Is Cancer History Associated With Assets, Debt, and Net Worth in the United States?

Is Cancer History Associated With Assets, Debt, and Net Worth in the United States? Background: Financial hardships experienced by cancer survivors have become a prominent public health issue in the United States. Few studies of financial hardship have assessed financial holdings, including assets, debts, and their values, associated with a cancer history. Methods: Using the 2008–2011 Medical Expenditure Panel Survey, we identified 1603 cancer survivors and 34 915 individuals age 18–64 years without a cancer history to assess associations between self-reported cancer history and assets, debts, and net worth. Distributions of self-reported asset and debt ownership, their values, and net worth were compared for adults with and without a cancer history with chi-square statistics. Multivariable ordered probit regression analysis was conducted to assess the association between cancer history and net worth using a two-sided Wald test. All analyses were stratified by age group (18–34, 35–44, 45–54, and 55–64 years). Statistical tests were two-sided. Results: Among those age 45–54 years, cancer survivors had a lower proportion of home ownership than individuals without a cancer history (59.0% vs 67.1%, P ¼ .0014) and were statistically significantly more likely to have negative net worth (– $3000) and less likely to have positive net worth ($3000). Cancer survivors were more likely to have debt than individuals without a cancer history, especially among those age 18–34 years (41.3% vs 27.1%, P < .001). Conclusions: Cancer history is associated with lower asset ownership, more debt, and lower net worth, especially in survivors age 45–54 years. Longitudinal studies of financial holdings will be important to inform development of interven- tions to reduce financial hardship. The number of cancer survivors is expected to grow from 15.5 health insurance. Consequently, the medical costs associated million in 2016 to 20.3 million by 2026 in the United States (1) with cancer have led to considerable financial hardship for due to an aging and growing population and increasing sur- survivors, especially the working-age population and their fami- vival resulting from improvements in early detection and lies (6). Recent studies estimated that between 13% and 34% of treatment. The cost of cancer treatment is also expected to working-age cancer survivors (age 18–64 years) report ever hav- increase over this time. The average launch price of a new ing to borrow money or go into debt because of cancer, its treat- therapeutic agent in oncology has increased by an average of ment, or lasting effects of treatment (6,7), about six times the $8500 per year (2), from 1995 to 2013, and currently novel anti- proportion of cancer survivors age 65 years and older (6). cancer drugs can cost more than $60 000 for a month of treat- Younger age (<40) is also associated with up to 10 times the ment (3). Cancer survivors and their family caregivers may rate of bankruptcy filings among cancer survivors (8). also experience limitations in ability to work (4,5), reducing Empirical studies have demonstrated that financial hardship household income and limiting access to employer-based is associated with lower adherence to cancer treatments (9–12), Received: November 21, 2017; Revised: January 19, 2018; Accepted: February 14, 2018 Published by Oxford University Press 2018. This work is written by US Government employees and is in the public domain in the US. This Open Access article contains public sector information licensed under the Open Government Licence v2.0 (http://www.nationalarchives.gov.uk/doc/open-government-licence/version/2/). 1of 7 Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 2of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 lower quality of life and perceived quality of care (13,14), and The comparison group included 34 915 individuals age 18–64 subsequently poorer health outcomes, including increased mor- years without a self-reported history of cancer. tality (15). Thus, financial hardship is also an increasingly sig- nificant social and public health issue. Measures As mentioned previously, material measures of financial hardship are typically measured as out-of-pocket spending by Sample Characteristics survivors for their medical care and health care in general and Sample characteristics for adults with and without a self- productivity loss, including lost income, missed work days, in- reported cancer history included age (18–34, 35–44, 45–54, and ability to participate in usual activities, or family members’ per- 55–64 years), sex, race/ethnicity (non-Hispanic white and other), sonal leave from work. However, the scope of financial marital status (married and not married/other), educational at- hardship in cancer survivors is not particularly well understood, tainment (<high school and high school), family size (1, 2 [eg, and few studies have assessed the underlying financial hold- husband and wife or parent and child], and 3þ), family income ings, including ownership of assets, debts, and their values, as as percentage of the federal poverty line (FPL; <399%, and novel measures of financial hardship of individuals following a 400%), employment (employed and not employed), health in- cancer diagnosis compared with those without a cancer history. surance type (private, any public, uninsured), and survey year. An older study used data from the 2002 Health and Retirement Comorbid conditions were identified with a series of questions Study to examine assets in cancer survivors and individuals about whether a doctor or other health professional ever told without a cancer history age 55 years and older, but information the person they had any MEPS priority condition, including ar- was not available about the working-age population younger thritis, asthma, diabetes, emphysema, coronary heart disease, than age 55 years, which is likely to be most vulnerable to the hypertension, stroke, angina, and high cholesterol. Comorbid risk of financial hardship (16). In addition, these data were col- conditions were categorized by the number of MEPS priority lected before more modern high-cost cancer care and did not in- medical conditions (0, 1, 2þ). Time since cancer diagnosis and clude any information about debt or net worth. Other studies of receipt of any cancer treatment during the year of the survey debt in cancer survivors have been limited by lack of a compari- were measured only for cancer survivors. son group without a cancer history (6,7) or have been conducted in defined geographic regions and were not representative of Measures of Wealth the US population (8). In this study, we used nationally repre- Measures of wealth included financial assets, debt, and net sentative data to assess the association between a cancer his- worth, which were estimated as previously described by tory and asset ownership, debts, and net worth among working- Bernard et al. (17). Briefly, total financial assets included owner- age individuals age 18–64 years. ship and value of first home, vehicles, checking and savings accounts, other financial assets (ie, money market funds, stocks, government and corporate bonds, mutual funds, certifi- Methods cates of deposit), individual retirement accounts (ie, IRA, Keogh and 401(k)), other properties (ie, second homes, rental real es- Data Source and Study Population tate, a business or farm, boats, trailer, or other recreational vehicles). Total debt included ownership and value of debts re- This cross-sectional study used data from the 2008–2011 lated to home, vehicles, second homes, rental real estate, a Medical Expenditure Panel Survey (MEPS) Household business or farm, boats or other recreational vehicles, credit Component sponsored by the Agency for Healthcare Research card balances, debts owed to medical providers, life insurance and Quality (AHRQ). The MEPS is an ongoing nationally repre- policy loans, loans from relatives and other significant sources. sentative survey of health insurance, access to care, utilization, Net worth was estimated as the value of all assets minus debts. and health care expenditures in the US civilian noninstitution- Given that economic resources are shared among family mem- alized population. In-person interviews are conducted with a bers, measures of assets, debt, and income were estimated at family member who typically responds for all family members the family level in MEPS. We categorized net worth using a in the household over a two-year period consisting of five $3000 cutoff (negative net worth  –$3000; –$2999  net worth rounds of interviews. Asset and debt data are collected in round þ$2999; and positive net worth þ$3000) to represent monthly 5 only and are considered restricted use (only available through median household income after approximately 30% tax deduc- the AHRQ Data Center); 2008–2011 are the most recently avail- tions in the period these data were collected, 2008–2011 (22–24). able years of the survey with edited and reliable asset and debt data (17). MEPS estimate of net worth and asset holdings com- pare well to the Survey of Income and Program Participation Analyses (SIPP), and MEPS is comparable along many dimensions to the Survey of Consumer Finances (SCF), (17,18). These survey years Descriptive statistics were calculated for all sample demo- had a combined average annual response rate ranging from 54% graphic and socioeconomic characteristics. Distributions of self- to 59%. More information about survey design and content is reported asset and debt ownership, their values, and net worth available from http://www.meps.ahrq.gov/mepsweb/. were compared for adults with and without a cancer history We identified 1603 working-age cancer survivors age 18–64 with chi-square statistics. Analyses were stratified by age group years based on responses to the question “Have you ever been (18–34, 35–44, 45–54, and 55–64 years) to reflect phases of the life told by a doctor or other health professional that you had cancer course with respect to debt and asset accumulation. or a malignancy of any kind?” Individuals who did not respond Multivariable ordered probit regression analysis was con- to this question were excluded from the study. Consistent with ducted to assess the association between cancer history and net prior studies, individuals who reported only a diagnosis of non- worth. We present unadjusted, partially adjusted, and fully ad- melanoma skin cancer were not classified as cancer survivors justed predicted probabilities of cancer history and each of the and were included in the comparison group (19–21). net worth categories to first evaluate the effects of patient Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 3 of 7 demographic characteristics and then the addition of socioeco- 45–54 years were statistically significantly more likely to have a nomic characteristics on these associations. Partially adjusted negative net worth and statistically significantly less likely to model covariates included sex, race/ethnicity, marital status, have a positive net worth than those individuals without a his- and number of comorbid conditions. Additional covariates in tory of cancer (1.5%, 95% confidence interval [CI] ¼ 0.2 to 2.7, the fully adjusted models were educational attainment, family and –4.7%, 95% CI ¼ –9.0 to –0.6, respectively). Cancer survivors size, family income as percentage of FPL, employment status, in the 18–34 age category were also more likely to have a nega- type of health insurance, and survey year. All estimates were tive net worth and less likely to have a positive net worth (2.3%, weighted to account for the MEPS complex survey design and 95% CI ¼ –0.2 to 4.7, and –5.8%, 95% CI ¼ –12.1 to 0.5, respec- survey nonresponse using Stata, version 13.1 (College Station, tively), although this was only statistically significant in unad- TX, USA). Two-sided P values were calculated using the Wald justed and partially adjusted models. test. Comparisons with P values of less than .05 were considered statistically significant. Discussion Results Using nationally representative data from the MEPS, we found that among those age 45–54 years, cancer survivors had a lower Compared with individuals without a cancer history, survivors proportion of home ownership than individuals without a can- were relatively older (47.3% vs 17.8% age 55–64 years), mostly fe- cer history. They were also more likely to have a negative net male (65.9% vs 49.8%), and less likely to be employed (64.2% vs worth, even after controlling for key sociodemographic charac- 76.1%) (Table 1). Cancer survivors were also more likely to have teristics, including educational attainment, family income as a two or more comorbid conditions (49.9% vs 23.9%) and less likely percentage of the FPL, employment status, and type of health to not have any comorbid conditions (25.9% vs 52.6%) than those insurance, as well as prevalence of other comorbid conditions, without a cancer history. Cancer survivors were less likely to be which are more common among cancer survivors. The propor- uninsured and more likely to have higher educational attainment tion of cancer survivors with debt was higher than individuals and higher income as a percentage of the FPL. Of the cancer survi- without a history of cancer, especially in the 18–34 years age vors, the clear majority were more than two years since first can- group. These findings suggest that, compared with those with- cer diagnosis (81.6%) and not receiving cancer treatment at the out a cancer history, working-age individuals with a history of time of the survey (81.8%). Therefore, the sample of cancer survi- cancer have less financial stability, even many years after a can- vors was comprised mainly of longer-term cancer survivors with cer diagnosis. Longitudinal population-based studies will be im- at least one additional chronic condition. portant to assessing causality, patterns of debt, assets, and net worth throughout the cancer survivorship experience and to informing the development of interventions to reduce financial Asset Ownership hardship. Asset ownership was higher among cancer survivors and indi- Our study contributes to a growing body of research (25) doc- viduals without a cancer history in the older age groups than umenting the financial hardships associated with a cancer diag- the younger age groups. Among the group age 45 to 54 years, nosis, including 1) material conditions that develop as a result the proportion reporting family home ownership was statisti- of out-of-pocket expenses and lower income from inability to cally significantly lower for cancer survivors than those without work, 2) psychological responses to costs associated with diag- a cancer history (59.0% vs 67.1%, P ¼ .0014) (Figure 1). Asset own- nosis, its treatment, and lasting effects of treatment, and 3) cop- ership was similar for cancer survivors and individuals without ing behaviors used to manage increased expenditures and a cancer history in the other age groups. reduced income during and following cancer care. These coping behaviors may include filling a prescription or delaying the start of a treatment (12,26), nonadherence to treatment (27), or aban- Debt donment of a therapy (28). Given the late- and long-term effects associated with treatment, increased risk of second cancers (29– Among those age 18 to 34 years, a statistically significantly 31) and other chronic conditions (32), and the need for contin- higher proportion of cancer survivors reported debt ownership ued surveillance, such behaviors may adversely affect health than individuals without a cancer history (41.3% vs 27.1%, P < outcomes among cancer survivors and result in higher medical .001) (Figure 1; Appendix Table 1). In the group age 45–54 years, expenditures and increased risk of mortality (13,15,33–36). cancer survivors also had a higher proportion of debt ownership Recent studies have highlighted the elevated economic burden than individuals without a cancer history, although the differ- associated with additional chronic conditions among cancer ence was only marginally significant (36.5% vs 32.1%, P ¼ .06). survivors (20,37). To our knowledge, our measures of financial hardship are novel, and the prevalence of both debt and assets or net worth has not been assessed using the MEPS or other Net Worth data sources for working-age adults with other conditions, such The summary of assets and debt values for cancer survivors as heart disease, or in adults identified as having high medical and individuals without a cancer history who reported owner- spending in the past year. As a result, it is unclear whether our ship is shown in Figure 2. No statistically significant differences findings are specific to cancer survivors or are also consistent were observed in the reported mean assets and debt values of for adults with other conditions having high medical costs com- cancer survivors compared with individuals without a history pared with similar adults without those conditions. These will of cancer in the same age groups. be important areas for additional research. Relative to those without a history of cancer, the unadjusted We found that assets and net worth, at the time of the sur- and adjusted likelihood of cancer survivors falling into three net vey, varied substantially by age group for both cancer survivors worth categories is depicted in Table 2. Cancer survivors age and individuals without a cancer history, with greatest home Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 4of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 Table 1. Characteristics of sample population ownership and asset value in the older age groups (ie, 45–54 and 55–64 years). While not unexpected, these data suggest that No history of considerations of life course are critical in the evaluation of fi- Cancer survivor cancer nancial hardship. (n ¼ 1603) (n ¼ 34 915) An older study examined assets and cancer history, based on the 2002 Health and Retirement Study in older Americans No. No. Chi- (55 years, mean age ¼ 68 and 69 years for men and women, re- (weighted %) (weighted %) square P spectively) (16). Cancer survivors were defined as those who Age, y were diagnosed four or more years before the survey and did 18–34 192 (10.6) 13 513 (38.2) <.01 not receive treatment for their cancer in the preceding two 35–44 241 (14.3) 7772 (21.4) <.01 years. This study referred to net worth but only measured the 45–54 457 (27.8) 7796 (22.6) <.01 sum of housing equity and other assets and, unlike our study, 55–64 713 (47.3) 5834 (17.8) <.01 did not incorporate debt into the measure. Further, the study Sex found that male cancer survivors and those without a history of Male 482 (34.1) 1 612 (50.2) <.01 cancer had similar income and assets, but differed somewhat in Female 1121 (65.9) 18 303 (49.8) <.01 net worth, although no association was observed between can- Race/ethnicity cer history and assets among females (16). In our study, we did White, non-Hispanic 1055 (80.6) 15 153 (64.4) <.01 not have a sufficient sample to stratify our measures by both Black, Hispanic, other 548 (19.4) 19 762 (35.6) <.01 age group and sex. Exploring the effects of key demographic Marital status and socioeconomic characteristics on asset and debt accumula- Married 882 (59.5) 17 759 (52.5) <.01 tion over the life course will be an important area for future Not married 721 (40.5) 17 156 (47.5) <.01 research. Education* The MEPS is the only nationally representative database Less than high school 280 (11.9) 8492 (17.2) <.01 containing specific questions devoted to asset and debt catego- High school grad/ 1319 (88.1) 26 264 (82.8) <.01 college ries (ie, home ownership, car ownership, debts related to home, Family size vehicles, and businesses) for adults of all ages and types of Single 292 (20.2) 4847 (17.1) <.01 health insurance, as well as the uninsured. However, these Dyad 579 (38.9) 7956 (26.6) <.01 unique data were only available through 2011, prior to imple- 3þ 732 (40.9) 22 112 (56.2) <.01 mentation of many changes in health insurance as part of the Family income, % FPL Affordable Care Act (ACA). Our results provide important base- <399 1017 (53.5) 24 293 (59.1) <.01 line information in evaluating the effects of the ACA and other 400þ 586 (46.5) 10 604 (40.9) <.01 insurance changes on debt accumulation, safeguarding of Employment status household assets, and risk of financial hardship for cancer sur- Employed 956 (64.2) 25 165 (76.1) <.01 vivors. Several provisions of the ACA are especially relevant to Retired/not employed/ 647 (35.8) 9750 (23.9) <.01 cancer survivors, including Medicaid eligibility expansions in other some states, allowing dependent children to remain on their Health insurance parents’ employment-based health insurance until age 26 years, Private 1029 (73.8) 21 266 (70.7) <.01 elimination of annual and lifetime limits on coverage for essen- Public only 357 (15.7) 5155 (10.5) <.01 tial health benefits, caps on out-of-pocket spending, and insur- Uninsured 217 (10.5) 8494 (18.8) <.01 ance premium tax credits and cost-sharing subsidies for No. of comorbid conditions individuals and families who meet eligibility requirements (7). 0 409 (25.9) 18 743 (52.6) <.01 As additional years of the MEPS asset module data become 1 364 (24.2) 7946 (23.5) .96 available, further investigations of the effects of changes in 2þ 830 (49.9) 8226 (23.9) <.01 health insurance coverage in working-age adults are warranted. Year 2008 343 (20.6) 7030 (20.9) .22 Although this study provides unique information about 2009 466 (29.2) 10 360 (27.9) .61 measures of wealth, including assets, debt, and net worth, and 2010 427 (27) 9276 (27.7) .95 will inform future studies of financial hardship, certain limita- 2011 367 (23.2) 8249 (23.4) .5 tions should be noted. First, not all MEPS participants completed Years since first cancer diagnosis the MEPS assets module in round 5, but the distributions of soci- 1 206 (13.1) N/A odemographic characteristics, including age, sex, race/ethnicity, 2–5 459 (28.8) marital status, educational attainment, and number of MEPS 6–9 267 (16.1) priority conditions, in our sample and in other studies of cancer 10–19 358 (22.7) survivors and individuals without a cancer history using the 20 213 (14) MEPS are similar (20). Second, our study data, including finan- Missing 100 (5.3) cial holdings and cancer history, were self-reported and there- In treatment† fore subject to recall bias and misclassification. Third, for Yes 274 (18.2) cancer survivors, we did not have information about stage at di- No 1329 (81.8) agnosis, types of treatment(s) received, cancer recurrence, and other clinical characteristics likely to be associated with treat- *Out of 1603 cancer survivors, four had missing values for educational attain- ment. Out of 34 915 individuals without a history of cancer, 159 had missing val- ment. Fourth, the small number of survivors with specific can- ues. FPL ¼ federal poverty level. cers in our sample precluded us from reporting the results by †In treatment was defined as receiving chemotherapy or radiation therapy for a cancer type. Patterns of asset and debt accumulation and em- cancer condition in either an outpatient or office-based setting or having a pre- ployment options may vary substantially by cancer type and scription for an antineoplastic medication. age at diagnosis. For example, survivors of pediatric cancers Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 5 of 7 Figure 1. Home and debt ownership by cancer status. A) Home ownership by cancer status. P ¼ .0014 for age group 45–54 years, chi-square test. B) Debt ownership by cancer status. P < .001 for age group 18–34 years, chi-square test. *P < .05. who received treatment during key developmental stages may trends), and we included survey year in our multivariable analy- have significant late and lasting effects of cancer and its treat- ses. As a result, we do not believe that any specific year had a ment that affect employment, whereas early-stage breast can- particular impact on our findings. Nonetheless, it is possible cer survivors with limited treatment may have few health that the economic downtown might have affected cancer survi- effects and little employment disruption. Fifth, because the can- vors more than other Americans in their assets, debts, and net cer diagnosis question refers to cancer or malignancy of any worth as survivors were more susceptible to changes in employ- kind, it may have included individuals with pre-invasive dis- ment and health insurance coverage or faced greater expenses ease. However, any misclassification of cancer history would due to their cancer diagnosis and greater comorbidity. Future likely bias our comparisons with individuals without a cancer studies are needed to confirm the magnitudes of difference ob- history to a null association. Sixth, although a previous study served in our study. Finally, as our study was cross-sectional surmised that poverty could place patients at greater risk for across MEPS interview years, we were unable to assess the certain types of cancer (38), this doesn’t seem to be the mecha- causal nature of the observed associations. nism in our sample. The cancer survivors in our study had Our cross-sectional study provides evidence of material higher insurance coverage, education level, and income. measures of financial hardship that working-age cancer survi- Further longitudinal studies are needed to tease apart whether vors and their households may experience. We assessed the as- cancer itself is the cause of poverty. Seventh, the most recent sociation between cancer history and the components of net years of the assets section available were 2008–2011, which may worth and identified the age groups most affected by a cancer not reflect current values. These years include the recent eco- diagnosis. We found that cancer history has an association with nomic downturn in the United States. However, the proportion asset ownership, debt, and net worth, especially in those age of cancer survivors and individuals without a cancer history in- 45–54 years. Longitudinal studies that assess the causality and cluded for each year were similar, our study compared cancer patterns of financial holdings throughout the cancer experience survivors and individuals without a cancer history (rather than are warranted. Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 6of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 Figure 2. Debts and assets values (2011 dollars) in cancer survivors and in individuals without a cancer history who reported ownership by age group. A) Owned assets and debt values in age group 18–34 years by cancer status. Error bars represent the 95% confidence interval. B) Owned assets and debt values in age group 35–44 years by cancer status. Error bars represent the 95% confidence interval. C) Owned assets and debt values in age group 45–54 years by cancer status. Error bars represent the 95% confidence interval. D) Owned assets and debt values in age group 55–64 years by cancer status. Error bars represent the 95% confidence interval. Other asset value refers to the values of all assets not including primary home and cars. Any asset value refers to the values of all assets. Table 2. Cancer survivors and net worth categories relative to individuals without a history of cancer Age group, y Unadjusted Mar eff (95% CI) Partially adjusted Mar eff (95% CI) Fully adjusted Mar eff (95% CI) Negative net worth  –$3000 18–34 0.034* (0.008 to 0.060) 0.034* (0.009 to 0.060) 0.023 (–0.002 to 0.047) 35–44 0.025 (–0.001 to 0.051) 0.019 (–0.006 to 0.044) 0.010 (–0.013 to 0.033) 45–54 0.026* (0.009 to 0.042) 0.023* (0.008 to 0.038) 0.015* (0.002 to 0.027) 55–64 0.002 (–0.009 to 0.013) 0.007 (–0.002 to 0.016) 0.005 (–0.002 to 0.012) –$2999  net worth þ$2999 18–34 0.046* (0.011 to 0.081) 0.048* (0.012 to 0.085) 0.035 (–0.003 to 0.073) 35 44 0.030 (–0.001 to 0.062) 0.025 (–0.008 to 0.059) 0.016 (–0.021 to 0.053) 45–54 0.038* (0.014 to 0.063) 0.041* (0.014 to 0.068) 0.033* (0.004 to 0.062) 55–64 0.004 (–0.019 to 0.026) 0.018 (–0.006 to 0.042) 0.017 (–0.009 to 0.043) Positive net worth þ$3000 18–34 –0.080* (–0.141 to –0.018) –0.083* (–0.145 to –0.020) –0.058 (–0.121 to 0.005) 35–44 –0.055 (–0.114 to 0.002) –0.045 (–0.104 to 0.015) –0.026 (–0.086 to 0.034) 45–54 –0.064* (–0.105 to –0.023) –0.063* (–0.105 to –0.022) –0.047* (–0.090 to –0.006) 55–64 –0.006 (–0.038 to 0.027) –0.025 (–0.058 to 0.008) –0.022 (–0.054 to 0.011) *P < .05. Marginal effect (Mar eff) shows the discrete change in probability when cancer status changes from 0 (without a history of cancer) to 1 (cancer survivor). Partially adjusted model covariates included sex, race/ethnicity, marital status, and number of comorbid conditions. Additional covariates in the fully adjusted models were educational attainment, family size, family income as percentage of the federal poverty level, employment status, type of health insurance, and survey year. CI ¼ confidence interval. Program, American Cancer Society, Atlanta, GA (XH); Office of Notes Health Policy, US Department of Health and Human Services, Affiliations of authors: Early Detection Branch, Division of Washington, DC (KRY). Cancer Prevention, National Cancer Institute, Bethesda, MD The authors are grateful to Dr. Paul F. Pinsky and Dr. Barry (MD); Surveillance Research Program, Division of Cancer Control Kramer (National Cancer Institute) and Dr. Didem Bernard and and Population Sciences, National Cancer Institute, Rockville, Mr. Ray Kuntz (AHRQ) for assistance in study design, data as- MD (DC); The Center for Health Research, Kaiser Permanente, sembly, and providing us with the most recent literature. The Portland, OR (MPB); Surveillance and Health Services Research findings and conclusions in this article are those of the authors Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 7 of 7 11. Farias AJ, Du XL. Association between out-of-pocket costs, race/ethnicity, and do not necessarily represent the official position of the and adjuvant endocrine therapy adherence among medicare patients with National Cancer Institute or the Department of Health and breast cancer. J Clin Oncol. 2017;35(1):86–95. Human Services, Kaiser Permanente, or the American Cancer 12. Zheng Z, Han X, Guy GP Jr, et al. Do cancer survivors change their prescription Society. The work was performed by federal employees, and drug use for financial reasons? Findings from a nationally representative sample in the United States. Cancer. 2017;123(8):1453–1463. Drs. Matthew P. Banegas and Xuesong Han donated their time. 13. Zafar SY, McNeil RB, Thomas CM, Lathan CS, Ayanian JZ, Provenzale D. Nonfederal employees did not receive funding for this project. Population-based assessment of cancer survivors’ financial burden and qual- Preliminary results from this study were presented at the 38th ity of life: A prospective cohort study. J Oncol Pract. 2015;11(2):145–150. 14. Lathan CS, Cronin A, Tucker-Seeley R, Zafar SY, Ayanian JZ, Schrag D. Annual Meeting of Society for Medical Decision Making confer- Association of financial strain with symptom burden and quality of life for ence in Vancouver, Canada, October 23–26, 2016. patients with lung or colorectal cancer. J Clin Oncol. 2016;34(15):1732–1740. 15. Ramsey SD, Bansal A, Fedorenko CR, et al. Financial insolvency as a risk fac- tor for early mortality among patients with cancer. J Clin Oncol. 2016;34(9): 980–986. Appendix 16. Norredam M, Meara E, Landrum MB, Huskamp HA, Keating NL. Financial sta- tus, employment, and insurance among older cancer survivors. J Gen Intern Med. 2009;24(Suppl 2):S438–S445. 17. Bernard DM, Banthin JS, Encinosa WE. Wealth, income, and the affordability Appendix Table 1. Asset ownership by age group of health insurance. Health Aff (Millwood). 2009;28(3):887–896. 18. Panel on Measuring Medical Care Risk in Conjunction with the New Without a Supplemental Income Poverty Measure, O’Grady MJ, Wunderlich GS. Medical Cancer survivor cancer history P Care Economic Risk : Measuring Financial Vulnerability From Spending on Medical (n ¼ 1603) (n ¼ 34 915) Care. Washington, DC: National Academies Press; 2012. 19. Yabroff KR, Guy GP Jr, Ekwueme DU, et al. Annual patient time costs associ- No. (weighted %) No. (weighted %) ated with medical care among cancer survivors in the United States. Med Care. 2014;52(7):594–601. Asset owner, age 18–34 y 20. Guy GP Jr, Ekwueme DU, Yabroff KR, et al. Economic burden of cancer survi- Home 42 (24.6) 2281 (21) .32 vorship among adults in the United States. J Clin Oncol. 2013;31(30): Transportation Vehicle 111 (63.6) 7131 (59.1) .26 3749–3757. Other assets 108 (64.4) 6826 (59.8) .32 21. Yabroff KR, Lawrence WF, Clauser S, Davis WW, Brown ML. Burden of illness in cancer survivors: Findings from a population-based national sample. J Natl Any assets 134 (77.4) 8934 (73.8) .34 Cancer Inst. 2004;96(17):1322–1330. Asset owner, age 35–44 y 22. US Census Bureau. Household income for states: 2010 and 2011. https:// Home 117 (58.6) 3829 (57.8) .83 www.census.gov/prod/2012pubs/acsbr11-02.pdf. Accessed January 25, 2017. Transportation Vehicle 178 (80.7) 5762 (79.7) .75 23. US Census Bureau. Household income for states: 2007 and 2008. https:// www.census.gov/prod/2009pubs/acsbr08-2.pdf. Accessed January 25, 2017. Other assets 159 (74.8) 5135 (74.8) .99 24. US Census Bureau. Household income for states: 2009 and 2010. https:// Any assets 203 (89.4) 6512 (88.8) .78 www.census.gov/prod/2011pubs/acsbr10-02.pdf. Accessed January 25, 2017. Asset owner, age 45–54 y 25. Altice CK, Banegas MP, Tucker-Seeley RD, Yabroff KR. Financial hardships ex- Home 237 (59) 4648 (67.1) .00* perienced by cancer survivors: A systematic review. J Natl Cancer Inst. 2017; 109(2):djw205. Transportation Vehicle 351 (82.4) 5866 (80.9) .50 26. Huntington SF, Weiss BM, Vogl DT, et al. Financial toxicity in insured patients Other assets 301 (71.9) 5362 (76.5) .08 with multiple myeloma: A cross-sectional pilot study. Lancet Haematol. 2015; Any assets 385 (89.0) 6696 (90.5) .26 2(10):e408–e416. Asset owner, age 55–64 y 27. Zullig LL, Peppercorn JM, Schrag D, et al. Financial distress, use of cost-coping strategies, and adherence to prescription medication among patients with Home 486 (73.4) 3846 (72.3) .59 cancer. J Oncol Pract. 2013;9(6 suppl):60s–63s. Transportation Vehicle 567 (83.6) 4422 (81.4) .19 28. Streeter SB, Schwartzberg L, Husain N, Johnsrud M. Patient and plan charac- Other assets 534 (80.7) 4245 (79.5) .47 teristics affecting abandonment of oral oncolytic prescriptions. J Oncol Pract. Any assets 641 (93.1) 5140 (92.0) .33 2011;7(3 suppl):46s–51s. 29. Berrington de Gonzalez A, Wong J, Kleinerman R, Kim C, Morton L, Bekelman JE. Risk of second cancers according to radiation therapy technique and mo- *P < .05. dality in prostate cancer survivors. Int J Radiat Oncol Biol Phys. 2015;91(2): 295–302. 30. Grantzau T, Overgaard J. Risk of second non-breast cancer after radiotherapy for breast cancer: A systematic review and meta-analysis of 762,468 patients. References Radiother Oncol. 2015;114(1):56–65. 31. Lee JS, DuBois SG, Coccia PF, Bleyer A, Olin RL, Goldsby RE. Increased risk of 1. American Cancer Society. Cancer Treatment & Survivorship Facts & Figures 2016- second malignant neoplasms in adolescents and young adults with cancer. 2017. Atlanta: American Cancer Society; 2016. Cancer. 2016;122(1):116–123. 2. Howard DH, Bach PB, Berndt ER, Conti RM. Pricing in the market for antican- 32. Bluethmann SM, Mariotto AB, Rowland JH. Anticipating the “Silver cer drugs. J Econ Perspect. 2015;29(1):139–62. Tsunami”: Prevalence trajectories and comorbidity burden among older can- 3. Bach PB. Why drugs cost so much. New York Times. January 14, 2015. cer survivors in the United States. Cancer Epidemiol Biomarkers Prev. 2016;25(7): 4. de Moor JS, Dowling EC, Ekwueme DU, et al. Employment implications of in- 1029–1036. formal cancer caregiving. J Cancer Surviv. 2017;11(1):48–57. 33. Ubel PA, Abernethy AP, Zafar SY. Full disclosure—out-of-pocket costs as side 5. Ekwueme DU, Yabroff KR, Guy GP Jr, et al. Medical costs and productivity effects. N Engl J Med. 2013;369(16):1484–1486. losses of cancer survivors—United States, 2008-2011. MMWR Morb Mortal 34. Zafar SY, Peppercorn JM, Schrag D, et al. The financial toxicity of cancer treat- Wkly Rep. 2014;63(23):505–510. ment: A pilot study assessing out-of-pocket expenses and the insured cancer 6. Yabroff KR, Dowling EC, Guy GP Jr, et al. Financial hardship associated with patient’s experience. Oncologist. 2013;18(4):381–390. cancer in the United States: Findings from a population-based sample of 35. Zheng Z, Yabroff KR, Guy GP Jr, et al. Annual medical expenditure and pro- adult cancer survivors. J Clin Oncol. 2016;34(3):259–267. ductivity loss among colorectal, female breast, and prostate cancer survivors 7. Banegas MP, Guy GP Jr, de Moor JS, et al. For working-age cancer survivors, in the United States. J Natl Cancer Inst. 2016;108(5):djv382. medical debt and bankruptcy create financial hardships. Health Aff (Millwood). 36. Darkow T, Henk HJ, Thomas SK, et al. Treatment interruptions and non- 2016;35(1):54–61. adherence with imatinib and associated healthcare costs: A retrospective 8. Ramsey S, Blough D, Kirchhoff A, et al. Washington State cancer patients analysis among managed care patients with chronic myelogenous leukae- found to be at greater risk for bankruptcy than people without a cancer diag- mia. Pharmacoeconomics. 2007;25(6):481–496. nosis. Health Aff (Millwood). 2013;32(6):1143–1152. 37. Guy GP Jr, Yabroff KR, Ekwueme DU, Rim SH, Li R, Richardson LC. Economic 9. Dusetzina SB, Winn AN, Abel GA, Huskamp HA, Keating NL. Cost sharing and burden of chronic conditions among survivors of cancer in the United States. adherence to tyrosine kinase inhibitors for patients with chronic myeloid J Clin Oncol. 2017;35(18):2053–2061. leukemia. J Clin Oncol. 2014;32(4):306–311. 38. Boscoe FP, Johnson CJ, Sherman RL, Stinchcomb DG, Lin G, Henry KA. The re- 10. Neugut AI, Subar M, Wilde ET, et al. Association between prescription co- lationship between area poverty rate and site-specific cancer incidence in payment amount and compliance with adjuvant hormonal therapy in the United States. Cancer. 2014;120(14):2191–2198. women with early-stage breast cancer. J Clin Oncol. 2011;29(18):2534–2542. Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Cancer Spectrum Oxford University Press

Is Cancer History Associated With Assets, Debt, and Net Worth in the United States?

Free
7 pages

Loading next page...
 
/lp/ou_press/is-cancer-history-associated-with-assets-debt-and-net-worth-in-the-L00KagayW5
Publisher
Oxford University Press
Copyright
Published by Oxford University Press 2018.
eISSN
2515-5091
D.O.I.
10.1093/jncics/pky004
Publisher site
See Article on Publisher Site

Abstract

Background: Financial hardships experienced by cancer survivors have become a prominent public health issue in the United States. Few studies of financial hardship have assessed financial holdings, including assets, debts, and their values, associated with a cancer history. Methods: Using the 2008–2011 Medical Expenditure Panel Survey, we identified 1603 cancer survivors and 34 915 individuals age 18–64 years without a cancer history to assess associations between self-reported cancer history and assets, debts, and net worth. Distributions of self-reported asset and debt ownership, their values, and net worth were compared for adults with and without a cancer history with chi-square statistics. Multivariable ordered probit regression analysis was conducted to assess the association between cancer history and net worth using a two-sided Wald test. All analyses were stratified by age group (18–34, 35–44, 45–54, and 55–64 years). Statistical tests were two-sided. Results: Among those age 45–54 years, cancer survivors had a lower proportion of home ownership than individuals without a cancer history (59.0% vs 67.1%, P ¼ .0014) and were statistically significantly more likely to have negative net worth (– $3000) and less likely to have positive net worth ($3000). Cancer survivors were more likely to have debt than individuals without a cancer history, especially among those age 18–34 years (41.3% vs 27.1%, P < .001). Conclusions: Cancer history is associated with lower asset ownership, more debt, and lower net worth, especially in survivors age 45–54 years. Longitudinal studies of financial holdings will be important to inform development of interven- tions to reduce financial hardship. The number of cancer survivors is expected to grow from 15.5 health insurance. Consequently, the medical costs associated million in 2016 to 20.3 million by 2026 in the United States (1) with cancer have led to considerable financial hardship for due to an aging and growing population and increasing sur- survivors, especially the working-age population and their fami- vival resulting from improvements in early detection and lies (6). Recent studies estimated that between 13% and 34% of treatment. The cost of cancer treatment is also expected to working-age cancer survivors (age 18–64 years) report ever hav- increase over this time. The average launch price of a new ing to borrow money or go into debt because of cancer, its treat- therapeutic agent in oncology has increased by an average of ment, or lasting effects of treatment (6,7), about six times the $8500 per year (2), from 1995 to 2013, and currently novel anti- proportion of cancer survivors age 65 years and older (6). cancer drugs can cost more than $60 000 for a month of treat- Younger age (<40) is also associated with up to 10 times the ment (3). Cancer survivors and their family caregivers may rate of bankruptcy filings among cancer survivors (8). also experience limitations in ability to work (4,5), reducing Empirical studies have demonstrated that financial hardship household income and limiting access to employer-based is associated with lower adherence to cancer treatments (9–12), Received: November 21, 2017; Revised: January 19, 2018; Accepted: February 14, 2018 Published by Oxford University Press 2018. This work is written by US Government employees and is in the public domain in the US. This Open Access article contains public sector information licensed under the Open Government Licence v2.0 (http://www.nationalarchives.gov.uk/doc/open-government-licence/version/2/). 1of 7 Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 2of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 lower quality of life and perceived quality of care (13,14), and The comparison group included 34 915 individuals age 18–64 subsequently poorer health outcomes, including increased mor- years without a self-reported history of cancer. tality (15). Thus, financial hardship is also an increasingly sig- nificant social and public health issue. Measures As mentioned previously, material measures of financial hardship are typically measured as out-of-pocket spending by Sample Characteristics survivors for their medical care and health care in general and Sample characteristics for adults with and without a self- productivity loss, including lost income, missed work days, in- reported cancer history included age (18–34, 35–44, 45–54, and ability to participate in usual activities, or family members’ per- 55–64 years), sex, race/ethnicity (non-Hispanic white and other), sonal leave from work. However, the scope of financial marital status (married and not married/other), educational at- hardship in cancer survivors is not particularly well understood, tainment (<high school and high school), family size (1, 2 [eg, and few studies have assessed the underlying financial hold- husband and wife or parent and child], and 3þ), family income ings, including ownership of assets, debts, and their values, as as percentage of the federal poverty line (FPL; <399%, and novel measures of financial hardship of individuals following a 400%), employment (employed and not employed), health in- cancer diagnosis compared with those without a cancer history. surance type (private, any public, uninsured), and survey year. An older study used data from the 2002 Health and Retirement Comorbid conditions were identified with a series of questions Study to examine assets in cancer survivors and individuals about whether a doctor or other health professional ever told without a cancer history age 55 years and older, but information the person they had any MEPS priority condition, including ar- was not available about the working-age population younger thritis, asthma, diabetes, emphysema, coronary heart disease, than age 55 years, which is likely to be most vulnerable to the hypertension, stroke, angina, and high cholesterol. Comorbid risk of financial hardship (16). In addition, these data were col- conditions were categorized by the number of MEPS priority lected before more modern high-cost cancer care and did not in- medical conditions (0, 1, 2þ). Time since cancer diagnosis and clude any information about debt or net worth. Other studies of receipt of any cancer treatment during the year of the survey debt in cancer survivors have been limited by lack of a compari- were measured only for cancer survivors. son group without a cancer history (6,7) or have been conducted in defined geographic regions and were not representative of Measures of Wealth the US population (8). In this study, we used nationally repre- Measures of wealth included financial assets, debt, and net sentative data to assess the association between a cancer his- worth, which were estimated as previously described by tory and asset ownership, debts, and net worth among working- Bernard et al. (17). Briefly, total financial assets included owner- age individuals age 18–64 years. ship and value of first home, vehicles, checking and savings accounts, other financial assets (ie, money market funds, stocks, government and corporate bonds, mutual funds, certifi- Methods cates of deposit), individual retirement accounts (ie, IRA, Keogh and 401(k)), other properties (ie, second homes, rental real es- Data Source and Study Population tate, a business or farm, boats, trailer, or other recreational vehicles). Total debt included ownership and value of debts re- This cross-sectional study used data from the 2008–2011 lated to home, vehicles, second homes, rental real estate, a Medical Expenditure Panel Survey (MEPS) Household business or farm, boats or other recreational vehicles, credit Component sponsored by the Agency for Healthcare Research card balances, debts owed to medical providers, life insurance and Quality (AHRQ). The MEPS is an ongoing nationally repre- policy loans, loans from relatives and other significant sources. sentative survey of health insurance, access to care, utilization, Net worth was estimated as the value of all assets minus debts. and health care expenditures in the US civilian noninstitution- Given that economic resources are shared among family mem- alized population. In-person interviews are conducted with a bers, measures of assets, debt, and income were estimated at family member who typically responds for all family members the family level in MEPS. We categorized net worth using a in the household over a two-year period consisting of five $3000 cutoff (negative net worth  –$3000; –$2999  net worth rounds of interviews. Asset and debt data are collected in round þ$2999; and positive net worth þ$3000) to represent monthly 5 only and are considered restricted use (only available through median household income after approximately 30% tax deduc- the AHRQ Data Center); 2008–2011 are the most recently avail- tions in the period these data were collected, 2008–2011 (22–24). able years of the survey with edited and reliable asset and debt data (17). MEPS estimate of net worth and asset holdings com- pare well to the Survey of Income and Program Participation Analyses (SIPP), and MEPS is comparable along many dimensions to the Survey of Consumer Finances (SCF), (17,18). These survey years Descriptive statistics were calculated for all sample demo- had a combined average annual response rate ranging from 54% graphic and socioeconomic characteristics. Distributions of self- to 59%. More information about survey design and content is reported asset and debt ownership, their values, and net worth available from http://www.meps.ahrq.gov/mepsweb/. were compared for adults with and without a cancer history We identified 1603 working-age cancer survivors age 18–64 with chi-square statistics. Analyses were stratified by age group years based on responses to the question “Have you ever been (18–34, 35–44, 45–54, and 55–64 years) to reflect phases of the life told by a doctor or other health professional that you had cancer course with respect to debt and asset accumulation. or a malignancy of any kind?” Individuals who did not respond Multivariable ordered probit regression analysis was con- to this question were excluded from the study. Consistent with ducted to assess the association between cancer history and net prior studies, individuals who reported only a diagnosis of non- worth. We present unadjusted, partially adjusted, and fully ad- melanoma skin cancer were not classified as cancer survivors justed predicted probabilities of cancer history and each of the and were included in the comparison group (19–21). net worth categories to first evaluate the effects of patient Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 3 of 7 demographic characteristics and then the addition of socioeco- 45–54 years were statistically significantly more likely to have a nomic characteristics on these associations. Partially adjusted negative net worth and statistically significantly less likely to model covariates included sex, race/ethnicity, marital status, have a positive net worth than those individuals without a his- and number of comorbid conditions. Additional covariates in tory of cancer (1.5%, 95% confidence interval [CI] ¼ 0.2 to 2.7, the fully adjusted models were educational attainment, family and –4.7%, 95% CI ¼ –9.0 to –0.6, respectively). Cancer survivors size, family income as percentage of FPL, employment status, in the 18–34 age category were also more likely to have a nega- type of health insurance, and survey year. All estimates were tive net worth and less likely to have a positive net worth (2.3%, weighted to account for the MEPS complex survey design and 95% CI ¼ –0.2 to 4.7, and –5.8%, 95% CI ¼ –12.1 to 0.5, respec- survey nonresponse using Stata, version 13.1 (College Station, tively), although this was only statistically significant in unad- TX, USA). Two-sided P values were calculated using the Wald justed and partially adjusted models. test. Comparisons with P values of less than .05 were considered statistically significant. Discussion Results Using nationally representative data from the MEPS, we found that among those age 45–54 years, cancer survivors had a lower Compared with individuals without a cancer history, survivors proportion of home ownership than individuals without a can- were relatively older (47.3% vs 17.8% age 55–64 years), mostly fe- cer history. They were also more likely to have a negative net male (65.9% vs 49.8%), and less likely to be employed (64.2% vs worth, even after controlling for key sociodemographic charac- 76.1%) (Table 1). Cancer survivors were also more likely to have teristics, including educational attainment, family income as a two or more comorbid conditions (49.9% vs 23.9%) and less likely percentage of the FPL, employment status, and type of health to not have any comorbid conditions (25.9% vs 52.6%) than those insurance, as well as prevalence of other comorbid conditions, without a cancer history. Cancer survivors were less likely to be which are more common among cancer survivors. The propor- uninsured and more likely to have higher educational attainment tion of cancer survivors with debt was higher than individuals and higher income as a percentage of the FPL. Of the cancer survi- without a history of cancer, especially in the 18–34 years age vors, the clear majority were more than two years since first can- group. These findings suggest that, compared with those with- cer diagnosis (81.6%) and not receiving cancer treatment at the out a cancer history, working-age individuals with a history of time of the survey (81.8%). Therefore, the sample of cancer survi- cancer have less financial stability, even many years after a can- vors was comprised mainly of longer-term cancer survivors with cer diagnosis. Longitudinal population-based studies will be im- at least one additional chronic condition. portant to assessing causality, patterns of debt, assets, and net worth throughout the cancer survivorship experience and to informing the development of interventions to reduce financial Asset Ownership hardship. Asset ownership was higher among cancer survivors and indi- Our study contributes to a growing body of research (25) doc- viduals without a cancer history in the older age groups than umenting the financial hardships associated with a cancer diag- the younger age groups. Among the group age 45 to 54 years, nosis, including 1) material conditions that develop as a result the proportion reporting family home ownership was statisti- of out-of-pocket expenses and lower income from inability to cally significantly lower for cancer survivors than those without work, 2) psychological responses to costs associated with diag- a cancer history (59.0% vs 67.1%, P ¼ .0014) (Figure 1). Asset own- nosis, its treatment, and lasting effects of treatment, and 3) cop- ership was similar for cancer survivors and individuals without ing behaviors used to manage increased expenditures and a cancer history in the other age groups. reduced income during and following cancer care. These coping behaviors may include filling a prescription or delaying the start of a treatment (12,26), nonadherence to treatment (27), or aban- Debt donment of a therapy (28). Given the late- and long-term effects associated with treatment, increased risk of second cancers (29– Among those age 18 to 34 years, a statistically significantly 31) and other chronic conditions (32), and the need for contin- higher proportion of cancer survivors reported debt ownership ued surveillance, such behaviors may adversely affect health than individuals without a cancer history (41.3% vs 27.1%, P < outcomes among cancer survivors and result in higher medical .001) (Figure 1; Appendix Table 1). In the group age 45–54 years, expenditures and increased risk of mortality (13,15,33–36). cancer survivors also had a higher proportion of debt ownership Recent studies have highlighted the elevated economic burden than individuals without a cancer history, although the differ- associated with additional chronic conditions among cancer ence was only marginally significant (36.5% vs 32.1%, P ¼ .06). survivors (20,37). To our knowledge, our measures of financial hardship are novel, and the prevalence of both debt and assets or net worth has not been assessed using the MEPS or other Net Worth data sources for working-age adults with other conditions, such The summary of assets and debt values for cancer survivors as heart disease, or in adults identified as having high medical and individuals without a cancer history who reported owner- spending in the past year. As a result, it is unclear whether our ship is shown in Figure 2. No statistically significant differences findings are specific to cancer survivors or are also consistent were observed in the reported mean assets and debt values of for adults with other conditions having high medical costs com- cancer survivors compared with individuals without a history pared with similar adults without those conditions. These will of cancer in the same age groups. be important areas for additional research. Relative to those without a history of cancer, the unadjusted We found that assets and net worth, at the time of the sur- and adjusted likelihood of cancer survivors falling into three net vey, varied substantially by age group for both cancer survivors worth categories is depicted in Table 2. Cancer survivors age and individuals without a cancer history, with greatest home Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 4of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 Table 1. Characteristics of sample population ownership and asset value in the older age groups (ie, 45–54 and 55–64 years). While not unexpected, these data suggest that No history of considerations of life course are critical in the evaluation of fi- Cancer survivor cancer nancial hardship. (n ¼ 1603) (n ¼ 34 915) An older study examined assets and cancer history, based on the 2002 Health and Retirement Study in older Americans No. No. Chi- (55 years, mean age ¼ 68 and 69 years for men and women, re- (weighted %) (weighted %) square P spectively) (16). Cancer survivors were defined as those who Age, y were diagnosed four or more years before the survey and did 18–34 192 (10.6) 13 513 (38.2) <.01 not receive treatment for their cancer in the preceding two 35–44 241 (14.3) 7772 (21.4) <.01 years. This study referred to net worth but only measured the 45–54 457 (27.8) 7796 (22.6) <.01 sum of housing equity and other assets and, unlike our study, 55–64 713 (47.3) 5834 (17.8) <.01 did not incorporate debt into the measure. Further, the study Sex found that male cancer survivors and those without a history of Male 482 (34.1) 1 612 (50.2) <.01 cancer had similar income and assets, but differed somewhat in Female 1121 (65.9) 18 303 (49.8) <.01 net worth, although no association was observed between can- Race/ethnicity cer history and assets among females (16). In our study, we did White, non-Hispanic 1055 (80.6) 15 153 (64.4) <.01 not have a sufficient sample to stratify our measures by both Black, Hispanic, other 548 (19.4) 19 762 (35.6) <.01 age group and sex. Exploring the effects of key demographic Marital status and socioeconomic characteristics on asset and debt accumula- Married 882 (59.5) 17 759 (52.5) <.01 tion over the life course will be an important area for future Not married 721 (40.5) 17 156 (47.5) <.01 research. Education* The MEPS is the only nationally representative database Less than high school 280 (11.9) 8492 (17.2) <.01 containing specific questions devoted to asset and debt catego- High school grad/ 1319 (88.1) 26 264 (82.8) <.01 college ries (ie, home ownership, car ownership, debts related to home, Family size vehicles, and businesses) for adults of all ages and types of Single 292 (20.2) 4847 (17.1) <.01 health insurance, as well as the uninsured. However, these Dyad 579 (38.9) 7956 (26.6) <.01 unique data were only available through 2011, prior to imple- 3þ 732 (40.9) 22 112 (56.2) <.01 mentation of many changes in health insurance as part of the Family income, % FPL Affordable Care Act (ACA). Our results provide important base- <399 1017 (53.5) 24 293 (59.1) <.01 line information in evaluating the effects of the ACA and other 400þ 586 (46.5) 10 604 (40.9) <.01 insurance changes on debt accumulation, safeguarding of Employment status household assets, and risk of financial hardship for cancer sur- Employed 956 (64.2) 25 165 (76.1) <.01 vivors. Several provisions of the ACA are especially relevant to Retired/not employed/ 647 (35.8) 9750 (23.9) <.01 cancer survivors, including Medicaid eligibility expansions in other some states, allowing dependent children to remain on their Health insurance parents’ employment-based health insurance until age 26 years, Private 1029 (73.8) 21 266 (70.7) <.01 elimination of annual and lifetime limits on coverage for essen- Public only 357 (15.7) 5155 (10.5) <.01 tial health benefits, caps on out-of-pocket spending, and insur- Uninsured 217 (10.5) 8494 (18.8) <.01 ance premium tax credits and cost-sharing subsidies for No. of comorbid conditions individuals and families who meet eligibility requirements (7). 0 409 (25.9) 18 743 (52.6) <.01 As additional years of the MEPS asset module data become 1 364 (24.2) 7946 (23.5) .96 available, further investigations of the effects of changes in 2þ 830 (49.9) 8226 (23.9) <.01 health insurance coverage in working-age adults are warranted. Year 2008 343 (20.6) 7030 (20.9) .22 Although this study provides unique information about 2009 466 (29.2) 10 360 (27.9) .61 measures of wealth, including assets, debt, and net worth, and 2010 427 (27) 9276 (27.7) .95 will inform future studies of financial hardship, certain limita- 2011 367 (23.2) 8249 (23.4) .5 tions should be noted. First, not all MEPS participants completed Years since first cancer diagnosis the MEPS assets module in round 5, but the distributions of soci- 1 206 (13.1) N/A odemographic characteristics, including age, sex, race/ethnicity, 2–5 459 (28.8) marital status, educational attainment, and number of MEPS 6–9 267 (16.1) priority conditions, in our sample and in other studies of cancer 10–19 358 (22.7) survivors and individuals without a cancer history using the 20 213 (14) MEPS are similar (20). Second, our study data, including finan- Missing 100 (5.3) cial holdings and cancer history, were self-reported and there- In treatment† fore subject to recall bias and misclassification. Third, for Yes 274 (18.2) cancer survivors, we did not have information about stage at di- No 1329 (81.8) agnosis, types of treatment(s) received, cancer recurrence, and other clinical characteristics likely to be associated with treat- *Out of 1603 cancer survivors, four had missing values for educational attain- ment. Out of 34 915 individuals without a history of cancer, 159 had missing val- ment. Fourth, the small number of survivors with specific can- ues. FPL ¼ federal poverty level. cers in our sample precluded us from reporting the results by †In treatment was defined as receiving chemotherapy or radiation therapy for a cancer type. Patterns of asset and debt accumulation and em- cancer condition in either an outpatient or office-based setting or having a pre- ployment options may vary substantially by cancer type and scription for an antineoplastic medication. age at diagnosis. For example, survivors of pediatric cancers Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 5 of 7 Figure 1. Home and debt ownership by cancer status. A) Home ownership by cancer status. P ¼ .0014 for age group 45–54 years, chi-square test. B) Debt ownership by cancer status. P < .001 for age group 18–34 years, chi-square test. *P < .05. who received treatment during key developmental stages may trends), and we included survey year in our multivariable analy- have significant late and lasting effects of cancer and its treat- ses. As a result, we do not believe that any specific year had a ment that affect employment, whereas early-stage breast can- particular impact on our findings. Nonetheless, it is possible cer survivors with limited treatment may have few health that the economic downtown might have affected cancer survi- effects and little employment disruption. Fifth, because the can- vors more than other Americans in their assets, debts, and net cer diagnosis question refers to cancer or malignancy of any worth as survivors were more susceptible to changes in employ- kind, it may have included individuals with pre-invasive dis- ment and health insurance coverage or faced greater expenses ease. However, any misclassification of cancer history would due to their cancer diagnosis and greater comorbidity. Future likely bias our comparisons with individuals without a cancer studies are needed to confirm the magnitudes of difference ob- history to a null association. Sixth, although a previous study served in our study. Finally, as our study was cross-sectional surmised that poverty could place patients at greater risk for across MEPS interview years, we were unable to assess the certain types of cancer (38), this doesn’t seem to be the mecha- causal nature of the observed associations. nism in our sample. The cancer survivors in our study had Our cross-sectional study provides evidence of material higher insurance coverage, education level, and income. measures of financial hardship that working-age cancer survi- Further longitudinal studies are needed to tease apart whether vors and their households may experience. We assessed the as- cancer itself is the cause of poverty. Seventh, the most recent sociation between cancer history and the components of net years of the assets section available were 2008–2011, which may worth and identified the age groups most affected by a cancer not reflect current values. These years include the recent eco- diagnosis. We found that cancer history has an association with nomic downturn in the United States. However, the proportion asset ownership, debt, and net worth, especially in those age of cancer survivors and individuals without a cancer history in- 45–54 years. Longitudinal studies that assess the causality and cluded for each year were similar, our study compared cancer patterns of financial holdings throughout the cancer experience survivors and individuals without a cancer history (rather than are warranted. Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 6of 7 | JNCI J Natl Cancer Inst, 2018, Vol. 0, No. 0 Figure 2. Debts and assets values (2011 dollars) in cancer survivors and in individuals without a cancer history who reported ownership by age group. A) Owned assets and debt values in age group 18–34 years by cancer status. Error bars represent the 95% confidence interval. B) Owned assets and debt values in age group 35–44 years by cancer status. Error bars represent the 95% confidence interval. C) Owned assets and debt values in age group 45–54 years by cancer status. Error bars represent the 95% confidence interval. D) Owned assets and debt values in age group 55–64 years by cancer status. Error bars represent the 95% confidence interval. Other asset value refers to the values of all assets not including primary home and cars. Any asset value refers to the values of all assets. Table 2. Cancer survivors and net worth categories relative to individuals without a history of cancer Age group, y Unadjusted Mar eff (95% CI) Partially adjusted Mar eff (95% CI) Fully adjusted Mar eff (95% CI) Negative net worth  –$3000 18–34 0.034* (0.008 to 0.060) 0.034* (0.009 to 0.060) 0.023 (–0.002 to 0.047) 35–44 0.025 (–0.001 to 0.051) 0.019 (–0.006 to 0.044) 0.010 (–0.013 to 0.033) 45–54 0.026* (0.009 to 0.042) 0.023* (0.008 to 0.038) 0.015* (0.002 to 0.027) 55–64 0.002 (–0.009 to 0.013) 0.007 (–0.002 to 0.016) 0.005 (–0.002 to 0.012) –$2999  net worth þ$2999 18–34 0.046* (0.011 to 0.081) 0.048* (0.012 to 0.085) 0.035 (–0.003 to 0.073) 35 44 0.030 (–0.001 to 0.062) 0.025 (–0.008 to 0.059) 0.016 (–0.021 to 0.053) 45–54 0.038* (0.014 to 0.063) 0.041* (0.014 to 0.068) 0.033* (0.004 to 0.062) 55–64 0.004 (–0.019 to 0.026) 0.018 (–0.006 to 0.042) 0.017 (–0.009 to 0.043) Positive net worth þ$3000 18–34 –0.080* (–0.141 to –0.018) –0.083* (–0.145 to –0.020) –0.058 (–0.121 to 0.005) 35–44 –0.055 (–0.114 to 0.002) –0.045 (–0.104 to 0.015) –0.026 (–0.086 to 0.034) 45–54 –0.064* (–0.105 to –0.023) –0.063* (–0.105 to –0.022) –0.047* (–0.090 to –0.006) 55–64 –0.006 (–0.038 to 0.027) –0.025 (–0.058 to 0.008) –0.022 (–0.054 to 0.011) *P < .05. Marginal effect (Mar eff) shows the discrete change in probability when cancer status changes from 0 (without a history of cancer) to 1 (cancer survivor). Partially adjusted model covariates included sex, race/ethnicity, marital status, and number of comorbid conditions. Additional covariates in the fully adjusted models were educational attainment, family size, family income as percentage of the federal poverty level, employment status, type of health insurance, and survey year. CI ¼ confidence interval. Program, American Cancer Society, Atlanta, GA (XH); Office of Notes Health Policy, US Department of Health and Human Services, Affiliations of authors: Early Detection Branch, Division of Washington, DC (KRY). Cancer Prevention, National Cancer Institute, Bethesda, MD The authors are grateful to Dr. Paul F. Pinsky and Dr. Barry (MD); Surveillance Research Program, Division of Cancer Control Kramer (National Cancer Institute) and Dr. Didem Bernard and and Population Sciences, National Cancer Institute, Rockville, Mr. Ray Kuntz (AHRQ) for assistance in study design, data as- MD (DC); The Center for Health Research, Kaiser Permanente, sembly, and providing us with the most recent literature. The Portland, OR (MPB); Surveillance and Health Services Research findings and conclusions in this article are those of the authors Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018 M. Doroudi et al. | 7 of 7 11. Farias AJ, Du XL. Association between out-of-pocket costs, race/ethnicity, and do not necessarily represent the official position of the and adjuvant endocrine therapy adherence among medicare patients with National Cancer Institute or the Department of Health and breast cancer. J Clin Oncol. 2017;35(1):86–95. Human Services, Kaiser Permanente, or the American Cancer 12. Zheng Z, Han X, Guy GP Jr, et al. Do cancer survivors change their prescription Society. The work was performed by federal employees, and drug use for financial reasons? Findings from a nationally representative sample in the United States. Cancer. 2017;123(8):1453–1463. Drs. Matthew P. Banegas and Xuesong Han donated their time. 13. Zafar SY, McNeil RB, Thomas CM, Lathan CS, Ayanian JZ, Provenzale D. Nonfederal employees did not receive funding for this project. Population-based assessment of cancer survivors’ financial burden and qual- Preliminary results from this study were presented at the 38th ity of life: A prospective cohort study. J Oncol Pract. 2015;11(2):145–150. 14. Lathan CS, Cronin A, Tucker-Seeley R, Zafar SY, Ayanian JZ, Schrag D. Annual Meeting of Society for Medical Decision Making confer- Association of financial strain with symptom burden and quality of life for ence in Vancouver, Canada, October 23–26, 2016. patients with lung or colorectal cancer. J Clin Oncol. 2016;34(15):1732–1740. 15. Ramsey SD, Bansal A, Fedorenko CR, et al. Financial insolvency as a risk fac- tor for early mortality among patients with cancer. J Clin Oncol. 2016;34(9): 980–986. Appendix 16. Norredam M, Meara E, Landrum MB, Huskamp HA, Keating NL. Financial sta- tus, employment, and insurance among older cancer survivors. J Gen Intern Med. 2009;24(Suppl 2):S438–S445. 17. Bernard DM, Banthin JS, Encinosa WE. Wealth, income, and the affordability Appendix Table 1. Asset ownership by age group of health insurance. Health Aff (Millwood). 2009;28(3):887–896. 18. Panel on Measuring Medical Care Risk in Conjunction with the New Without a Supplemental Income Poverty Measure, O’Grady MJ, Wunderlich GS. Medical Cancer survivor cancer history P Care Economic Risk : Measuring Financial Vulnerability From Spending on Medical (n ¼ 1603) (n ¼ 34 915) Care. Washington, DC: National Academies Press; 2012. 19. Yabroff KR, Guy GP Jr, Ekwueme DU, et al. Annual patient time costs associ- No. (weighted %) No. (weighted %) ated with medical care among cancer survivors in the United States. Med Care. 2014;52(7):594–601. Asset owner, age 18–34 y 20. Guy GP Jr, Ekwueme DU, Yabroff KR, et al. Economic burden of cancer survi- Home 42 (24.6) 2281 (21) .32 vorship among adults in the United States. J Clin Oncol. 2013;31(30): Transportation Vehicle 111 (63.6) 7131 (59.1) .26 3749–3757. Other assets 108 (64.4) 6826 (59.8) .32 21. Yabroff KR, Lawrence WF, Clauser S, Davis WW, Brown ML. Burden of illness in cancer survivors: Findings from a population-based national sample. J Natl Any assets 134 (77.4) 8934 (73.8) .34 Cancer Inst. 2004;96(17):1322–1330. Asset owner, age 35–44 y 22. US Census Bureau. Household income for states: 2010 and 2011. https:// Home 117 (58.6) 3829 (57.8) .83 www.census.gov/prod/2012pubs/acsbr11-02.pdf. Accessed January 25, 2017. Transportation Vehicle 178 (80.7) 5762 (79.7) .75 23. US Census Bureau. Household income for states: 2007 and 2008. https:// www.census.gov/prod/2009pubs/acsbr08-2.pdf. Accessed January 25, 2017. Other assets 159 (74.8) 5135 (74.8) .99 24. US Census Bureau. Household income for states: 2009 and 2010. https:// Any assets 203 (89.4) 6512 (88.8) .78 www.census.gov/prod/2011pubs/acsbr10-02.pdf. Accessed January 25, 2017. Asset owner, age 45–54 y 25. Altice CK, Banegas MP, Tucker-Seeley RD, Yabroff KR. Financial hardships ex- Home 237 (59) 4648 (67.1) .00* perienced by cancer survivors: A systematic review. J Natl Cancer Inst. 2017; 109(2):djw205. Transportation Vehicle 351 (82.4) 5866 (80.9) .50 26. Huntington SF, Weiss BM, Vogl DT, et al. Financial toxicity in insured patients Other assets 301 (71.9) 5362 (76.5) .08 with multiple myeloma: A cross-sectional pilot study. Lancet Haematol. 2015; Any assets 385 (89.0) 6696 (90.5) .26 2(10):e408–e416. Asset owner, age 55–64 y 27. Zullig LL, Peppercorn JM, Schrag D, et al. Financial distress, use of cost-coping strategies, and adherence to prescription medication among patients with Home 486 (73.4) 3846 (72.3) .59 cancer. J Oncol Pract. 2013;9(6 suppl):60s–63s. Transportation Vehicle 567 (83.6) 4422 (81.4) .19 28. Streeter SB, Schwartzberg L, Husain N, Johnsrud M. Patient and plan charac- Other assets 534 (80.7) 4245 (79.5) .47 teristics affecting abandonment of oral oncolytic prescriptions. J Oncol Pract. Any assets 641 (93.1) 5140 (92.0) .33 2011;7(3 suppl):46s–51s. 29. Berrington de Gonzalez A, Wong J, Kleinerman R, Kim C, Morton L, Bekelman JE. Risk of second cancers according to radiation therapy technique and mo- *P < .05. dality in prostate cancer survivors. Int J Radiat Oncol Biol Phys. 2015;91(2): 295–302. 30. Grantzau T, Overgaard J. Risk of second non-breast cancer after radiotherapy for breast cancer: A systematic review and meta-analysis of 762,468 patients. References Radiother Oncol. 2015;114(1):56–65. 31. Lee JS, DuBois SG, Coccia PF, Bleyer A, Olin RL, Goldsby RE. Increased risk of 1. American Cancer Society. Cancer Treatment & Survivorship Facts & Figures 2016- second malignant neoplasms in adolescents and young adults with cancer. 2017. Atlanta: American Cancer Society; 2016. Cancer. 2016;122(1):116–123. 2. Howard DH, Bach PB, Berndt ER, Conti RM. Pricing in the market for antican- 32. Bluethmann SM, Mariotto AB, Rowland JH. Anticipating the “Silver cer drugs. J Econ Perspect. 2015;29(1):139–62. Tsunami”: Prevalence trajectories and comorbidity burden among older can- 3. Bach PB. Why drugs cost so much. New York Times. January 14, 2015. cer survivors in the United States. Cancer Epidemiol Biomarkers Prev. 2016;25(7): 4. de Moor JS, Dowling EC, Ekwueme DU, et al. Employment implications of in- 1029–1036. formal cancer caregiving. J Cancer Surviv. 2017;11(1):48–57. 33. Ubel PA, Abernethy AP, Zafar SY. Full disclosure—out-of-pocket costs as side 5. Ekwueme DU, Yabroff KR, Guy GP Jr, et al. Medical costs and productivity effects. N Engl J Med. 2013;369(16):1484–1486. losses of cancer survivors—United States, 2008-2011. MMWR Morb Mortal 34. Zafar SY, Peppercorn JM, Schrag D, et al. The financial toxicity of cancer treat- Wkly Rep. 2014;63(23):505–510. ment: A pilot study assessing out-of-pocket expenses and the insured cancer 6. Yabroff KR, Dowling EC, Guy GP Jr, et al. Financial hardship associated with patient’s experience. Oncologist. 2013;18(4):381–390. cancer in the United States: Findings from a population-based sample of 35. Zheng Z, Yabroff KR, Guy GP Jr, et al. Annual medical expenditure and pro- adult cancer survivors. J Clin Oncol. 2016;34(3):259–267. ductivity loss among colorectal, female breast, and prostate cancer survivors 7. Banegas MP, Guy GP Jr, de Moor JS, et al. For working-age cancer survivors, in the United States. J Natl Cancer Inst. 2016;108(5):djv382. medical debt and bankruptcy create financial hardships. Health Aff (Millwood). 36. Darkow T, Henk HJ, Thomas SK, et al. Treatment interruptions and non- 2016;35(1):54–61. adherence with imatinib and associated healthcare costs: A retrospective 8. Ramsey S, Blough D, Kirchhoff A, et al. Washington State cancer patients analysis among managed care patients with chronic myelogenous leukae- found to be at greater risk for bankruptcy than people without a cancer diag- mia. Pharmacoeconomics. 2007;25(6):481–496. nosis. Health Aff (Millwood). 2013;32(6):1143–1152. 37. Guy GP Jr, Yabroff KR, Ekwueme DU, Rim SH, Li R, Richardson LC. Economic 9. Dusetzina SB, Winn AN, Abel GA, Huskamp HA, Keating NL. Cost sharing and burden of chronic conditions among survivors of cancer in the United States. adherence to tyrosine kinase inhibitors for patients with chronic myeloid J Clin Oncol. 2017;35(18):2053–2061. leukemia. J Clin Oncol. 2014;32(4):306–311. 38. Boscoe FP, Johnson CJ, Sherman RL, Stinchcomb DG, Lin G, Henry KA. The re- 10. Neugut AI, Subar M, Wilde ET, et al. Association between prescription co- lationship between area poverty rate and site-specific cancer incidence in payment amount and compliance with adjuvant hormonal therapy in the United States. Cancer. 2014;120(14):2191–2198. women with early-stage breast cancer. J Clin Oncol. 2011;29(18):2534–2542. Downloaded from https://academic.oup.com/jncics/article-abstract/2/2/pky004/5026705 by Ed 'DeepDyve' Gillespie user on 21 June 2018

Journal

JNCI Cancer SpectrumOxford University Press

Published: Apr 20, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off