Lack of health insurance is associated with a variety of adverse health outcomes.1,2 A cancer diagnosis may compound these challenges as cancer survivors may face increased need for diagnostic, therapeutic, and supportive care services across multiple settings. Lack of insurance among cancer survivors has been associated with worse access to guideline-consistent therapy,3 high out-of-pocket burdens,4 reduced access to surveillance,5 and increased morbidity and mortality.6 Ensuring access to insurance coverage for cancer survivors is of critical importance to improve cancer-related outcomes. Before the enactment of the Affordable Care Act (ACA), an estimated 14.7% of cancer survivors aged below 65 years (1.4 million) lacked insurance.7 This situation is largely due to loss of employment-based insurance and/or because of unavailable or unaffordable health insurance in the individual market due to history of cancer.8,9 Most state Medicaid programs provided coverage only for adults with dependent children, low-income seniors, pregnant women, and the disabled,10 and because most cancer survivors are not of child bearing age11 relatively few qualified on this basis. State Medicaid programs provided coverage for treatment of breast or cervical cancer, but coverage was limited in scope and duration.12
Several ACA coverage provisions implemented in January 2014, were intended to increase demand, and improve access to health insurance, including insurance for cancer survivors.13 The ACA established health insurance Marketplaces which offer individual plans with standard levels of coverage and provide income-based tax credits to reduce premiums and cost-sharing subsidies. The ACA also prohibited basing premium prices on health history in many private insurance markets. Potentially most important, the ACA expanded Medicaid to adults without dependent children at higher incomes. Although 5 states plus Washington, DC undertook early expansions (between 2010 and 2012) as permitted by the ACA or through a Section 1115 waiver, 19 states14 implemented Medicaid expansions in 2014, with an additional 7 states joining by 2016.10 Among uninsured cancer survivors aged below 65 years, an estimated 31% would be eligible for tax credits to purchase insurance in the Marketplaces and another 30% would become Medicaid-eligible under the ACA.7 The ACA included an individual coverage mandate with tax penalties, which was expected to increase coverage across eligibility categories. Evidence from the first 2 years of the program indicates that the ACA was associated with reduced uninsured rates among adults not eligible for Medicare, with estimates ranging from 7.8 to 14.9 percentage points.14,15 The decrease was evident in all states and sociodemographic groups though more pronounced for adults in states that expanded Medicaid,16,17 in persons with low socioeconomic status, and in racial and ethnic minorities.18,19 The decline in uninsured rates nationally comes from a 5.5 percentage point increase in private coverage, and a 2.2 percentage point increase in public (Medicaid) enrollment.20
Previous studies documented that Medicaid expansion was associated with reduced uninsured rates among adults with mental health and other chronic conditions 16,21,22 and newly diagnosed cancers.23,24 However, little is known about how insurance coverage for adult cancer survivors changed after implementation of the ACA. Furthermore, most estimates of coverage increases reflect averages across the income spectrum, rather than specific criteria for coverage or subsidies, or are limited to poor adults, comparing those in states that expanded Medicaid to those in nonexpansion states. In this study, we begin to fill the information gap along these 2 dimensions. We used data from a large nationally representative survey that captures information on both health status and insurance to assess changes in coverage for cancer survivors before and after implementation of the ACA. We assessed changes overall, and specific to the type of coverage or subsidy for which the cancer survivor is newly eligible. We hypothesized that the percent uninsured would decrease with ACA implementation particularly among those expansion eligible for Medicaid and those eligible for premium subsidies, relative to those with incomes > 400% federal poverty guideline (FPG), who faced the insurance mandate, but otherwise did not benefit from subsidized coverage. We examined characteristics of cancer survivors who remained uninsured after implementation of the ACA. Finally, we compared cancer survivors and adults without cancer. We further hypothesize that the effects of ACA coverage provisions would be larger for adults with a cancer history, as they are expected to have greater demand for coverage, and the ACA includes a provision to reduce coverage barriers for individuals with preexisting health conditions.
We pooled data from 2012 to 2015 National Health Interview Survey (NHIS), a household survey representative of the civilian, noninstitutionalized US population. The NHIS includes ∼100,000 people in 40,000 households each year, and collects information on demographic and socioeconomic characteristics, health insurance, health status, access to care, and health care utilization. A sample adult from each household provides more detailed information on health conditions, labor force participation, and other characteristics. We accessed confidential geographic identifiers available through the National Center for Health Statistics Research Data Center, and linked state-year–specific policies on Medicaid eligibility.10
We selected adults aged 19 through 64 years who reported ever being told they had cancer (including melanoma but excluding other skin cancer types), and a comparison group without a cancer history.
The key outcome in this study was insurance coverage at the time of the survey. We created indicators for private insurance (including employment related, civilian military, and privately purchased), public insurance [Medicaid, Children’s Health Insurance Program (CHIP), other government non-Medicare], and uninsured. Although Medicaid and privately purchased coverage are the 2 categories that we expect to be affected most directly by the ACA coverage provisions, we used broader measures (for example, augmenting our Medicaid measure with responses about Children’s Health Insurance Program and other government coverage), as individuals may be confused about the source of their insurance coverage. The public and private coverage indicators were not mutually exclusive. The key independent variables were eligibility for Medicaid before ACA expansion in 2014, and eligibility for various types of coverage (Medicaid, private insurance purchased through the Marketplace) once the ACA coverage provisions were implemented in 2014. There is no observable measure of insurance eligibility in the NHIS. To assess eligibility, we linked state-specific Medicaid eligibility rules to each individual based on the state of residence and survey year. The eligibility rules specify the lower and upper income thresholds, expressed as a percent of FPG. Using data on income and family structure from the NHIS we constructed measures of income as a percent of FPG, employment status, and presence of a dependent child for each sample adult. We compared the characteristics for each sample adult to the relevant state and time period–specific thresholds to assess likely eligibility. We assessed eligibility for the Marketplace premium subsidies and cost-sharing reductions based on state-specific lower thresholds (138% FPG in states expanding Medicaid, 100% FPG in nonexpansion states), but the upper income eligibility thresholds were applied nationally. We used information on current insurance (employer-sponsored insurance or Medicare) or offers of insurance from a current employer to further assess eligibility for Marketplace premium subsidies. Table 1 indicates the income eligibility groups used and mechanism for ACA coverage effects. For individuals living in states with early Medicaid expansions (between 2010 and 2012) we assessed eligibility and assigned them to the pre-ACA eligibility category; in sensitivity analyses we excluded them.
We described sample characteristics, patterns of eligibility, and trends in coverage pre-post implementation of ACA coverage provisions overall and by eligibility category, comparing adults with and without a cancer history using Student’s t or χ2 tests. Pooling data from the pre-ACA and post-ACA periods, with stratification by cancer history, we estimated multivariable linear probability models (ordinary least squares) with the dichotomous dependent variable defined as being uninsured. The models included a categorical measure of ACA eligibility status, an indicator for observations in the post-ACA period, and an interaction between the 2. The interaction terms capture the association between eligibility category in the postperiod and coverage. The analyses also controlled for current adult age, sex, race, marital status, presence of other chronic conditions, region, and metropolitan statistical area size. We included county-specific and year-specific measures of the unemployment rate to control for variation in the economy geographically and over time. We tested for parallel trends, finding limited deviations from the assumption (see Appendix, Supplemental Digital Content 1, http://links.lww.com/MLR/B530 for estimates and discussion). Finally, we examined characteristics of adults who remained uninsured in 2014–2015, comparing those with and without cancer.
We conducted analyses using Stata 14. All analyses were weighted and SEs were adjusted to address the complex survey design of the NHIS. All statistical tests were 2-sided with an alpha of 0.05. Selected additional details related to the sample selection, measurement and analysis are provided in the Appendix (Supplemental Digital Content 1, http://links.lww.com/MLR/B530).
Over all study years (2012–2015), and after excluding observations with missing values for key variables, our sample of adult cancer survivors (n=4115, weighted N=6.9 million annually) was older (46.2% vs. 18.4% aged 56–64 y), without dependent children (77.9% vs. 64.7%), and with family income >400% FPG (42.2% vs. 36.9%), when compared with adults without a cancer history (n=89,015, weighted N=162.1 million) (Table 2). As a result of the different socioeconomic characteristics, adults with a cancer history were less likely to have been eligible for Medicaid before the expansions (4.4% vs. 6.6% for adults without cancer), less likely to have affordable alternative coverage (27.0% vs. 29.6%) and more likely to exceed the 400% FPG maximum for premium tax credits (42.7% vs. 37.1%). All differences were significant at P<0.001 (Table 2).
Before implementation of ACA coverage provisions, 12.4% of cancer survivors were uninsured, 68.4% reported private coverage, and 15.4% reported public insurance (Fig. 1). Pre-post changes reveal a decline of 4.7 percentage points (P<0.001) in the uninsured rate, and a 3.3 percentage point increase in public coverage (P=0.026) (Fig. 1). Adults without a cancer history had a higher uninsured rate (17.2%) before ACA implementation, a similar private coverage rate (70.5%), but a lower percent with public coverage (11.0%). Pre-post changes reveal a significant decline in the uninsured rate of 5.4 percentage points, and increases in both private and public coverage of 3.1 and 2.6 percentage points, respectively.
Uninsured rate decreases in the targeted eligibility groups were substantially larger. For example, among cancer survivors who became newly eligible for Medicaid, the uninsured rate declined from 25.8% to 7.4% (P<0.001); the uninsured rate trended downward (from 43.8% to 32.5%, P=0.071) among those eligible for Marketplace subsidies (full estimates are presented in Appendix Table 2, Supplemental Digital Content 1, http://links.lww.com/MLR/B530). Notably, among adults without a cancer history, the uninsured rate declined significantly for all eligibility categories, with magnitudes ranging from 1.6 percentage points among adults with alternative affordable coverage to 19.7 percentage points among adults eligible for premium subsidies in the insurance Marketplace (P<0.001).
The estimated effects of the ACA coverage provisions on the uninsured rates by eligibility category, adjusted for demographic characteristics, health status, geographic characteristics and unemployment rate are presented in Table 3. Full model results are presented in Appendix Table 3 (Supplemental Digital Content 1, http://links.lww.com/MLR/B530). Relative to the high-income category (income over 400% FPG), the interaction terms between eligibility category and post-ACA indicate that cancer survivors experienced significant decreases in the probability of being uninsured among those previously eligible for Medicaid [−8.4 percentage points; 95% confidence interval (CI), −15.6 to −1.3], newly eligible for Medicaid (−16.7 percentage points; 95% CI, −24.5 to −9.0). There was also an estimated decrease for those subsidy eligible (−11.3 percentage points; 95% CI, −23.5 to 0.8) but it did not reach statistical significance (P=0.069). For adults without a cancer history, the point estimates associated with eligibility for Medicaid pre-ACA and Medicaid expansion eligible are smaller than for cancer survivors. In contrast, the estimated effects associated with premium subsidy eligibility are larger for adults without a cancer history. The difference in estimated effects of the ACA implementation for cancer survivors and adults without a cancer history trends toward significance (P=0.086) only for expansion-eligible adults (we provide estimates for the pooled sample of cancer survivors and adults without a cancer history as Appendix Table 4, Supplemental Digital Content 1, http://links.lww.com/MLR/B530). Sensitivity analyses that exclude early expansion and late adopting states (Appendix Table 5, Supplemental Digital Content 1, http://links.lww.com/MLR/B530) reveal small differences in estimated effect sizes but no qualitative changes in results.
Despite the gains in coverage, ∼528,000 adults with a cancer history and 19.1 million without a cancer history remained uninsured in the post-ACA observation period nationally. Among uninsured cancer survivors, 32.7% reported being in fair/poor health, and 36.2% reported limitations associated with chronic conditions (Table 4). Over half (55.2%) of the uninsured were identified as likely eligible for either Medicaid (11.5%) or subsidized Marketplace plans (43.7%); whereas 16.2% were in the eligibility gap, 9.9% had alternative affordable coverage options, and 18.6% had income over 400% FPG. A slightly larger proportion of uninsured adults without a cancer history(58.9%) were eligible for Medicaid or Marketplace subsidies.
Insurance coverage provides access to care and can protect all enrolled, including cancer survivors, from overwhelming financial burden associated with care.4,6 The results of this study provide new information and insights into how coverage changed after implementation of the ACA coverage provisions for adult cancer survivors. Specifically, the 4.7 percentage point reduction in the uninsured rate for cancer survivors translates into a substantial 38% decrease in the percentage uninsured. Furthermore, the gains in coverage are substantially larger when we focus on the subgroups who became eligible for specific provisions of the ACA. We find that the largest absolute gains in coverage for cancer survivors occurred among those newly eligible for Medicaid, representing an18.3 percentage point or 71.1% relative decrease in the percent uninsured.
Compared with adults without a cancer history, cancer survivors had somewhat different patterns of coverage at baseline. After adjusting for demographic and other characteristics, the point estimates for the ACA coverage expansion effects suggested greater coverage gains for cancer survivors eligible for Medicaid, but smaller effects for those eligible through premium subsidies, relative to adults without a cancer history. However, the difference in effects on coverage between cancer survivors and those without a cancer history only trended toward significance for the Medicaid expansion eligible, and otherwise did not approach significance. We had hypothesized greater gains in coverage for cancer survivors, due to their elevated need for care and historical barriers to gaining private insurance, but the study results only partially support that hypothesis. There are several possible reasons. First, the NHIS provides a prevalent sample of adults with a cancer history, including many who are long-term survivors, and are likely quite heterogeneous with respect to their need for ongoing cancer care. This may dampen the magnitude of the estimated associations. The NHIS does not provide information that would allow us to limit our sample of cancer survivors requiring active treatment or surveillance. Second, cancer survivors showed evidence of lower uninsured rates prior to the ACA implementation. Despite barriers to coverage, it is possible that some cancer survivors used all means (paying large premiums, buying in to high-risk pools, and/or adjusting labor force participation by the survivor or spouse) to sustain or gain coverage, resulting in a smaller than expected increase associated with eligibility for Marketplace subsidies. Our NHIS measure of private insurance did not distinguish between plans from different sources.
Despite increases in coverage, 7.7% of individuals with a cancer history still did not have insurance after the ACA coverage provisions were implemented, representing just under two-thirds of the pre-ACA levels. A large portion of those uninsured were not eligible for insurance expansions. In particular, 16% of adults with a cancer history had incomes below the federal poverty guideline, but because they lived in states that did not expand Medicaid, they remained ineligible for Medicaid after January 2015. Although additional states expanded Medicaid in 2016, the likely impact on cancer survivors is unclear. An additional 10% met income eligibility requirements for Marketplace premium subsidies, but because they had employer-sponsored insurance offers that were deemed to be affordable alternatives, they could not take advantage of those subsidies. Some of these individuals may be unable to afford the employment-related insurance offered, particularly if they need to purchase coverage for family members, a problem known as the “family glitch.”25 Several policy strategies have been considered to address this barrier to coverage,26 which could be considered as part of a broader effort to strengthen access to health insurance.
This is the first study, to our knowledge, to compare changes in insurance coverage among adults aged below 65 years with a cancer history before and after the implementation of 2 major provisions (expansion of Medicaid and Marketplace insurance exchange with subsides) of the ACA using nationally representative survey data. A number of studies have examined the effects of the ACA expansions on the general population, finding gains similar to those of our noncancer history subgroup.15,18,20 One previous study by Parsons et al27 focused on the coverage effects of the 2010 expansions that allowed young adults up to age 26 to be covered on a parent’s family policy. That study found higher coverage rates among young adults newly diagnosed with cancer compared with a slightly older group who could not be covered under a parent policy. Several recent studies have examined changes in insurance coverage for newly diagnosed cancer patients23,24 but they have not yet assessed long-term effects on treatment, and there is no relevant comparison with adults without cancer.
Despite many strengths to our study, the analysis is subject to several limitations. First, the NHIS data are based on self-report of a number of key measures used in the analysis including income, health status, and health insurance, and hence, are subject to reporting error. There is no direct measure of eligibility for Medicaid or Marketplace premium subsidies in the NHIS or any other dataset, so it was necessary to generate an algorithm to assess likely eligibility, similar to prior studies evaluating insurance expansions.7,28,29 As the algorithm used self-reported information on income and family structure, there may be errors in the eligibility assignments, biasing some of the estimated associations toward the null. Second, our sample size for adults with a cancer history may have limited statistical power when assessing coverage changes by eligibility category. Finally, the analysis examines changes in coverage during the 2 years before and after implementation of the Medicaid expansions and insurance Marketplaces in January 2014, but we are unable to demonstrate a definitive causal relationship. We note that our multivariable analyses control for county-year–specific unemployment rates, which might otherwise explain some changes in insurance coverage. In addition, our study design explicitly compares changes in coverage for adults by eligibility category. Results consistent with hypothesized effects of eligibility category provide strong support for a causal relationship.
The implementation of the ACA was associated with substantial reductions in the uninsured rate for adult cancer survivors aged below 65 years. Evidence from previous adult coverage expansions, and early evidence from the ACA coverage provisions suggest important and potentially lifesaving shifts to earlier stage cancer diagnoses,30–34 improved access to cancer treatment,35 and better surveillance and survivorship care. Given the potentially longer term time frame to observe some health benefits, it is essential to monitor the ongoing effects of health care reform on insurance coverage and impacts on cancer stage at diagnosis, treatment and outcomes and to continue to explore ways to improve cancer care.
The authors wish to acknowledge helpful comments from Robin Yabroff, PhD, MBA, on earlier versions of the manuscript.
1. Ayanian JZ, Weissman JS, Schneider EC, et al. Unmet health needs of uninsured adults in the United States. JAMA. 2000;284:2061–2069.
2. McWilliams JM. Health consequences of uninsurance among adults in the United States: recent evidence and implications. Milbank Q. 2009;87:443–494.
3. Coburn N, Fulton J, Pearlman DN, et al. Treatment variation by insurance
status for breast cancer patients. Breast J. 2008;14:128–134.
4. Bernard DS, Farr SL, Fang Z. National estimates of out-of-pocket health care expenditure burdens among nonelderly adults with cancer: 2001 to 2008. J Clin Oncol. 2011;29:2821–2826.
5. Sabatino SA, Thompson TD, Richardson LC, et al. Health insurance
and other factors associated with mammography surveillance among breast cancer survivors
: results from a national survey. Med Care. 2012;50:270–276.
6. Marlow NM PA, Bian J, Ward EM, et al. The Relationship Between Insurance
Coverage and Cancer Care: A Literature Synthesis. Research Triangle Park, NC: RTI International; 2009.
7. Davidoff AJ, Hill SC, Bernard D, et al. The affordable care act and expanded insurance eligibility
among nonelderly adult cancer survivors
. J Natl Cancer Inst. 2015;107:djv181.
8. Mehnert A, de Boer A, Feuerstein M. Employment challenges for cancer survivors
. Cancer. 2013;119(suppl 11):2151–2159.
9. Kirchhoff AC, Kuhlthau K, Pajolek H, et al. Employer-sponsored health insurance
coverage limitations: results from the Childhood Cancer Survivor Study. Support Care Cancer. 2013;21:377–383.
10. Kaiser Family Foundation, ed. Kaiser Commission on Medicaid
and the Uninsured, Trends in Medicaid
and CHIP eligibility
over time. Kaiser Family Foundation Washington, DC; 2016.
11. Miller KD, Siegel RL, Lin CC, et al. Cancer treatment and survivorship statistics, 2016. CA: Cancer J Clin. 2016;66:271–289.
12. Rosenbaum S. The ACA: implications for the accessibility and quality of breast and cervical cancer prevention and treatment services. Public Health Rep. 2012;127:340–344.
13. Virgo KS, Burkhardt EA, Cokkinides VE, et al. Impact of health care reform
legislation on uninsured and Medicaid
-insured cancer patients. Cancer J. 2010;16:577–583.
14. Wherry LR, Miller S. Early coverage, access, utilization, and health effects associated with the Affordable Care Act Medicaid
Expansions: a quasi-experimental study. Ann Intern Med. 2016;164:795–803.
15. Sommers BD, Gunja MZ, Finegold K, et al. Changes in self-reported insurance
coverage, access to care, and health under the Affordable Care Act. JAMA. 2015;314:366–374.
16. McMorrow S, G JA, Long SK, et al. Medicaid
expansion increased coverage, improved affordability, and reduced psychological distress for low-income parents. Health Aff. 2017;36:808–818.
17. McMorrow S, Kenney GM, Long SK, et al. Uninsurance among young adults continues to decline, particularly in Medicaid
expansion states. Health Aff. 2015;34:616–620.
18. Buchmueller TC, Levinson ZM, Levy HG, et al. Effect of the Affordable Care Act on racial and ethnic disparities in Health Insurance
Coverage. Am J Public Health. 2016;106:1416–1421.
19. Chen J, Vargas-Bustamante A, Mortensen K, et al. Racial and ethnic disparities in health care access and utilization under the Affordable Care Act. Med Care. 2016;54:140–146.
20. Cohen RA MM, Zammitti EP. NCHS. Health Insurance
coverage: early release of estimates from the National Health Interview Survey, 2015. National Health Interview Survey Early Release Program. Hyattsville, MD: National Center for Health Statistics; 2016.
21. Wen H, Druss BG, Cummings JR. Effect of Medicaid
expansions on Health Insurance
Coverage and access to care among low-income adults with behavioral health conditions. Health Serv Res. 2015;50:1787–1809.
22. Jessica P, Vistnes P, Lipton B, et al. STATISTICAL BRIEF #490: Uninsurance and Insurance
Transitions Before and After 2014: Estimates for US, Non-Elderly Adults by Health Status, Presence of Chronic Conditions and State Medicaid
Expansion Status. Rockville, MD: Agency for Healthcare Research and Quality; 2016.
23. Jemal A, Lin CC, Davidoff AJ, et al. Changes in Insurance
Coverage and stage at diagnosis among nonelderly patients with cancer after the Affordable Care Act. J Clin Oncol. 2017;35:3906–3915.
24. Soni A, Sabik LM, Simon K, et al. Changes in insurance
coverage among cancer patients under the Affordable Care Act. JAMA Oncol. 2017;4:122–124.
25. Buettgens M, Dubay L, Kenney GM. Marketplace subsidies: changing the ‘Family Glitch’ reduces family health spending but increases government costs. Health Aff. 2016;35:1167–1175.
26. Nowak SA, Saltzman E, Cordova A. Alternatives to the ACA's Affordability Firewall. Rand Health Q. 2016;5:3.
27. Parsons HM, Schmidt S, Tenner LL, et al. Early impact of the Patient Protection and Affordable Care Act on insurance
among young adults with cancer: analysis of the dependent insurance
provision. Cancer. 2016;122:1766–1773.
28. Davidoff AJ, Garrett AB, Makuc DM, et al. Medicaid
-eligible children who don't enroll: health status, access to care, and implications for Medicaid
enrollment. Inquiry. 2000;37:203–218.
29. Davidoff AJ, Yemane A, Hill I. Public insurance eligibility
and enrollment for special health care needs children. Health Care Financing Rev. 2004;26:119–135.
30. Amini A, Jones BL, Yeh N, et al. Disparities in disease presentation in the four screenable cancers according to health insurance
status. Public Health. 2016;138:50–56.
31. Mahal BA, Aizer AA, Ziehr DR, et al. The association between insurance
status and prostate cancer outcomes: implications for the Affordable Care Act. Prostate Cancer Prostatic Dis. 2014;17:273–279.
32. Han X, Zhu S, Tian Y, et al. Insurance
status and cancer stage at diagnosis prior to the Affordable Care Act in the United States. J Registry Manag. 2016;41:143–151.
33. Lissenden B, Yao NA. Affordable Care Act changes to medicare led to increased diagnoses of early-stage colorectal cancer among seniors. Health Aff. 2017;36:101–107.
34. Han X, Zang Xiong K, Kramer MR, et al. The Affordable Care Act and cancer stage at diagnosis among young adults. J Natl Cancer Inst. 2016;108:djw058.
35. Robbins AS, Han X, Ward EM, et al. Association between the Affordable Care Act dependent coverage expansion and cervical cancer stage and treatment in young women. JAMA. 2015;314:2189–2191.