Quality of life of caregivers at the end of their child’s pediatric cancer treatment: cancer-specific worry and material hardship

Abstract

Objective

Emerging research suggests that completion of pediatric cancer treatment can be challenging for caregivers given shifting roles and responsibilities. Lower caregiver quality of life (QOL) has been associated with cancer-related variables, higher cancer caregiving strain, and more household material hardship during pediatric cancer treatment. Caregiver QOL at the end of treatment has not been fully investigated but has implications for child and family well-being. Using a psycho-oncology framework, this study aimed to understand the cumulative burden of household material hardship and cancer-related factors on caregiver QOL at the end of treatment.

Methods

Caregivers (N = 143) of children (M_age=10.51 years) within 1 year of ending active cancer treatment completed self-report questionnaires assessing their QOL, cancer-specific worry, and material hardship (e.g., housing, insurance). Total months of active cancer treatment were extracted from the medical record. Hierarchical linear regression was used to test the relative effects of length of treatment, material hardship, and cancer-specific worry on caregiver QOL.

Results

Cancer-specific worry was significantly associated with and accounted for significant variance in caregiver QOL, above and beyond the length of treatment and material hardship. Caregivers who endorsed more cancer-specific worry had poorer QOL. Material hardship was also significantly associated with caregivers’ QOL, but length of treatment was not.

Conclusions

Caregivers with sufficient resources and less worry about cancer have higher QOL. Findings highlight the importance of end-of-treatment screening of caregivers’ subjective cancer-specific worry in addition to material hardship, irrespective of their cancer-related stressors, for ongoing psychosocial support.

oncology, quality of life, parents, parent psychosocial functioning, health disparities and inequities

Completing pediatric cancer treatment is often perceived by families and clinicians as a celebratory milestone for patients and their families (Wakefield et al., 2010). Yet, previous qualitative work demonstrates that the end of treatment is also a stressful period in the cancer trajectory for caregivers as they adjust to changes in their daily routines and children’s medical care (Jones et al., 2022; Wakefield et al., 2011). Children receive continued health care, but the frequency of appointments decreases, and caregivers have reduced contact with medical teams, resulting in distress and difficulty making post-treatment medical decisions (Wakefield et al., 2011). Caregivers report a need for more emotional and financial support, including contact with psychosocial providers, and additional information about expectations for their child’s health care needs as their child transitions off pediatric cancer treatment (Karst et al., 2018; Wakefield et al., 2013). Many caregivers worry about managing their child’s care and endorse ongoing concern for their child’s health and fear of relapse after treatment is completed (Hobbie et al., 2010; Lemos et al., 2020). Quality of life (QOL), a holistic construct that captures caregiver well-being and functioning, is critical to assess because it is associated with children’s outcomes such that higher caregiver QOL is associated with better child QOL (Al Ghriwati et al., 2021; Bakula et al., 2020; Eiser et al., 2005; Kazak & Barakat, 1997). Despite the importance of caregiver QOL and reported challenges in the transition off treatment, little is known about caregiver QOL during this period.

Fletcher and colleagues (2012) describe a model for organizing the numerous stress-related and contextual factors associated with caregiver well-being, including QOL, for adults with cancer that is applicable to caregivers of children with cancer including during the transition off treatment. The model indicates primary and secondary stressors, appraisal and cognitive-behavioral responses that predict caregiver QOL. Primary stressors are illness-related factors that initiate the stress process such as disease, prognosis, and duration of treatment as well as caregiving demands (e.g., managing disease symptoms, side effects). Pediatric cancer research has demonstrated that poorer child health and functioning, greater treatment intensity, and less time since diagnosis are associated with poorer caregiver QOL both while children are on and off treatment (Al Ghriwati et al., 2021; Klassen et al., 2008).

Secondary stressors include those that are affected by caregiving over time (e.g., role changes, changes in family structure, changes in finances and employment, and schedule disruptions). Caregivers of children with cancer often experience secondary stressors such as disruption to employment and income (Bona et al., 2014, 2016; Miedema et al., 2008; Warner et al., 2014), and household material hardship related to cancer disproportionately affects families with lower incomes (Bona et al., 2014). Further, lower income and more financial burden are associated with poorer caregiver psychological adjustment and poorer caregiver QOL during pediatric cancer treatment (Bemis et al., 2015; Creswell et al., 2014; Galtieri et al., 2022; Klassen et al., 2008), though less is known about the association between secondary stressors and caregiver QOL at the end of treatment.

For appraisal and cognitive-behavioral responses, it is important to consider the role of subjective experiences on family outcomes, such as the perceived burden of caregiving and beliefs and worries about future outcomes, as they likely negatively impact caregiver QOL (Fletcher et al., 2012; Varni et al., 2004). Research with caregivers of children with cancer has demonstrated that higher cancer caregiving strain during pediatric cancer treatment is associated with poorer caregiver QOL (Klassen et al., 2011). Caregiver appraisals, such as cancer-specific worry that is operationalized as caregivers’ subjective appraisal about or thoughts of children’s cancer-related outcomes (e.g., worry that medical treatment is working) and future functioning (e.g., worry about long-term side effects of medication), may have a significant impact on caregivers’ emotional functioning and subsequently how they perceive and manage the challenges of transitioning off treatment.

Caregiver QOL may be affected by cancer-related stressors and subjective appraisal of stressors during the dynamic transition off active cancer treatment, which is characterized by distress and uncertainty (Jones et al., 2022). The aim of the current study was to evaluate the associations of primary stressors (i.e., length of treatment), secondary stressors (i.e., material hardship), and subjective appraisal (i.e., cancer-specific worry) with caregiver QOL utilizing standardized screening measures easily applied in clinical care. We hypothesized that longer treatment, more material hardship, and more cancer-specific worry would be associated with poorer QOL for caregivers. From a health equity perspective, it is essential to consider not only cancer-specific risk factors (treatment, worry) but also the material hardship associated with childhood cancer and the range of responsibilities caregivers must manage outside of the hospital to better identify specific targets for and timing of psychosocial interventions tailored to each family’s needs.

Methods

Participants and procedures

As part of a clinical program, caregivers of children ages 0–24 years within 1 year of ending pediatric cancer treatment participated in a psychosocial screening and education program to assess patient/family needs as children transitioned off active cancer treatment (see Perez et al., 2023). Participation was voluntary. Eligibility for the clinical education and screening program included: being a caregiver of a child with a cancer diagnosis within 1 year of the end of active treatment for that diagnosis and English-speaking. Completion of treatment was determined by chart review and confirmed with the patient’s social worker before families were contacted. During a scheduled follow-up medical appointment, a psychosocial team member met with caregivers to review an end-of-treatment education handbook, provide caregivers with self-report screening questionnaires via an electronic tablet, and address family needs and questions.

Patient demographics and information about the child’s diagnosis, treatment, and length of treatment were extracted from their medical record. IRB exemption (19-017086) was granted to utilize the clinical screening data for research purposes. The data underlying this article cannot be shared publicly due to the privacy of individuals who were willing to share their clinical information for quality improvement purposes. The data will be shared on reasonable request to the corresponding author.

Measures

Caregiver quality of life

Caregiver QOL was assessed using the Caregiver Quality of Life subscale from the PedsQL Family Impact Module (PedsQL FIM; Varni et al., 2004). Caregiver QOL is measured using 23 items that assess physical, emotional, social, and cognitive functioning. Caregivers rated statements about themselves on a 5-point Likert scale from 0 (never) to 4 (almost always) based on the last month. Per Varni et al. (2004), raw scores were reverse coded and transformed to a 0–100 scale as follows: 0 = 100, 1 = 75, 2 = 50, 3 = 25, 4 = 0. Overall QOL was calculated by averaging the transformed score of all physical, emotional, social, and cognitive subscale items, with higher scores indicating better QOL. Internal consistency of the scale was excellent (α = 0.95).

Cancer-specific worry

Cancer-specific worry was assessed using the Worry subscale from the PedsQL FIM (Varni et al., 2004). This subscale includes 5 items about caregivers’ worries related to their child’s cancer (e.g., “I worry about whether or not my child’s medical treatments are working”; “I worry about my child’s future”). Caregivers rated how much of a problem each statement has been for them over the past month on a 5-point Likert scale from 0 (never) to 4 (almost always). Raw scores were reverse coded and transformed to a 0–100 scale as follows: 0 = 100, 1 = 75, 2 = 50, 3 = 25, 4 = 0, and cancer-specific worry was determined by averaging the transformed ratings, with lower scores indicating more worry. Internal consistency of the scale was good (α = 0.86).

Material hardship

Material hardship was assessed using the National Comprehensive Cancer Network Problem List: Practical Problems checklist (NCCN Guidelines, 2016), which is a 6-item self-report screener of concrete resource needs: childcare, housing, financial, transportation, work/school, and treatment decisions. Caregivers responded “yes/no” to indicate whether a resource has “been a problem” in the past week. Number of items endorsed was summed, with higher scores indicating more material hardship. Internal consistency of the scale was questionable (α = 0.63).

Data analytic strategies

Descriptive statistics (mean, standard deviation, frequency) of key variables as well as sample characteristics were computed using SPSS version 28. To avoid issues of multicollinearity, a single variable was chosen to represent the primary stressor in the model. To determine which variable to include as the primary stressor variable in the model, associations between cancer-related variables (diagnosis, treatment type, length of treatment, time since diagnosis) were calculated using analysis of variance, Pearson’s r, and chi-square tests as appropriate based on the type of data. Relationships between cancer-related variables and caregiver QOL were also examined. Given that length of treatment was significantly associated with other cancer-related variables and no significant correlations between cancer-related variables and QOL were identified (Table 1), length of treatment was selected as the single cancer variable to simplify the model. Hierarchical linear regression, based on Fletcher and colleagues’ (2012) theory, tested the relative effects of primary stressors (length of treatment), secondary stressors (material hardship) and caregiver subjective appraisal (cancer-specific worry) on caregiver overall QOL. Length of treatment was entered on the first block, material hardship was added on the second block, and cancer-specific worry was entered on the third block. Variables were standardized prior to running the hierarchical regression for ease of interpretation. Change in R² was used to determine whether predictor variables accounted for added variance in QOL. Power analyses were run a posteriori, given that data were part of a clinical program. The full model demonstrated a strong power of 1.00, likely due to the large sample size (N = 143) and the minimal number of predictors (3) included in the regression model (Onwuegbuzie & Leech, 2004).

Table 1

Associations among cancer-related variables and caregiver quality of life (QOL).

χ2
	Treatment type
Diagnosis group	χ2=41.22***
Leukemia/lymphoma
Solid tumors
CNS tumors
One-way ANOVAs
		Months since diagnosis	Months since end of treatment	Length of treatment	Caregiver QOL
Diagnosis group		F(2,138) = 14.14***	F(2,138) = 0.35	F(2,138) = 14.89***	F(2,129) = 0.61
Leukemia/lymphoma
Solid tumors
CNS tumors
Treatment type		F(3,137) = 4.38**	F(3,137) = 0.12	F(3,137) = 3.99**	F(3,129) = 0.73
Chemotherapy only
Both chemotherapy and Radiation
Radiation only
Correlations
			Months since end of treatment	Length of treatment	Caregiver QOL
Months since diagnosis			r = 0.12	r = 0.96**	r =−0.121
Months since end of treatment				r=−0.17*	r = 0.012
Length of treatment					r =−0.123

Note. To examine the associations among these variables, given that some data are categorical (i.e., diagnosis group, treatment type) and other data are continuous (i.e., months since diagnosis, months since end of treatment, length of treatment, caregiver QOL), statistical analyses were chosen based on type of data for different sets of variables.

*

p < 0.05,

**

p < 0.01,

***

p<.001.

Results

Caregivers were 80% mothers, 17% fathers, and 3% other (e.g., grandmother). Children were treated for solid tumors, leukemia, lymphoma, and central nervous system (CNS) tumors and were on average 5 months (SD = 3.24) from the end of treatment (Table 2). On average, caregivers overall QOL was rated as 71.99 (SD = 17.95; Table 2). Prior research using the PedsQL FIM has demonstrated that the average QOL of caregivers of children with chronic illnesses is 67 (SD = 16; Medrano et al., 2013); approximately 36% of caregivers in the current sample reported their QOL to be at a level below a score of 67. Children’s average length of treatment was 12.62 months (SD = 11.37), the family average material hardship was 0.54 (SD = 1.00), and the caregiver average worry was 60.39 (SD = 24.08; Table 2). Mean ratings suggest that on average families in this sample experienced minimal material hardship and typical levels of worry. Many families endorsed zero problems of material hardship (70%), but for those who did experience material hardship, endorsed items were challenges with work (21%), finances (13%), childcare (8%), housing (5%), transportation (4%), and treatment decisions (3%).

The first block of the hierarchical regression tested the association between length of cancer treatment and caregiver overall QOL, independent of other predictors (Table 3). Length of treatment was not significantly associated with caregiver QOL (p >.05). The second block of the regression model tested the effect of length of treatment and material hardship on caregiver overall QOL. Material hardship, as measured by total number of practical problems, was significantly associated with caregiver QOL when accounting for length of treatment (β =−0.35, p < 0.001; Table 3), such that more material hardship was associated with poorer QOL. Finally, the third block tested the cumulative contributions of length of treatment, material hardship, and cancer-specific worry on caregiver overall QOL. Cancer-specific worry was significantly associated with caregiver QOL above and beyond the length of treatment and material hardship, such that caregivers who endorsed more cancer-specific worry had poorer QOL [β = 0.65, F(3,130)=49.29, p < 0.001; Table 3]. Cancer-specific worry accounted for significant variance in caregiver QOL above and beyond length of treatment and material hardship (ΔR²=0.40).

Table 2

Sample characteristics.

	N (%)
Caregiver role
Mother	114 (80)
Father	24 (17)
Other	5 (3)
Race
White	102 (72)
Black/African American	23 (16)
Other	8 (6)
Indian	5 (3)
Asian	4 (3)
Ethnicity
Non-Hispanic	67 (91)
Hispanic	7 (9)
Diagnosis type
Leukemia/lymphoma	68 (48)
Solid tumor	50 (35)
CNS tumor	24 (17)
Treatment type
Chemotherapy only	107 (76)
Both chemotherapy and radiation	26 (18)
Radiation only	8 (5)
CAR T-cell therapy	1 (1)
Child age (years) M (SD)	10.51 (5.8)
Child gender
Female	80 (55)
Male	43 (45)

	M (SD)	Minimum	Maximum
Months since diagnosis	17.64 (11.28)	2	50
Months since end of treatment	5.02 (3.24)	0	12
Months on treatment	12.61 (11.37)	1	41
Caregiver QOL	71.99 (17.95)	0	100
Material hardship rating	0.54 (1.00)	0	4
Cancer-specific worry rating	60.39 (24.08)	0	100

Note. Given that this data was part of a clinical quality improvement project, patient demographic information was extracted from the electronic health record. Patient race and ethnicity were limited to the categorical options offered through the electronic record. Thus, information regarding “other” race is unknown.

Table 3

Hierarchical linear regression model predicting caregiver QOL.

Variable	B	SE	Β 95% CI [LL, UL]	F (df)	R2	p	ΔR2
Block 1				2.03 (1, 132)	0.17	.16
Constant	−0.01	0.09	[−0.18, 0.17]
Length of treatment	−0.13	0.09	[−0.31, 0.05]
Block 2				9.94 (2, 131)	0.13	<.001	0.12
Constant	−0.01	0.08	[−0.17, 0.15]
Length of treatment	−0.10	0.09	[−0.27, 0.07]
Material hardship	0.35***	0.08	[−0.51, −0.18]
Block 3				49.29 (3, 130)	0.52	<.001	0.40
Constant	0.00	0.06	[−0.12, 0.12]
Length of treatment	0.04	0.07	[−0.17, 0.09]
Material hardship	0.20**	0.06	[−0.32, −0.07]
Cancer-related worry	0.65***	0.06	[0.53, 0.77]

**

p <.01,

***

p <.001.

Discussion

The current study aimed to expand the literature on caregiver QOL as their child transitions off active cancer treatment, a time in the cancer trajectory often filled with changes and uncertainty, by identifying contributing factors of the cancer experience. Guided by a model of stressors related to caregiver QOL in adult cancer (Fletcher et al., 2012), results identified that material hardship (secondary stressor) and caregivers’ level of worry about their children’s cancer (subjective appraisal) were associated with their overall QOL. Caregivers who experienced greater material hardship and more cancer-specific worry reported poorer QOL. Notably, the length of the child’s cancer treatment (a proxy for cancer-related primary stressors) was not associated with caregiver QOL for this sample, which differs from what is suggested by Fletcher et al. (2012). Similar to other studies (e.g., Barakat et al., 2006), this finding may suggest that caregiver perceived severity and intensity of treatment are more likely to be related to their QOL outcomes rather than objective cancer-related variables such as length of treatment. When the combined effect of all three variables was analyzed, caregivers’ cancer-specific worry was the strongest predictor of their QOL, above and beyond material hardship and length of treatment. Thus, caregiver perception of treatment and their child’s cancer outcomes is important to understand in the context of their QOL. Importantly, findings suggest that access to material resources as well as strategies to manage worry may serve as protective factors for caregiver well-being, regardless of how long children were undergoing cancer treatment.

This study highlights the need to continue to assess material hardship throughout the cancer trajectory and support caregivers in accessing concrete resources (Wiener et al., 2015). Prior work in pediatric cancer has described the financial burden of treatment, which includes increased expenses due to treatment costs combined with loss of income due to changes in employment from the demands of cancer caregiving (Pelletier & Bona, 2015; Santacroce et al., 2018; Tsimicalis et al., 2011). Greater material hardship may limit or interfere with the caregiver’s ability to engage consistently in their child’s follow-up cancer care after treatment (van Breeschoten et al., 2017; Zheng et al., 2016). Efforts to provide more equitable healthcare should focus on screening families for material hardship throughout the cancer trajectory to identify families in need of specific resources (e.g., housing, childcare, transportation). Particular attention should be paid to families who endorse challenges maintaining work and difficulties with finances, as these are among the most commonly reported challenges associated with food and housing insecurity, and linked to cancer outcomes (Bona et al., 2014, 2016). Increasing access to resources and reducing material burdens will likely promote better caregiver QOL, consistent with recent studies demonstrating positive outcomes from providing families with needed resources (Umaretiya et al., 2021).

The results of this study suggest a role for interventions targeting caregiver cancer-specific worry near the end of their child’s treatment to promote better caregiver QOL. One concerning finding is that approximately one third of caregivers in this sample endorsed below-average QOL, indicating that even families who endorse minimal material hardship (such as those in this sample) are at risk for poorer QOL. Caregivers of children with cancer who express more uncertainty about their child’s illness are more likely to engage in poorer coping and ineffective problem-solving, which can lead to poorer caregiver psychological adjustment (Basile et al., 2021; Perez et al., 2018; Szulczewski et al., 2017). Interventions might include cognitive-behavioral approaches to assist caregivers in identifying their beliefs and expectations around a “new normal” in their family’s life post-treatment as well as skills to manage worries. One example of such intervention is Promoting Resilience in Stress Management for Parents (PRISM-P), a cognitive behavioral therapy approach adapted specifically for caregivers of children with chronic illnesses (Yi-Frazier et al., 2017). This intervention includes skill-building modules directed at managing stress and coping (e.g., breathing, relaxation, planning, goal setting, overcoming obstacles) and building resilience through cognitive restructuring and benefit finding, and has been associated with more positive caregiver adjustment (Yi-Frazier et al., 2017). Similarly, Mullins et al. (2012) adapted a cognitive behavioral therapy intervention to support caregivers of children with cancer to identify thoughts associated with their child’s cancer, specifically worry thoughts associated with illness uncertainty, and to engage in cognitive restructuring of those thoughts to reduce distress. Finally, the Bright IDEAS problem-solving skills training intervention may support caregivers in navigating challenges as family’s transition off-treatment and has been shown to reduce anxiety (Sahler et al., 2013). Interventions that target caregivers’ worries may be particularly important to help caregivers adapt and improve their QOL as their child transitions off cancer treatment. Of note, because these interventions have been conducted for caregivers of childrenn on treatment, there is value in the evaluation of their benefit during the transition off treatment.

To understand additional psychosocial factors that may moderate the effect of worry and material hardship on caregiver QOL, future research may investigate variables such as coping or distress tolerance, as well as other family and contextual factors (described by Fletcher et al., 2012) to improve caregiver QOL. For example, coping may be a strong protective factor that buffers the impact of worry on caregiver QOL, as caregivers who utilize more coping skills may be able to better address stressors (Compas et al., 2015). It is possible that caregivers who report less cancer-specific worry, and consequently better QOL at the end of treatment, are those who utilize effective coping skills. Other caregivers or family members may also act as a protective buffer against the impact of worry and material hardship on caregiver QOL. Prior pediatric cancer studies have demonstrated the protective effects of secondary caregivers on family functioning (e.g., family cohesion, parenting behaviors; Keim et al., 2021). Households with two caregivers may be at lower risk for financial strain and therefore lower risk for caregiver psychological distress (Galtieri et al., 2022).

Future work may aim to further investigate the interaction between subjective and objective factors and the effect on QOL. Primary stressors or cancer factors were not associated with caregiver QOL. This outcome is consistent with the pediatric cancer literature reporting that cancer-related variables, such as diagnosis and treatment intensity, are not correlated with caregiver and child psychosocial outcomes (Alderfer et al., 2009; Barakat et al., 1997). Furthermore, prior work has similarly demonstrated that caregiver worry and subjective appraisal are stronger predictors of distress in pediatric cancer populations relative to objective cancer-related variables (e.g., diagnosis; Stuber et al., 1997). This suggests that secondary stressors and subjective appraisals (as proposed by Fletcher et al., 2012), including worry, may be more relevant to caregiver QOL than objective, cancer-related variables.

There are limitations to the current study to consider. Additional data on primary stressors (e.g., specific caregiving demands of cancer) associated with pediatric cancer were not included as predictors of caregiver QOL due to limitations in the clinical information available to analyze. As noted, length of cancer treatment was not associated with caregiver QOL, but it is possible that more nuanced aspects of pediatric cancer treatment, such as days hospitalized, or other treatment-related variables may relate to QOL as families transition off cancer treatment. The clinical education and screening program included only English-speaking caregivers due to materials not yet being translated into other languages and a majority of caregivers identified as white. Families whose primary language is a language other than English are more likely to have lower income, less likely to be insured, and have significantly longer hospital stays (Flores & Tomany-Korman, 2008; Levas et al., 2011; Scheurer et al., 2018). Thus, it is necessary to examine the association between material hardship and caregiver QOL specifically for families who speak languages other than English. The clinical screening program has since translated materials into Spanish with plans for Arabic and other languages. Future work, in both research and clinical settings, should include questionnaires and educational written materials that are translated to a family’s language as well as use interpreters to facilitate communication and address family’s needs. Another limitation of the current project is that the majority of the caregivers in this sample were mothers, and findings may be specific to predictors of mothers’ QOL at the end of pediatric cancer treatment. Understanding of predictors of fathers’ QOL as their child transitions off active treatment is necessary given that fathers often serve as second caregivers in a different role from mothers in cancer caregiving responsibilities (Nicholas et al., 2016).

Consistent with the Psychosocial Standards of Care in Pediatric Cancer, this study highlights the need to continue to screen for and address material hardship, caregiver worry, and caregiver QOL as children transition off active treatment (Wiener et al., 2015). A portion of caregivers (36%) in this sample endorsed lower QOL than caregivers of children with chronic illness more generally, even as their child ends cancer treatment, providing evidence that caregiver QOL should be regularly assessed throughout the cancer trajectory. It is important for clinicians to use standardized measures of caregiver QOL, such as the PedsQL FIM (Varni et al., 2004), that can be incorporated into clinical practice to screen for families at risk for poorer outcomes and in need of more support as their child transitions off active cancer treatment. Implementation of standardized screening of families’ material needs (e.g., childcare, employment, housing) in primary care settings leads to more referrals to, access to, and utilization of community resources (Garg et al., 2015). Similarly, in pediatric cancer populations, families who are screened for psychosocial needs receive more services (Kazak et al., 2011), promoting better and more equitable family and patient outcomes. From a health equity lens, ongoing screening is important to identify patients and families who are most at risk for poorer outcomes and to more appropriately allocate resources to promote healthy outcomes for all.

Author Contributions

Liana Galtieri (Conceptualization [equal], Formal analysis [lead], Writing—Original draft [lead], Writing—Review & editing [equal]), Megan N. Perez (Conceptualization [equal], Data curation [equal], Formal analysis [supporting], Methodology [equal], Supervision [equal], Writing—Review & editing [equal]), Lamia P. Barakat (Conceptualization [equal], Funding acquisition [lead], Investigation [equal], Methodology [equal], Resources [lead], Supervision [equal], Writing—Review & editing [equal])

Funding

This work was supported by the La Speranza Charitable Foundation to the Children’s Hospital of Philadelphia Cancer Center.

Conflicts of interest

None declared.

Acknowledgments

The authors wish to thank the caregivers who completed screening during cancer clinic visits and Heather Zukin and Rebecca Madden, clinical research coordinators, for administering the end of treatment education and screening program.