Financial hardship in childhood cancer survivors treated with hematopoietic cell transplant: A report from the Childhood Cancer Survivor Study

Introduction

Advances in hematopoietic cell transplant (HCT) techniques and supportive care have resulted in an expanding population of HCT survivors. [1–2] In the United States (US), it is estimated that there will be more than a half-million long-term survivors of adult and childhood cancers treated with HCT by the year 2030, including approximately 70,000 survivors who underwent HCT at <18 years of age. [3] Survivors treated with HCT are at elevated risk for chronic health conditions and psychosocial challenges which often impact employment, health insurance status, educational attainment, and income, culminating in financial hardship. [4–10]

Financial hardship has been identified as an important social outcome in survivors treated with HCT, but there is limited evidence to understand the overall burden among childhood cancer survivors who received HCT. [11] Prevalence of financial hardship in survivors treated with HCT from prior studies has ranged from 20 to 70%, in large part due to the heterogeneity of definitions used for financial hardship. [7–8] Additionally, these studies do not provide detailed prevalence of domain-specific financial hardship such as material hardship, behavioral hardship, and psychological hardship. [12] It is known that long-term survivors of childhood cancer treated with HCT are at a higher risk of chronic health conditions compared to cancer survivors not treated with HCT (non-HCT) likely due to the higher intensity of therapy, the need for prolonged immune suppression, and the risk of graft vs. host disease (GVHD). [13] Therefore, it is essential to understand the risk of long-term financial hardship experienced by cancer survivors treated with HCT compared to cancer survivors treated with non-HCT therapy. This information may help clinicians and researchers develop targeted interventions to screen for and mitigate financial hardship.

Between 2017–2019, Childhood Cancer Survivor Study (CCSS) sent a cross-sectional survey to the participants to assess financial hardship across various domains such as material hardship or financial sacrifices, behavioral hardship, and psychological hardship. In the recently completed analysis using this survey, long-term survivors of childhood cancer were noted to have significantly higher hardship compared to sibling controls across all three domains. [14] However, this study did not highlight the differences in financial hardship for survivors who had undergone HCT. To address this knowledge gap, we used the same cohort and aimed to specifically describe the overall burden of financial hardship among adult survivors of childhood cancer who have undergone HCT and identify sociodemographic characteristics as well as treatment exposures and complications following treatment associated with financial hardship. We also compared the financial hardship in those who had undergone an HCT with both non-HCT cancer survivors and sibling controls. We hypothesized that the financial hardship scores across all three domains would be significantly higher among adult survivors of childhood cancer that have undergone HCT compared to non-HCT cancer survivors and sibling controls.

Methods

Study Design and Participants

The CCSS is a retrospectively constructed cohort study with prospective follow-up of 5-year survivors of childhood cancers diagnosed < 21 years of age between January 1, 1970, and December 31, 1999, at one of 31 institutions in the US and Canada, that includes a randomly selected population of nearest aged siblings. Description of the study design, methods, and CCSS population has been published previously. [15–16] Information about financial hardship was collected through a survey administered between 2017–2019 to CCSS survivors and siblings selected by a stratified random sampling with strata of age (<35, 35–44, 45–71 years) and primary cancer diagnosis. Survivors with primary cancer diagnosis of Wilms tumor, soft tissue sarcoma, bone cancer were excluded from this analysis because of their small sample sizes (<5) in HCT population. All participants provided informed consent. Treatment information was obtained from the survivors’ treating institution by trained medical record abstractors using a defined protocol. The protocols and questionnaires utilized by the CCSS underwent approval by the Institutional Review Boards of all collaborating centers.

Primary Outcome Measures

Financial hardship was characterized by 21 items depicted in the Appendix (20 items for the sibling version after omitting one item about cancer follow-up care) derived from several US national surveys (National Health Interview Survey, Medical Expenditures Panel Survey Experiences with Cancer, Behavioral Risk Factor Surveillance System) as described previously. [14] Additional details of the survey instrument can be found on the CCSS website (www.ccss.org). The previous CCSS analysis used factor analysis to determine that each of these items loaded highly onto one of 3 domains (material hardship or financial sacrifices, behavioral hardship, and psychological hardship) except two items not loading highly onto any of the 3 domain. [14] The material hardship or financial sacrifices domain (8 items) queried respondents about financial sacrifices resulting in reduced spending on specific items (e.g., vacation or leisure, large purchases) in the past year. The behavioral hardship domain (8 items) queried respondents regarding forgoing specific aspects of medical care (e.g., yearly visit to your primary care doctor) due to cost in the past year. The psychological hardship domain (3 items) queried respondents regarding worry or stress about having adequate money for health care and important purchases (e.g., to pay rent or mortgage). The two items that did not load highly onto a specific factor included having a history of being sent to debt collection or of filing for bankruptcy.

We also examined other financial outcomes as depicted in the Appendix, specifically: health insurance costs (e.g., annual deductible) and concerns (e.g., concerned about the ability to be able to maintain health insurance coverage) and employment concerns and work schedule changes (e.g., taking time off, changing jobs, changing to a more flexible schedule, avoiding promotion, worry about forced retirement, and remaining at a job to avoid loss of health insurance or so-called ‘job lock’). Assets (e.g., home ownership, and value) and debts (e.g., mortgages or loans on home, and associated balance, credit cards, car loans, medical provider debt, life insurance policy loans, or loans from relatives) were also assessed.

Independent Variables

Survivors’ sociodemographic, cancer diagnosis, treatment exposures, and complications after completion of therapy were evaluated as risk factors for financial hardship. Survivors’ sociodemographic characteristics included: age, sex, race/ethnicity, educational attainment, marital status, household income, health insurance status, and employment status at the time of follow-up survey completion. Survivors’ diagnosis, treatment exposures including total body irradiation (TBI) exposure and busulfan exposure, as well as complications after completion of therapy including chronic health conditions, late relapse, and development of a subsequent malignant neoplasm (SMN) were self-reported. We were unable to discriminate type of HCT (allogeneic versus autologous). Survivors’ chronic health conditions were characterized according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE criteria): mild (grade 1), moderate (grade 2), severe or disabling (grade 3), or life-threatening (grade 4). [17]

Statistical Analysis

The analysis aimed to compare financial hardship between survivors of childhood cancer treated with HCT, survivors treated without HCT, and sibling controls. The standardized score of each domain of financial hardship was calculated by adding the item responses and dividing by the standard deviation among siblings. Calculation of the standardized score in the behavioral hardship domain was adjusted to account for the fact that survivors had 8 items while siblings had 7 items. Descriptive statistics were calculated for the individual items of financial hardship within each domain (material hardship/financial sacrifices, behavioral hardship, psychological hardship) and the two individual items (sent to debt collection, filing for bankruptcy). Descriptive statistics were also generated for the other financial outcomes described above such as health insurance costs and concerns, employment concerns and work schedule changes, as well as assets and debts. Frequency counts and proportions were also depicted for the two additional items (sent to debt collection and bankruptcy).

Crude and adjusted differences in mean domain-specific standardized scores were also calculated for survivors treated with HCT compared with non-HCT survivors, as well as survivors treated with HCT compared with siblings. Specifically, financial hardship domains as well as the other financial outcomes were compared using general linear models for each domain-specific standardized score and multivariable logistic regression for the two additional binary items between survivors of childhood cancer treated with HCT and survivors treated with non-HCT therapy, adjusting for age, sex, race/ethnicity and cancer diagnosis. The outcomes were also compared between survivors treated with HCT and sibling controls, adjusting for age, sex, and race/ethnicity, using generalized estimating equations (GEE) modification of the linear model for each domain-specific standardized score and of the logistic regression model for the two additional binary items to account for potential intra-family correlation between the survivor and siblings from the same family.

The other financial outcomes (health insurance costs and concerns, employment concerns and work schedule changes, assets, and debts) were also compared using multivariable logistic regression with adjustment for age, sex, race/ethnicity and cancer diagnosis between HCT therapy survivors and non-HCT therapy survivors, and with adjustment for age, sex and race/ethnicity between HCT survivors and siblings.

To further evaluate the impact of HCT status on financial hardship across the three aforementioned domains with additional independent variables, general linear models with sociodemographic factors, cancer diagnosis, treatment exposures, and complications after treatment including chronic health conditions, late relapse, and development of a SMN were used. Adjusted mean differences and odds ratios accompanied by 95% confidence intervals (CI) were reported where applicable. A p-value <0.05 was considered statistically significant. Analyses were carried out using the SAS statistical software (SAS Institute, Cary, NC).

Results

Consort diagrams for survivors treated with and without HCT are provided in Supplemental Figure 1A, and for those for siblings in Supplemental Figure 1B. A total of 2,844 survivors (133 survivors treated with HCT and 2,711 without HCT) and 1,027 sibling controls were included in the study. Characteristics of the study population are provided in Table 1. The median age of survivors treated with HCT, without HCT, and siblings at the time of survey completion was 35 years (range: 20–61 years), 39 years (range: 19–67 years), and 46 years (range: 21–69 years), respectively. Survivors treated with HCT and non-HCT survivors were an average of 25 years (range: 18–43 years), and 32 years (range: 18–49 years) from diagnosis; respectively. A greater proportion of survivors treated with HCT (60%), compared to non-HCT survivors (36%) and sibling controls (14%), reported a severe or life-threatening chronic health condition (Grade 3 or 4), with the majority (87%) reporting multiple (≥2) chronic health conditions.

Table 1.

Characteristics of Study Participants

	HCT survivors (n=133)		Non-HCT survivors (n=2,711)		p-value*	Siblings (N=1,027)		p-value**
	No. of Survivors	%	No. of Survivors	%		No. of Siblings	%
Age at completion of questionnaires, years					<0.001			<0.001
18–30	44	33.1	400	14.8		57	5.6
30–39	48	36.1	1,008	37.2		239	23.3
>40	41	30.8	1,303	48.1		731	71.2
Median Age	35.3		39.4			46.4
Range	20.2 – 60.8		18.8 – 66.6			21.1 – 69.2
Sex					0.33			0.64
Male	59	44.4	1,320	48.7		435	42.4
Female	74	55.6	1,391	51.3		592	57.6
Race/ethnicity					0.027			<0.001
Non-Hispanic White	103	77.4	2,346	86.5		916	89.2
Non-Hispanic Black	9	6.8	113	4.2		15	1.5
Hispanic/Latino	13	9.8	169	6.2		32	3.1
Other	8	6.0	83	3.1		64	6.2
Household Income					0.037			<0.001
<$20,000	12	11.3	233	10.5		30	3.3
$20,000–59,999	40	37.7	613	27.5		169	18.6
$60,000–99,999	30	28.3	601	27.0		221	24.3
>= $100,000	24	22.6	782	35.1		489	53.8
Education					0.045			<0.001
High school (HS) or below HS	38	28.6	565	20.9		128	12.5
Some college	34	25.6	625	23.1		185	18.0
College graduate and above	61	45.9	1,515	56.0		712	69.5
Marital Status					<0.001			<0.001
Married	50	38.5	1,501	55.8		749	74.2
Not married	80	61.5	1,188	44.2		261	25.8
Employment Status					0.080			0.002
Employed	86	70.5	2,008	77.3		844	83.4
Unemployed	36	29.5	589	22.7		168	16.6
Health Insurance Status					0.100			<0.001
None	12	9.3	217	8.2		46	4.6
Private	89	69.0	2,006	75.9		866	86.3
Public	27	20.9	366	13.9		50	5.0
Canadian	1	0.8	53	2.0		41	4.1
Diagnosis					<0.001
Leukemia	91	68.4	1,093	40.3
Brain tumor	7	5.3	564	20.8
Hodgkin Lymphoma	6	4.5	475	17.5
Non-Hodgkin Lymphoma	7	5.3	330	12.2
Neuroblastoma	22	16.5	249	9.2
Time Since Diagnosis, years
Median time	24.8		31.9
Range	17.9 – 42.9		17.7 – 48.7
Total Body Irradiation					<0.001
Yes	69	54.8	20	0.8
No	57	45.2	2,526	99.2
Busulfan					<0.001
Yes	24	18.0	4	0.2
No	109	82.0	2,593	99.8
Any Chronic Health Condition					<0.001			<0.001
Grade <3	53	39.8	1,746	64.4		882	85.9
Grade=3 or 4	80	60.2	965	35.6		145	14.1
Number of Chronic Health Conditions					<0.001			<0.001
<2	18	13.5	948	35.0		595	57.9
>=2	115	86.5	1,763	65.0		432	42.1
Late Relapse / Subsequent Malignant Neoplasm					0.081
Yes	16	12.0	212	7.8
No	117	88.0	2,499	92.2

^*Comparison of HCT-treated survivors and non-HCT cancer survivors using T-test for continuous variables and Chi-square test for categorical variables. Bolded p-values are <0.05.

^**Comparison of HCT treated survivors and siblings adjusted for potential within-family correlation. Bolded p-values are <0.05.

Nearly two thirds of survivors treated with HCT reported financial hardship impacting at least one domain which was not statistically different compared to non-HCT survivors (61% vs. 60%, p-value=0.70), but was statistically significantly higher than siblings (61% vs. 48%, p-value=0.037). Within specific items and compared to non-HCT survivors, a higher proportion of survivors treated with HCT reported reduced spending on home improvement (24.1% vs.17.5%, p-value=0.030), reduced spending on basics such as food and clothing (21.1% vs. 14.4%, p-value=0.044), having forgone eyeglasses (20.3% vs. 12.2%, p-value=0.021), and worry or stress about having enough money to pay household utilities (37.6% vs. 27.7%, p-value=0.034) (Table 2). Similarly, compared to siblings, higher proportion of survivors treated with HCT reported reduced spending on large purchases, home improvement, and basics, paying off bills over time, problems paying medical bills, having forgone prescription medicine and eyeglasses, and worry or stress about having enough money to pay rent or mortgage, buy nutritious meals, and pay household utilities (all p-values<0.05). However, after adjusting for age, sex, race/ethnicity, and cancer diagnosis, the differences in mean standardized scores between survivors treated with and without HCT and by domain were not statistically significant (Table 3). In contrast, the adjusted mean standardized scores in all domains were significantly higher among survivors treated with HCT than that of siblings.

Table 2.

Proportion of Study Participants Reporting Financial Hardship Items in HCT Survivors, Non-HCT survivors, and Siblings

	HCT survivors	Non-HCT survivors	p-value*	Siblings	p-value**
	No. (%)	No. (%)		No. (%)
Material Hardship / Financial Sacrifices ***	61 (45.9)	1191 (43.9)	0.85	347 (33.8)	0.022
Reduced spending on vacation/leisure	35 (26.3)	635 (23.4)	0.43	187 (18.2)	0.16
Reduced spending on large purchases	32 (24.1)	467 (17.2)	0.045	129 (12.6)	0.011
Reduced spending on home improvement	32 (24.1)	475 (17.5)	0.030	138 (13.4)	0.013
Used savings set aside for other purposes	25 (18.8)	439 (16.2)	0.39	134 (13.0)	0.079
Reduced spending on basics	28 (21.1)	390 (14.4)	0.044	86 (8.4)	0.001
Paying off medical bills over time	39 (29.3)	819 (30.2)	0.52	206 (20.1)	0.021
Made a change to living situation	11 (8.3)	172 (6.3)	0.43	34 (3.3)	0.045
Problems paying medical bills	31 (23.3)	540 (19.9)	0.33	124 (12.1)	0.001
Behavioral Hardship ***	49 (36.8)	876 (32.3)	0.39	235 (22.9)	0.002
Forgone any needed medical care	20 (15.0)	359 (13.2)	0.79	75 (7.3)	0.049
Forgone yearly visit to your primary care	14 (10.5)	280 (10.3)	0.92	59 (5.7)	0.12
Foregone specialist	17 (12.8)	311 (11.5)	0.60	80 (7.8)	0.13
Foregone follow-up care	12 (9.0)	193 (7.1)	0.52
Foregone prescription medicine	20 (15.0)	318 (11.7)	0.37	78 (7.6)	0.027
Foregone eyeglasses	27 (20.3)	330 (12.2)	0.021	109 (10.6)	0.003
Forgone mental health counseling	13 (9.8)	228 (8.4)	0.97	51 (5.0)	0.17
Foregone dental care	27 (20.3)	599 (22.1)	0.52	163 (15.9)	0.20
Psychological Hardship ***	62 (46.6)	1087 (40.1)	0.38	282 (27.5)	0.002
Worry / stress about having enough money to pay rent or mortgage	50 (37.6)	889 (32.8)	0.51	241 (23.5)	0.014
Worry / stress about having enough money to buy nutritious meals	50 (37.6)	721 (26.6)	0.046	162 (15.8)	<0.001
Worry / stress about having enough money to pay household utilities	50 (37.6)	750 (27.7)	0.034	175 (17.0)	<0.001
Other Items
History of being sent to debt collection	37 (27.8)	752 (27.7)	0.84	217 (21.1)	0.050
History of filing for bankruptcy	5 (3.8)	215 (7.9)	0.11	74 (7.2)	0.20

^*Comparisons between HCT and non-HCT survivors adjusted for age at questionnaire, sex, race and cancer diagnosis group. Bolded p-values are <0.05.

^**Comparisons between HCT survivors and sibling controls adjusted for age at questionnaire, sex and race. GEE used to account for potential within-family correlation. Bolded p-values are <0.05.

^***N (%) for any items with affirmative response

Table 3.

Average Difference in Standardized Domain-Specific Financial Hardship Scores and Single-Item Proportions across Groups

	HCT and non-HCT				HCT and Siblings
Domains	Crude Mean Difference (95% CI)	P-value	Adjusted Mean Difference (95% CI)*	P-value	Crude Mean Difference (95% CI)	P-value	Adjusted Mean Difference (95% CI)**	P-value
Material Hardship / Financial Sacrifices	0.16 (−0.04, 0.37)	0.12	0.18 (−0.05, 0.41)	0.13	0.40 (0.22, 0.59)	<0.001	0.41 (0.17, 0.66)	<0.001
Behavioral Hardship	0.12 (−0.11, 0.35)	0.31	0.07 (−0.18, 0.32)	0.57	0.39 (0.20, 0.58)	<0.001	0.39 (0.12, 0.65)	0.004
Psychological Hardship	0.25 (0.05, 0.46)	0.016	0.19 (−0.04, 0.42)	0.10	0.55 (0.37, 0.74)	<0.001	0.45 (0.22, 0.69)	<0.001
Other Items	Percent (95% CI)		Percent (95% CI)		Percent (95% CI)		Percent (95% CI)
History of being sent to debt collection***	0.1 (−7.7, 7.9)	0.98	0.2 (−7.6, 8.1)	0.95	6.7 (−0.8, 14.1)	0.079	5.3 (−2.8, 13.5)	0.20
History of filing for bankruptcy***	−4.2 (−8.8, 0.5)	0.079	−2.7 (−6.2, 0.7)	0.12	−3.4 (−8.0, 1.1)	0.14	−1.1 (−5.0, 2.8)	0.59

^*Adjusted for age at questionnaire, sex, race/ethnicity and cancer diagnosis group. Bolded p-values are <0.05.

^**Adjusted for age at questionnaire, sex and race/ethnicity. GEE used to account for potential within-family correlation. Bolded p-values are <0.05.

^***Percentage differences were reported for these binary items. Bolded p-values are <0.05.

Compared to non-HCT survivors, a significantly greater proportion of survivors treated with HCT reported that their spouses stayed in jobs to avoid loss of health insurance (25.8% vs. 15.9%, p-value=0.036) (Table 4). A similar difference was also noted when comparing siblings and survivors treated with HCT (25.8% vs. 14.5%, p-value=0.006). Asset accumulation such as home ownership was also lower in survivors treated with HCT (23.3%) compared to non-HCT survivors (41.7%) and sibling controls (62.0%).

Table 4.

Proportion of Study Participants Reporting Health Insurance / Employment Costs / Concerns and Assets / Debts Items in HCT Survivors, Non-HCT survivors, and Siblings

	HCT Survivors	Non-HCT survivors	p-value*	Siblings	p-value**
Health Insurance & Employment Costs/ Concerns	No. (%)	No. (%)		No. (%)
Annual deductible for plan			0.84		0.085
<$1,300	30 (25.6)	595 (25.0)		228 (24.9)
$1,301–$2,600	21 (17.9)	424 (17.8)		190 (20.8)
>=$2,600	21 (17.9)	603 (25.4)		302 (33.0)
Don’t know	45 (38.5)	755 (31.8)		194 (21.2)
Concern maintain insurance next 12 months			0.97		0.18
Not concerned	82 (70.1)	1,694 (70.6)		703 (76.2)
A little concerned	19 (16.2)	365 (15.2)		136 (14.8)
Moderately concerned	9 (7.7)	183 (7.6)		47 (5.1)
Greatly concerned	7 (6.0)	159 (6.6)		36 (3.9)
Taking time off, change jobs, change to flexible schedule, avoid promotion, worry about forced retirement, job lock, stay to avoid loss of insurance			0.35		0.72
Yes	50 (37.6)	1,097 (40.5)		381 (37.1)
No	83 (62.4)	1,614 (59.5)		646 (62.9)
Spouse stays in job to avoid loss of insurance			0.036		0.006
Yes	16 (25.8)	271 (15.9)		115 (14.5)
No	46 (74.2)	1,431 (84.1)		678 (85.5)
Assets / Debts	No. (%)	No. (%)		No.(%)
Survivor owning home			0.008		<0.001
Yes	31 (23.3)	1,130 (41.7)		637 (62.0)
No	102 (76.7)	1,581 (58.3)		390 (38.0)
Spouse owning home			0.044		<0.001
Yes	24 (18.0)	823 (30.4)		460 (44.8)
No	109 (75.2)	1,888 (69.6)		567 (55.2)
Someone else owning home			0.24		0.002
Yes	33 (24.8)	416 (15.3)		71 (6.9)
No	100 (75.2)	2,295 (84.7)		956 (93.1)
Mortgage or loans and outstanding balance			0.96		0.073
None	33 (46.5)	722 (43.1)		268 (35.5)
<$50,000	6 (8.5)	150 (9.0)		55 (7.3)
$50,000–$250,000	22 (31.0)	569 (34.0)		268 (35.5)
>=$250,000	10 (14.1)	233 (13.9)		164 (21.7)
Other debt (credit cards, car loans, medical providers, life insurance policy loans, relatives, etc. and outstanding balance)			<0.001		<0.001
None	44 (38.6)	814 (33.6)		317 (33.8)
<$50,000	52 (45.6)	1,267 (52.3)		484 (51.5)
$50,000–$250,000	18 (15.8)	310 (12.8)		129 (13.7)
>=$250,000	0 (0.0)	32 (1.3)		9 (1.0)

^*Comparisons between HCT and non-HCT survivors adjusted for age at questionnaire, sex, race/ethnicity and cancer diagnosis group. Bolded p-values are <0.05.

^**Comparisons between HCT survivors and sibling controls adjusted for age at questionnaire, sex and race/ethnicity. GEE used to account for potential within-family correlation. Bolded p-values are <0.05.

In further-adjusted multivariable analysis, HCT status (HCT versus non-HCT therapy) was not associated with material, behavioral, or psychological hardship (Table 5). Sociodemographic factors including lower household income, lower educational attainment, and unemployment were associated with at least one domain of financial hardship in all survivors. Lack of health insurance was also associated with higher financial hardship across all three domains. Higher number of chronic health conditions (≥2) were also associated with higher financial hardship across all three domains.

Table 5.

Adjusted^* Multivariable Models of Predictors of Financial Hardship Reported by Study Participants

	Material Hardship Financial Sacrifices Domain		Behavioral Hardship Domain		Psychological Hardship Domain
	Adjusted mean difference (95% CI)	p-value	Adjusted mean difference (95% CI)	p-value	Adjusted mean difference (95% CI)	p-value
Treatment
HCT	Referent		Referent		Referent
Non-HCT	−0.09 (−0.34, 0.15)	0.47	−0.03 (−0.28, 0.22)	0.81	−0.12 (−0.33, 0.10)	0.28
Socio-demographics
Household Income
<$20,000	0.69 (0.48, 0.91)	<.001	1.07 (0.84, 1.29)	<.001	1.48 (1.29, 1.67)	<.001
$20,000–$59,000	0.92 (0.79, 1.06)	<.001	0.77 (0.63, 0.91)	<.001	1.16 (1.04, 1.28)	<.001
$60,000–$99,000	0.50 (0.37, 0.62)	<.001	0.28 (0.15, 0.41)	<.001	0.44 (0.33, 0.54)	<.001
>=$100,000	Referent		Referent		Referent
Education
Highschool / Below	−0.05 (−0.19, 0.09)	0.51	0.03 (−0.12, 0.17)	0.72	0.14 (0.01, 0.26)	0.029
Some College	0.15 (0.03, 0.27)	0.018	0.14 (0.02, 0.27)	0.028	0.21 (0.11, 0.32)	<.001
College Grad / Above	Referent		Referent		Referent
Marital Status
Married	Referent		Referent		Referent
Divorced/Separated/Widowed	0.06 (−0.13, 0.25)	0.53	0.31 (0.11, 0.50)	0.002	0.08 (−0.08, 0.25)	0.33
Never married	−0.23 (−0.35, −0.11)	<.001	−0.06 (−0.19, 0.06)	0.33	−0.14 (−0.24, −0.03)	0.013
Employed
Employed -Full	Referent		Referent		Referent
Employed -Part	−0.01 (−0.17, 0.15)	0.89	0.05 (−0.12, 0.21)	0.59	−0.03 (−0.17, 0.11)	0.69
Unemployed	−0.02 (−0.16, 0.12)	0.76	0.01 (−0.14, 0.15)	0.92	−0.18 (−0.30, −0.06)	0.004
Health Insurance
None	0.33 (0.14, 0.53)	<.001	1.43 (1.23, 1.63)	<.001	0.38 (0.21, 0.55)	<.001
Private	Referent		Referent		Referent
Public	0.02 (−0.17, 0.21)	0.82	−0.06 (−0.26, 0.13)	0.53	0.17 (0.01, 0.34)	0.043
Canadian	−0.30 (−0.64, 0.04)	0.080	0.07 (−0.28, 0.42)	0.70	0.14 (−0.16, 0.43)	0.37
Post-treatment complications
Any Chronic Health Condition
Max Grade <3	Referent		Referent		Referent
Max Grade 3 or 4	0.11 (−0.01, 0.22)	0.071	0.02 (−0.10, 0.14)	0.80	0.00 (−0.10, 0.10)	0.99
No. Chronic Health Conditions
<2	Referent		Referent		Referent
>=2	0.27 (0.16, 0.39)	<.001	0.22 (0.10, 0.34)	<.001	0.18 (0.08, 0.28)	<.001
Late Relapse / SMN
Yes	0.11 (−0.08, 0.29)		−0.13 (−0.32, 0.06)	0.17	0.00 (−0.16, 0.16)	0.98
No	Referent	0.26	Referent		Referent

Abbreviations: 95%CI, 95% confidence interval; SMN, subsequent malignant neoplasm.

^*Adjusted for age at questionnaire, sex, race and cancer diagnosis group. Bolded p-values are <0.05.

Discussion

To our knowledge, this is the first study to examine self-reported financial hardship and its individual domains among long-term survivors treated with HCT compared to those treated without HCT and siblings of cancer survivors. In our cohort, financial hardship was not greater among survivors treated with HCT when compared to non-HCT survivors; however, it was higher in HCT survivors when compared to sibling controls. While an incremental impact of HCT over non-HCT treatment was not observed for overall hardship as measured by domain specific scores, significant differences were noted for specific items (e.g., financial sacrifices related to reduced spending on home improvement and basics such as food) and survivors treated with HCT also reported greater insurance and employment concerns associated with financial hardship as well as fewer assets. These findings suggest that the experience of financial hardship among survivors treated with HCT may be different and worthy of further characterization when compared to survivors treated without HCT. It is possible that we did not find significant difference in overall hardship because of low numbers of HCT survivors and lack of information about allogeneic vs. autologous transplants since autologous transplants do tend to behave more like chemotherapy only patients. Potentially low number of allogeneic HCTs also made it difficult to tease out the potential incremental impact of complications such as graft vs. host disease on financial hardship.

Our results also suggest that financial hardship is an important aspect of the survivorship experience regardless of HCT status even many years after diagnosis and treatment, indicating the need for continued screening. [18] Given the many challenges associated with the diagnosis and treatment of childhood cancer as well as a need for HCT, survivors may possess a certain degree of uncertainty about their future as well as worry about recurrence and associated financial hardship. [19] Survivors in general more often reported worry and or stress focused on the ability to pay for necessities. This worry and/or stress may be related to previous experiences of financial hardship that occurred during diagnosis and treatment, and perceptions of financial vulnerability, or may be related to current experiences of financial hardship. [20] Interventions for cancer patients and survivors such as educational interventions and financial navigation programs to help build financial resilience and the perceived ability to deal with financial uncertainties and crises are being developed and tested. [21–24] Such interventions may be adapted for adult survivors of childhood cancers who remain at a life-long high risk of health complications and psychosocial consequences such as employment/ schooling and financial challenges to help them maintain a locus of control over their finances.

Other related financial outcomes differed when comparing survivors treated with HCT to non-HCT survivors including aspects of insurance coverage, employment, as well as assets and debts. Interestingly, a greater proportion of survivors treated with HCT reported that their spouse stayed in a job to avoid losing health insurance coverage for the family. [25] In the US, employer-based coverage is common as group plans and those who are subsidized by employers often ensure cheaper premiums than individual plans. [26] Many individuals fear the cost of medical care which may drive the need to stay in a less desirable job. We also found that other aspects of financial stability, such as home ownership, were less common among survivors treated with HCT. Home ownership is a key asset and this gap between survivors treated with and without HCT represents an important area for understanding whether survivors treated with HCT may be less likely to achieve socio-economic independence.

Chronic health conditions were associated with higher financial hardship risk regardless of HCT status. This is concerning when considering the burden of chronic health conditions among survivors treated with HCT. Nearly two thirds of survivors treated with HCT will develop a chronic health condition and half will develop a severe or life-threatening chronic health condition. [27–28] Survivors’ health problems may create ongoing financial concerns (e.g., lost wages, inability to work, frequent visits for medical care etc.). These findings underscore the importance of assessing both the medical needs of survivors treated with HCT and their ability to cope and manage stress relating to their own personal health to ensure financial health. Survivor sociodemographic factors including household income were associated with financial hardship. This is not surprising as survivors with lower socioeconomic status may have a greater perceived vulnerability to financial problems and may in turn endorse greater concerns with respect to financial hardship.

This study is based on a survey of survivors of childhood cancer diagnosed between 1970 and 1999. Since the late 1990s, the prognosis for pediatric cancer patients who have undergone HCT has improved and HCT techniques (e.g., greater use of reduced intensity conditioning regimens) and supportive care have evolved. Thus, the impact of HCT on financial outcomes might vary with era of therapy. Given the small size of our sample, we were not able to adjust for this in our analyses. The evolution in HCT practice may also impact the generalizability of the conclusions drawn from analysis of this cohort. Moreover, only a single timepoint assessment of financial hardship was completed and details on financial hardship prior to treatment initiation were unavailable. Nonetheless, the significantly higher financial hardship among HCT survivors at nearly three decades since diagnosis highlight the lasting effects of treatment and treatment-related complications could have on survivors. Detailed HCT characteristics are not currently available in the CCSS (e.g., type of HCT including allogeneic versus autologous HCT, conditioning regimen intensity, complications such as GVHD); efforts to merge the CCSS with other databases (e.g., Center for International Blood and Marrow Transplant Research) are underway. The survey was also administered prior to the COVID-19 pandemic which may have had a differential impact on outcomes such as insurance, employment, and financial hardship. Lastly, participants who provided responses to the financial hardship items may differ from those who do not. Survivors with missing data, the number of which was small, were excluded from analysis. Despite these limitations, this is the largest study of financial hardship among long-term adult survivors treated with HCT performed in childhood with comparisons to non-HCT survivors and siblings.

Our results show that a significant proportion of childhood cancer survivors report ongoing concerns regarding financial hardship well into adult survivorship when comparisons are made to sibling controls. HCT by itself is not associated with an overall difference in financial burden when compared with non-HCT treatment. These results underscore the importance of identifying and meeting financial needs through routine screening and management of financial hardship in at-risk survivors of childhood cancer. These individuals may benefit from interventions to prevent and mitigate the impact of financial hardship. Our results also highlight the need for policy-level changes for employment and insurance coverage for the childhood cancer survivors with a goal to avoid discrimination and provide equitable access to financial resources.

Highlights:

• Overall, adult survivors of childhood cancer treated with hematopoietic cell transplantation (HCT) do not endorse greater financial hardship when compared to non-HCT survivors.

• However, differences between survivors who received HCT and siblings were significant.

• Surveillance and intervention may be necessary for all survivors regardless of HCT status.