The psychosocial cost burden of cancer: A systematic literature review

Abstract Background and Objective Psychosocial costs, or quality of life costs, account for psychological distress, pain, suffering and other negative experiences associated with cancer. They contribute to the overall economic burden of cancer that patients experience. But this category of costs remains poorly understood. This hinders opportunities to make the best cancer control policy decisions. This study explored the psychosocial cost burden associated with cancer, how studies measure psychosocial costs and the impact of this burden. Methods A systematic literature review of academic and grey literature published from 2008 to 2018 was conducted by searching electronic databases, guided by the Institute of Medicine’s conceptualization of psychosocial burden. Results were analyzed using a narrative synthesis and a weighted proportion of populations affected was calculated. Study quality was assessed using the Ottawa‐Newcastle instrument. Results A total of 25 studies were included. There was variation in how psychosocial costs were conceptualized and an inconsistent approach to measurement. Most studies measured social dimensions and focused on the financial consequences of paying for care. Fewer studies assessed costs associated with the other domains of this burden, including psychological, physical, and spiritual dimensions. Fourty‐four percent of cancer populations studied were impacted by psychosocial costs and this varied by disease site (38%‐71%). Two studies monetized the psychosocial cost burden, estimating a lifetime cost per case ranging from CAD$427753 to CAD$528769. Studies were of varying quality; 60% of cross‐sectional studies had a high risk of bias. Conclusions Consistency in approach to measurement would help to elevate this issue for researchers and decision makers. At two‐thirds of the total economic burden of cancer, economic evaluations should account for psychosocial costs to better inform decision‐making. More support is needed to address the psychosocial cost burden faced by patients and their families.


| INTRODUCTION
The rising costs of cancer treatment and supportive care 1 are a concern for health care systems, patients, and their families. Due to the large economic burden of cancer care, it is important to have an accurate estimate of the costs associated with cancer and a good understanding of who bears those costs. 2 Cost-of-illness studies can help translate the adverse effects of diseases into dollars, which is one input to support decision-making. This information is crucial to help set future health budgets, to help aid in the allocation of scarce resources and, ultimately, supports decision-making for cancer control systems.
Generally, the economic burden of cancer care has been described as three broad categories: direct costs, indirect costs, and psychosocial costs 3 (Figure 1). Direct costs are those that include the use of resources for medical and non-medical care as well the time spent obtaining such care. Indirect costs are those that result from the loss of resources and opportunities due to cancer. Psychosocial costs have generally been defined as intangible costs associated with cancer, such as pain and suffering. 4 Psychosocial costs are entirely borne by patients and their families and are the least well understood category of the three.
The psychosocial burden of cancer refers to the myriad ways that cancer and the cancer experience impact patients and their families, including caregivers. This burden is conceptualized as resulting from impacts on four domains: physical, psychological, spiritual, and social well-being ( Figure 2). A recent study found most cancer survivors in Canada face ongoing and unmet needs related to psychological (90%), physical (80%), and practical (50%) challenges. 5 These challenges and their collective burden can affect the quality of life of patients and family members involved in their care. 5 For this reason, psychosocial costs or the psychosocial cost burden have been described as synonymous with quality of life costs. 3,6 Psychosocial costs represent the additional cost to individuals' well-being that is associated with cancer. A seminal definition describes psychosocial costs as follows: In other literature, psychosocial costs have been described as an intangible cost due to the challenges of assigning a monetary value. 8 These costs are distinct from, and in addition to, the direct costs that may be paid for using additional health care services to address psychosocial issues, for example, to see a psychologist. Psychosocial costs are also related to, but distinct, from the indirect costs borne by society due to lost productivity that may result from an unresolved symptom burden during treatment that prevents a patient from working.
These challenges with the definition of psychosocial costs have resulted in a mixed approach to measurement. Some studies approach measurement of these costs using generic assessments of quality of life that may not capture the full realm of psychosocial burden F I G U R E 1 Contributors to the economic burden of cancer (from 3 ) according to the four key domains, namely: physical, psychological, spiritual, and social well-being. While some of this impact may in fact be captured in generic quality of life assessments, it is unclear whether and to what extend the generic tools are able to capture the full breadth and impact of the psychosocial burden of illness.
There is also a varied approach to measurement of psychosocial costs. The inconsistent approach to examining these costs in the literature creates missed opportunities for leveraging these learnings to inform decision-making and resource allocation. Furthermore, despite recognition of the additional burden caused by psychosocial costs raised in the Hosdgeson and Meiners 1982 paper, the methods to advance to measurement and use of psychosocial cost data have been scare, particular in the health economics literature.
Understanding the scope and scale of the cancer burden borne by patients and families is important to properly and accurately measure the overall economic burden of cancer to society. It is also required if we are to take seriously our commitments and value placed on patient-centered care within the cancer control system. Moreover, this information is necessary to inform decisionmaking around best models of care to support patients and the patient experience throughout the cancer trajectory. It is also critical for monitoring the potential impact-even inadvertent-of shifts in cost out of the system that may result in an economic burden for patients.
The objective to this review was to answer the following research In this review, we drew on the definition of psychosocial burden from the Institute of Medicine's conceptual framework. We aimed to understand and describe the estimated costs associated with this burden, either in monetary terms or when defined as intangible costs.
Psychosocial consequences of cancer can constitute a cost to patients and caregivers that may or may not have been monetized. These consequences also have potential to lead to direct economic impacts for patients and their families, which also may or may not be monetized.
This review intentionally casts its scope broadly to capture all these potential conceptualizations of psychosocial costs, aligned with the definition of psychosocial costs provided by Hodgson 1982. 7 A detailed search strategy was developed and informed by the Institute of Medicine's conceptual framework for the psychosocial burden of cancer, 9 with input from a medical librarian and content experts (see online supplement in Data S1). This framework was used as it reflects the general consensus on the main features of the psy-  The reference lists of all included papers were reviewed to identify potentially relevant papers. We also consulted experts to identify other relevant studies. Only studies in English were considered.
We aimed to include all evidence on the psychosocial cost burden associated with cancer in all populations. We included pediatric and adult cancers to understand whether there were important differences in the cost estimates, the approaches to measurement used across studies as well as the extent to which this issue was measured across different patient populations. For pediatric cancers, the caregiver's/parent's costs, reflect a main source of economic burden associated with these types of cancers. In addition, and importantly, patient and caregiver costs are ultimately borne and shared at the household-level. For these reasons, both adult and pediatric populations were considered within scope of this review.
The titles and abstracts of all retrieved papers were independently screened for eligibility by two individuals and, subsequently, the full text articles of the remaining articles, using pre-specified inclusion criteria. Articles were included if they were published between the years 2008 and 2018, were a primary research study that focused on measuring or describing the psychosocial cost burden associated with cancer, and were conducted in a population of patients at any point along the cancer care continuum (from diagnosis to palliative and end-of-life care, including survivorship). Articles were excluded if they were not written in English. Consistency in the screening process between individuals was ensured by having each person independently review between 20 and 50 papers and then compare their results with each other, discussing and resolving any discrepancies and consulting with the research team as needed.
A form was developed to guide the data extraction process. The form captured information on study characteristics (eg, jurisdiction, setting, design, data sources, or tools used) measurement/definition of outcomes, key demographic characteristics of the research population (eg, age, proportion female, cancer site, point of care along the cancer care continuum, perspective), dimensions of the psychosocial burden (physical, spiritual, social, and psychological) and its measurement, and key findings.
We assessed the risk of bias in the cross-sectional and cohort studies using the Ottawa-Newcastle quality appraisal tool. 10 The tool assessed study quality using a "star system," from three broad perspectives: the selection of the study population; the comparability of the study population; and the ascertainment of either the exposure or outcome of interest.
The results were first analyzed using a narrative synthesis that focused on distilling the main dimensions of the psychosocial cost burden that were the focus of the studies, how they were measured, the factors that contributed to the psychosocial cost burden, and notable differences in this burden between cancer populations. Studies were categorized according to how they conceptualized the psychosocial costs burden, including costs associated with physical, spiritual, social, and psychological well-being ( Figure 2). Subsequently, all estimates of the proportion of the study population who experienced a psychosocial cost burden, regardless of how measured, were synthesized and a weighted average of the proportion and 95% confidence intervals were calculated for each cancer type, by study population (adult and pediatric) and for all studies. The purpose of the weighted average was to describe how prevalent the psychosocial cost burden was across the published studies. Given the heterogeneity in how the costs were estimated, cost data were not synthesized.
This systematic review is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (online appendix). This was a systematic review, so ethics approval was not required. The Ottawa-Newcastle tool was used to assess the risk of bias in cross-sectional studies (20/26 studies) (Figure S1-online supplement in Data S1). There was a high risk of selection bias in 95% of the studies due to the recruitment of convenience or non-representative samples.

| Summary of included studies
There was also a high risk of bias related to the measurement of outcomes in 76% of studies, as most studies did not conduct a blinded measurement of outcomes or use a gold standard or validated tool. However, most studies completed appropriate analyses that adjusted for potential confounding variables, resulting in 75% of studies being assessed as having a low risk of bias in the comparability of results. Taken together, the overall quality of the evidence in this review was assessed as being low.
However, as the focus of this study was on understanding the state of the literature on this topic, the evidence is still deemed to provide useful insights to address the main research questions.

| What is the psychosocial cost burden associated with cancer and how is it measured?
There were 24 of 25 studies that focused on the social dimension of the psychosocial burden of cancer. The focus of these studies was on measuring the prevalence of negative financial consequences or impacts associated with the private (ie, out-of-pocket) costs of cancer.
There was inconsistency in the terminology, definitions, and measurement of outcomes related to financial consequences associated with cancer. Most studies measured the negative consequences to patients and families in terms of the financial burden (n = 10), financial strain or stress (n = 4), or financial toxicity (n = 3) reported by patients and their families. There was also wide variation in the tools used to measure these outcomes, with no gold standard tool acknowledged and few validated tools used to measure the outcomes ( Table 2). The cost to patients due to social support needs or poor social functioning was also investigated in a minority of studies.
The burden on patients associated with psychological and physical consequences of cancer and their impact on patients' well-being was investigated in six studies and were described in terms of quality of life costs. In these studies, quality of life costs were explicitly conceptualized as the cost to patients of lost opportunities for social engagement reduced social functioning and the impact on their intrinsic value of health. 6,20 They were also measured as the cost to individuals due to pain, suffering, and loss of enjoyment associated with poor health due to cancer. 20 Two instruments that were used to quantify quality of life costs included the SF-12 35 (Table 2).
There were no studies that measured the psychosocial cost burden of cancer in a holistic way, accounting for all four of the dimensions, and none of the studies explicitly measured the costs associated with the spiritual dimension of this burden.

| What is the prevalence and impact of the psychosocial cost burden?
The proportion of the study populations reporting psychosocial costs associated with cancer varied between studies, from 16% of patients reporting financial distress in a study of individuals with breast, esophageal, kidney, lung, pancreatic, prostate, sarcoma, testicular, or uterine T A B L E 2 Outcomes and tools used to measure the psychosocial cost burden of cancer cancers 11 to as high as 71% of patients reporting either a significant financial burden or financial toxicity associated with treatment for colorectal cancer 28 and multiple myeloma. 19 There was a lower prevalence of this cost burden found in the studies of pediatric populations, ranging from 27% 12 to 35% 16 of study populations affected. As most studies were conducted in the United States, where the majority of individuals are not covered under a publicly funded health insurance system and often face significant out-ofpocket costs associated with cancer, we looked at the prevalence of the cost burden in non-US studies and found the estimates ranged between 32% 30 and 47% 27 for the studies conducted in Ireland. The prevalence of this cost burden was not estimated in the two studies conducted in Canada. Figure 5 shows the results of the weighted estimates of the prevalence of the psychosocial cost burden by cancer type. The highest prevalence was found in study populations with multiple myeloma (71%, 95% CI: 54%-87%), and the lowest estimate was found in studies that included multiple cancers types but predominately focused on breast, colorectal, lung, and prostate cancers (38%, 95% CI: 30%-47% tions is to adopt a health care system perspective to best inform health system decision-making. 48 However, this review highlights the importance of adopting a broader societal perspective in economic evaluations to account for the significant psychosocial impact that is associated with cancer and borne by patients and their families. A recent systematic review of health interventions that included the patient perspective found no studies included intangible or psychosocial costs, 49 further supporting that these costs are not routinely incorporated. Economic evaluations that account for the psychosocial burden associated with cancer should also be an input to support informed decision-making, as this information will allow decisionmakers to weigh up the true value and cost-effectiveness of new cancer interventions.

| Study limitations
This study has limitations. We only included studies published in the last 10 years to ensure that, to the extent possible, the costs and policy contexts reflected contemporary circumstances. We also limited our inclusion criteria to English studies only. It is possible that we missed relevant studies; however, key experts were engaged to ensure seminal papers were not missed. The study quality of most studies was low. This limits the extent to which conclusions can be drawn but emphasizes the need for a research agenda to improve the measurement, breadth, and quality of studies on the psychosocial costs of cancer across settings.

| CONCLUSION
The prevalence of the psychosocial cost burden found in this review supports a need for greater attention to address these costs, account for them in the context of economic evaluations and improve the psychosocial support programs available to cancer patients and their families. Elevating the importance of psychosocial costs is needed to drive efforts to develop more routine approaches to measurement as well as efforts to use these data to support decision making to lessen the economic burden faced by cancer patients and their families. Advancing our understanding of the psychosocial costs of cancer will help to meet commitments to patient-centred care and research by improving our understanding of the full breadth of costs borne by patients and areas where supportive care can best mitigate this burden.

ACKNOWLEDGEMENTS
This review was conducted by the Canadian Partnership Against Cancer (the Partnership). The Partnership is the steward of the Canadian Strategy on Cancer Control which provides a roadmap to improve equity in the cancer system and to deliver world-class cancer care to everyone in Canada, while focusing on a sustainable healthcare system for the future. The results of this review will inform efforts to strengthen supports available to cancer patients, aligned with the key priorities and actions outlined in the Strategy.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.