Abstract
Objective:
By limiting household expenditures on rent and utilities and connecting individuals to health services, federal housing assistance programs could facilitate health care access among low-income cancer survivors. We examined the association between receipt of rental assistance and health care access among low-income adult cancer survivors.
Methods:
We used 2019-2023 National Health Interview Survey data on adults aged ≥20 years (1) with a history of cancer diagnosis, (2) with a family income-to-poverty ratio <2, and (3) who were renters. We used propensity score weighting to address differences in observed demographic characteristics between rental assistance recipients and nonrecipients. We used logistic regression analyses to examine the associations of rental assistance receipt with outcome variables.
Results:
Lack of receipt of rental assistance was significantly associated with higher odds of delaying medical care due to cost (odds ratio [OR] = 1.90; 95% CI, 1.07-3.40) and experiencing medical financial hardship (OR = 1.85; 95% CI, 1.21-2.80), as well as lower odds of being covered by health insurance (OR = 0.04; 95% CI, 0.01-0.18).
Conclusion:
Receipt of rental assistance may help improve health care access among low-income adult cancer survivors. Our findings are important in relation to a shortage of affordable housing in the United States and highlight the need for efforts to expand housing assistance.
Keywords
Advances in the early detection and treatment of cancer are contributing to an increasing trend in cancer survival in the United States. 1 In the United States, cancer survivors numbered about 18.1 million in 2022, a number projected to rise to 21.6 million by 2030 and 26 million by 2040. 1 A large proportion of these individuals are long-term survivors: in 2022, about 70% had lived ≥5 years after diagnosis and 11% for ≥25 years after diagnosis. 1 Cancer survivors have multifaceted medical needs that include surveillance for cancer recurrence and malignancies, monitoring for late effects of cancer treatment, management of other chronic conditions, and general preventive care. 2 Thus, timely access to cancer-related and routine health care is crucial to meet the long-term health needs of this population.2,3
Despite their need for comprehensive medical care, cancer survivors may experience barriers to accessing health services. Financial burden and distress resulting from cancer treatment 4 may limit their ability to afford ongoing or future medical care. Cancer survivors also may experience uninsurance because of cancer-related employment disruptions and/or loss of government-sponsored health insurance following their initial cancer treatment. 5 Using Behavioral Risk Factor Surveillance System data, Chino et al 6 found a decline in the proportion of insured cancer survivors from 2016 to 2019. Furthermore, compared with adults without a history of cancer, cancer survivors were more likely to forego medical care because of cost and to experience challenges affording prescription medications and other health care services.5,7 Improving continuity of medical care is essential to support the physical and psychosocial well-being of cancer survivors.
Housing insecurity, characterized by lacking stable, safe, and affordable housing, could impede access to health care among cancer survivors. 8 In 2017, about 17% of cancer survivors experienced moderate to high levels of housing insecurity, 9 a condition likely exacerbated by the COVID-19 pandemic. 10 In 2022, about half of renter households were considered to be cost-burdened, as defined by spending more than 30% of their income on rent and utilities. 11 Furthermore, the share of cost-burdened renter households increased by about 9.0 percentage points from 2001 to 2022. 11 Federal housing assistance programs, in the form of public housing, tenant-based housing such as the Housing Choice Vouchers, and other initiatives, serve to mitigate housing challenges faced by low-income households. 12 In addition, in response to housing instability resulting from the COVID-19 pandemic, the US Department of Treasury implemented the Emergency Rental Assistance Program to help eligible households with payment of rent and utility expenses. 10 By limiting rent and utility burdens and facilitating connections to health care services,13 -15 participation in these housing assistance programs may improve health care access, particularly among cancer survivors who report considerable financial distress and greater medical expenditures than those without a cancer diagnosis.16,17 However, evidence supporting this relationship is needed, especially in light of the limited supply of affordable housing in the United States. 18
The objective of this study was to examine the association between the receipt of rental assistance and health care access, medical financial hardship, and health insurance coverage among low-income adult cancer survivors.
Methods
Data Source and Study Sample
We used National Health Interview Survey (NHIS) data from 2019 through 2023. The NHIS is a series of annual, cross-sectional household surveys that collect data on health status, health care access, health behaviors, and sociodemographic characteristics of the noninstitutionalized US civilian population.19,20 Because of data collection challenges resulting from the COVID-19 pandemic, changes were made to the survey operations. Specifically, from late March through June 2020, NHIS shifted from in-person to all-telephone interviews. Starting May 2021, the NHIS returned to regular survey procedures, with the first contact being made in-person and follow-up being completed by telephone. 21 The percentage of interviews conducted at least partially by telephone in 2022 and 2023 was lower than that in 2020 and 2021 but higher than that in 2019. More details are available elsewhere.22 -24
For this study, we included adults aged ≥20 years who met the following criteria: (1) reported ever being told by a doctor or health professional that they had cancer, (2) had a family income-to-poverty ratio of <2, and (3) were renters. The family income-to-poverty ratio cutoff of 2 is consistent with prior research on rental assistance.25,26 While the US Department of Housing and Urban Development (HUD) uses area median income (AMI) to determine eligibility for rental assistance, 27 a family income of <200% of the federal poverty level approximately corresponds to the HUD definition for low-income households (ie, ≤80% of AMI) in many areas.28 -30 Adults with imputed income and/or missing data on receipt of rental assistance were excluded. Verbal consent was obtained for participation in NHIS. The NHIS procedures and protocols were approved by the National Center for Health Statistics Research Ethics Review Board. Because we used a publicly available, deidentified database, the Washington State University Institutional Review Board (IRB) considered this study as not human subjects research and did not require IRB approval.
Independent Variable: Receipt of Rental Assistance
Adults were classified as receiving rental assistance if they responded affirmatively to the question, “Are you or anyone in your family paying lower rent because the federal, state, or local government is paying part of the cost?” A negative response to this question classified adults as not receiving rental assistance.
Outcome Variables: Health Care Access, Medical Financial Hardship, and Health Insurance Coverage
An affirmative response to the question “During the past 12 months, have you delayed getting medical care because of the cost?” identified adults who delayed medical care due to cost, whereas a negative response identified those who did not delay medical care due to cost. An affirmative response to the question, “During the past 12 months, was there any time when you needed medical care but did not get it because of the cost?” identified individuals with unmet need for medical care. Similarly, adults were classified as having delayed mental health care and unmet need for mental health care based on their responses to questions on getting counseling or therapy from a mental health professional. 31 An affirmative response to the question, “In the past 12 months, did you/anyone in the family have problems paying or were unable to pay any medical bills?” identified individuals from families experiencing medical financial hardship. Health insurance coverage status was classified as not covered or covered.
Covariates
Demographic characteristics used as covariates included age (20-64 or ≥65 years, consistent with the commonly used threshold to define nonelderly and elderly adults 32 ), sex (male or female), race and ethnicity (Hispanic, non-Hispanic Black, non-Hispanic White, other), education level (high school/less than high school, some college/associate’s degree, bachelor’s degree or higher), citizenship status (US citizen or non–US citizen), marital status (married/living with a partner or widowed/divorced/separated/never married), employment status (employed or unemployed 26 ), rurality (urban or rural 33 ), health status (poor or excellent/very good/good/fair), and family income-to-poverty ratio. Survey period was dichotomized based on the survey year as prepandemic (2019) or pandemic (2020-2023). These covariates were chosen a priori based on previous research.25,33 -38
Statistical Analyses
We used propensity score (PS) weighting to minimize potential selection bias by addressing observable differences between rental assistance recipients and nonrecipients. 39 To estimate the probability of receiving rental assistance (ie, PS), we applied a logistic regression model using rental assistance receipt as the dependent variable and the aforementioned covariates as the predictor variables. Using the estimated PSs, we calculated weights as 1/PS for households that received rental assistance and 1/(1 – PS) for those not receiving rental assistance. 9 We then divided the weights by the mean weight to obtain stabilized weights. 40 We used sampling weights to estimate PSs, 41 and we used the product of sampling weights and PS weights 41 in the models examining the association of rental assistance receipt with outcome variables.
We examined differences in demographic characteristics by receipt of rental assistance using the Rao–Scott χ2 test and linear regression, before and after PS weighting. We used the Rao–Scott χ2 test, weighted by the product of sampling and PS weights, to examine differences in outcome variables by receipt of rental assistance. P < .05 was considered statistically significant. We used logistic regression analyses, weighted by the product of sampling weight and PS weight and adjusted for covariates, to examine the association between receipt of rental assistance and health care access, medical financial hardship, and health insurance coverage. Because of the potential impact of rental assistance receipt on health status, 42 the use of health status as a covariate could have attenuated the study associations. Thus, we used additional logistic regression analyses, weighted by the product of sampling weight and PS weight, to examine associations between receiving rental assistance and outcome variables, adjusting for the aforementioned covariates, with the exception of health status. We conducted additional logistic regression analyses to examine the study associations, adjusting for demographic covariates and survey year fixed effects (in the place of survey period). Analyses accounted for the complex sampling design of NHIS, including stratification and clustering, and were performed by using SAS version 9.4 (SAS Institute, Inc).
Results
Demographic Characteristics
The primary study sample included 1338 adults, of whom 1322 had complete data on the study covariates and were included in the models examining associations between receipt of rental assistance and outcome variables. All 1322 adults had complete data on health care access and health insurance, and 1321 adults had complete data on medical financial hardship (Table 1). Prior to PS weighting, we found significant differences in age, sex, race and ethnicity, education level, marital status, employment status, and family income-to-poverty ratio between rental assistance recipients and nonrecipients. After PS weighting, no significant differences in demographic characteristics remained between the 2 groups.
Defined as adults from households with a ratio of family income to poverty threshold <2.
Data represented as unweighted number (%) or mean ± SE of the mean. Data source: National Health Interview Survey. 19
Values are presented as unweighted numbers, and proportions are weighted by sampling weight.
Because of missing data, cell sizes for some demographic characteristics do not sum to the group totals.
Rao–Scott χ2 test and linear regression were used to examine differences in demographic characteristics by receipt of rental assistance. P < .05 was considered significant.
Values are presented as unweighted numbers, and proportions are weighted by product of sampling weight and propensity score weight.
Includes non-Hispanic Asian only, non-Hispanic American Indian and Alaska Native (AI/AN) only, non-Hispanic AI/AN and any other group, and other single and multiple races.
Receipt of Rental Assistance, Health Care Access, Medical Financial Hardship, and Health Insurance Coverage in Low-Income Adult Cancer Survivors
The proportions of rental assistance recipients who delayed medical care due to cost (7.2% vs 13.6%), experienced medical financial hardship (19.7% vs 30.8%), and did not have health insurance coverage (0.6% vs 10.0%) were significantly lower than the corresponding proportions for nonrecipients (Table 2). We observed no significant differences in other outcomes (ie, experiencing unmet need for medical care, delayed mental health care due to cost, and unmet need for mental health care) by receipt of rental assistance.
Defined as adults from households with a ratio of family income to poverty threshold <2.
Data source: National Health Interview Survey. 19
Values are presented as unweighted numbers, and proportions are weighted by product of sampling weight and propensity score weight.
Using the Rao–Scott χ2 test to examine differences in health care measures by receipt of rental assistance. P < .05 was considered significant.
Due to missing data, cell sizes do not sum to the total in the group that received rental assistance.
Due to cell count <5, the P value must be interpreted with caution.
When the reference category was adults who did not delay medical care due to cost, lack of receipt of rental assistance was significantly associated with 90% higher odds of delaying medical care due to cost (odds ratio [OR] = 1.90; 95% CI, 1.07-3.40) (Table 3). Lack of receipt of rental assistance was also significantly associated with 84% higher odds of experiencing medical financial hardship (OR = 1.84; 95% CI, 1.21-2.80). Compared with rental assistance recipients, low-income adult cancer survivors who did not receive assistance had 96% lower odds of being covered by health insurance (OR = 0.04; 95% CI, 0.01-0.18). Receipt of rental assistance was not significantly associated with the odds of experiencing unmet need for medical care, delaying mental health care due to cost, and experiencing unmet need for mental health care.
Logistic regression models adjusted for respondents’ age, sex, race and ethnicity, US citizenship, education level, marital status, employment status, rurality, health status, ratio of family income to poverty, and survey period. Regression models weighted by product of sampling weight and propensity score weight.
Defined as adults from households with a ratio of family income to poverty threshold <2.
Data source: National Health Interview Survey. 19
Receipt of housing assistance was the reference category.
Additional Analyses
Logistic regression analyses, adjusted for all covariates except health status, did not change the overall findings. Lack of receipt of rental assistance was significantly associated with higher odds of delaying medical care due to cost (OR = 1.94; 95% CI, 1.09-3.44) and experiencing medical financial hardship (OR = 1.83; 95% CI, 1.20-2.79). Lack of receipt of rental assistance also was significantly associated with lower odds of being covered by health insurance (OR = 0.05; 95% CI, 0.01-0.18); however, associations with other outcome variables were not significant. Logistic regression analyses, adjusted for demographic covariates and survey year fixed effects (in the place of survey period), also did not change the overall findings. Of note, while lack of receipt of rental assistance was significantly associated with higher odds of delaying medical care due to cost (OR = 1.91; 95% CI, 1.06-3.47) and experiencing medical financial hardship (OR = 1.85; 95% CI, 1.23-2.79), it was associated with lower odds of being covered by health insurance (OR = 0.04; 95% CI, 0.01-0.17). Associations with other outcome variables were not significant.
Discussion
In this nationally representative household survey, low-income adult cancer survivors who did not receive rental assistance had significantly higher odds of delaying medical care due to cost and experiencing medical financial hardship than low-income adult cancer survivors who did receive rental assistance. Lack of receipt of rental assistance (vs receiving assistance) was also significantly associated with lower odds of being covered by health insurance. Findings from this study suggest that receipt of rental assistance could be associated with lower rates of medical financial hardship and better health care access and health insurance coverage among low-income adult cancer survivors. These findings have implications for the long-term health of cancer survivors, given the importance of surveillance for cancer recurrence 43 and the role of optimal health care access in promoting survival.44,45
Our findings align with those of another study in a general population of adults, indicating that receipt of rental assistance is associated with lower rates of unmet need for care due to cost and uninsurance rates. 36 Our findings also are somewhat comparable with another study indicating that receipt of housing assistance is related to a lowered risk of medical financial hardship among cancer survivors 46 ; however, our study focused on low-income cancer survivors because they are more likely to be eligible for rental assistance than higher-income cancer survivors. The downstream effects of rental assistance on health care access may be partly attributed to better housing affordability achieved via participation in housing programs. By limiting rent to about 30% of a household’s income, 14 rental assistance could free up financial resources that can be allocated to health care. 47 In addition to providing housing assistance, housing programs also could connect individuals to other social and health services. 13 This “income effect” and “gateway effect” of HUD-housing programs could explain the current study findings related to associations between housing assistance and delaying medical care due to cost and experiencing medical financial hardship. By facilitating access to health services, rental assistance could help individuals, including those undergoing or recovering from cancer treatment, receive more appropriate care.
Gaining stable and affordable housing also could allow rental assistance recipients to focus on their health care needs rather than housing concerns. Notably, frequent relocations, a commonly used indicator of housing instability, have been associated with disruptions in social safety-net program participation, including Medicaid. 48 For example, because mail is the primary method of communicating information, such as renewal information, to Medicaid enrollees, 49 frequent changes in addresses could result in Medicaid notices being returned to the agency. Because some states do not have a policy related to follow-up on returned mail, 50 housing instability may lead to the termination of Medicaid benefits and, subsequently, increase the risk of being uninsured. In contrast, a stable physical address, via housing program participation, could help enrollees maintain their health insurance status.
In response to the COVID-19 pandemic, the Coronavirus Aid, Relief, and Economic Security Act and the Emergency Rental Assistance Program included funding for rental assistance. 10 Although our study did not explicitly distinguish individuals who received HUD housing assistance from those who received assistance via pandemic relief funding, receipt of emergency rental assistance could have independently or jointly improved health care access among cancer survivors.
The shortage of affordable housing in the United States has resulted in severe housing cost burden, especially among low-income households, 18 and created extensive waitlist time among households applying to receive assistance. 51 Consequently, housing programs have underserved the populations of interest, with only about 1 in 4 eligible households receiving federal housing assistance. 52 Efforts to expand housing programs could be crucial to meet the housing needs of low-income adult cancer survivors.
Limitations
This study had several limitations. First, this study was limited by its cross-sectional design, which did not allow for inferring causal relationships between receipt of rental assistance and health care access, medical financial hardship, and health insurance coverage among cancer survivors. Second, each variable also was self-reported and subject to social desirability bias and recall bias; however, it was found in a study that used NHIS data that self-reporting did not bias results between rental assistance and health. 53 Third, while NHIS assesses current receipt of rental assistance, most of the outcome measures reflected health care access in the past 12 months. The question used to assess receipt of housing assistance did not capture duration, stability, or type of assistance. Fourth, because associations between rental assistance and health care access, medical financial hardship, and health insurance coverage could vary by program type, lack of differentiation among the types of housing assistance could have masked the heterogeneity in effects. Fifth, the associations between rental assistance and study outcomes could also have differed by site of cancer; however, our study did not stratify the associations by site of cancer because of limited sample sizes. Sixth, the NHIS does not query participants on cancer treatment and current cancer status. Lack of adjustment of the analyses for treatment and cancer status could have biased the associations toward the null. Seventh, because housing programs could vary substantially across states, lack of adjustment for state fixed effects also could have biased the study associations. Finally, models examining associations between receipt of rental assistance and mental health care access and health insurance coverage may have been biased because of a low prevalence of unmet need for mental health care and uninsurance in the study sample. 54
Despite these limitations, the study findings had several strengths. For one, the findings were strengthened by the use of a nationally representative survey on sociodemographic and health characteristics. Second, PS weighting could have improved causal inferences of the findings by adjusting for selection bias and confounders.55,56 Future studies examining the study associations by site of cancer could extend our findings.
Conclusion
In this nationally representative survey of low-income adults with a history of cancer diagnosis, lack of receipt of rental assistance (vs receiving rental assistance) was significantly associated with higher odds of delaying medical care due to cost and experiencing medical financial hardship and lower odds of being covered by health insurance. Study findings suggest that receipt of rental assistance could be related to better health care access among low-income adult cancer survivors. These findings have important implications, given the role of health care access in meeting the health needs of cancer survivors. Expanding housing programs so that individuals receive rental assistance in a timely manner could help improve health care access among cancer survivors.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.