living with and beyond cancer.

https://doi.org/10.1371/journal.pone.0286308.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Exemplar quotations illustrating the analytical themes.

https://doi.org/10.1371/journal.pone.0286308.t003

The notion of cognitive work, or time spent thinking about cancer management, was present across most studies. Cancer was always present in the background, and effort was expended trying to engage in other meaningful activities to distract from cancer or to maintain a “positive mental attitude”.

Cognitive work also involved sense-making activities, such as reading about cancer or its treatment and assimilating personal medical records. Participants in several studies described additional time spent planning their activities to account for problematic symptoms or potential complications, for example, thinking about the location of toilets for those with unpredictable bowel symptoms or incontinence, or carrying medicines with them. Financial challenges associated with or caused by cancer and its treatment also required significant planning and administrative effort.

Individuals with cancer engaged in the practical work of attending appointments, taking medicines regularly, and managing symptoms. Examples included using wraps or manual lymphatic drainage to manage lymphoedema or making up weekly pill boxes to improve medication adherence. Many tasks involved a combination of practical and cognitive work, for example, concealing a stoma might involve planning what to wear, and then the practical tasks of using tape to mask the stoma. Several studies described participants making efforts to avoid the topic of cancer being raised by others during social or daily activities. Individuals were motivated to conceal physical signs of their cancer treatment to maintain a sense of identity and normality, to maintain their sense of physical attractiveness, and to mitigate social stigma or judgements about having cancer.

Relational work for those with cancer involved enlisting the support of others (such as healthcare professionals, friends, and significant others) to assist with the work of health management. Many studies [34, 37, 46, 47, 49, 59, 62, 65, 69, 76, 78] noted that cancer and cancer treatment altered the dynamics of personal relationships by reducing sexual intimacy or changing roles within relationships. For example, friends and loved ones had to take on caregiving roles, and individuals with cancer struggled to fulfil the family responsibilities they had valued prior to their diagnosis (e.g., looking after children, helping with household chores). It took significant effort to maintain important relationships during cancer treatment and beyond.

Theme two: Context, motivation and reward

A recurring finding across most studies was that cancer caused significant biographical disruption. It was not only a threat to life, but also threatened the individual’s sense of identity and current way of life, altering their sense of “normality”. Fear was a common response to cancer, including fear of death and fear of recurrence in cancer survivors. Cancer was associated with loss, including loss of physical functioning, loss of independence/increased reliance on others, loss of meaningful activities, financial loss, and loss of social relationships or changed relationship dynamics (exemplar quotations are in Table 3). Fear, the wish to prolong life or prevent recurrence, and the desire to maintain or re-establish personal identity/sense of normality were strong motivators for participation in the work of cancer management (see also Table 2).

Theme three: Resource management

Individuals used personal and healthcare system resources to manage cancer. Key personal resources were time; finances; existing knowledge and skills (such as communication skills, or the ability to find, assimilate, and appraise health information); and social networks. Most studies emphasised the importance of social support for psychological well-being and for delegating certain health care tasks (exemplar quotations are in Table 3) [34, 37, 42, 44–47, 49, 54, 60, 61, 63, 65, 66, 69, 76, 79].

Another less tangible personal resource was “energy”, which was a concept that was present across several studies [37, 42, 50, 63, 65, 66, 72, 79]. Energy was treated as a resource which had to be carefully managed. For example, one individual with problematic bowel symptoms and a stoma described the need to balance food intake to maintain energy, whilst trying to control the timing of bowel movements [63]. Others managed energy levels by pacing activities or altering routines/avoiding certain tasks (see exemplar quotation, Table 3).

Healthcare system resources that were particularly valued were healthcare professional time, expertise, supportive professional relationships, and the provision of clear and tailored information. Supportive contact with healthcare professionals was reassuring.

Theme four: Treatment burden as resource mismatch

Individuals were burdened by health care when there was a mismatch between the need for a specific resource and the availability or accessibility of that resource. The most prominent negative consequence of treatment workload (see Table 2 and exemplar quotations, Table 3) was that time expended undertaking health care work took away time available for meaningful family or personal activities/time spent with support networks [34, 46, 64, 68, 73].

Several studies [34, 44, 47, 48, 67, 75] discussed the impact of financial resources being depleted due to cancer. Paying for cancer-related expenses meant that some individuals were struggling to afford necessities, were in debt, had lost savings, or were experiencing worry. Balancing finances and managing material resources took time and energy (exemplar quotations, Table 3).

Healthcare services contributed to burden when they were difficult to access or navigate. Those from minority ethnic groups [34, 35] or experiencing language barriers, and individuals dealing with economic hardship [48, 75] faced particular barriers to accessing healthcare services (exemplar quotations are in Table 3).

Healthcare services also contributed to burden when individuals perceived that their healthcare professional lacked knowledge of a particular problem or was dismissive of their concerns [39, 49, 71, 78]. The idea of abandonment by the healthcare system was mentioned by participants in a number of studies [32, 39, 54, 60, 71], particularly at transition points in care such as discharge after an operation or discharge from specialist services.

The influence of cancer type and stage

There were many similarities in treatment burden across cancer types but also some differences. Severe and persistent symptoms contributed to treatment burden. Individuals with ongoing bowel symptoms, incontinence or stomas after colorectal cancer described embarrassment and social isolation [53, 55, 60, 61, 63], resulting in limited help and support from their social networks. Stigma, embarrassment, and self-blame were prominent findings in studies of individuals with lung cancer who smoked [64, 66, 68], and there were examples of individuals becoming socially isolated due to the perception of being judged negatively by peers [68].

Individuals with stomas, incontinence, and ongoing bowel symptoms after colorectal cancer discussed the practical difficulties of integrating physical exercise into their lives, despite acknowledging this as an important aspect of survivorship care [60, 61].

The perception of “abandonment” and lack of support from the health system was mentioned frequently in studies of breast cancer survivors with lymphoedema [39, 71, 78]. There was a sense that clinicians treating breast cancer were focused on improving prognosis, lacked knowledge of lymphoedema management, and under-estimated the negative impact of lymphoedema and its management on breast cancer survivors’ quality of life.

The finding that time spent on health care related activities was a source of burden was mainly noted in studies of patients with advanced and poor prognosis cancers [46, 64, 73]. Individuals who were aware of having limited life expectancies valued meaningful activities and perceived the time they spent on health care to be more burdensome.

Quality of included studies

Most studies were rated moderately good according to the modified CASP checklist (see S3 Table). A frequent finding was that the researcher had not reflected on their role or position and how this might have influenced their research. Some studies did not give detailed descriptions of individual participant demographics, for example, few studies listed individuals’ socioeconomic status, and some quotations were not labelled with the demographic details of the individual who made the statement. Some papers provided a lot of raw data and participant quotations with little attempt to synthesise this data within the paper. Three papers [50, 63, 66] did not make an explicit statement about receipt of ethical approval.

Main findings

Living with and beyond cancer involves cognitive, practical, and relational work for patients and necessitates the use of a range of personal and healthcare system resources. Cancer management is often perceived positively, in terms of taking control, engaging in practices that could extend life or prevent recurrence, improving psychological well-being and returning to “normality”. Treatment burden occurs when resources are consumed, unavailable or difficult to access.

Chronic symptoms that persist after cancer treatment, such as bowel symptoms, lymphoedema, pain and fatigue contribute to treatment burden. Individuals who must carefully balance financial resources and those without strong social support networks or who have become isolated by their disease and its treatments are also particularly at risk of treatment burden. For those with advanced disease, time becomes more precious and is a valued resource. The consumption of time by health workload can be perceived as a burden.

Health system factors can contribute to perceptions of treatment burden when there are barriers to access, or when services do not meet individuals’ needs. Those facing language and economic barriers seem to be particularly at risk. The notion of “abandonment” at transition points in care emphasises that individuals require expert support with the work of self-management. The shift in responsibility for health care workload from professionals to patients at transition points may lead to patient burden.

Comparison with existing literature

Several theories and conceptual models have been used by researchers to better understand treatment burden across a range of conditions [4, 7, 8, 80, 81]. Our analysis involved a thematic approach that was not theory driven, however our findings do fit well with existing theory. Normalisation Process Theory (NPT) [81] identifies factors that facilitate or prevent the incorporation of interventions into everyday life and has been used to understand how patients integrate illness management work into their lives. NPT describes coherence work, relational work, enacting work and reflecting work. These categories align closely with the cognitive, relational, and practical work outlined in this review.

The Theory of Patient Capacity describes the factors that can influence a person’s ability to manage their health [80]. This includes the biographical disruption associated with illness, the mobilisation of material and personal resources, and social functioning. Additionally, the realization of work is described as a driver for feeling successful and increasing confidence, and the environment in which health is managed is cited as an important influencer of patient capacity [80]. These factors align with our themes ‘context, motivation and reward’ and ‘resource management’.

The Cumulative Complexity Model explains how the balance between workload and capacity can influence outcomes, with illness and treatment burden increasing if workload outweighs capacity [7]. This aligns with our theme ‘treatment burden as resource mismatch’. We found that managing resource challenges (e.g., making trade-offs in energy used or pain endured to perform an activity, or deciding how to use finite financial resources) was a major source of work for individuals living with and beyond cancer.

Burden of Treatment Theory models the relationship between the patient, their social networks and health services [4]. This theory is echoed across all four of our themes, with cancer patients working hard to mobilize resources and interact with both their social networks and healthcare providers to perform the tasks of health management.

One strong theme in the cancer literature which does not seem to be as strongly echoed in reports of treatment burden in other diseases is the sense of fear associated with cancer. Fear of death and fear of recurrence were prominent in most of the studies reviewed, and self-management work takes place in this context. Indeed, individuals’ perceptions of severity and threat to life in cancer may serve as strong motivators for undertaking cancer-related health management. This may be one explanation for the mainly positive perceptions of the benefits of self-management activities (Table 2).

Multimorbidity contributes to treatment burden [82] but the management of comorbidities as a source of treatment burden did not feature as a prominent finding in any of the 48 studies reviewed here. This is surprising because over 75% of people with cancer have at least one other chronic medical condition [83] and most of the studies in this review set out with broad aims such as examining “challenges”, “concerns”, and “experiences” in individuals with cancer (see S2 Table). This review focused on cancer as an index condition, and many of the participants were experiencing significant ongoing cancer-related illness burden. Participants and researchers may have chosen to focus on cancer management in isolation.

Two reviews have specifically examined multimorbidity management in individuals with cancer [84, 85]. Cavers et al. [84] found that multimorbidity could increase the complexity of medicines management. Corbett et al. [85] found that the combination of old age and multimorbidity complicated self-management after cancer, and that older individuals prioritised the management of the health condition which was having the greatest negative impact on independent living [85]. It seems likely that multimorbidities compete for and consume finite personal resources and contribute to the “resource mismatch” that could lead to treatment burden in individuals with cancer.

A key finding in this review is that consumption of patients’ time can lead to treatment burden. There is increasing recognition of “time toxicity” in cancer care, which is conceptualised as time spent coordinating care and in frequent visits to healthcare facilities [86].

Some healthcare systems have introduced patient navigation and nurse case manager roles in which a dedicated person or team helps patients to navigate complex systems, coordinates care, and helps patients to manage transitions in care [87]. Navigators and case managers can improve patient satisfaction and measures of patient experience with their care [87, 88]. Evidence about their effectiveness on reducing hospitalisation, use of emergency services and overall cost effectiveness is less clear [87, 89]. This review suggests that coordinating care and visiting healthcare facilities form only a small part of the work of managing cancer. Patients’ time is consumed by cognitive, relational, and other practical processes related to cancer management. Many of these processes take place away from healthcare settings.

Strengths and limitations

The inclusion of 48 original research studies with over 1000 individual participants from 11 different countries allowed for a detailed exploration of the subject and provided new insights into treatment burden in cancer. The synthesis goes beyond a simple aggregation or summary of existing literature, developing key analytical themes and concepts which were consistently and repeatedly identified across different cancer types and in different settings. The use of at least two independent reviewers for screening and data extraction added rigour.

There were limitations to the review. Treatment burden is a relatively new concept in cancer care, and only one study was identified that specifically set out to examine treatment burden after cancer [64]. The results of this review are derived from heterogeneous studies, with wide ranging aims and scope. There is a risk that results and participant quotations from these studies have been taken out of context. However, the use of a specific definition of treatment burden at the outset of the review, along with strict inclusion and exclusion criteria, ensured that all the included studies examined treatment burden. It was also reassuring that the same themes were present across diverse studies.

Breast and colorectal cancer were over-represented compared to lung and prostate cancer in this review. There may be specific aspects of prostate and lung cancer self-management work that have not been fully explored here. Furthermore, it was beyond the scope of this review to explore caregiver perceptions of treatment burden. Caregiver experiences are likely to give additional insights into treatment burden. This is an important avenue for further enquiry.

Implications for research and practice

Most of the resources needed to manage cancer, from finances to social networks and individual “energy”, are not fixed and are likely to change over time. The demands placed on patients and their families are also likely to change during different stages in the disease and treatment trajectory. Importantly, many of the factors identified in this review that contribute to treatment burden are potentially modifiable.

Healthcare professionals should be aware of treatment burden and should consider carefully what they are asking or expecting patients to do. There may be simple ways of reducing patient/caregiver workload, for example, conducting some consultations remotely (to reduce travel time), rationalising medication dosing regimens, and reducing the administrative burden of health care for patients. Formal, professional psychological support that addresses fear associated with cancer, and support with behaviour change could also reduce the cognitive workload of cancer self-management.

Financial toxicity during cancer care is likely to vary based on the levels of social inequality in a country, and in Government provision for health insurance and welfare benefits. It may be beyond individual health care practitioners to be able to influence welfare and economic policy, but it is essential that practitioners are aware that patients who face financial pressures during cancer treatment experience treatment burden, and that treatment burden can modify important health outcomes. Professionals should specifically ask about financial burden related to disease management and be able to sign post to relevant local resources.

A careful balance needs to be struck between the supportive, reassuring nature of frequent contact between patients and the healthcare system with the burden of time spent on health care. Data from this review suggest that individuals with distressing symptoms or symptoms that limit quality of life appreciate support from health care professionals, particularly during transition points in care, such as discharge from specialist clinic follow up. There may be an important role for planned review by specialist nurses and primary care practitioners around these transition points to minimise the sense of abandonment. Conversely, individuals with poor prognosis cancer consider time as a precious and scarce resource and professionals should consider how they might limit the demands made on patients’ time for health-related tasks. Future research might quantify the time spent on health care and the factors that influence this (e.g., rurality, multimorbidity, cancer type/stage/treatment received). Quantifying “time toxicity” could be an important step in understanding and reducing treatment burden after cancer.