Policy Points:
In a national survey of approximately 8,000 adults in 2015, we found large income‐ and race‐based disparities in perceived health care quality, affordability, and use of emergency departments.
Lack of health insurance is one factor that contributes to worse health care experiences among lower‐income Americans and racial/ethnic minorities, but it only explains a small to moderate portion of these disparities.
While the Affordable Care Act has led to significant improvements in health care access and affordability, large gaps remain. Repeal of the law would undo much of this progress, but even if the law remains in effect, policymakers need to address other social determinants that contribute to ongoing income‐ and race‐based disparities in health care.
Context
The Affordable Care Act (ACA) has reduced the US uninsured rate to a historic low. But coverage is only one of many factors contributing to race‐ and income‐based disparities in health care access, affordability, and quality.
Methods
Using a novel 2015 national survey of more than 8,000 Americans, we examined disparities between low‐income and high‐income adults and between racial/ethnic minorities and whites. We conducted a series of regression analyses, starting with models that only took into account income or race, and then sequentially adjusted for health insurance, state of residence, demographics, and health status. We examined self‐reported quality of care, cost‐related delays in care, and emergency department (ED) use due to lack of available appointments. Then we used multivariate regression to assess respondents’ views of whether quality and affordability had improved over the past 2 years and whether the ACA was helping them.
Findings
Quality of care ratings were significantly worse among lower‐income adults than higher‐income adults. Only 10%‐25% of this gap was explained by health insurance coverage. Cost‐related delays in care and ED use due to lack of available appointments were nearly twice as common in the lowest‐income group, and less than 40% of these disparities was explained by insurance. There were significant racial/ethnic gaps: reported quality of care was worse among blacks and Latinos than whites, with 16%‐70% explained by insurance. In contrast to these disparities, lower‐income and minority groups were generally more likely than whites or higher‐income adults to say that the ACA was helping them and that the quality and/or affordability of care had improved in recent years.
Conclusions
Our post–health reform survey shows ongoing stark income and racial disparities in the health care experiences of Americans. While the ACA has narrowed these gaps, insurance expansion alone will not be enough to achieve health care equity.
Introduction
Disparities in us health care are a source of considerable public health and policy concern, with substantial evidence that minorities and low‐income Americans experience greater barriers to care and worse health outcomes across numerous measures.1, 2 At the same time, the United States is currently in the midst of the largest overhaul of the health care system in more than 50 years, with the passage and implementation of the Affordable Care Act (ACA). Evidence shows that the ACA expanded health insurance to nearly 20 million individuals and brought the uninsured rate to an all‐time low.3 Whether—and how much—this expansion of coverage narrowed disparities in health care is unclear.
Cross‐sectional studies from before the ACA demonstrate that coverage is just one aspect of disparities in health care experienced by racial/ethnic minorities and those with low incomes. Even among those without insurance, access to a regular source of care and health care utilization rates differ significantly among racial and ethnic groups,4, 5 with studies suggesting contributions from factors such as educational attainment, language barriers, citizenship, and neighborhood effects.6, 7
Previous health insurance expansions have a mixed record in terms of improving equity. Often called the model for the ACA, Massachusetts’ 2006 health reform led to improved access to outpatient care for vulnerable populations in the state, including non‐elderly adults living in low‐ and middle‐income areas, elderly adults, and non‐elderly Hispanic adults.8 Some studies found that the state's policy reduced disparities. For instance, the state's reform was associated with a significant decrease in mortality and a narrowing of disparities, with mortality improvements largest among nonwhites and those living in poorer counties.9 Another survey‐based study found that improvements in self‐reported health after Massachusetts health reform were largest for lower‐income adults and minorities.10 However, not all research found a reduction in disparities after the state's reform. In these studies, even though vulnerable populations in Massachusetts experienced improvements in cost‐related barriers and coverage, similar or larger gains were observed among white and nonpoor groups, resulting in no significant progress toward the elimination of racial disparities for many outcomes.11, 12
State‐level Medicaid expansions preceding the ACA have, by definition, disproportionately benefited lower‐income individuals, since they are the ones eligible for the program. Evidence of Medicaid's impact on racial disparities, however, is less clear. Large Medicaid expansions in the early 2000s in New York, Maine, and Arizona were associated with significant reductions in all‐cause mortality, as compared to demographically similar neighboring states that did not expand Medicaid. These gains were greatest among racial and ethnic minorities and residents of poorer counties, suggesting that state Medicaid expansions may reduce mortality disparities among vulnerable groups.13 Other studies of Medicaid expansions found improvements in access to care and self‐reported health, but did not provide information on how these effects varied by race or socioeconomic status.14, 15
Researchers also examined the impact of the ACA's 2010 dependent coverage provision, which allowed adults to remain on their parents’ health plans through age 26, on disparities among young adults. Studies indicate significant gains in insurance coverage and reduced out‐of‐pocket spending, but mixed progress when it comes to narrowing disparities. Among young adults ages 19‐25, the dependent coverage provision increased private coverage for men and women, for most racial and ethnic minorities, for those with limited English proficiency, and for those with and without citizenship.16 However, net gains were greater for whites than for other racial or ethnic minorities,17 and one study found evidence that the policy primarily benefited higher‐income families.18
While much of the research on disparities has focused on race and ethnicity, gaps in health care coverage and access related to income are also of significant concern. Moreover, widening income inequality19—combined with the steady rise of health care costs over the past several decades20—poses particular challenges for health care access, which the ACA in part was designed to mitigate.21
Since the beginning of the ACA's major insurance expansions in 2014, several studies demonstrated larger coverage gains among lower‐income groups and minorities, with some concurrent improvements in access to primary care and affordability of care.22, 23, 24, 25, 26 For instance, one study found that reductions in the uninsured rate among blacks and Latinos were nearly twice as large as those among whites.23 Meanwhile, the uninsured rate for those living below the poverty level fell from 33% in 2013 to 25% by 2016, compared to a much smaller drop, from 12% to 8%, among those with incomes between 250% and 400% of the poverty level.26
While many of these prior studies have used pre‐post comparisons or quasi‐experimental study designs to evaluate the effect of coverage expansions on disparities, as noted earlier, other studies have used multivariate cross‐sectional approaches to evaluate the extent to which baseline income and racial disparities in access to care and health care quality can be attributed to insurance differences across groups. These comparisons indicate that coverage plays a significant role in these gaps, but is not the only factor at play.6 However, to our knowledge, there has been little post‐ACA analysis of the remaining disparities in health care—particularly in terms of perceived health care quality—and how much of a role health insurance coverage still plays in these gaps.
Our study objectives were (1) to examine disparities based on race/ethnicity and income in perceived health care quality, access to care, and affordability of care, using a post‐ACA sample of adults; (2) to estimate what proportion of these disparities could be attributed to differences in health insurance coverage across groups; and (3) to compare perceptions across groups of how the ACA and recent trends have affected these outcomes.
Methods
Survey Data
Our study data are from the “Patients’ Perspectives on Health Care in the United States” survey, a project conducted by the Harvard T.H. Chan School of Public Health, the Robert Wood Johnson Foundation, and National Public Radio.27 The survey was a random‐digit dialing telephone survey (of both cell phones and landlines), fielded by the research firm SSRS. Interviews were available in English and Spanish, and calls were completed between September 8 and November 9, 2015, among adults ages 18 or older. In each contacted household, one eligible respondent was selected at random to participate in the survey. The study contained 8 different subsamples, each with roughly 1,000 respondents. The first group was a nationally representative sample in all 50 states and the District of Columbia. The other samples were from 7 states: Florida, Kansas, New Jersey, Ohio, Oregon, Texas, and Wisconsin. These states were selected to represent a geographically and demographically diverse group of states that have not been studied extensively by other polls and represent a range of policy environments related to the Affordable Care Act.
The final sample contained 1,002 adults in the national sample and 7,036 adults total in the 7 states. The study oversampled African‐American/blacks, Latinos, and adults with annual household incomes of less than $25,000. The overall response rate was 15%, calculated according to the American Association for Public Opinion Research's Response Rate 3 definition.28 Data from each of the 8 subsamples were weighted by cell phone/landline use and demographics (eg, sex, age, race/ethnicity, education, and household income) to reflect the appropriate population, based on data from the US Census Bureau and National Health Interview Survey. Further details about the survey design are available in the Appendix (available online).
The survey collected data on demographic information, personal health care experiences, perceptions of health care in their respective states, and changes in these measures over the past year. The survey's chief advantages for our research purposes were its timeliness, enabling analysis of outcomes nearly 2 years into the implementation of the ACA, and the use of several health care–related domains that are not typically covered by federal surveys.
Study Outcomes
We assessed several outcomes related to perceived quality and affordability of care, ED use due to lack of available appointments (as a measure of health care access), and perceptions of the ACA.
For quality, we asked respondents, “Overall, how would you rate the health care you receive?” on a 4‐point scale (excellent, good, fair, or poor). We then asked whether the quality of care had gotten better, worse, or stayed the same over the past 2 years.
For affordability, we asked respondents whether they had needed health care in the past 2 years but did not get it because they could not afford that care. We also asked whether their care had become more affordable, less affordable, or stayed the same over the past 2 years.
For ED use, we asked whether they had used the ED in the past 2 years and then, among those with an ED visit, whether the primary reason was that “other facilities were not open or you could not get an appointment.” We focus on this particular outcome, rather than on any ED use, since numerous factors influence ED use that are not likely to be related to health insurance (such as transportation issues, availability of paid sick leave, and geographic proximity). We focus on appointment availability as a meaningful assessment of access to outpatient care.
Finally, we asked each respondent, “Would you say the Affordable Care Act, also called Obamacare, has directly helped you, directly hurt you, or has it not had a direct impact?”
Covariates
Several of our models included covariates as described below. Covariates were selected based on the Andersen revised behavioral model for access to health care.29 The factors that increase one's likelihood of using medical care, which Andersen terms “predisposing characteristics,” include sex, age, education, and race and ethnicity. We used insurance information, income, and state of residence as our main indicators of “enabling resources”—that is, those factors that affect one's ability to obtain health care services. Finally, we added self‐reported health status (on a 5‐point scale) as a proxy measure of one's need for medical care.
Statistical Analysis
We analyzed our data in several steps. First, we assessed for the presence of disparities in our study outcomes in unadjusted models by race/ethnicity and separately by income. Race/ethnicity was categorized into white non‐Latino, black non‐Latino, Latino, and other/missing. Household income was categorized as less than $25,000 per year, $25,001‐$50,000, $50,001‐$100,000, greater than $100,000, and income not reported. These models were simple linear probability models, with whites as the omitted reference group for race/ethnicity, and the highest income group omitted for income. Thus, the coefficients for each group identify the disparity relative to whites or those earning more than $100,000, respectively, without adjustment for any other demographic or health‐related differences. This baseline unadjusted measure of disparity is consistent with the Institute of Medicine's recommendations and with prior research.1, 30
Next, we added health insurance information as a covariate to our regression models, in the following categories: Medicaid, employer‐sponsored insurance, Medicare, ACA Marketplace insurance, other coverage, and uninsured. This model produces coefficients that indicate the disparities that remain based on income or race/ethnicity, after adjustment for differences in health insurance coverage across groups. We then present a fully adjusted model that includes the following complete list of covariates: age, sex, education, income, race/ethnicity, self‐reported health status, state of residence, and health insurance. By comparing the coefficients across these models, we are able to assess the contribution of health insurance differences to disparities for each of our study outcomes.
Then, using the outcomes related to changes over time, we used our multivariate linear model to assess what factors were associated with improved quality and affordability in the last 2 years and with a perception that the ACA had directly helped the respondent. These outcomes were each coded on a 3‐point scale from negative to positive (eg, quality of care has gotten worse, stayed the same, or improved).
For all analyses, we separately analyzed the nationally representative sample (n = 1,002) and the 7‐state sample (n = 7,036). Both analyses used survey weights to approximate the target population in each sample; the 7 states were weighted based on the 2014 population size in each state according to the US Census Bureau's American FactFinder. Thus, for each outcome and model, we produced a national estimate and an aggregated 7‐state estimate. All regressions used a linear model to provide straightforward estimates of the magnitude of change for each outcome across subgroups; for assessments of disparities, odds ratios or other nonlinear estimates are more difficult to interpret. However, we tested the robustness of our results to those obtained using predicted probabilities from a logistic model and the results were quite similar. We also compared the results when splitting our 7‐state sample into Medicaid expansion and non‐expansion states.
The study investigators had access only to deidentified survey data, and the protocol was exempted from review by the Harvard Institutional Review Board. Analyses used Stata 14.0.
Results
Perceived Quality of Care
Tables 1 and 2 present disparities by income and race/ethnicity, respectively, for the proportion of adults reporting that the quality of their overall health care was “fair” or “poor.”
Table 1.
Disparities in Receipt of Fair or Poor Care by Incomea
| MODEL 1: Unadjusted Difference | MODEL 2: Adjusted for Health Insurance | MODEL 3: Adjusted for Health Insurance, Race, State, Health Status, and Demographics (age, sex, education) | |||||
|---|---|---|---|---|---|---|---|
| Income Group (Annual) | Sample Size (n) | US (n = 975) | 7 States (n = 6,883) | US (n = 975) | 7 States (n = 6,883) | US (n = 975) | 7 States (n = 6,883) |
| Less than $25,000 | 2,490 | 29.1%*** | +17.1%*** | 26.0%*** | +13.2%*** | 20.1%*** | +6.6** |
| $25,001‐$50,000 | 1,651 | 10.4%*** | +11.3%*** | 9.2%** | +8.6%*** | 7.8%* | +5.0%** |
| $50,001‐$100,000 | 1,669 | 6.9%* | +3.6%* | 7.1%* | +2.8% | 9.1%** | +2.3% |
| Greater than $100,000 | 1,150 | Reference | Reference | Reference | Reference | Reference | Reference |
| Income Not Reported | 898 | 11.7%** | +6.2%*** | 10.4%** | +4.4%* | 9.2%* | +0.7% |
All results report percentage‐point differences relative to the highest annual income group (Greater than $100,000; the reference group for this category). Outcome mean for highest income group was 6.5% for the national sample and 11.0% for the 7‐state sample. Sample excludes those with missing data for the outcome variable.
*P<.10, **P<.05, ***P<.01.
Table 2.
Disparities in Receipt of Fair or Poor Care by Race/Ethnicitya
| MODEL 1: Unadjusted Difference | MODEL 2: Adjusted for Health Insurance | MODEL 3: Adjusted for Health Insurance, Race, State, Health Status, and Demographics (age, sex, education) | |||||
|---|---|---|---|---|---|---|---|
| Race/Ethnicity | Sample Size (n) | US (n = 975) | 7 States (n = 6,883) | US (n = 975) | 7 States (n = 6,883) | US (n = 975) | 7 States (n = 6,883) |
| White non‐Latino | 5,196 | Reference | Reference | Reference | Reference | Reference | Reference |
| Black non‐Latino | 845 | 11.1%** | 4.9%* | 9.3%* | 2.8% | 7.2% | −0.6% |
| Latino | 1,136 | 11.9%** | 7.4%*** | 8.3% | 2.2% | 1.5% | −3.3% |
| Other/missing | 681 | 12.9%** | 6.3%** | 13.4%** | 4.7%* | 10.9%** | −3.0% |
All results report percentage‐point differences relative to whites (the reference group for this category). Outcome mean for whites was 14.7% for the national sample and 17.8% for the 7‐state sample. Sample excludes those with missing data for the outcome variable.
*P<.10, **P<.05, ***P<.01.
In Table 1, the unadjusted data (Model 1) demonstrates significantly worse care at lower incomes for each step along the income distribution. At the extremes, those in the lowest income group reported receiving fair or poor care at a rate 29.1 percentage points higher than those in the highest income group for the national sample and 17.1 percentage points in the 7‐state sample (both P < .01). For comparison, in the highest income group, only 6.5% of the national sample and 11.0% of the 7‐state sample reported fair or poor quality of care. Adjustment for health insurance status (Model 2) reduced these disparities only slightly, to 26.0 and 13.2 percentage points, respectively. This implies that health insurance explained just 11% to 23% of the disparity for the lowest‐income group compared to the highest group. Meanwhile, full multivariate adjustment (Model 3) still left large residual disparities for lower‐income adults in both the national and the 7‐state samples.
In Table 2, we see significant disparities in receipt of fair or poor care based on race/ethnicity, though smaller than the disparities based on income. Blacks reported rates of fair/poor care 11.1 percentage points (P < .05) and 4.9 percentage points (P < .10) higher than whites in the national and the 7‐state samples, respectively, while Latinos experienced disparities for this measure of 11.9 (P < .05) and 7.4 percentage points (P < .01), compared to whites. Again, adjustment for health insurance (Model 2) narrowed these gaps somewhat, but only by 16% for blacks and 30% for Latinos in our national sample. In the 7‐state model, however, insurance played a larger role, eliminating 43% of the black‐white disparity and 70% of the white‐Latino disparity. After full multivariate adjustment (Model 3), no statistically significant disparities remained for blacks and Latinos.
Problems Affording Needed Care
Tables 3 and 4 present disparities by income and race/ethnicity, respectively, for the proportion of adults reporting that they had not obtained needed medical care due to cost in the previous 2 years.
Table 3.
Disparities in Health Care Not Obtained Due to Costs, by Incomea
| MODEL 1: Unadjusted Difference | MODEL 2: Adjusting for Health Insurance | MODEL 3: Adjusted for Health Insurance, Race, State, Health Status, and Demographics (age, sex, education) | |||||
|---|---|---|---|---|---|---|---|
| Income Group (Annual) | Sample Size (n) | US (n = 962) | 7 States (n = 6,804) | US (n = 962) | 7 States (n = 6,804) | US (n = 962) | 7 States (n = 6,804) |
| <$25,000 | 2,312 | 10.4%*** | 10.8%*** | 6.5%** | 7.7%*** | 6.3%* | 8.0%*** |
| $25,001‐$50,000 | 1,593 | 9.2%** | 11.5%*** | 7.5%** | 8.9%*** | 7.8%** | 10.2%*** |
| $50,001‐$100,000 | 1,637 | 1.9% | 7.5%*** | 2.0% | 6.9%*** | 2.0% | 7.7%*** |
| Greater than $100,000 | 1,136 | Reference | Reference | Reference | Reference | Reference | Reference |
| Income not reported | 818 | 0.8% | 5.8%*** | −0.9% | 4.7%** | 1.1% | 6.6%*** |
All results report percentage‐point differences relative to the highest annual income group (Greater than $100,000; the reference group for this category). Outcome mean for highest income group was 3.9% for the national sample and 3.4% for 7‐state sample. Sample excludes those with missing data for the outcome variable.
*P<.10, **P<.05, ***P<.01.
Table 4.
Disparities in Health Care Not Obtained Due to Costs, by Race/Ethnicitya
| MODEL 1: Unadjusted Difference | MODEL 2: Adjusted for Health Insurance | MODEL 3: Adjusted for Health Insurance, Race, State, Health Status, and Demographics (age, sex, education) | |||||
|---|---|---|---|---|---|---|---|
| Race/Ethnicity | Sample Size (n) | US (n = 962) | 7 States (n = 6,804) | US (n = 962) | 7 States (n = 6,804) | US (n = 962) | 7 States (n = 6,804) |
| White non‐Latino | 5,146 | Reference | Reference | Reference | Reference | Reference | Reference |
| Black non‐Latino | 835 | 1.8% | 0.7% | 0.1% | −0.7% | −0.3% | −3.0% |
| Latino | 1,118 | 1.4% | −1.0% | −1.1% | −6.6%*** | −3.6% | −9.3%*** |
| Other/missing | 667 | −0.7% | 2.5% | −0.2% | 1.0% | −4.0% | 0.3% |
All results report percentage‐point differences relative to whites (the reference group for this category). Outcome mean for whites was 8.7% for the national sample and 11.3% for the 7‐state sample. Sample excludes those with missing data for the outcome variable.
*P<.10, **P<.05, ***P<.01.
In Table 3, we see large income‐based disparities. The lowest income group reported rates 10.4 percentage points greater of skipping needed care due to cost in the national sample and 10.8 percentage points in the 7‐state sample, compared to the highest income group. Those earning between $25,000 and $50,000 experienced similarly large disparities. Adjusting for health insurance shrank these gaps somewhat, reducing the disparities between the highest and lowest income groups by 38% in the national sample and 29% in the 7‐state sample. In contrast to these results, Table 4 does not indicate any significant racial/ethnic disparities in skipping needed care due to cost in the unadjusted models. In fact, in the fully adjusted model, rates of skipping needed care were actually lower among Latinos than whites (−9.3 percentage points, P < .01) in the 7‐state sample, though there are no significant differences in the national sample. In part this may relate to the much smaller sample size for the national analysis compared to the state analysis.
Appendix Table 1 (available online) presents coefficients for the other covariates in the full model. These results indicate that quality of care and ability to afford care were generally better among those with Medicaid, Medicare, or employer‐sponsored insurance than the uninsured. Non‐elderly adults were more likely to report fair/poor quality care or affordability problems than were adults over age 65. Adults in fair/poor health were much more likely to report poor quality of care and cost‐related delays in care, compared to adults in excellent health. Educational status and gender were inconsistent predictors of these outcomes.
ED Use Due to Lack of Available Appointments
Appendix Tables 2 and 3 (available online) present disparities by income and race/ethnicity for the proportion of adults reporting that they had visited the ED in the past 2 years because they could not obtain an outpatient appointment in time. Rates were 5.7 (national sample) and 3.1 percentage points (7‐state sample) higher for the lowest‐income group than the highest income group, though only the latter was statistically significant. These estimates dropped by 22%‐26% after adjustment for insurance type. Meanwhile, disparities for this measure were larger for blacks versus whites, with a gap of 10 percentage points (P < .10) in the national sample and 4.5 percentage points (P < .05) in the 7‐state sample. Roughly 10% of these disparities were explained by health insurance type. In contrast to blacks, Latinos had similar or even lower rates of ED visits due to lack of appointments than did whites.
Perceived Changes in Health Care Over Time
Table 5 presents results from multivariate regressions assessing respondents’ perceptions of how their health care has changed over the prior 2 years. Each coefficient shows the changes on a 3‐point scale (where −1 is getting worse, 0 is unchanged, and +1 is improving), compared to the reference group in each category. In both the national and 7‐state samples, we find consistent evidence that blacks and Latinos were far more likely than whites to report that the ACA had personally helped them (P < .01); on average, whites felt the law had hurt them, while nonwhites reported that it had helped them. Those with Medicaid were also much more likely to report that the ACA had helped them, as were those with Marketplace coverage in the 7‐state sample; meanwhile, those without health insurance felt the law had hurt them on average. The lowest‐income group also reported more favorable views toward the ACA in the 7‐state sample, even after adjustment for health insurance and race/ethnicity.
Table 5.
Impact of Income, Race, and Health Insurance on Perceived Changes in Health Care
| ACA Has Directly Helped vs Hurt Youb, c | Change in Quality of Care Over Past 2 Yearsc | Costs of Health Care Have Become More Affordable Over Past 2 Yearsc | ||||
|---|---|---|---|---|---|---|
| Variablea | US (n = 466) | 7 States (n = 3,383) | US (n = 974) | 7 States (n = 6,861) | US (n = 961) | 7 States (n = 6,821) |
| Income | ||||||
| <$25,000 | .069 | .173*** | .037 | .008 | .190** | .157*** |
| $25,001‐$50,000 | .082 | .054 | .036 | −.009 | −.018 | .096*** |
| $50,001‐$100,000 | −.113 | −.004 | −.031 | −.036 | −.017 | .019 |
| >$100,000d | Reference | Reference | Reference | Reference | Reference | Reference |
| (−.049) | (−.176) | (.069) | (.054) | (−.264) | (−.339) | |
| Not reported | −.350*** | .114* | −.100 | .036 | −.012 | .115** |
| Race | ||||||
| White non‐Latinod | Reference | Reference | Reference | Reference | Reference | Reference |
| (−.225) | (−.191) | (.013) | (−.020) | (−.307) | (−.343) | |
| Black non‐Latino | .482*** | .427*** | .236*** | .267*** | .165** | .319*** |
| Latino | .323*** | .142*** | .168*** | .146*** | .100 | .209*** |
| Other | .183* | .143** | −.044 | −.032 | −.018 | .068 |
| Insurance | ||||||
| Medicaid | .467*** | .326*** | .095 | .126** | .284*** | .246*** |
| Employer‐based | .026 | −.010 | −.014 | .071* | −.030 | .029 |
| Medicare | .016 | .153** | −.109 | .111** | .037 | .196*** |
| Marketplace | .215 | .599*** | .001 | .120 | .163 | .197* |
| Other coverage | .164 | .113 | −.037 | .042 | .094 | .164*** |
| Uninsuredd | Reference | Reference | Reference | Reference | Reference | Reference |
| (−.179) | (−.198) | (.090) | (−.040) | (−.312) | (−.363) | |
Regressions adjust for income, race, insurance, education, age, sex, health status, and state of residence.
Question was only asked of half the survey sample.
All questions were rated on a 3‐point scale (Yes/Better, No Effect/No Change, or No/Worse), with positive numbers indicating better outcomes.
Adjusted mean for each reference group is listed in parentheses, using the Stata “margins” command.
*P<.10, **P<.05, ***P<.01.
Blacks and Latinos were significantly more likely than whites to say that the quality of their health care had gotten better over the past 2 years. On average, whites reported little to no change in quality, in contrast to the significant improvements reported by minorities. Medicaid beneficiaries also reported improving quality of care in the 7‐state sample. Meanwhile, overall affordability declined for all racial groups, but less so for blacks and Latinos than for whites. Those with Medicaid coverage were more likely to say that their health care had become more affordable than did those without insurance or those with employer‐based coverage. Lower‐income respondents were more likely than their higher‐income peers to report that their care had become more affordable in the past 2 years (P < .05), though they did not report any significant changes in quality of care.
Medicaid Expansion Versus Non‐Expansion
We repeated our main analyses using the 7‐state sample stratified into expansion versus non‐expansion; the national sample was not large enough to support this analysis. Appendix Tables 4–8 (available online) report these results. Rates of fair/poor care were higher in non‐expansion states than in expansion states (12.6% vs 8.0% for the highest‐income group and 18.5% vs 16.0% for whites). The pattern of disparities between income groups and between whites versus blacks were similar in the two groups of states, though white‐Latino disparities were smaller in non‐expansion states. Patterns of income and racial disparities in cost‐related delays in care were similar across expansion and non‐expansion states.
However, larger differences were evident for changes in these outcomes over time, based on expansion status (online Appendix Table 8). Lower‐income adults in expansion states were much more likely to report that the ACA had directly helped them, compared to lower‐income adults in non‐expansion states. Medicaid recipients in expansion states were much more likely to report that the ACA had helped them—by 3 times as much—compared to those in non‐expansion states; meanwhile, Marketplace recipients rated the ACA more highly in non‐expansion states (where a higher share of low‐income adults are eligible for Marketplace subsidies in lieu of Medicaid). Medicaid beneficiaries in expansion states were also much more likely to report improving affordability of care than were their counterparts in non‐expansion states. Patterns by race/ethnicity did not differ dramatically across expansion versus non‐expansion states—in both groups of states, blacks and Latinos were more likely than whites to say that their care was improving and becoming more affordable.
Discussion
In our 2015 survey of nearly 8,000 Americans, we find large racial and economic disparities in affordability of medical care, perceived quality of care, and access to timely outpatient care. Thus, even 2 years into the largest expansion of health insurance in 50 years, inequality remains a fundamental attribute of American health care. We find evidence that health insurance coverage can help narrow some of these gaps, and minorities and adults with lower incomes tend to feel most positively about the ACA and recent changes in their own health care. Moreover, Medicaid expansion was generally associated with larger changes in favor of lower‐income and minority groups. But even so, health insurance coverage only explains a small to moderate portion of the ongoing disparities in affordability, quality, and access.
Most of the disparities we document here are evident in both the national sample and in the 7‐state sample and exist for both comparisons of whites versus nonwhites and higher‐income versus lower‐income adults. However, for most outcomes, the differences were larger across income groups than racial/ethnic groups. For instance, the lowest‐income group reported receiving fair or poor quality of care at a rate nearly 30 percentage points greater than the top income group (representing a nearly fivefold increase), while blacks and Latinos reported receiving fair or poor care at rates 11–12 percentage points higher than whites. Moreover, the income‐based disparities persisted to a greater extent after multivariate adjustment than did racial/ethnic disparities. In this context, the ACA's income‐based approach to coverage expansion is likely to improve equity, which is consistent with other evidence on the law to date.22, 24, 26
Previous research documented the major gains in coverage under the ACA, with larger gains for minorities and low‐income adults.23, 24, 25 Our findings add to our understanding of these issues through the use of a novel survey, which included consumer‐rated health care quality and reasons for ED use, which have not been characterized in prior ACA‐related research, as well as our comparison of both a nationally representative sample and a more in‐depth examination of 7 diverse states.
The largest gaps we observed were for perceived quality of health care. How might poverty and race affect quality, even after controlling for health insurance coverage? Among those with insurance, cost‐sharing requirements have increased consistently over recent years leading to a problem of “underinsurance”31 that likely places disproportionate burdens on lower‐income groups. These financial barriers—which are evident in our data as well—may interfere with their ability to get the care they desire and see the providers they consider to be of high quality. For racial/ethnic minorities, lack of cultural competence, language barriers, and mistrust of health care institutions due to historic abuses may all contribute to worsened perceptions of health care quality.32, 33
We find less pronounced but still significant disparities in reliance on the ED due to a lack of available appointments for blacks and lower‐income adults. Some of this is mediated by insurance coverage—not only whether one is uninsured but also by the type of insurance, consistent with previous work showing lower provider participation rates in Medicaid than in private insurance34 and with some studies showing higher ED use associated with Medicaid coverage.35 Interestingly, we did not find elevated rates of ED usage due to lack of appointments among Latinos, compared to whites, which is consistent with previous research showing lower overall utilization rates including in the ED among both native‐born and noncitizen Latinos.7, 36
Our analysis of attitudes toward the ACA and perceived changes in health care over time provide a silver lining in these large disparities. Lower‐income adults and racial/ethnic minorities were much more likely than other groups to report that their care has become more affordable in the past 2 years, and similar progress is evident for blacks and Latinos regarding quality of care and whether they felt the ACA had directly helped them. In multivariate models, health insurance itself was also a strong predictor of attitudes toward the ACA, with those who have Medicaid or Marketplace coverage the most likely to report that the law had helped them. These general patterns are consistent with other polling data on the health reform law.37, 38
Limitations
Our study relies on self‐reported measures of quality, affordability, and access. These may be influenced by a variety of cultural and economic factors, and perceptions of these factors may themselves be subject to disparate interpretations across the lines of race/ethnicity and income. All of our outcomes are also subject to potential recall bias or social desirability bias. However, the general patterns we detect here are consistent with a large body of evidence that points to the existence of fundamental inequities in health care,1 suggesting that these are not just subjectively perceived differences or measurement error.
Our approach of adding health insurance status to an unadjusted model as the first set of covariates means that these variables may capture both direct effects of insurance as well as some confounding factors tightly associated with insurance (such as state of residence, citizenship, or age). Thus, if anything, the difference between Models 1 and 2 in our findings may overstate the actual contribution of health insurance to the disparities we find in our data. Our results can therefore be seen as the upper bound of how much insurance expansion might close these gaps in quality, affordability, and access.
Our survey is also subject to potential sources of nonsampling error, including nonresponse bias, question wording, and ordering effects. Nonresponse in random‐digit dialing telephone surveys produces some known biases in survey‐derived estimates because participation tends to vary for different subgroups of the population. To compensate for these known biases and for variations in probability of selection within and across households, as well as the relatively low response rate, we weighted to population benchmarks using federal survey data, which has been shown to mitigate the potential for nonresponse bias and produce estimates that closely resemble results from government in‐person surveys.39, 40, 41
Finally, while our analysis includes outcomes explicitly related to the ACA, we are not conducting a quasi‐experimental evaluation of the health reform law comparable to many of the studies discussed in the introduction. Rather, we are attempting to decompose health care disparities into underlying contributing factors, with health insurance coverage as the key variable of interest. On a related note, our questions often asked about health care experiences over the prior 2 years; given our survey's timing in the fall of 2015, the time frame spanned both pre‐ and post‐ACA periods. For questions about overall experiences over the prior 2 years, the time frame covered a mixture of pre‐ and post‐ACA experiences. For questions about changes in health care experiences over the past 2 years, the time frame is relatively consistent with evaluating changes concurrent with the ACA's implementation.
Policy Implications and Conclusions
Long‐standing disparities in health care access, quality, and affordability continue in the post‐ACA period, with lower‐income families and racial/ethnic minorities generally experiencing more cost‐related barriers to care, worse perceived health care quality, and more difficulty obtaining needed appointments. While lower‐income and minority respondents were generally more supportive of the ACA and reported improving trends in these outcomes relative to higher‐income and white adults, our analysis suggests that health insurance only explains a small to moderate portion of the baseline disparities.
The reasons for these remaining gaps in care are not completely clear and are beyond the scope of our data to directly assess. However, others have suggested 2 broad explanations: (1) structural and (2) social determinants. The first category suggests that there are not enough accessible and high‐quality health‐related resources in many low‐income and minority communities.42, 43 Factors relevant to this argument include provider shortages and limited facilities for advanced diagnostic testing or treatment. Increased coverage only mitigates a share of these shortfalls. Potential policy solutions to these challenges include expanding not only coverage but also financial support for safety net providers, such as federally qualified health centers and safety net hospitals. While the ACA did temporarily increase federal grants to health centers, more sustained and predictable long‐term funding streams could be even more helpful to a permanent expansion of health care capacity in disadvantaged urban neighborhoods and rural areas.44 As for hospitals, the ACA's planned cuts to Medicaid Disproportionate Share Hospital payments may hamper efforts to close some of these disparities in quality and access.45 Another area of policy that is a necessary complement to coverage expansion is the health care workforce. Increasing financial incentives and programs to practice in underserved settings, such as the National Health Service Corps,46 and making concerted efforts to train a racially and ethnically diverse provider workforce47 could potentially improve the availability and quality of care for vulnerable populations. Whether these approaches would be successful in narrowing disparities is unclear and worthy of future study. However, given the results of the 2016 election, it is unlikely that any expansion in these programs is in the offing and, rather, that the main policy debate of the coming year is likely to be whether to maintain the ACA's coverage gains at all.
The other broad explanation for persistent disparities relates to the broader circumstances that lower‐income and minority individuals are more likely to face, with challenges such as inadequate public transportation, substandard housing, decreased availability of healthy food and safe exercise opportunities, and physical environments less conducive to good health.48, 49 A growing body of research—both domestically and internationally—suggests that more public spending on social services can yield a higher return for health outcomes than solely focusing on health care.50, 51 A recent effort by Medicaid officials—the new Accountable Health Communities grant program—to explore interventions geared at social factors influencing health is a critical next step in this approach.52 Whether the new administration will continue this program is unclear for now.
In reality, both explanations probably contribute significantly to these remaining noninsurance‐mediated disparities. Our results point to continuing gaps and the need for a policy and research agenda that extends beyond simply the expansion of insurance coverage. Coverage expansion may help narrow these gaps somewhat, which means that a potential repeal of the ACA poses significant risk particularly to low‐income groups and racial/ethnic minorities. But even if the new administration and its Republican allies in Congress ultimately maintain the ACA in some form, the law's coverage expansion should not be considered the primary solution to racial and socioeconomic disparities in health care. Additional policy attention will be needed to address these serious problems in the post‐ACA era.
Supporting information
Appendix: Survey Methodology
Appendix Table 1. Full Multivariate Model (Model 3) for Study Outcomes
Appendix Table 2. Disparities in ED Visits Due to Lack of Available Appointments, by Income
Appendix Table 3. Disparities in ED Visits Due to Lack of Available Appointments, by Race/Ethnicity
Appendix Table 4. Disparities in Receipt of Fair or Poor Care by Income in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 5. Disparities in Receipt of Fair or Poor Care by Race/Ethnicity in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 6. Disparities in Health Care Not Obtained Due to Costs by Income in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 7. Disparities in Health Care Not Obtained Due to Costs by Race/Ethnicity in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 8. Impact of Income, Race, and Health Insurance on Perceived Changes in Health Care in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Funding/Support
This project was supported by a research grant from the Robert Wood Johnson Foundation.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr. Sommers reports grants from the Agency for Healthcare Research and Quality, grants from the National Institute for Health Care Management, grants from the Commonwealth Fund, personal fees and nonfinancial support from the University of Michigan, personal fees and nonfinancial support from MedPAC, nonfinancial support from America's Health Insurance Plan, personal fees and nonfinancial support from Johns Hopkins University, and personal fees and nonfinancial support from the University of Pennsylvania outside the submitted work. Dr. Sommers served as a senior advisor in the US Department of Health & Human Services from September 2011 to June 2016.
Acknowledgments: None.
References
- 1. Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: Institute of Medicine; 2002. [PubMed] [Google Scholar]
- 2. Meyer PA, Yoon PW, Kaufmann RB; Centers for Disease Control and Prevention . Introduction: CDC Health Disparities and Inequalities Report—United States, 2013. MWMR Suppl. 2013;62(3):3‐5. [PubMed] [Google Scholar]
- 3. Cohen RA, Martinez ME, Zammitti EP. Health Insurance Coverage: Early Release of Estimates From the National Health Interview Survey, 2015. Hyattsville, MD: National Center for Health Statistics; 2016. [Google Scholar]
- 4. Hargraves JL. The insurance gap and minority health care, 1997‐2001. Track Rep. 2002;(2):1‐4. [PubMed] [Google Scholar]
- 5. Ayanian JZ, Weissman JS, Schneider EC, Ginsburg JA, Zaslavsky AM. Unmet health needs of uninsured adults in the United States. JAMA. 2000;284(16):2061‐2069. [DOI] [PubMed] [Google Scholar]
- 6. Lillie‐Blanton M, Hoffman C. The role of health insurance coverage in reducing racial/ethnic disparities in health care. Health Aff (Millwood). 2005;24(2):398‐408. [DOI] [PubMed] [Google Scholar]
- 7. Bustamante AV, Chen J. Health expenditure dynamics and years of U.S. residence: analyzing spending disparities among Latinos by citizenship/nativity status. Health Serv Res. 2012;47(2):794‐818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Zhu J, Brawarsky P, Lipsitz S, Huskamp H, Haas JS. Massachusetts health reform and disparities in coverage, access and health status. J Gen Intern Med. 2010;25(12):1356‐1362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sommers BD, Long SK, Baicker K. Changes in mortality after Massachusetts health care reform: a quasi‐experimental study. Ann Intern Med. 2014;160(9):585‐593. [DOI] [PubMed] [Google Scholar]
- 10. Courtemanche CJ, Zapata D. Does universal coverage improve health? The Massachusetts experience. J Policy Anal Manage. 2014;33(1):36‐69. [DOI] [PubMed] [Google Scholar]
- 11. Hanchate AD, Lasser KE, Kapoor A, et al. Massachusetts reform and disparities in inpatient care utilization. Med Care. 2012;50(7):569‐577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Van Der Wees PJ, Zaslavsky AM, Ayanian JZ. Improvements in health status after Massachusetts health care reform. Milbank Q. 2013;91(4):663‐689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Sommers BD, Baicker K, Epstein AM. Mortality and access to care among adults after state Medicaid expansions. N Engl J Med. 2012;367(11):1025‐1034. [DOI] [PubMed] [Google Scholar]
- 14. Baicker K, Taubman SL, Allen HL, et al. The Oregon experiment—effects of Medicaid on clinical outcomes. N Engl J Med. 2013;368(18):1713‐1722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. DeLeire T, Dague L, Leininger L, Voskuil K, Friedsam D. Wisconsin experience indicates that expanding public insurance to low‐income childless adults has health care impacts. Health Aff (Millwood). 2013;32(6):1037‐1045. [DOI] [PubMed] [Google Scholar]
- 16. Chen J, Bustamante AV, Tom SE. Health care spending and utilization by race/ethnicity under the Affordable Care Act's dependent coverage expansion. Am J Public Health. 2015;105 Suppl 3:S499‐S507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. O'Hara B, Brault MW. The disparate impact of the ACA‐dependent expansion across population subgroups. Health Serv Res. 2013;48(5):1581‐1592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Shane DM, Ayyagari P. Will health care reform reduce disparities in insurance coverage?: Evidence from the dependent coverage mandate. Med Care. 2014;52(6):528‐534. [DOI] [PubMed] [Google Scholar]
- 19. Saez E, Zucman G. Wealth inequality in the United States since 1913: evidence from capitalized income tax data. Q J Econ. 2016;131(2):519‐578. [Google Scholar]
- 20. National Health Expenditures per Capita, 1960‐2023 . Kaiser Family Foundation website. http://kff.org/health-costs/slide/national-health-expenditures-per-capita-1960-2023. Published October 23, 2014. Accessed December 15, 2016.
- 21. Clark CR, Ommerborn MJ, A Coull B, Pham do Q, Haas JS. Income inequities and Medicaid expansion are related to racial and ethnic disparities in delayed or forgone care due to cost. Med Care. 2016;54(6):555‐561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Chen J, Vargas‐Bustamante A, Mortensen K, Ortega AN. Racial and ethnic disparities in health care access and utilization under the Affordable Care Act. Med Care. 2016;54(2):140‐146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Sommers BD, Gunja MZ, Finegold K, Musco T. Changes in self‐reported insurance coverage, access to care, and health under the Affordable Care Act. JAMA. 2015;314(4):366‐374. [DOI] [PubMed] [Google Scholar]
- 24. Buchmueller TC, Levinson ZM, Levy HG, Wolfe BL. Effect of the Affordable Care Act on racial and ethnic disparities in health insurance coverage. Am J Public Health. 2016:106(8):1416‐1421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Shartzer A, Long SK, Anderson N. Access to care and affordability have improved following Affordable Care Act implementation; problems remain. Health Aff (Millwood). 2016;35(1):161‐168. [DOI] [PubMed] [Google Scholar]
- 26. Collins SR, Gunja MZ, Doty MM, Beutel S. Americans’ Experiences with ACA Marketplace and Medicaid Coverage: Access to Care and Satisfaction . http://www.commonwealthfund.org/publications/issue-briefs/2016/may/ACA-tracking-survey-access-to-care-and-satisfaction. Published May 25, 2016. Accessed December 15, 2016. [PubMed]
- 27. Patients’ Perspectives on Health in the United States: A Look at Seven States and the Nation. National Public Radio, Robert Wood Johnson Foundation, Harvard T.H; Chan School of Public Health; 2016. http://www.rwjf.org/en/library/research/2016/02/patients-perspectives-on-health-care-in-the-united-states.html. Published February 2016. Accessed December 15, 2016. [Google Scholar]
- 28. American Association for Public Opinion Research (AAPOR) . Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 9th ed http://www.aapor.org/AAPOR_Main/media/publications/Standard-Definitions20169theditionfinal.pdf. Oakbrook Terrace, IL: AAPOR; 2016. Accessed December 15, 2016. [Google Scholar]
- 29. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;36(1):1‐10. [PubMed] [Google Scholar]
- 30. Cook BL, McGuire TG, Zaslavsky AM. Measuring racial/ethnic disparities in health care: methods and practical issues. Health Serv Res. 2012;47(3 pt 2):1232‐1254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Schoen C, Collins SR, Kriss JL, Doty MM. How many are underinsured? Trends among U.S. adults, 2003 and 2007. Health Aff (Millwood). 2008;27(4):w298‐w309. [DOI] [PubMed] [Google Scholar]
- 32. Rodriguez MA, Bustamante AV, Ang A. Perceived quality of care, receipt of preventive care, and usual source of health care among undocumented and other Latinos. J Gen Intern Med. 2009;24 (Suppl 3):508‐513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Brandon DT, Isaac LA, LaVeist TA. The legacy of Tuskegee and trust in medical care: is Tuskegee responsible for race differences in mistrust of medical care? J Natl Med Assoc. 2005;97(7):951‐956. [PMC free article] [PubMed] [Google Scholar]
- 34. Decker SL. In 2011 nearly one‐third of physicians said they would not accept new Medicaid patients, but rising fees may help. Health Aff (Millwood). 2012;31(8):1673‐1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Taubman SL, Allen HL, Wright BJ, Baicker K, Finkelstein AN. Medicaid increases emergency‐department use: evidence from Oregon's health insurance experiment. Science. 2014;343(6168):263‐268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Ku L, Matani S. Left out: immigrants’ access to health care and insurance. Health Aff (Millwood). 2001;20(1):247‐256. [DOI] [PubMed] [Google Scholar]
- 37. Sommers BD, Maylone B, Nguyen KH, Blendon RJ, Epstein AM. The impact of state policies on ACA applications and enrollment among low‐income adults in Arkansas, Kentucky, and Texas. Health Aff (Millwood). 2015;34(6):1010‐1018. [DOI] [PubMed] [Google Scholar]
- 38. Dugan A. Despite enrollment success, healthcare law still unpopular. Gallup. May 29, 2014. http://www.gallup.com/poll/170750/despite-enrollment-success-healthcare-law-remains-unpopular.aspx. Accessed December 15, 2016. [Google Scholar]
- 39. Keeter S, Kennedy C, Dimock M, Best J, Craighill P. Gauging the impact of growing nonresponse on estimates from a national RDD telephone survey. Public Opin Q. 2006;70(5):759‐779. [Google Scholar]
- 40. Davern M. Nonresponse rates are a problematic indicator of nonresponse bias in survey research. Health Serv Res. 2013;48(3):905‐912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Pew Research Center . Assessing the Representativeness of Public Opinion Surveys. Washington, DC: Pew Research Center; 2012. http://www.people-press.org/files/legacy-pdf/Assessing%20the%20Representativeness%20of%20Public%20Opinion%20Surveys.pdf. Accessed December 15, 2016. [Google Scholar]
- 42. Gaskin DJ, Dinwiddie GY, Chan KS, McCleary RR. Residential segregation and the availability of primary care physicians. Health Serv Res. 2012;47(6):2353‐2376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Ko M, Ponce NA. Community residential segregation and the local supply of federally qualified health centers. Health Serv Res. 2013;48(1):253‐270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Shin P, Sharac J, Rosenbaum S. Community health centers and Medicaid at 50: an enduring relationship essential for health system transformation. Health Aff (Millwood). 2015;34(7):1096‐1104. [DOI] [PubMed] [Google Scholar]
- 45. Neuhausen K, Spivey M, Kellermann AL. State politics and the fate of the safety net. N Engl J Med. 2013;369(18):1675‐1677. [DOI] [PubMed] [Google Scholar]
- 46. Saxton JF, Johns MM. Grow the US National Health Service Corps. JAMA. 2009;301(18):1925‐1926. [DOI] [PubMed] [Google Scholar]
- 47. Mensah MO, Sommers BD. The policy argument for healthcare workforce diversity. J Gen Intern Med. 2016;31(11):1369‐1372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Berkman LF, Kawachi I, eds. Social Epidemiology. 2nd ed New York, NY: Oxford University Press; 2014. [Google Scholar]
- 49. Williams DR, Collins C. Racial residential segregation: a fundamental cause of racial disparities in health. Public Health Rep. 2001;116(5):404‐416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Bradley EH, Elkins BR, Herrin J, Elbel B. Health and social services expenditures: associations with health outcomes. BMJ Qual Saf. 2011;20(10):826‐831. [DOI] [PubMed] [Google Scholar]
- 51. Bradley EH, Canavan M, Rogan E, et al. Variation in health outcomes: the role of spending on social services, public health, and health care, 2000‐09. Health Aff (Millwood). 2016;35(5):760‐768. [DOI] [PubMed] [Google Scholar]
- 52. Bachrach D, Guyer J, Levin A. Medicaid Coverage of Social Interventions: A Road Map for States. New York, NY: Milbank Memorial Fund; 2016. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix: Survey Methodology
Appendix Table 1. Full Multivariate Model (Model 3) for Study Outcomes
Appendix Table 2. Disparities in ED Visits Due to Lack of Available Appointments, by Income
Appendix Table 3. Disparities in ED Visits Due to Lack of Available Appointments, by Race/Ethnicity
Appendix Table 4. Disparities in Receipt of Fair or Poor Care by Income in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 5. Disparities in Receipt of Fair or Poor Care by Race/Ethnicity in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 6. Disparities in Health Care Not Obtained Due to Costs by Income in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 7. Disparities in Health Care Not Obtained Due to Costs by Race/Ethnicity in 7‐State Sample, Medicaid Expansion vs Non‐Expansion
Appendix Table 8. Impact of Income, Race, and Health Insurance on Perceived Changes in Health Care in 7‐State Sample, Medicaid Expansion vs Non‐Expansion