Severity of Financial Toxicity for Patients Receiving Palliative Radiation Therapy

Introduction: Financial toxicity has negative implications for patient well-being and health outcomes. There is a gap in understanding financial toxicity for patients undergoing palliative radiotherapy (RT). Methods: A review of patients treated with palliative RT was conducted from January 2021 to December 2022. The FACIT-COST (COST) was measured (higher scores implying better financial well-being). Financial toxicity was graded according to previously suggested cutoffs: Grade 0 (score ≥26), Grade 1 (14-25), Grade 2 (1-13), and Grade 3 (0). FACIT-TS-G was used for treatment satisfaction, and EORTC QLQ-C30 was assessed for global health status and functional scales. Results: 53 patients were identified. Median COST was 25 (range 0-44), 49% had Grade 0 financial toxicity, 32% Grade 1, 15% Grade 2, and 4% Grade 3. Overall, cancer caused financial hardship among 45%. Higher COST was weakly associated with higher global health status/Quality of Life (QoL), physical functioning, role functioning, and cognitive functioning; moderately associated with higher social functioning; and strongly associated with improved emotional functioning. Higher income or Medicare or private coverage (rather than Medicaid) was associated with less financial toxicity, whereas an underrepresented minority background or a non-English language preference was associated with greater financial toxicity. A multivariate model found that higher area income (HR .80, P = .007) and higher cognitive functioning (HR .96, P = .01) were significantly associated with financial toxicity. Conclusions: Financial toxicity was seen in approximately half of patients receiving palliative RT. The highest risk groups were those with lower income and lower cognitive functioning. This study supports the measurement of financial toxicity by clinicians.


Introduction
Palliative radiotherapy (RT) is routinely indicated for solid tumors that have become symptomatic.Tumors warranting palliation can occur anywhere in the body, with common areas being bones, brain, spine, and chest. 1,24][5][6] Treatment of the patient necessarily entails consideration of the burden of treatmentrelated costs, which is known as financial toxicity. 7Financial burden can extend from out-of-pocket expenses, time away from work, travel and parking fees, and caregiving needs, and it is likely to be linked to health-related quality of life (HRQoL), patient well-being, and possibly patient mortality.
Palliative RT treatments are generally provided to those with incurable disease, often at the end of life when patients are physically and financially suffering. 3Understanding the patient-reported experience is now critical for drug development assessment by the Food and Drug Administration (FDA) and a major interest of the NCI. 8,9Efforts have focused on HRQoL and treatment toxicity reports. 10ittle is known about patients' experience of financial toxicity and its impact on HRQoL with palliative RT.The gap in understanding financial toxicity represents an important unmet need, since economic loss and bankruptcy are common among patients with metastatic cancer. 11espite a declining incidence of many cancers, due to an aging population, longer survival after diagnosis, and more expensive therapies, the national cost of care for cancer patients is projected to grow 30% from 2015 to 2030. 12he goal of this study was to determine the degree of financial toxicity in a population undergoing palliative RT, as well as identify relationships between financial toxicity and HRQoL in this patient group.

Patients and Procedures
A review of patients referred for palliative RT at an NCIdesignated Comprehensive Cancer Center was conducted from January 2021 to December 2022.Patients were included if they were treated with palliative intent RT for any cancer diagnosis and prior to treatment had completed a financial toxicity survey that is routinely administered in the course of clinical care.This was a retrospective study approved by the institutional review board.

Measures
Financial toxicity was determined with the FACIT-COST (COST), a validated 12-item self-report measure of financial well-being yielding scores between 0 and 44 (Table 1). 13igher scores imply better financial well-being.Financial toxicity scores were graded according to previously suggested cutoffs for the COST: Grade 0 (score ≥26), Grade 1 (14-25), Grade 2 (1-13), and Grade 3 (0). 14,15n addition, the 8-item FACIT-TS-G (Version 4) was reviewed from the first follow-up appointment to assess treatment satisfaction.To determine HRQoL, the EORTC QLQ-C30, a 30-item patient-reported outcome measure that consists of a score of global health status and functional scales, which include physical functioning, role functioning, emotional functioning, cognitive functioning, and social functioning domains (Version 3.0). 16,17Higher scores for the functioning scales and global health status denote better functioning.
Demographic, clinical, and treatment variables were extracted from the electronic medical record.Hospital admissions within the prior 12 months were determined.Socioeconomic status was further characterized by area income, defined as the median family income in the patient's census tract using 2020 Census Data from the Federal Financial Institutions Examination Council. 18

Data Analytic Approach
Summary scores for the COST and FACIT-TS-G and scores from the EORTC global health status and functional scales were computed. 19ultiple imputations by chained equations using predictive mean matching were computed and used for incomplete survey responses.Spearman's rank correlation coefficient, Kruskal-Wallis testing, and linear regressions were conducted to determine associations between COST and demographic, clinical, and patient-reported outcome (PRO) variables.
Multivariate logistic regression was done for the presence of financial toxicity (ie, COST <26), with the initial model incorporating any variable significant (P < .1)from univariate modeling.The final model was determined from backward stepwise selection using Akaike Information Criterion (AIC) corresponding to a relative likelihood of 5% and confirmed with likelihood ratio testing.Internal validation was done by generating 100 bootstrap samples with replacement, and model performance was assessed with area under the curve (c-index).Optimism-corrected performance was calculated by subtracting the mean difference between bootstrap and test performance from the c-index.
The Kaplan-Meier method was used to estimate survival from the time of survey, and Cox proportional hazards modeling was used to measure associations with overall survival (OS).

Results
In total, 53 patients were identified who had completed PRO surveys.Median age was 62 years, 51% were female, 54% white race, and 89% spoke English as their preferred language (Table 2).Insurance coverage included 49% commercial payers, 30% Medicare, and 21% Medicaid.The median area income was $98,958 (range $32,303-$190,833), and median distance to the treatment center was 11.6 miles (range 2.4-78.6 miles).
Examination of demographic factors revealed that higher income was inversely associated with financial toxicity (rho = À.45,P = .0006)(Figure 2).Having Medicare (P = .003)or private (P = .001)coverage (rather than Medicaid) was associated with less financial toxicity, whereas having an underrepresented minority background (P = .001)or having a non-English language preference (rho = .39,P = .003)was associated with greater financial toxicity.
Multivariable logistic regression for the presence of financial toxicity was done (Table 3).The final model after backward selection on AIC included higher area income (HR .80,P = .007),and higher cognitive functioning (HR .96,P = .01).Likelihood ratio testing was performed (P = .17,indicating goodness of fit).Internal validation was done with bootstrap sampling, with an initial c-index of .813,and optimism corrected c-index of .796.

Discussion
Financial toxicity and financial hardship were seen in approximately half of patients receiving palliative RT (51% and 45%, respectively).Patient-reported global health status and functional domains correlated with financial toxicity.On multivariate analysis, area income and cognitive functioning were independently associated with financial toxicity.
In the initial validation studies of the COST measure, a model was trained on 233 patients with stage IV cancer receiving chemotherapy, and validated on 367 patients with a diagnosis of thyroid cancer. 13Financial toxicity was seen in 58% of patients in both the training and validation sets, and it explained 7.6% of the variability in HRQoL.There are no known studies using the COST measure specific to the population receiving palliative RT; however, in 1 study of 63 patients who received RT for head and neck cancer, the median COST was 26.5. 20Similar to that of current study, financial toxicity was associated with area income, and those at higher risk were also more likely to miss clinic visits, require infusions, and require a feeding tube.In a researcherdesigned survey of 157 patients, RT was associated with certain aspects of financial toxicity, including time away from one's job, loss of income, and additional costs with transportation. 21Such results highlight the financial vulnerabilities of RT patients, as radiation treatment requires daily sessions over days or weeks, which poses a risk to maintaining income unique from other cancer treatments.Interestingly, cognitive functioning remained significantly associated with financial toxicity in multivariate analysis.The cognitive functioning domain pertains to concentration and memory, and the potential predictive value for financial toxicity requires further validation.
Nationally, the cost of cancer care for the last 12 months of life for patients makes up about 28% of costs overall. 12Costconscious approaches to care and financial services are particularly relevant for this group.Since there is tremendous overlap between the therapeutic goals of palliative RT and services from palliative care specialists, it is ripe for organizational integration. 22For example, the multidisciplinary clinic model has shown success in shortening radiation schedules and boosting the portion who receive palliative care at the end-of-life. 23However, in 1 study less than half of patients treated with palliative RT were also receiving palliative care. 24ble 1.FACIT-COST v2 Survey Items.Respondents Were Asked to Indicate a Response as it Applied to the Past 7 Days.For Each Question, the Allowed Responses Were 'not at all', 'a Little', 'Somewhat', 'Quite a Bit', and 'Very Much'.Copyright 2014, FACIT and The University of Chicago.FT1.I know that I have enough money in savings, retirement, or assets to cover the costs of my treatment FT2.My out-of-pocket medical expenses are more than I thought they would be FT3.I worry about the financial problems I will have in the future as a result of my illness or treatment FT4.I feel I have no choice about the amount of money I spend on care FT5.I am frustrated that I cannot work or contribute as much as I usually do FT6.I am satisfied with my current financial situation FT7.I am able to meet my monthly expenses FT8.I feel financially stressed FT9.I am concerned about keeping my job and income, including work at home FT10.My cancer or treatment has reduced my satisfaction with my present financial situation FT11.I feel in control of my financial situation FT12.My illness has been a financial hardship to my family and me Although financial toxicity is understudied in the population receiving RT, physicians are taking note of the importance of this issue.In a survey of 210 radiation oncologists, 53% were "very concerned" with treatment-related costs negatively affecting their patients. 21This may represent a missed opportunity, since there tends to be a disconnect between the patient and physician experience of cost discussions. 25But when physicians engage in communication around costs, it can result in a decrease in expenditures and improvement in medical advice adherence. 26,27One effective step to meet this need will involve education in cost discussions across levels of practice and training. 28educed access to in-person care related to the COVID-19 pandemic has exacerbated financial burden for some patients.In 1 cross-sectional study, those participating in telehealth visits were significantly more likely to be worried about their future financial problems. 29In an online survey study, anxiety about COVID-19 correlated with higher levels of financial toxicity. 30In the current study, there was no observed association between the date of the survey and COST scores.However, the entire study occurred after the beginning of the pandemic.Additionally, it is not clear how financial strain and toxicity has changed as the course of the pandemic has evolved.Serial measurements will be needed to better assess the impact of COVID-19 and provide a more accurate picture of trends over time. 313][34][35][36][37] This study reproduces findings in other research demonstrating that those from racial and ethnic minority groups are at higher risk for financial toxicity. 32,38Thus, the current study highlights the importance of measuring financial toxicity in these high-risk groups.
Cancer treatments are expensive to patients, as noted by a survey of patients who did not receive any financial assistance, in which average out-of-pocket expenditures were $708 per month. 39For patients who utilized copayment assistance programs, expenditures were lower, but they were more likely to reduce spending on basics like food and clothing (51%), borrow money or used credit to pay for medications (42%), and partially fill prescriptions (24%).Patients with multiple physical or mental health comorbidities also have more healthcare-associated costs and out-of-pocket expenditures. 40,41As a result, long-term survivors of cancer risk being faced with large medical bills, with 1 survey of 4517 patients finding 34% suffered debt due to their cancer. 11or those who file bankruptcy, healthcare outcomes and mortality are higher. 42proximately half of patients in the current study had government sponsored insurance (Medicare, Medicaid).4][45][46][47] In fact, financial toxicity is significantly greater outside the US, in countries with lower average incomes. 48While COST is widely-used, and a psychometrically validated measure of financial toxicity, its design and use maybe specific to the US healthcare context. 31ork is needed to be able to extend this measurement tool to other populations. 49ne of the strengths of this study is the socioeconomic diversity of the sample.However, a notable limitation is that the survey was administered solely in English.Additionally, all patients seen in the radiation oncology clinic had some form of health insurance, including Medicaid given on an emergency basis for those hospitalized with a new cancer diagnosis.The population of uninsured and non-English speaking patients likely represent a group at even higher risk for financial toxicity, and generalizability of the current study to those populations is not advised.In addition, the suggested grading system and validation of financial toxicity utilizing COST was done with the FACT-G and presented in abstract form. 15While convenient, the grading system has generally not used clinically.Other limitations include selection bias, response bias, and measurement bias due to the single-institutional design, the fact that the COST survey was designed from a sample of insured participants in the US, and that area income may not adequately describe the income for the studied population.Lastly, the sample size was limited by inclusion of patients at a single cancer center, and model overfitting was possible despite the attempts at internal validation.

Conclusions
In conclusion, financial toxicity and financial hardship are prevalent in the population receiving palliative radiation.The highest risk groups are those with lower income and lower cognitive functioning.This study supports the hypothesis that financial toxicity is a common and unique adversity that should be measured in clinics seeing cancer patients.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Figure 1 .
Figure 1.Distribution of (A) COST, (B) Financial Hardship (COST question 12), (C) treatment satisfaction (FACIT-TS-G), and (D) EORTC global health status/QoL and functional scales.Lower COST indicates higher financial toxicity, higher financial hardship indicates higher toxicity, higher treatment satisfaction and EORTC functional scales indicate improved quality of life.

Table 2 .
Clinical and Demographic Characteristics of Study Population.

Table 3 .
Associations Between a Lower COST (higher financial toxicity) and Clinical Variables.Lower COST (Positive Rho or Positive Beta) Correspond to More Financial Toxicity.Non-Parametric Models were Included Spearman's Rank Correlations and Kruskal-Wallis Tests.Univariate Linear Regressions were used for Parametric Modeling.Multivariate Logistic Regression was for the Presence of Financial Toxicity (Grade 1-3), with the Initial Model Incorporating any Variable Significant (P < .1)From Univariate Modeling.The Final Model was Determined From Backward Stepwise Selection Using AIC.