|Home | About | Journals | Submit | Contact Us | Français|
Although hematopoietic cell transplantation (HCT) is the only curative therapy for many advanced hematologic cancers, little is known about the financial hardship experienced by HCT patients, nor the association of hardship with patient-reported outcomes. We mailed a 43-item survey to adult patients approximately 180 days post first autologous or allogeneic HCT at three high-volume centers. We assessed decreases in household income, difficulty with HCT-related costs such as need to relocate or travel, and two types of hardship: “hardship_1” (reporting one or two of the following: dissatisfaction with present finances, difficulty meeting monthly bill payments, or not having enough money at the end of the month), and “hardship_2” (reporting all three). Patient-reported stress was measured with the Perceived Stress Scale (PSS-4), and seven-point scales were provided for perceptions of overall quality of life (QOL) and health. 325 of 499 surveys (65.1%) were received. The median days since HCT was 173; 47% underwent an allogeneic HCT, 60% were male, 51% were > 60 years old, and 92% were white. Overall, 46% reported income decline post-HCT, 56% reported “hardship_1” and 15% “hardship 2.” In multivariable models controlling for income, those reporting difficulty paying for HCT-related costs were more likely to report financial hardship (OR 6.9 [3.8, 12.3]). “Hardship_1” was associated with QOL below the median (OR 2.9 [1.7, 4.9]), health status below the median (OR 2.2 [1.3, 3.6]), and stress above the median (OR 2.1 [1.3, 3.5]). In this sizable cohort of HCT patients, financial hardship was prevalent, and associated with worse QOL and higher levels of perceived stress. Interventions to address patient financial hardship—especially those that ameliorate HCT-specific costs—are likely to improve patient-reported outcomes.
Hematopoietic cell transplantation (HCT) is the only curative therapy for many advanced blood cancers. While it is a resource-intensive procedure associated with high costs to the medical system,1–8 little is known about how financial hardship manifests for patients and their families. Familial financial hardship may arise from income decrement post-HCT, insurance deductibles or co-insurance, and/or from additional HCT-specific costs. For example, many patients must temporarily relocate near a HCT center, or otherwise have difficulty with transportation for frequent clinic appointments. Moreover, the post-HCT recovery period can be lengthy, and often leads to extended time away from work.9,10 Wage replacement for patients is not universally available, as 45% of all workers in the United States are without any paid sick days.11 Caregivers can also experience financial hardship due to travel expenses, time away from work, and accommodation for patient treatment.12–14
Several retrospective studies have suggested that patients experience significant changes in quality of life (QOL) after HCT,15–20 that financial concerns are likely prevalent among these patients before and after the procedure,21,22 and that these concerns may affect patient-reported outcomes. 9,10,22 Financial hardship after HCT may also contribute to poor post-HCT treatment adherence23 and may even compromise HCT outcomes such as survival.24 Given the extensive support typically available to HCT patients (physicians, nurses, social workers, and financial resource specialists), if financial hardship is shown to be prevalent and affect patient-reported outcomes, such data would not only be important for HCT patients but be educative for those taking care of patients with other resource-intensive malignancies with less support.
Lower socioeconomic status has been associated with poor survival for blood cancers,25,26 and also specifically with worse post-HCT survival.27 While pre-HCT measures of socioeconomic status such as household income are undoubtedly related to post-HCT financial hardship, we posit that they are not the same. Indeed, even when household income is controlled, measures of financial hardship may remain significantly associated with health outcomes.28,29 Some patients with lower income levels might be partially protected from financial hardship due to pre-existing supports such as access to government benefits, while others with relatively higher pre-HCT incomes may have more extensive financial commitments and thus may be highly susceptible to new post-HCT costs and income changes.
Through a mailed survey design, we aimed to characterize financial hardship post-HCT and its association with outcomes such as perceived stress, QOL, and overall health by specifically asking patients and their families about their finances while they were going though the post-HCT period. We hypothesized that familial hardship would be prevalent among patients post-HCT, and more intense among those who reported HCT-specific factors such as high transportation costs. We also hypothesized that those with higher levels of financial hardship would report worse QOL and overall health, and higher levels of perceived stress.
We undertook a cross-sectional observational study to characterize the financial experience of patients at six-months post-HCT and its association with self-reported outcomes such as QOL, health status, and perceived stress. All English-speaking adult patients (≥18 years) who had survived to 150 days post-allogeneic or autologous HCT were eligible to participate. We screened for eligible patients at three HCT sites (Dana-Farber Cancer Institute [DFCI], Mayo Clinic Arizona [MCA]), and Roswell Park Cancer Institute [RPCI]); of note, Spanish-speaking patients were also eligible to participate at MCA.
Starting in June of 2014, a mailing including an invitation letter, a questionnaire, and an opt-out card was sent to all patients who had reached 150 days post-HCT, a monthly process that continued until January of 2015. After two weeks of non-response to the initial mailing, and a medical review confirming that the patient was still alive, a second packet (letter, questionnaire, and opt-out card) was mailed. After four weeks of non-response, the physician investigators at two sites (DFCI and MCA) made follow-up phone calls. The Institutional Review Boards at all sites approved the study methods.
A 43-item self-administered questionnaire assessing financial hardship, household income, employment status, and insurance type was created specifically for this study. The questionnaire also collected self-reported demographic information, including month and year of cancer diagnosis that led to HCT and self-reported health outcomes. The instrument (Appendix A) was developed after a structured literature review, a focus group of HCT nurses and social workers at DFCI (n=10), and seven in-depth cognitive debriefing interviews with patients at DFCI who had undergone HCT. The survey included self-reported questions relating to overall QOL, health, and perceived stress. The invitation letter, questionnaire, and opt-out card were translated into Spanish for MCA using a professional translation service. All sites used the same questionnaire with changes only to the names of institutions and site-specific programs and/or benefits.
Three financial distress questions were used.30 The first was: “How satisfied are you with your family’s present financial situation?” (range from 1 to 5, where 1=not satisfied at all, and 5=completely satisfied). The coding for this question was reversed so that higher numbers represented more dissatisfaction, then converted to a dichotomous variable (“unsatisfied finances”) defined as reporting a level 3 or greater. The second question was: “How difficult is it for you/your family to meet monthly payments on your bills?” (range from 1 to 5, where 1= not at all difficult, and 5=extremely difficult). This question was dichotomized into the following categories: “difficulty with payments” defined as reporting a level 3 or greater versus “no difficulty” reporting a level less than 3. The third question was “How do your family’s finances usually work out at the end of the month?” (choices were “some money left over,” “just enough money,” or “not enough money.”)31 This scale was converted to a dichotomous variable (“not enough money” versus other). The variable “hardship_1” was operationalized as experiencing any one or two of the three hardship questions; “hardship_2,” a more extreme measure of financial distress, was operationalized as experiencing all three.
Current monthly familial household income was asked as a categorical question, with the following categories: “less than $1,000,” “$1,000 to $2,999,” “$3,000 to $4,999,” “$5,000 to $6,999,” and “$7,000 or more.” This variable was converted into “low income” (less than $3,000), “middle income” ($3,000 to $6,999), and “high income” ($7,000 or more). Of note, these categories roughly correspond to US household income distributions, where about 30 percent fall into the lower- and upper-income ranges and 40 percent fall in the middle-income range.32 The survey also included a question regarding changes in monthly familial household income at the time of response as compared to before HCT; responses were converted to a binary variable (“income_decline”). The survey clearly specified that familial income could be from any source, not just wages.
Patients were also asked to rank, on a scale of 1 (not at all difficult) to 5 (extremely difficult): “How difficult has it been for you to pay for the following costs” (1) temporarily relocate closer to HCT center; (2) transportation to and from appointments (gas, parking, etc); (3) transplant-related changes at home (cleaning, special foods, etc). These were converted to binary variables (“relocate,” “transit,” and “changes at home,”) and considered difficult for respondents who reported level 3 or higher. The variable “HCT_costs” was defined as positive if patients experiencing any of the three HCT costs.
The survey included a seven-point self-assessment scale to report overall QOL (higher scores indicated better QOL) and a seven-point self-assessment scale to report overall health (higher scores indicated better health); these two questions used similar wording as the global health status domain of the EORTC QLQ-C30, v.3.33 Perceived stress was measured with the Perceived Stress Scale-4 (PSS-4; scores range from 0 to 16 with higher scores indicating higher stress), a validated short-form psychometric scale consisting of four questions.34 Measures were converted to dichotomous variables using their median score.
Demographic and clinical characteristics were summarized using frequency, percent (categorical variables) and median and range (continuous variables). Chi-square tests and Kruskal-Wallis test were used to analyze associations between patient characteristics andhardship_1, hardship_2, income_decline, and HCT_costs. Next, multivariable logistic regression models were fit for potential covariates of hardship_1, hardship_2 and income_decline (including HCT type, race, gender, age, income, marital status, education, time since diagnosis, HCT_costs and distance to HCT center ≥ 75th percentile). Additional models were fit to assess the relationship between hardship_1, hardship_2 and income_decline and patient self-reported QOL, health, and perceived stress (controlling for the same covariates except for HCT_costs, as we expected this variable to be at least moderately associated with hardship_1, hardship_2 and income_decline). Analyses were performed with STATA (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP).
Of 499 eligible patients, 325 completed surveys, for an overall response rate of 65.1% (71% at DFCI; 55% at RPCI; and 56% at MCA). Analysis of available demographics (gender, race, age, HCT type, and days since HCT day 0) revealed only two significant distinctions between responders and non-responders: responders were more likely to be older (median age 61 vs. 57) and to be white (92% vs. 81%). Characteristics of the respondents are presented in Table 1. The median months from first cancer diagnosis to HCT day zero was 7.8. Nearly half of respondents (47%) received an allogeneic transplant, and the median number of days since HCT was 173. Respondents represented 20 different states, with the median distance to HCT center of 54 miles (75th percentile = 122 miles).
About half of respondents (46%) experienced a decline in their household income post-HCT (Table 2). The majority (56%) reported hardship_1, and 15% reported hardship_2. More specifically, nearly half (49%) were unsatisfied with their present financial situation, 42% faced difficulty meeting monthly payments, and 19% did not have enough money at the end of the month. In bivariable analyses, time since diagnosis was not associated with any hardship measure. Outside of lower monthly income, which was associated with hardship_1, hardship_2 and income_decline, other effects were mixed. For example, patients who lived a distance ≥ 75th percentile were more likely to report hardship_1, but not hardship_2 or income_decline.
A majority of patients (51%) were positive for HCT_costs (Table 3). Forty-one percent had difficulty paying for transportation, about a third (36%) had difficulty meeting costs associated with changes in the home environment, and one in five (19%) had difficulty paying to relocate closer to their HCT center. In multivariable models controlling for income, reporting any of these HCT-related cost increased the odds of hardship_1 by a factor of 6.9, hardship_2 by a factor of 3.8, and income_decline by 2.4 compared to those reporting little to no difficulty in HCT-related costs; in contrast, neither distance ≥ 75th percentile nor time since diagnosis had an association (Table 4). Finally, of the specific HCT-related costs, in a multivariable model including all three, patients reporting transportation difficulty had a 2.0 times higher odds of reporting income_decline (p=.03).
The odds of reporting QOL below the median was 2.9 times higher among those who experienced hardship_1, and 2.2 times higher for those reporting hardship_2 (Table 5), compared to those not reporting any hardship. The odds of reporting overall health status below the median was 2.2 times higher among those that experienced hardship_1 compared to those not reporting any hardship. Patients reporting hardship_1 had 2.1 times higher odds of reporting a higher level of perceived stress, and those with hardship_2 had a 3.1 times higher odds of reporting higher stress, compared to those not reporting any hardship. Finally, no decrease in QOL or overall health was seen for those with income_decline, although those reporting it were 2.1 times as likely to report stress above the median, compared to those who did not. Of note, neither time since diagnosis nor distance ≥ 75th percentile had an association with any of the PROs.
The majority of patients in this sizable multi-site study reported financial hardship at six months post-HCT. Difficulty meeting any HCT-related costs (HCT_costs) was associated with all three hardship variables (hardship_1, hardship_2 and income_decline), and the most prevalent HCT cost reported was difficulty paying for transportation to HCT centers (41%). Patients with any hardship (hardship_1) reported worse QOL, overall health, and higher levels of stress. Our data suggest that among this population of advanced hematologic malignancy patients who all have insurance coverage, financial hardship is common and has a significant association with worse patient-reported outcomes. Moreover, the fact that we found these results while controlling for income suggests that experiencing financial hardship may be distinct from having low income.
Finding financial hardship in a post-HCT patient population is not in itself a reason for action. After all, having cancer and needing treatment necessitates time off from work with its risk of missed wages, as well as many incidental costs for patients and their families. In determining how to intervene, it may make sense to focus on those patients who report both hardship and compromised PROs. Indeed, had we found no association between hardship and worse PROs, we might reasonably conclude that current post-HCT ameliorative schemes are working very well in limiting the effects of hardship on how patients experience their disease and QOL. This was not the case.
A prior pilot study assessing patient and caregiver costs in the first three months after HCT (n=30)14 found that the median out-of-pocket cost was over $2,000 (range: $199 to $13,769), and that patients/caregivers who required temporary lodging had higher expenses compared with those who did not. Our data corroborate these post-HCT personal financial consequences in a large multi-insitutional sample, and reinforce that those who have to relocate or travel to HCT centers face higher costs. Another single-institution study of allogeneic HCT survivors assessed a median of 2.3 years after the procedure found that 47% reported ongoing financial burden.21 The current anlaysis reveals familial financial hardship to be prevalent shortly after both allogeneic and autologous HCT, and that it is associated with worse PROs during this period.
Where patients live in relation to their HCT center has been shown to affect overall survival,24,35 although in our analysis, distance alone was not significant associated with hardship in multivariable models. We found that difficulty with transportation costs was prevalent and associated with income decline even in multivariable models controlling for the other HCT-related costs. Moreover, difficulty with these costs was more likely among those who underwent allogeneic versus autologous HCT. The sicker pre-HCT nature of allogeneic patients, their more intense follow-up needs, and possibility of complications such as graft versus host disease may explain this differential. Innovative models of care delivery that include some local care after HCT or even telehealth methods could potentially ameliorate this hardship specifically for the allogeneic population. In addition, wider insurance coverage for ostensibly non-medical expenses such as transportation to clinic appointments and associated lodging should be encouraged.
While the majority of patients in our cohort reported financial hardship at six months after HCT, our study cannot determine whether hardship was present prior to HCT, nor what its effect may have been in relation to PROs at earlier and potentially clinically important times. Ideally, financial hardship and outcomes would be studied in a quasi-experimental design with measurements taken pre- and post-HCT; as we did not know the prevalence of hardship after HCT, nor whether patients would be highly willing to answer such questions while preparing for the procedure, we chose a cross-sectional design at six months. This limits our conclusions. On the other hand, the fact that almost half of our respondents reported a decrease in income post-HCT does suggest that hardship intensified due to HCT itself. In addition, our finding that PROs were worse for those experiencing financial hardship after HCT is compelling, even if those same patients already had financial hardship before HCT.
Our study has additional limitations. Although implemented at three sites, our patients represent a small subset of the overall HCT population in the United States, and the majority of patients were from one center (DFCI). On the other hand, despite regional and racial differences in patient population at the three sites, income loss, financial hardship and extreme hardship were similar. Second, although we report a strong response rate from sick patients on a difficult topic, it is possible that our analysis suffered from non-response and social desirability bias. Third, due to constraints on survey length, while we used a well-validated measure of perceived stress (PSS-4), our QOL and overall health measures were limited in scope. Finally, data from the Center for International Blood and Marrow Transplantation Research illustrate that in 2012, for those with a known racial classification, 85% of those receiving HCT were white;36 in contrast, our cohort had more white patients (92%). While this difference is small, certain aspects of HCT are vulnerable to racial-ethnic disparities,37 and this may limit the generalizability of our results.
There has been a growing interest in oncology28,38 to address familial financial hardship associated with treatment. Our analyses reveal that financial concerns are prevalent in the year after HCT, and likely affect patients’ QOL. We were surprised to find the majority of patients we contacted were willing to answer questions about their finances and income, and hope such willingness can be leveraged into a screening tool to identify those most likely to face hardship before HCT. While such a strategy would allow for the provision of innovative models of care tailored for those most at risk, a critical balance would have to be met so that patients at higher financial risk are not discouraged from accessing this potentially life-saving procedure.
Funding: U54 Comprehensive Partnership to Reduce Cancer Health Disparities, University of Massachusetts Boston and Dana-Farber/Harvard Cancer Center.
Disclosures: The authors have no relevant financial interests to disclose.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.