Demystifying Financial Toxicity Measures: Semi-structured Interviews Comparing COST-FACIT and ENRICh among Cancer Survivors on Targeted Oral Anticancer Medications (TOAMs)

Abstract

Purpose:

Despite increasing efforts to address financial toxicity among patients with cancer, approaches to best identify patients for targeted intervention remain unclear. We compared qualitative accounts of financial toxicity among patients receiving targeted oral anticancer medications (TOAMs) to their scores in two screening tools, the COmprehensive Score for financial Toxicity-Functional Assessment of Chronic Illness Therapy (COST-FACIT) and the Economic Strain and Resilience in Cancer (ENRICh).

Methods:

We conducted mixed-methods analysis of adult patients at a comprehensive cancer center. Patients had confirmed cancer diagnoses and orders for TOAMs in the last three months. Participants completed surveys (n=180), and a subset with COST-FACIT scores indicating financial toxicity participated in semi-structured interviews (n=20). We applied the Framework Method to examine qualitative accounts and developed joint matrices to assess how well each measure identified financial toxicity.

Results:

Qualitative accounts highlighted two elements of financial toxicity, financial strain and financial worry. For participants with extreme accounts of financial strain and financial worry (n=6), COST-FACIT and ENRICh scores indicated medium to high levels of financial toxicity. For the rest (n=14), qualitative accounts of financial strain and financial worry were not completely captured by COST-FACIT and/or ENRICh.

Conclusion:

Our findings suggest a single measure may not adequately identify financial toxicity and differentiate between strain and worry. A stepwise, wholistic approach may better identify and connect patients to appropriate resources.

Implications for Cancer Survivors:

Interventions that address financial toxicity in this group may need to use more than one measures to identify and route patients to targeted intervention.

Introduction

Financial toxicity (FT) has been defined as the unplanned financial burden and distress experienced by patients due to cancer treatment,¹ and FT affects approximately 50% of cancer survivors.² The concept of FT was first introduced by Zafar and Abernathy in 2013 and focused on the patient-level impacts of objective financial burden and subjective financial distress.³ Altice and colleagues expanded the concept of FT in a 2016 systematic review. They proposed a typology dividing FT into three domains: material conditions, psychological response, and coping behaviors.² Witte et al.’s influential 2019 systematic review further unpacked FT by examining formal and self-designed measures of FT from 40 separate studies.⁴ They identified six domains of FT: active financial spending, use of passive financial resources, psychosocial responses, support seeking, coping with care, and coping with ones’ lifestyle.

Much of the existing research on FT involves patients undergoing active treatment with a foreseeable completion date. Likewise, measures used to identify patients affected by FT were developed with this group in mind. Less is known about FT experienced by cancer patients who receive Targeted Oral Anticancer Medications (TOAMs) on a continuous basis and rely on these costly drugs for extended periods with no clear end date.^5–9 These longer-term cancer treatments can have unique and devastating financial consequences.^8,9 This subset of cancer survivors often rely on ongoing cancer therapy, including TOAMs, to manage slowly progressing disease and prevent relapse.^5–8,10 Patients incur the high and increasing costs of TOAMs for extended periods of time, managing cancer as a chronic condition. Cancer survivors that remain on longer-term cancer treatment present unique challenges in this field, as definitions and measurements surrounding FT were largely developed with active treatment and a foreseeable completion date in mind.

We aim to compare qualitative accounts of FT to Comprehensive Score for financial Toxicity – Functional Assessment of Chronic Illness Therapy (COST-FACIT) and the Economic Strain and Resilience in Cancer (ENRICh) scores among patients receiving TOAMs. Our study rationale stems from three considerations. First, comparing qualitative accounts in this understudied group to scores on the COST-FACIT, the most widely used measure of FT,¹¹ allows our findings to contribute to a large and growing body of research. The COST-FACIT was developed and tested among patients with advanced cancer of various primary tumor types who were largely insured and middle class in the U.S. The COST-FACIT has been validated in 10 different languages across 12 different countries with patients at various stages of treatment.¹¹ However, it is unclear whether any patients were those who were being treated with no clear end date.

Second, our study builds on our prior work on the ENRICh measure to examine how this newer, more comprehensive tool identifies FT differently from its more established and focused predecessor. Members of our team sought to develop a measure that assessed multiple dimensions of FT and not solely the “maladaptive” behaviors (e.g., delaying care). The Economic Strain and Resilience in Cancer (ENRICh)^12–15 was designed to capture cancer patients’ levels of risk for financial strain and resilience from in-depth qualitative interviews with 62 cancer survivors.¹⁶ The original, 52-item, measure was developed and validated using data from general oncology patients with cancer staged I to IV seeking care at a comprehensive cancer center or a safety-net hospital.¹⁴ In recent years, a more practice-ready 15-item¹³ screening tool along with translated versions have been developed and validated in Spanish among patients with cancer living the U.S. and in Chinese among patients without cancer living in China.^12,17

Third, our comparison expands existing knowledge on how both measures will capture FT among cancers patients on longer-term treatment with no clear end date. For our study, we will use the terms “financial strain” and “financial worry” to describe FT. We conceptualize financial strain as the experience of financial burden (e.g., economic impact of cancer treatment such as out-of-pocket costs, dept, lost income, or percentage of income spent on healthcare) and behavioral coping (e.g., delaying care). Financial worry is conceptualized as the psychological impact of FT that could eventually lead to anxiety of distress.

Methods

Study Design and Population

We employed a sequential mixed methods approach¹⁸ by first collecting survey data on subjective FT and then collecting qualitative data to triangulate survey findings. We completed 20 semi-structured interviews, ensuring we reached thematic saturation with no new changes needed to the codes or their definitions.^19–21 We chose COST-FACIT because it is a commonly used measure of patient-reported FT in clinical settings.¹¹

All participants were recruited from a tertiary cancer center in a major metropolitan area in the southern United States from 2021 to 2023. Eligible patients were ≥18 years, fluent in English, diagnosed with cancer at any stage, and prescribed TOAMs within the last three months. Participants were identified through medical record review. Data analysis was completed in 2024.

Study Procedures

We contacted eligible patients via email, phone and/or electronic health record messaging. Consented participants completed surveys via the Research Electronic Data Capture (REDCap) data collection and management platform.²² Participants who scored ≤25 points (median score) on the COST-FACIT participated in a semi-structured interview within three months of completing the survey. The median score was selected as prior studies at the time of this project used this cutoff to identify individuals experiencing FT.^23,24 The ENRICh measure was not used for sample selection, since clear cutoffs for FT had not been established at the start of this project.

We conducted 60-minute, in-depth interviews focusing on participants’ accounts of financial strain and financial worry, here after referred to as “financial strain and worry,” related to their diagnosis and treatment as well as their access to resources (e.g., employment, support, economic status prior to diagnosis), means of coping (e.g., behavioral, stress management), and perceived financial outlook (e.g., how they think their finances will be in the future or how long they expect to be able to afford their housing, healthcare, etc…). All interview data were recorded and transcribed verbatim by members of the research team and a professional transcription service. All procedures were approved by and in compliance with the University of Texas MD Anderson Cancer Center’s Institutional Review Board.

Survey Measures

COmprehensive Score for financial Toxicity (COST-FACIT).

The COST-FACIT is a 12-item, validated measure of financial distress (Cronbach’s α ranging from 0.85 to 0.92; ICC = 0.8, with 95% CI, 0.57–0.92).^23,24 Each item lists a statement regarding the last 7 days. The respondent indicates their degree of agreement with the statement on a five-point Likert scale ranging from 0 (not at all) to 4 (very much). Negative items are reverse-scored, and the first 11 items are summed to calculate total scores ranging from 0 to 44, with lower scores indicating higher levels of FT and higher scores indicating lower levels of FT.

Economic StraiN and Resilience in Cancer (ENRICh).

The long-form ENRICh measure has 52-items and assesses four domains of FT (financial wellbeing, work and daily needs, resources and support, and other sources of help) on a 10-point Likert scale, with higher scores indicating worse FT.¹⁴ Scores on the 52-item ENRICh measure correlated with COST-FACIT (r = −0.82, P<0.001) and differed between patients from a safety-net hospital and a tertiary cancer center (4.9 vs 2.1, P < 0.001), thus demonstrating known group validity.¹⁴ The subscales were found to be reliable with a high internal consistency (Cronbach α = 0.78 to 0.94). We calculated scores and central tendencies on the original 52-item measure and the shorter 15-item measure¹³ (Table S1).

Participant Characteristics.

We collected data on patients’ self-reported age, gender, race, ethnicity, insurance coverage, employment status, and income category. We also collected patient’s cancer type and stage.

Quantitative Analysis

We conducted descriptive analyses and calculated means and standard deviations (or medians and interquartile ranges) for continuous variables and frequencies and proportions for categorical variables. We used nonparametric bootstrapping with 1,000 replications to compare the participant characteristics of the full sample with the qualitative interview sample. Quantitative analyses were conducted using Stata 18.0.

Qualitative Data Collection and Analysis

Using the Framework Method,²⁵ two members developed emergent codes and independently coded transcripts using ATLAS.ti version 23.²⁶ We used a matrix to document when data saturation was reached (rows=codes; columns=interviews) by documenting when a code was created and modified (i.e., change in definition). We created a summary table with thematic codes listed as columns and participants as rows. Each cell included a rich description of the theme for a given participant and direct quotes illustrating key elements of our description. We then developed two analytic matrices using the original definition of FT to examine participant accounts—one focused on financial strain and the other on financial worry.

In each analytic matrix, we summarized participant descriptions of any sources of financial strain or worry, the severity of these financial strains or worries reported, and any protective factors to reduce the impact of these sources of financial strain or worry. Based on these summaries, we classified participants into three groups based on severity of financial strain or worry (higher, medium, or lower). Those who were classified as having higher strain and/or worry were likely to report multiple sources of strain (e.g., job loss, housing loss) and/or worry (e.g., fear or anxiety) that was persistent. Those who were classified as having medium strain and/or worry were likely one of two possible situations: 1) reported at least one source of strain and worry that was not too high or too low or 2) had some high sources and some low sources (e.g., high material burden with a strong support system or coping mechanisms to offset the burden). Those who were classified as having lower strain and/or worry did not report any significant sources of strain and/or worry (e.g., strong health insurance with low fear or anxiety about their financial future). Two trained team members reviewed and recorded classifications for both matrices separately, and the team met to compare classifications. The first author resolved any conflicts with input from the research team, and an additional team member reviewed interviews a final time to verify classifications.

Triangulation of Quantitative and Qualitative Data to Classify Financial Toxicity.

We integrated our qualitative classifications with quantitative analyses results for each interviewed participant. Our triangulation approach assumed qualitative categorization was the most accurate depiction of FT given the first-hand, in-depth and contextualized nature of the interview accounts.

We used a color-coded heatmap informed by Abraham and Dohan’s²⁷ visualization approach, with blue hues indicating greater FT. Our heatmap summarized the qualitative classifications and quantitative survey results from COST-FACIT (full scale) and ENRICh (full scale and subscales) for each participant. We sorted participants by ascending COST-FACIT score. Lastly, we used the heatmap to examine conflicts or potential misclassification, and reviewed participant interview data to identify themes that would shed light on these discrepancies.

Results

Sample Description

We contacted 1,047 potentially eligible patients (see eligibility criteria above); 116 participants consented, and 108 completed the survey. We invited those who had a COST-FACIT score of ≤25 points for the semi-structured interviews, which resulted in 29 invitations and 20 completed semi-structured interviews. We reached thematic saturation by the 12th interview (Figure 1).

Figure 1: Study Flow Diagram.

The diagram depicts the flow of the study including eligibility, enrollment, and participation.

The full study sample (Table 1) had a median age of 62.4 years (interquartile range: 51.7 – 69.1 years). Of those, a majority identified as White (84%), married (74%), and college educated (90%). At the time of the study, slightly less than half of participants reported being employed full-time (42%) and a similar number reported being retired (42%). More than half of the patients reported annual household income of $75,000 or greater (65%) and were insured through their employer (54.6%) and/or Medicare (40.7%).

Table 1.

Description of full sample and interview subsample

	Full sample (n=108)	Interview subsample (n=20)	p-value
	n(%)	n(%)
Race and Ethnicity
Asian or Asian American	3 (2.8%)
Black or African American	11 (10.2%)	7 (35.0%)	<0.01
Hispanic origin	9 (8.3%)	3 (15.0%)	0.36
Other race or more than one race	3 (2.8%)	1 (5.0%)	0.33
White	91 (84.3%)	12 (60.0%)	0.01
Age, median (IQR)	62.4 (51.7–69.1)	60.2 (49.7–66.5)	0.34
Sex assigned at birth*
Female	48 (44.4%)	10 (50.0%)	0.653
Male	58 (53.7%)	10 (50.0%)	0.653
Gender identity*
Woman	49 (45.4%)	10 (50.0%)	0.689
Man	57 (52.8%)	10 (50.0%)	0.755
Primary cancer types
Solid	68 (63.0%)	15 (75.0%)	>0.596
Blood	40 (37.0%)	5 (25.0%)	>0.225
Educational attainment
Less than High School	2 (1.9%)	1 (5.0%)	0.468
High School or GED	9 (8.3%)	1 (5.0%)	0.469
Some College, Associates, or Trade School	32 (29.6%)	10 (50.0%)	0.038
College Degree (BS, BA)	34 (31.5%)	8 (40.0%)	0.382
More Than College	30 (27.8%)
Marital status
Married	80 (74.1%)	11 (55.0%)	0.05
Divorced	12 (11.1%)	5 (25.0%)	0.11
Separated	2 (1.9%)	2 (10.0%)	0.157
Widowed	3 (2.8%)	1 (5.0%)	0.61
Single, never been married	8 (7.4%)	1 (5.0%)	0.593
Cohabitating	2 (1.9%)
Employment status (not mutually exclusive)
Full-time (35+ hours/week)	45 (41.7%)	10 (50.0%)	0.41
Part-time (1–34 hours/week)	10 (9.3%)	1 (5.0%)	0.39
Unemployed and looking for work	1 (0.9%)	1 (5.0%)	0.31
Temporarily laid off or on leave	1 (0.9%)
Retired	45 (41.7%)	7 (35.0%)	0.36
Disability	10 (9.3%)	1 (5.0%)	0.49
Homemaker	2 (1.9%)	1 (5.0%)	0.45
Other	2 (1.9%)
Health Insurance type (not mutually exclusive)
Employer-provided	59 (54.6%)	13 (65.0%)	0.29
Individual plan+	19 (17.6%)	1 (5.0%)	0.02
Medicare	44 (40.7%)	7 (35.0%)	0.55
Medicaid or other state insurance	2 (1.9%)
Military/Tricare/VA	4 (3.7%)
Uninsured	1 (0.9%)
Other	3 (2.8%)	1 (5.0%)	0.59
Annual household income from all sources
$0 to $19,999	7 (6.5%)	2 (10.0%)	0.145
$20,000 to $49,999	18 (16.7%)	5 (25.0%)	0.119
$50,000 to $74,999	13 (12.0%)	2 (10.0%)	0.748
$75,000 to $99,999	23 (21.3%)	4 (20.0%)	0.739
$100,000 to $199,999	26 (24.1%)	5 (25.0%)	0.879
$200,000+	21 (19.4%)	2 (10.0%)	0.916
Household size supported by income*
1 (just you)	18 (16.7%)	4 (20.0%)	0.695
2 people	49 (45.4%)	5 (25.0%)	0.027
3 people	14 (13.0%)	5 (25.0%)	0.159
4 or more people	26 (24.1%)	6 (30.0%)	0.542

⁺Individual health insurance plans could represent plans purchased through the Marketplace or direct purchase through the health insurance company.

^*Discrepancies in % due to missing data.

Table 2 summarizes key measures of financial burden. The full sample had a mean COST-FACIT score of 23.6 points. Nearly 41% of our sample indicated that they were worse or much worse off financially since starting cancer treatment. Bootstrapping analyses comparing the full sample to the qualitative subsample found that the full sample was more likely to identify as White (estimated difference = 24%, SE=0.1, p=0.01, 95%CI [0.06,0.43]), less likely to identify as Black or African American (estimated difference = −29%, SE=0.1, p<0.01, 95%CI [−0.47,−0.11]), and more likely to have an individual health insurance plan (estimated difference = 13%, SE=0.06, p=0.02, 95%CI [0.02,0.23]).

Table 2.

Financial toxicity reported by full sample and interview subsample^*.

	Full sample (n=108)	Interview subsample (n=20)
	n (%)	n (%)
COST-FACIT score, mean median (IQR)a	23.6, 24.0 (15–32)	12.8, 13 (7–19)
Meets threshold for toxicity (COST-FACIT score ≤25)	62 (59.0%)	20 (100.0%)
52-Item ENRICh total, mean median (IQR)b	3.1, 2.5 (0.8–4.9)	4.2, 3.3 (2.4–6.4)
Finances Domain, mean, median (IQR)	2.2, 1.3 (0.3–3.3)	5.4, 5.8 (3.6–7.5)
Personal Resources Domain, mean, median (IQR)	2.1, 1.5 (0.5–3.2)	4.5, 3.9 (2.7–7.2)
How much has having cancer and/or treatment used up your overall finances? (standalone item on the 52-item ENRICh on a 10 point Likert scale with 0=not at all used up, 10=used up a great deal)	3.9, 3.0 (0–8)	6.4, 8 (3–9.5)
Compared to before you had cancer, how is your financial situation? (standalone item on the 52-item ENRICh)
Much better	4 (3.7%)	1 (5.0%)
Better	16 (14.8%)	2 (10.0%)
Nearly the same	36 (33.3%)	2 (10.0%)
Worse	33 (30.6%)	9 (45.0%)
Much worse	11 (10.2%)	5 (25.0%)

^*Discrepancies in % due to missing data;

^aCOST-FACIT scores range from 0 to 44 with lower scores indicating greater financial toxicity;

^b52-item ENRICh scores range from 0 to 10 with higher scores indicating greater financial toxicity.

Qualitative Classification

Table 3 presents examples of participants categorized as higher, medium/unclear, or lower financial strain or worry, based on interview data. We found a mix of accounts reflecting higher (n=6), medium/unclear (n=7), and lower (n=7) levels of financial strain and higher (n=4), medium/unclear (n=9), and lower (n=7) level of financial worry. Overall, qualitative classifications of financial strain and worry coincided with one another for nearly half of our interviewees. In the nine cases where these did not coincide, five reported higher strain and four reported higher worry.

Table 3.

Illustrative Quotes Representing the Categories of Financial Strain and Worry

Qualitative Category		Illustrative Quote
Financial Strain	Financial Worry
Higher	Higher	"so I’m just hoping once the bill gets so far up there, they won’t cancel treatment, like, you gotta pay this before we’ll see you again. That’s my biggest concern." (ID-314, COST-FACIT=15, ENRICH-52=1.1, 2.0 years since diagnosis) “I just recently did bloodwork last month. Um, and it—I mean, it was well over 18 months, probably 20 months at that rate, um, [sniff] and the reason why I waited so long was because of a financial hardship. But now that my—you know, my deductible was $7,200, I was like, “If I go in and just do the bare minimum, um, just to make sure that I’m okay because I’m just going to continue to take the medication, then I can get by another year.” (ID-1210, COST-FACIT=8, ENRICH-52=2.8, 17.8 years since diagnosis)
Higher	Medium	"I am currently just getting disability and a small pension, which makes it a little difficult to live in today’s time when everybody wants an arm or leg for, you know, the smallest amount of space." (ID-77, COST-FACIT = 9, ENRICh-52 = 6.7, 1.8 years since diagnosis)
Higher	Lower	None
Medium	Higher	"[It’s] the physical pain, the mental anguish, your financial, financial, like ruin, you know, could be imminent, you know, depending on what [health insurance provider] does one day, you know, or you could be dead because they won’t approve the CAT scan." (ID-557, COST-FACIT = 4, ENRICh-52 = 6.7, 2.4 years since diagnosis)
Medium	Medium	“My parents right now are just paying for like my bills that I have. So right now, I am really throwing away, let’s be honest, like even my collections’ bills, ‘cause when I even open those, it’s like 17 grand […]So, I mean, it’s stressful, but to be honest I throw it away at this point because I’m not working, and I don’t have income.” (ID-884, COST-FACIT = 22, ENRICh = 0.9, 3.2 years since diagnosis)
Medium	Lower	“It [insurance] covers some of them [medications]. I mean it’s gonna leave me with some copayments, so any time I get prescriptions I’ll- I’ll try to see if the VA has that type of medicine also too, ’cause that’ll—I mean it’s not free, but it’s substantially a whole lot cheaper than it would be like going like to Walgreens to get the drugs that I get from there on a daily basis […] Uh, I have a good family support system. It’s, um, mostly my oldest daughter. Everything—anything get rough, I’ll holler at her.” (ID-1258, COST-FACIT = 13.2, ENRICh = 3.6, 2.1 years since diagnosis)
Low	High	None
Low	Medium	Well, you go up to the counter, of even [Hospital A], and they say, “for your physical therapy, that’ll be a $25 copay.” Well, you know, it’s—I’m in physical therapy because of the [cancer] medicines that affected my knees or for my lymphoedema. I didn’t have lymphoedema before I had cancer. Well, you know- and you look at them I say, “I shouldn’t have to pay for this!” And it’s every time. It’s not a one off. It is every time I go and speak to that representative or whatever you call them. (ID-751, COST = 13, ENRICh = 2.8, 4.6 years since diagnosis)
Low	Low	“But since I was working, I had good insurance, and I was getting a pension from other job that I retired from. That helped me a whole bunch. But if I had been retired before, and not working it probably would- it probably would’ve hurt my finances big time because I was- what I got, when I found out this, I had a hurricane come to my house. OK, so we were working on my house trying to get a stem cell transplant. And you know it was just too much stuff. It’s two hurricanes at one time and, I mean, the stem cell in the hospital... It was just heck...But I went- I got through it. I got through it. But the only reason all the reason why, is because I had good insurance, I had, I had income. I had the income coming in.” (ID-918, COST-FACIT = 23, ENRICh = 8.2, 2.4 years since diagnosis)

Triangulation of Integrated Data

The larger number of dark blue cells across the rows in the top third of the heatmap (Figure 2) suggests that those with the lowest COST-FACIT scores (i.e., highest FT levels) generally report ENRICh scores and interview accounts at higher levels of financial strain and worry. However, the heatmap also suggests there were inconsistencies across FT measures and qualitative categorizations in the remainder of the sample.

Figure 2: Comparison of COST-FACIT, ENRICh, and Qualitative Data.

Darker shades indicate higher financial toxicity. For COST-FACIT, participants’ total score on the 12-item COST-FACIT is listed under “Total,” and their COST-FACIT score grouped into High (0–13), Medium (>13–23), and Low (>23) levels of toxicity is listed under “Categorical.” Under the 15- and 52-Item ENRICh headers, “Total” represents the averaged score from each respective index. “Overall Impact” is a standalone variable from the 52-item ENRICh under the 4th domain (Other Sources of Help); “Used up Overall Finances” is a standalone variable from 52-item ENRICh under the 4th domain (Other Sources of Help). For the Qualitative Classifications: “Financial Strain” is categorized as High, Medium, or Low need for intervention to address financial strain from patient interviews; “Financial Worry” is categorized as High, Medium, or Low need for intervention to address financial worry from patient interviews. *Missing response for Case ID: 257.

A Single Measure May Not Identify Those with Financial Strain and Worry.

Some participants had relatively low and moderate ENRICh scores despite having qualitative accounts of higher financial strain and worry and COST-FACIT scores consistent with higher FT levels. Their interviews included accounts of rationing medication and seeking assistance in accessing food and housing and described high degrees of worry, even becoming visibly distressed over the course of the interview. One such participant (a 40 year-old White, non-Hispanic woman with COST-FACIT = 8 and 52-Item ENRICh = 2.9) experienced decades of financial strain exacerbated by her diagnosis. Her financial strain due to cancer was inextricably linked to her challenging life circumstances. She managed her cancer diagnosis since late adolescence, causing her to take lower-paying jobs for decades to secure health insurance.

I am a single mom, so I’ve had to basically chase a job that offered some type of medical benefits, which, in turn, probably led me down a path where I had to take a job for lesser pay. […] I worked for a small-community, um, municipality and took a job because they offered medical benefits, getting paid probably $3 an hour less than what I could have gone out and gotten a different job […]. I do work at [small organization] currently, and [small organization] does not offer medical benefits, um, yeah, so therefore, I’ve had to do the governmental, uh, benefits that are offered [the Health Insurance Marketplace]. (ID-1210)

A few participants reported qualitative accounts of moderate financial strain or worry that were identified by ENRICh but not COST-FACIT. These participants often described a high degree of stress and disappointment surrounding their cancer costs that was difficult to manage. Some expressed uncertainty about their future financial health, despite not having financial difficulty prior to diagnosis and treatment. For instance, one participant, a 63-year-old Black, non-Hispanic man who retired from a job in the financial sector (COST-FACIT = 25 and 52-Item ENRICh = 2.45) described rising costs of Medicare as a major source of financial worry.

“So you know, you- you’re up to over $300 in just medical premiums for yourself alone, excluding prescription medication...and now I don’t have any income coming in. So, I can’t do anything to make that money grow any faster than it’s growing. That’s my biggest concern. That keeps me up at night.[…] I told Medicare when I realized that they were going to—they were going to put a subsidy on top of my premium—that they would double it, you know, […] we saved money all our lives so we could enjoy retirement. I don’t want to spend all that money.” (ID 427)

COST-FACIT and ENRICh Do Not Distinguish between Financial Strain and Worry.

By design, scores in both measures are calculated without distinction for financial strain or worry. Several participants who were correctly identified as having high or moderate levels of FT via COST-FACIT and ENRICh shared qualitative accounts consistent with low toxicity in one domain. For example, a 46 year-old Black, non-Hispanic man described a moderate level of financial strain but lower financial worry (COST-FACIT = 6 and 52-Item ENRICh = 5.5). He was placed on unpaid FMLA for two years due to his symptoms and treatment regimen, depleting his savings to meet basic needs and travel to seek treatment. However, he described various forms of behavioral coping: support seeking (Well, it is on the specialty program, but they are the ones [oncologist] that, um, told me about the copay card), and negotiating and deferring payments.

You have to make sure that you, uh, have the means to still get here and get the treatment that you need […] Maybe make a payment arrangement, maybe defer somethin’ for a while, kinda like, uh, what’s those things, robbin’ Peter to pay Paul. […] Uh, most-most of the time it’s-it’s-it’s worked out, it’s worked out. (ID 1370)

On the other hand, a few participants were relatively financially stable at the time of the interview but were having difficulty coping with the financial uncertainty brought on by their diagnosis. For example, they described significant worries about their future financial health but were not currently strained financially. This was the case for one participant (57-year-old White, non-Hispanic woman who worked in accounting, COST-FACIT = 4 and 52-item ENRICh = 6.67). She expressed gratitude as she described the coverage from her employer-based health insurance, and she also described extensive behavioral coping including budgeting and financial management. However, these resources and coping strategies were not enough to quell her sense of precarity. She noted,

It’s so overwhelming sometimes…the physical pain, the mental anguish, your financial, financial, like ruin, you know, could be imminent, you know, depending on what [health insurance provider] does one day, you know, or you could be dead because they won’t approve the CAT scan […] I guess, like it’ll be disconcerting if anything else strikes while I’m dealing with this.[…] And I’ll have a couple of good years to recuperate before this thing will maybe come back.” (ID 557)

Like this participant, several others indicated high levels of FT that were more closely tied to their sense of financial worry about the future rather than their current circumstances. They wanted information and support including retirement planning, retirement insurance planning, and more simplified, actionable cost communication from their insurers.

Discussion and Conclusion

Our study is the first to compare two quantitative measures of FT to qualitative interview data. Although we found evidence that the COST-FACIT and ENRICh likely detect FT for those experiencing very high levels of financial strain and worry, we also found that they may have limitations for patients experiencing moderate to low levels of financial strain and/or worry, as discerned in qualitative interviews.

Altice and colleagues characterized FT in three domains (material hardship, psychological burden and cost-related behavior changes).² Witte and colleagues expanded the framework to six subdomains encompassing perception and response to FT (active spending, utilization of passive financial resources, psychosocial responses, seeking support for financial assistance, coping with costs of care, and adapting lifestyle to manage financial burdens). The COST-FACIT addresses three domains from the Witte Framework (Material: financial planning; Material: financial resources; Psycho-social: affect). We found that the COST-FACIT measure identified those with financial worry and may serve as an initial screening measure. Both measures may miss financial worry associated with an uncertain financial future in that much of the prior work has been conducted with patients that have a clear end date for their treatment or have completed their treatment.^{1,2,4,14,17,23,24,28} In the case of patients receiving TOAMs, there is no clear end date to when they will stop needing their anti-cancer medications.

In comparison, the ENRICh measure addresses all six domains in Witte’s framework. However, it cannot differentiate between financial strain and worry nor the source of strain. For instance, patients with substantial pre-existing financial challenges may not distinguish cancer-related sources of financial worry when completing ENRICh. In addition, certain ENRICh domains may not be relevant to them. Domains like “Personal Resources” assume participants have access to income, credit, savings, assets, and banking and may not apply to them. The measure could include “not applicable” as a response option.

There has been a growing effort to develop and implement interventions that reduce the burden of FT over the cancer care continuum.²⁸ Research studies have initiated interventions as diverse as financial navigation, education, and cost communication.²⁸ Implementing findings from this growing body of work will require, among other things, evidence-based recommendations for selecting measures of FT. Neither the COST-FACIT nor ENRICh scales are designed to differentiate between “financial strain” (actual financial hardship) and “financial worry” (anxiety about finances). This distinction is clinically important because these two constructs require different or combined intervention approaches.

Now may be the time to re-examine these measures and critically assess what we are trying to achieve with these measures. Our findings suggest that a multi-step process is needed to 1) identify those who are worried about their financial health and/or experiencing financial strain, 2) differentiate those who would benefit from education or more tangible interventions (e.g., financial navigation) to address their financial strain, 3) implement appropriate the appropriate intervention. Thus, we would suggest a stepwise approach with Step 1: screening patients with COST-FACIT to identify those “distressed” about their finances, Step 2: administering a more comprehensive measure, such as the 52-item ENRICh measure, to identify possible sources of FT, and Step 3: offering financial navigation to clarify the type of intervention needed (e.g., education only or instrumental support). We propose to use COST-FACIT as the initial screener because it is well tested and commonly used in a variety of settings and languages.^2,4,28 The other rationale rests on our opinion that patients that are not experiencing “financial worry” may not need an intervention to mitigate FT. Additionally, if they are not worried about their finances, they may not be receptive to any intervention. The suggested approach may not be suitable for all patients as it may be important to identify the level of financial strain and worry at the time of cancer diagnosis to better support patients during the cancer journey.

Several limitations warrant discussion. First, we did not directly ask participants to clarify discrepancies between their quantitative responses during their interviews. However, our robust analytical approach with multiple team members and triangulation generated rich insights. Second, our study administered the 52-item of ENRICH, which is a longer version than what has been used in prior publications.^14,15 However, our findings support the use of the longer version to highlight the more nuanced sources of FT. While we were unable to find evidence-based cutoffs for FT in the ENRICh measure at the time of the analysis, future research could replicate these methods once these benchmarks have been established. Third, our participants’ experiences may not reflect broader U.S. patterns among patients receiving TOAMs and may contain non-response bias. We used a non-representative sampling approach, selection criteria requiring a COST-FACIT score of ≤25 from a large urban cancer center. Patients with COST-FACIT scores above 25 may have also experienced financial strain and worry in ways that were distinct or difficult to capture with our measures. Because we recruited patients who spoke English, our findings may not apply to non-English speakers, who may be at risk for financial toxicity. All our participants had or obtained sufficient resources to seek care at a large, urban cancer center. Since we did not compare characteristics of participating patients to those who declined, there may be unknown bias in our findings. Fourth, our study had low participation rates, likely due to our recruitment approach. Eligible patients were identified via electronic health record review and sent recruitment emails with a survey link. We did not recruit patients in connection with upcoming appointments, nor did we have clinical providers discuss the study with patients. Prior research suggests that in-person recruitment,²⁹ especially with a dedicated clinical recruiter,³⁰ may improve patient participation rates while mitigating burden on clinical staff. Finally, we acknowledge that any measure of FT would face substantial challenges capturing “financial worry,” as we observed qualitatively. Although financial worries affected our participants in the present, they were largely focused on the future state of their finances and costs.

In summary, our study confirms that FT is complex and multifaceted, and a single measure has limitations in identifying patients at risk for FT and triaging them to the interventions required to mitigate their FT.

Data Availability

The data supporting this study are not publicly available due to patient confidentiality and privacy. However, data may be made available upon reasonable request from the authors with appropriate regulatory processes.