[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
Payer Proportion of Non-Medicare Hospitalizations for Major Cardiovascular Events by Expansion Status, 2009-2014
Payer Proportion of Non-Medicare Hospitalizations for Major Cardiovascular Events by Expansion Status, 2009-2014

The dotted vertical line after year 2013 represents implementation of the Affordable Care Act that included Medicaid expansion. Data presented are for the completion of each year. Data for nonexpansion states in 2009 and 2010 do not include the state of Indiana. After Affordable Care Act implementation, there was a significant change in the yearly uninsured and Medicaid payer proportions in expansion states.

Figure 2.
Relative Changes in Proportion of Uninsured, Medicaid, and Private Insurance Non-Medicare Hospitalizations for Major Cardiovascular Events by State, 2013-2014
Relative Changes in Proportion of Uninsured, Medicaid, and Private Insurance Non-Medicare Hospitalizations for Major Cardiovascular Events by State, 2013-2014

A, Uninsured hospitalizations vs Medicaid and private insurance in expansion states. B, Uninsured hospitalizations vs Medicaid and private insurance in nonexpansion states. Concurrent with the decrease in uninsured hospitalizations, the proportion of Medicaid hospitalizations increased in most expansion states whereas this pattern was not seen consistently in the nonexpansion states. AR indicates Arkansas; AZ, Arizona; CA, California; CO, Colorado; FL, Florida; HI, Hawaii; IA, Iowa; IL, Illinois; IN, Indiana; KS, Kansas; KY, Kentucky; MD, Maryland; ME, Maine; MO, Missouri; NC, North Carolina; NJ, New Jersey; NM, New Mexico; NV, Nevada; NY, New York; OK, Oklahoma; OR, Oregon; RI, Rhode Island; SC, South Carolina; TN, Tennessee; TX, Texas; UT, Utah; VT, Vermont; WA, Washington; WI, Wisconsin; and WY, Wyoming.

Table 1.  
All Hospitalizations for Major Cardiovascular Events by State Medicaid Expansion Status Before ACA (2009-2013) and After ACA (2014) Medicaid Expansion
All Hospitalizations for Major Cardiovascular Events by State Medicaid Expansion Status Before ACA (2009-2013) and After ACA (2014) Medicaid Expansion
Table 2.  
Change in Payer Proportion for Non-Medicare Hospitalizations for Major Cardiovascular Events by State Medicaid Expansion Status Before ACA (2009-2013) and After ACA (2014) Medicaid Expansion
Change in Payer Proportion for Non-Medicare Hospitalizations for Major Cardiovascular Events by State Medicaid Expansion Status Before ACA (2009-2013) and After ACA (2014) Medicaid Expansion
Table 3.  
In-Hospital Mortality for Non-Medicare Hospitalizations of Major Cardiovascular Events by State Medicaid Expansion Status, 2009-2014
In-Hospital Mortality for Non-Medicare Hospitalizations of Major Cardiovascular Events by State Medicaid Expansion Status, 2009-2014
1.
Kochanek  KD, Murphy  SL, Xu  J, Tejada-Vera  B.  Deaths: final data for 2014.  Natl Vital Stat Rep.2016;65(4):1-122. PubMedGoogle Scholar
2.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
3.
Centers for Disease Control and Prevention, National Center for Health Statistics. National Hospital Discharge Survey: number and rate of discharges by first-listed diagnostic categories, 2010. https://www.cdc.gov/nchs/data/nhds/2average/2010ave2_firstlist.pdf. Accessed May 8, 2017.
4.
Centers for Disease Control and Prevention, National Center for Health Statistics. Table 105: no health insurance coverage among persons under age 65, by selected characteristics: United States, selected years 1984-2014. https://www.cdc.gov/nchs/data/hus/2015/105.pdf. Accessed May 8, 2017.
5.
Medford-Davis  LN, Fonarow  GC, Bhatt  DL,  et al.  Impact of insurance status on outcomes and use of rehabilitation services in acute ischemic stroke: findings from Get With the Guidelines–Stroke.  J Am Heart Assoc. 2016;5(11):e004282. doi:10.1161/JAHA.116.004282PubMedGoogle ScholarCrossref
6.
James  ML, Grau-Sepulveda  MV, Olson  DM,  et al.  Insurance status and outcome after intracerebral hemorrhage: findings from Get With the Guidelines–Stroke.  J Stroke Cerebrovasc Dis. 2014;23(2):283-292. doi:10.1016/j.jstrokecerebrovasdis.2013.02.016PubMedGoogle ScholarCrossref
7.
Kapoor  JR, Kapoor  R, Hellkamp  AS, Hernandez  AF, Heidenreich  PA, Fonarow  GC.  Payment source, quality of care, and outcomes in patients hospitalized with heart failure.  J Am Coll Cardiol. 2011;58(14):1465-1471. doi:10.1016/j.jacc.2011.06.034PubMedGoogle ScholarCrossref
8.
Mathews  R, Wang  TY, Honeycutt  E,  et al; TRANSLATE-ACS Study Investigators.  Persistence with secondary prevention medications after acute myocardial infarction: insights from the TRANSLATE-ACS study.  Am Heart J. 2015;170(1):62-69. doi:10.1016/j.ahj.2015.03.019PubMedGoogle ScholarCrossref
9.
Ng  DK, Brotman  DJ, Lau  B, Young  JH.  Insurance status, not race, is associated with mortality after an acute cardiovascular event in Maryland.  J Gen Intern Med. 2012;27(10):1368-1376. doi:10.1007/s11606-012-2147-9PubMedGoogle ScholarCrossref
10.
Carman  KG, Eibner  C, Paddock  SM.  Trends in health insurance enrollment, 2013-15.  Health Aff (Millwood). 2015;34(6):1044-1048. doi:10.1377/hlthaff.2015.0266PubMedGoogle ScholarCrossref
11.
Uberoi  N, Finegold  K, Gee  E; Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. ASPE Issue Brief: health insurance coverage and the Affordable Care Act, 2010-2016. https://aspe.hhs.gov/pdf-report/health-insurance-coverage-and-affordable-care-act-2010-2016. Published March 2016. Accessed May 8, 2017.
12.
Kaiser Family Foundation. Status of state action on the Medicaid expansion decision as of January 1, 2017. http://www.kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/. Accessed May 8, 2017.
13.
Sommers  BD, Blendon  RJ, Orav  EJ, Epstein  AM.  Changes in utilization and health among low-income adults after Medicaid expansion or expanded private insurance.  JAMA Intern Med. 2016;176(10):1501-1509. doi:10.1001/jamainternmed.2016.4419PubMedGoogle ScholarCrossref
14.
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:10.1001/jama.2015.8421PubMedGoogle ScholarCrossref
15.
Martinez ME, Ward BW. Health care access and utilization among adults aged 18–64, by poverty level: United States, 2013–2015. https://www.cdc.gov/nchs/products/databriefs/db262.htm. Accessed May 8, 2017.
16.
Okoro  CA, Zhao  G, Fox  JB, Eke  PI, Greenlund  KJ, Town  M.  Surveillance for health care access and health services use, adults aged 18-64 years—Behavioral Risk Factor Surveillance System, United States, 2014.  MMWR Surveill Summ. 2017;66(7):1-42. doi:10.15585/mmwr.ss6607a1PubMedGoogle ScholarCrossref
17.
Miller  S, Wherry  LR.  Health and access to care during the first 2 years of the ACA Medicaid expansions.  N Engl J Med. 2017;376(10):947-956. doi:10.1056/NEJMsa1612890PubMedGoogle ScholarCrossref
18.
Nikpay  S, Buchmueller  T, Levy  HG.  Affordable Care Act Medicaid expansion reduced uninsured hospital stays in 2014.  Health Aff (Millwood). 2016;35(1):106-110. doi:10.1377/hlthaff.2015.1144PubMedGoogle ScholarCrossref
19.
DeLiere  T, Joynt  K, McDonald  R; Department of Health and Human Services, Office of Assistant Secretary for Planning and Evaluation. ASPE Issue Brief: impact of insurance expansion on hospital uncompensated care costs in 2014. https://aspe.hhs.gov/pdf-report/impact-insurance-expansion-hospital-uncompensated-care-costs-2014. Published September 24, 2014. Accessed May 8, 2017.
20.
Freedman  S, Nikpay  S, Carroll  A, Simon  K.  Changes in inpatient payer-mix and hospitalizations following Medicaid expansion: evidence from all-capture hospital discharge data.  PLoS One. 2017;12(9):e0183616. doi:10.1371/journal.pone.0183616PubMedGoogle ScholarCrossref
21.
Xu  J, Murphy  SL, Kochanek  KD, Arias  E.  Mortality in the United States, 2015.  NCHS Data Brief. 2016;(267):1-8. PubMedGoogle Scholar
22.
Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP). Introduction to the HCUP State Inpatient Databases (SID). https://www.hcup-us.ahrq.gov/db/state/siddist/SID_Introduction.jsp. Published February 2018. Accessed July 13, 2018.
23.
Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP). Why the NIS should not be used to make State-level estimates. https://www.hcup-us.ahrq.gov/db/nation/nis/nis_statelevelestimates.jsp. Published January 2016. Accessed July 13, 2018.
24.
US Department of Health and Human Services, Agency for Healthcare Research and Quality. HCUPnet. Healthcare Cost and Utilization Project. https://hcupnet.ahrq.gov/. Accessed July 17, 2017.
25.
Yeh  RW, Sidney  S, Chandra  M, Sorel  M, Selby  JV, Go  AS.  Population trends in the incidence and outcomes of acute myocardial infarction.  N Engl J Med. 2010;362(23):2155-2165. doi:10.1056/NEJMoa0908610PubMedGoogle ScholarCrossref
26.
Kokotailo  RA, Hill  MD.  Coding of stroke and stroke risk factors using International Classification of Diseases, revisions 9 and 10.  Stroke. 2005;36(8):1776-1781. doi:10.1161/01.STR.0000174293.17959.a1PubMedGoogle ScholarCrossref
27.
Roumie  CL, Mitchel  E, Gideon  PS, Varas-Lorenzo  C, Castellsague  J, Griffin  MR.  Validation of ICD-9 codes with a high positive predictive value for incident strokes resulting in hospitalization using Medicaid health data.  Pharmacoepidemiol Drug Saf. 2008;17(1):20-26. doi:10.1002/pds.1518PubMedGoogle ScholarCrossref
28.
Psaty  BM, Delaney  JA, Arnold  AM,  et al.  Study of cardiovascular health outcomes in the era of claims data: the Cardiovascular Health Study.  Circulation. 2016;133(2):156-164. doi:10.1161/CIRCULATIONAHA.115.018610PubMedGoogle ScholarCrossref
29.
Gunasekara  FI, Richardson  K, Carter  K, Blakely  T.  Fixed effects analysis of repeated measures data.  Int J Epidemiol. 2014;43(1):264-269. doi:10.1093/ije/dyt221PubMedGoogle ScholarCrossref
30.
Davis  MM, Gebremariam  A, Ayanian  JZ.  Changes in insurance coverage among hospitalized nonelderly adults after Medicaid expansion in Michigan.  JAMA. 2016;315(23):2617-2618. doi:10.1001/jama.2016.6303PubMedGoogle ScholarCrossref
31.
US Census Bureau. Annual estimates of the resident population: April 1, 2010, to July 1, 2017: 2017 population estimates. https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=PEP_2017_PEPANNRES&src=pt. Accessed July 13, 2018.
32.
Gebreab  SY, Davis  SK, Symanzik  J, Mensah  GA, Gibbons  GH, Diez-Roux  AV.  Geographic variations in cardiovascular health in the United States: contributions of state- and individual-level factors.  J Am Heart Assoc. 2015;4(6):e001673. doi:10.1161/JAHA.114.001673PubMedGoogle ScholarCrossref
33.
Blavin  F.  Association between the 2014 Medicaid expansion and US hospital finances.  JAMA. 2016;316(14):1475-1483. doi:10.1001/jama.2016.14765PubMedGoogle ScholarCrossref
34.
Dranove  D, Garthwaite  C, Ody  C.  Uncompensated care decreased at hospitals in Medicaid expansion states but not at hospitals in nonexpansion states.  Health Aff (Millwood). 2016;35(8):1471-1479. doi:10.1377/hlthaff.2015.1344PubMedGoogle ScholarCrossref
35.
Chokshi  DA, Chang  JE, Wilson  RM.  Health reform and the changing safety net in the United States.  N Engl J Med. 2016;375(18):1790-1796. doi:10.1056/NEJMhpr1608578PubMedGoogle ScholarCrossref
36.
Agarwal  SK, Wruck  L, Quibrera  M,  et al.  Temporal trends in hospitalization for acute decompensated heart failure in the United States, 1998-2011.  Am J Epidemiol. 2016;183(5):462-470. doi:10.1093/aje/kwv455PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    2 Comments for this article
    Another sign of the effect of the ACA
    Frederick Rivara, MD, MPH | University of Washington
    This is another study showing that the ACA does help to decrease the rates of uninsured hospitalizations for serious conditions through the policy of Medicaid expansion. Next step is to look at the longer term effects on mortality.
    CONFLICT OF INTEREST: Editor in chief, JAMA Network Open
    Would we expect an impact on in-hospital mortality?
    Steven Bradley |
    As noted in the excellent commentary for this paper, probably not in the short term....but there may be longer term population effects as more patients get the care they need. More work to be done.
    CONFLICT OF INTEREST: None Reported
    Original Investigation
    Health Policy
    August 24, 2018

    Association of State Medicaid Expansion With Rate of Uninsured Hospitalizations for Major Cardiovascular Events, 2009-2014

    Author Affiliations
    • 1Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
    • 2Division of Cardiology, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
    JAMA Network Open. 2018;1(4):e181296. doi:10.1001/jamanetworkopen.2018.1296
    Key Points

    Question  Did uninsured hospitalizations for major cardiovascular events and in-hospital mortality vary by state-level policy decisions on the implementation of the Affordable Care Act Medicaid expansion?

    Findings  In this cohort study, difference-in-differences analysis of more than 3 million non-Medicare hospitalizations from the inpatient databases of 30 states found that expansion states had a significant reduction in the proportion of uninsured hospitalizations for major cardiovascular events within 1 year of Affordable Care Act Medicaid expansion compared with nonexpansion states. There were no changes in in-hospital mortality rates in expansion or nonexpansion states.

    Meaning  Further research is necessary to determine how state-level policy regarding Medicaid expansion could differentially affect cardiovascular outcomes.

    Abstract

    Importance  Cardiovascular disease is the leading primary diagnosis among all hospital discharges, and insurance status is associated with patient outcomes. The association of state-level policy decisions regarding the Affordable Care Act (ACA) Medicaid expansion with rates of uninsured hospitalizations for major cardiovascular events and in-hospital mortality has not been investigated to date.

    Objective  To investigate whether the rates of uninsured hospitalizations for major cardiovascular events and in-hospital mortality varied by state-level policy on ACA Medicaid expansion.

    Design, Setting, and Participants  For this cohort study, difference-in-differences analysis of data from the Healthcare Cost and Utilization Project State Inpatient Databases of 30 US states on 524 848 non-Medicare hospitalizations in 2014 and a mean of 516 811 non-Medicare hospitalizations per year from 2009 to 2013 was performed for major cardiovascular events (defined as a composite of acute myocardial infarction, stroke, and heart failure) from January 1, 2009, through December 31, 2014. Analyses were completed September 1, 2017.

    Exposure  State Medicaid expansion as of January 1, 2014.

    Main Outcomes and Measures  Comparison of mean payer mix proportions (uninsured, Medicaid, and privately insured) and in-hospital mortality between expansion and nonexpansion states for the years preceding the ACA Medicaid expansion (2009-2013) and the year after the ACA Medicaid expansion (2014).

    Results  Of the 801 819 hospitalizations in the 17 expansion states in 2014, 428 503 (53.4%) patients were men, 514 036 (64.1%) were white, and 365 797 (45.6%) were aged 65 to 84 years. Of 719 459 hospitalizations in the 13 nonexpansion states in 2014, 383 311 (53.3%) patients were men, 492 136 (68.4%) were white, and 335 781 (46.7%) were aged 65 to 84 years. There were 281 184 non-Medicare hospitalizations for major cardiovascular events in the 17 expansion states and 243 664 non-Medicare hospitalizations in the 13 nonexpansion states in 2014. In multivariable regression analyses, the expansion states had a significant 5.8–percentage point decrease in the proportion of uninsured hospitalizations after Medicaid expansion relative to the nonexpansion states (adjusted difference-in-differences estimate, −0.058; 95% CI, −0.075 to −0.042; P < .001). The expansion states also had a significant 8.4–percentage point increase in the Medicaid share after Medicaid expansion relative to the nonexpansion states (adjusted difference-in-differences estimate, 0.084; 95% CI, 0.065 to 0.102; P < .001). In-hospital mortality did not change significantly after Medicaid expansion in either the expansion states (before ACA, 3.8% vs after ACA, 3.7%) or the nonexpansion states (4.0% vs 4.0%).

    Conclusions and Relevance  States that expanded Medicaid during the ACA implementation had a significantly greater reduction in the proportion of uninsured hospitalizations for major cardiovascular events compared with the nonexpansion states. This study suggests that expansion status was not associated with in-hospital mortality rates in the first year after ACA implementation.

    Introduction

    Cardiovascular disease (CVD) is the leading primary hospital discharge diagnosis and the most common cause of death in the United States.1,2 Coronary heart disease, stroke, and congestive heart failure represent the most common causes of CVD-related hospitalization.3 Individuals with markers of low socioeconomic status bear a disproportionate burden of CVD, including coronary heart disease and stroke, and are more likely to be uninsured.2,4 Access to adequate care is essential in major cardiovascular (CV) events. Insurance status is associated with clinical outcomes.5-9

    It has been reported that since the first open enrollment period in October 2013, the Affordable Care Act (ACA) has resulted in millions of previously uninsured Americans acquiring health insurance coverage.10-12 The health insurance exchanges, subsidies, employer requirements for coverage, and Medicaid eligibility expansion have each contributed to these changes; however, not all states have opted to expand Medicaid.12 Previous data suggest that the ACA has improved access to health care clinicians and helped to reduce financial barriers to care, with the most significant gains occurring in expansion states—especially among low-income adults.13-17 Previous studies have examined ACA-related shifts in insurance payer mix for all hospitalizations in general and reported a significant reduction in the proportion of uninsured hospital discharges in Medicaid expansion states vs nonexpansion states.18,19 However, to our knowledge, the association between state-level policy regarding Medicaid expansion and uninsured hospitalizations for major CV events has not been investigated.

    Whereas it might be expected that there would be fewer uninsured CVD-related hospitalizations after ACA implementation, it is important to document whether this was observed. It is also important to quantify the magnitude of change given the prominent role of CVD in overall public health and as the leading source of US medical expenditures.2 This is especially the case given data suggesting that uniformity of changes in payer proportions across different types of hospitalizations should not be presumed.20

    Given the potential for CVD-related morbidity along with data suggesting mortality may have recently increased,21 understanding the association of particular components of the ACA with uninsured hospitalizations for major CV events has potentially significant health policy implications—especially in the non-Medicare population, in whom insurance coverage is less uniform and often closely linked to socioeconomic status. Furthermore, there is currently a dearth of data on the possible associations of expanded insurance coverage with CV outcomes. We sought to investigate whether the rates of uninsured hospitalizations for major CV events varied by state-level policy on Medicaid expansion. We further examined rates of in-hospital mortality over time as a potential indicator of changes in severity of presentation and level of care provided that may have been associated with Medicaid expansion.

    Methods

    For this study, we used publicly available data for January 1, 2009, through December 31, 2014, from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project State Inpatient Databases.22 Each participating Healthcare Cost and Utilization Project State Inpatient Database22 provided a census of actual inpatient hospitalizations specific to each state; this was distinct from the National Inpatient Sample,23 which provided estimates of hospitalizations and was not intended for state-level analyses. We collected data through the Healthcare Cost and Utilization Project Network24 online query engine, which provided aggregated patient information at the state level by year, including data on age, sex, race/ethnicity, payer (uninsured, Medicaid, Medicare, or privately insured) and residential classification (rural vs urban). Data were available through December 31, 2014, which was the first full year after implementation of ACA provisions including the individual insurance mandate and the Medicaid expansion. We chose 2009 as the beginning of our study period since this year preceded the signing of the ACA into law in 2010. This study was approved by the institutional review board at Northwestern University, Chicago, Illinois. Because we used publicly available deidentified data, the study was exempt from review by the Northwestern University institutional review board. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline to the extent possible allowed by this type of analysis. Analyses were completed September 1, 2017, incorporating data available publicly through this period.

    We identified hospitalizations for major CV events, defined as a composite of diagnoses that included acute myocardial infarction, stroke, and heart failure. These were based on prespecified International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes listed as the principal discharge diagnosis. We identified all hospitalizations in each state that had available data with the following ICD-9 diagnosis codes listed as the principal diagnosis on discharge: acute myocardial infarction (410.x0 and 410.x1), stroke (430, 431, 433.x1, 434.x1, and 435-436), and heart failure (402.x1, 404.x1, 404.x3, and 428.x). These ICD-9 diagnosis codes have shown good positive predictive value for these conditions when listed as the principal discharge diagnosis.25-28

    Data were publicly available for 35 states. We excluded states that did not have hospitalization data available in 2014 (Massachusetts and New Hampshire) or that did not have relevant demographic data available for any year (Nebraska and Minnesota). We categorized states by their Medicaid expansion status in 2014, excluding Michigan where Medicaid expansion did not become effective until April 1, 2014 (rather than January 1, 2014). The state of Indiana did not expand Medicaid until after 2014 and was categorized as a nonexpansion state. Data from the 17 states (Arizona, Arkansas, California, Colorado, Hawaii, Illinois, Iowa, Kentucky, Maryland, Nevada, New Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, and Washington) that did expand (expansion states) and the 13 states (Florida, Indiana, Kansas, Maine, Missouri, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Utah, Wisconsin, and Wyoming) that did not expand (nonexpansion states) by 2014 were aggregated together as expansion and nonexpansion states. The final sample consisted of all hospitalizations for major CV events from 2009 through 2014 for 29 of these 30 states, and from 2011 through 2014 for 1 state (Indiana) that did not have 2009 and 2010 data available. We did not exclude Indiana since it had one of the larger populations and sufficient data were available for analysis of associations before ACA.

    Statistical Analysis

    To assess the possible association of the ACA Medicaid expansion with uninsured hospitalizations for major CV events, we first calculated the proportion of the total hospitalizations for the expansion and nonexpansion states by insurance payer type. Because expansion of coverage was not expected to change in the Medicare population—which provided near-universal coverage for US adults older than 65 years even prior to the ACA—we limited our analyses to non-Medicare hospitalizations by subtracting Medicare-payer hospitalizations from the total number of hospitalizations. From this denominator of non-Medicare hospitalizations, we calculated the proportions of uninsured, Medicaid, and privately insured payer types.

    Using a difference-in-differences approach, we constructed multivariable linear regression models to compare the mean payer mix from the years preceding Medicaid expansion (2009-2013) and the year after Medicaid expansion (2014) between expansion and nonexpansion states. Separate models were estimated for each payer status and adjusted for the time-varying state-level demographics (proportions of individuals who were female, <65 years, non-Hispanic white, and living in a rural location) of those hospitalized for major CV events. All models included a fixed effect for each state to account for state-level variation and were weighted by the total number of hospital discharges. Fixed-effects models also account for state-level correlation by calculating a separate, fixed intercept for each state.29 We also calculated the unadjusted change in hospitalization payer mix from 2013 to 2014 for each state and presented these data graphically. In sensitivity analyses (eTable in the Supplement), we repeated the difference-in-differences models to include the Medicare hospitalizations for major CV events since Medicaid expansion would not be expected to affect the proportion of Medicare hospitalizations. To assess for a possible association between Medicaid expansion and clinical outcomes, we also calculated the total number of in-hospital deaths among non-Medicare hospitalizations in the expansion and nonexpansion states for each study year. We then calculated the overall in-hospital mortality rate for each study year to investigate whether there was a difference over time by expansion status. Two-sided P values less than .05 were deemed to be statistically significant. All statistical analyses were performed using SAS, version 9.4 (SAS Institute Inc).

    Results

    The demographic data for all major CV hospitalizations in expansion and nonexpansion states in the periods before ACA (2009-2013) and after ACA (2014) are presented in Table 1. Of the 801 819 hospitalizations in the expansion states in 2014, 428 503 (53.4%) patients were men, 514 036 (64.1%) were white, and 365 797 (45.6%) were aged 65 to 84 years. Of 719 459 hospitalizations in the nonexpansion states in 2014, 383 311 (53.3%) patients were men, 492 136 (68.4%) were white, and 335 781 (46.7%) were aged 65 to 84 years. In the 17 expansion states, the mean total number of non-Medicare hospitalizations during the period before ACA was 278 612, and the total number of non-Medicare hospitalizations in the first year after ACA (2014) was 281 184. In the 13 nonexpansion states, the mean total number of non-Medicare hospitalizations during the period before ACA was 238 199, and the total number of non-Medicare hospitalizations in the first year after ACA was 243 664. Overall, there were more hospitalizations in rural locations in the nonexpansion states than in the expansion states.

    Payer mix for each year from 2009 to 2014 for non-Medicare hospitalizations in the expansion and nonexpansion states is presented in Figure 1. After ACA implementation, there was a significant change in the yearly uninsured and Medicaid proportions of non-Medicare hospitalizations in expansion states. Throughout the study period, the proportion of uninsured hospitalizations for each year was consistently higher and the proportion of Medicaid hospitalizations consistently lower in the nonexpansion states compared with the expansion states, with an increase in these differences after ACA implementation (Figure 1). Payer mix composition differed by type of major CV event (myocardial infarction, stroke, or heart failure), but similar shifts in payer mix for uninsured and Medicaid proportions of non-Medicare hospitalizations for each type of major CV event were seen after ACA implementation (eFigure in the Supplement).

    In crude unadjusted analyses, after ACA implementation, the uninsured proportion of non-Medicare hospitalizations decreased by 5.4 percentage points after ACA vs before ACA (mean before ACA: 36 422 of 278 612 [13.1%] vs after ACA: 21 564 of 281 184 [7.7%]), whereas the Medicaid proportion increased by 9.5 percentage points (before ACA: 71 268 of 278 612 [25.6%] vs after ACA: 98 819 of 281 184 [35.1%]) in the expansion states. In the nonexpansion states, similar differences were not seen. The uninsured proportion of non-Medicare hospitalizations before ACA was 49 212 of 238 199 (20.7%) and after ACA was 48 787 of 243 664 (20.0%), whereas the Medicaid proportion before ACA was 43 998 of 238 199 (18.2%) and after ACA was 45 278 of 243 664 (18.6%).

    In regression analyses, the uninsured proportion of non-Medicare hospitalizations decreased significantly by 5.0 percentage points after ACA vs before ACA (difference estimate, −0.050; 95% CI, −0.062 to −0.038; P < .001) (Table 2), whereas the Medicaid proportion increased significantly by 10.2 percentage points (0.102; 95% CI, 0.088 to 0.116; P < .001). In the nonexpansion states, the uninsured proportion did not change significantly (difference estimate, 0.003 [or 0.3 percentage points]; 95% CI, −0.011 to 0.017), nor did the Medicaid proportion (0.009 [or 0.9 percentage points]; 95% CI, −0.003 to 0.020), consistent with the findings in the raw analyses.

    In the multivariable difference-in-differences regression analyses, the expansion states had a significant 5.8–percentage point decrease in the uninsured proportion of non-Medicare hospitalizations after expansion relative to the nonexpansion states (adjusted difference-in-differences estimate, −0.058; 95% CI, −0.075 to −0.042; P < .001) (Table 2) along with a significant 8.4–percentage point increase in the proportion of Medicaid hospitalizations after expansion relative to the nonexpansion states (0.084; 95% CI, 0.065 to 0.102; P < .001). There was no significant change in the private share of hospitalizations in the expansion states relative to the nonexpansion states (−0.007 [or −0.7 percentage points]; 95% CI, −0.029 to 0.016; P = .54)

    In sensitivity analyses with Medicare hospitalizations included in the population, the expansion states still had a significant decrease in proportion of uninsured hospitalizations after expansion compared with the nonexpansion states (adjusted difference-in-differences estimate, −0.019 [or −1.9 percentage points]; 95% CI, −0.024 to −0.014; P < .001) along with a significant increase in proportion of Medicaid hospitalizations after expansion compared with the nonexpansion states (0.031 [or 3.1 percentage points]; 95% CI, 0.025 to 0.038; P < .001). As would be expected, there were no significant changes in the proportion of Medicare hospitalizations in the expansion states or the nonexpansion states after Medicaid expansion (eTable in the Supplement).

    Figure 2 presents unadjusted individual state-level changes in payer mix between 2013 and 2014. The proportion of uninsured hospitalizations decreased in most of the individual states but these changes tended to be greater in the expansion states, concurrent with increases in Medicaid share, whereas this pattern was not seen consistently in the nonexpansion states.

    The percentages of in-hospital deaths per year among non-Medicare hospitalizations in the expansion and nonexpansion states are presented in Table 3. In regression analyses, there was no significant change in in-hospital mortality after ACA implementation in either the expansion states (3.7% in 2014 vs 3.8% over the 2009-2013 period; P = .30) or the nonexpansion states (4.0% in 2014 vs 4.0% over the 2009-2013 period; P = .39), nor was there a significant difference in in-hospital deaths in the expansion states relative to the nonexpansion states when comparing the periods before and after ACA (adjusted difference-in-differences estimate, −0.001 [or −0.1 percentage points]; 95% CI, −0.002 to 0.001; P = .42).

    Discussion
    Principal Findings

    In this quasi-experimental difference-in-differences cohort study, we examined the potential associations between the ACA implementation and the rates of uninsured hospitalizations for 3 of the top causes of CVD-related morbidity and mortality in the United States. We found that, after full implementation of the major provisions of the ACA in the beginning of 2014, there was a substantially greater decline in the proportion of uninsured hospitalizations for the major CV events under study in states that expanded Medicaid vs those states that did not. The distinct change in the proportion of hospitalizations that were uninsured and covered by Medicaid in the expansion states suggested that significant decreases in uninsured hospitalizations for major CV events were associated with Medicaid expansion. That these findings were already present within 1 year of ACA implementation suggested that the changes in payer mix were immediate. To our knowledge, our study is the first to document these observations for major CV events and to quantify the magnitude of these potential associations. This is important given that one could not presume that these changes in payer proportions would have been uniformly present across different types of hospitalizations.20 Notably, these changes occurred without an immediate influence on in-hospital mortality.

    Present Study in Context

    Our findings on payer mix for CV hospitalizations were consistent with previous analyses examining the potential association of Medicaid expansion on coverage status for all non-Medicare hospitalizations.18,19,30 Our study extended previous analyses in 3 significant ways. First, we characterized these payer results by specifically focusing on hospitalizations for CVD, the leading cause of death and hospitalization in the United States. We investigated which aspects of the ACA may have had the most significant immediate effect on coverage for uninsured patients who would be hospitalized for major CV events. It has been reported that millions of individuals in the United States have acquired private health insurance coverage since the full implementation of the ACA.10,11 Nevertheless, our analyses did not show a substantial difference between the expansion and nonexpansion states in the private insurance share of CV discharges after expansion. Thus, our data suggested that, of the uninsured individuals who would be hospitalized for a major CV event and who acquired insurance coverage at that time, most qualified as low-income status and received access to Medicaid in the first year after full ACA implementation. Second, our study examined this focused question of whether there were associations between changes in payer mix for CVD hospitalizations and Medicaid expansion with hospitalization data from a large number of states, which together represented nearly three-quarters of the US population.31 Finally, we examined whether state expansion status was associated with in-hospital mortality for these major CV events, an important consideration when evaluating the potential association of insurance status with outcomes.

    Previous data have shown that significant variations in CV health between states were contributed to not only by individual factors but also state-level factors, including policy.32 Previous data suggested that the ACA had led to improvements among low-income non-Medicare patients, such as having a primary care clinician, accessing routine medical care, and not forgoing medications because of cost—with Medicaid expansion being associated with the most substantial improvements.14-16 While previous studies have shown that insurance status can be tied to hospital care and outcomes after major CV events, there was an overall dearth of data directly comparing the CV outcomes of hospitalized patients with Medicaid coverage with those who were uninsured.5-7,9 We observed that in-hospital mortality did not change significantly after Medicaid expansion in our study. Greater decreases in uninsured hospitalizations did not translate to decreases in hospital death rates—which may not have been unexpected after only 1 year given that insurance coverage may take time to influence overall health status. However, in-hospital mortality also did not increase significantly despite the fact that ACA expansion provided coverage for more low-income persons and that lower socioeconomic status was associated with a higher prevalence of CVD and worse outcomes.2 Adequately exploring any potential associations of expanded coverage will require further research using more long-term data beyond what we had available for this study.

    A major consideration of our findings is the potential implications of cost at the macroeconomic and microeconomic levels. Cardiovascular disease is the leading source of medical expenditures in the United States, with direct medical costs projected to approach nearly $1 trillion by 2030, of which more than half is currently the result of inpatient hospitalizations.2 By nature of near-universal Medicare coverage for people 65 years or older, it is nonelderly patients with CVD who are more likely to be uninsured. That Medicaid expansion was associated with a significant reduction in the proportion of uninsured hospitalizations may have had important out-of-pocket cost implications for low-income patients who would have been previously uninsured but now had access to Medicaid.

    There was evidence that hospitals may have benefited from Medicaid expansion through decreases in uncompensated care costs.33,34 In contrast, the ACA also had provisions to decrease the disproportionate share hospital allotments, which supplemented the income of hospitals that take care of underinsured patients.35 In addition, any decrease in cost burden for individuals and hospitals because of the Medicaid expansion might be expected to shift to the states and the federal government. Thus, discussions of economic impact and costs must take into account all of these factors in addition to effects on hospital, state, and federal expenditures. Adequate time is unlikely to have passed to draw conclusions on the association of the ACA with costs related to CVD—the unclear direction of federal policy on health insurance coverage adds complexity to what is already a complicated assessment. Given the prominent role of CVD in overall public health, and as the leading source of medical expenditures, it will be important that future study takes into account the prominent role of CVD in addressing these questions.

    Limitations

    Strengths of our study included the use of a large representative sample of the top 3 causes of CVD-related hospitalizations in the United States. Our analyses were based on data from the Healthcare Cost and Utilization Project State Inpatient Databases,22 each of which provided a census of all inpatient hospitalizations by state. We note that the change in the sampling design for the National Inpatient Sample23 after 2011 did not apply to our analyses. Nevertheless, our findings should be viewed with other limitations in mind. We used ICD-9 diagnosis codes, which identified CVD as the principal diagnosis, to create our composite outcome. Despite good positive predictive value for identifying acute myocardial infarction, heart failure, and stroke, use of ICD-9 diagnosis codes for these, as with other diagnoses, can have suboptimal sensitivity.26,28,36 It is possible that we underestimated the total number of hospitalizations for these events. In addition, our data did not represent individuals but rather encounters, and rehospitalizations were not distinguished from primary hospitalizations. Data were only available for 1 year after ACA implementation; thus, we could not determine whether the associations noted in this study would have continued to the present. We also could not reliably analyze the changes in total number of hospitalizations for major CV events after ACA implementation—another important consideration in the assessment of the ACA influence on CVD, which was beyond the scope of this study. We used data from a large number of states that together represented nearly three-quarters of the US population; nevertheless, we could not rule out the possibility that if data from additional states were available, it might have significantly changed our results. We adjusted for time-varying, state-level demographics and accounted for variation at the state level using a fixed-effects modeling strategy. Nevertheless, there may be some residual confounding based on time-varying, state-level health or economic trends that we were unable to adjust for because we did not have data specific to the hospitalized population.

    Conclusions

    States that chose to expand Medicaid as part of the ACA implementation had a significantly greater reduction in the proportion of uninsured hospitalizations for major CV events compared with the nonexpansion states, but expansion status was not associated with in-hospital mortality in the first postexpansion year. Further research is necessary to determine how state-level policy decisions regarding ACA implementation could differentially influence short-term and long-term CV outcomes before, during, and after hospitalization.

    Back to top
    Article Information

    Accepted for Publication: May 24, 2018.

    Published: August 24, 2018. doi:10.1001/jamanetworkopen.2018.1296

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2018 Akhabue E et al. JAMA Network Open.

    Corresponding Author: Ehimare Akhabue, MD, Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, 680 N Lake Shore Dr, Ste 1400, Chicago, IL 60611 (ehimare.akhabue@northwestern.edu).

    Author Contributions: Drs Akhabue and Pool had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Akhabue, Yancy.

    Acquisition, analysis, or interpretation of data: Akhabue, Pool, Greenland, Lloyd-Jones.

    Drafting of the manuscript: Akhabue.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Akhabue, Pool.

    Supervision: Yancy, Greenland, Lloyd-Jones.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: This research was supported in part by grant 15SFDRN25080331 from the American Heart Association.

    Role of the Funder/Sponsor: The American Heart Association had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Additional Contributions: This study examined hospitalizations using discharge data from State Inpatient Databases of the following states: Arizona, Arkansas, California, Colorado, Florida, Hawaii, Illinois, Indiana, Iowa, Kansas, Kentucky, Maine, Maryland, Missouri, Nevada, New Jersey, New Mexico, New York, North Carolina, Oklahoma, Oregon, Rhode Island, South Carolina, Tennessee, Texas, Utah, Vermont, Washington, Wisconsin, and Wyoming. We thank these states and the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project for making these data publicly available.

    References
    1.
    Kochanek  KD, Murphy  SL, Xu  J, Tejada-Vera  B.  Deaths: final data for 2014.  Natl Vital Stat Rep.2016;65(4):1-122. PubMedGoogle Scholar
    2.
    Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
    3.
    Centers for Disease Control and Prevention, National Center for Health Statistics. National Hospital Discharge Survey: number and rate of discharges by first-listed diagnostic categories, 2010. https://www.cdc.gov/nchs/data/nhds/2average/2010ave2_firstlist.pdf. Accessed May 8, 2017.
    4.
    Centers for Disease Control and Prevention, National Center for Health Statistics. Table 105: no health insurance coverage among persons under age 65, by selected characteristics: United States, selected years 1984-2014. https://www.cdc.gov/nchs/data/hus/2015/105.pdf. Accessed May 8, 2017.
    5.
    Medford-Davis  LN, Fonarow  GC, Bhatt  DL,  et al.  Impact of insurance status on outcomes and use of rehabilitation services in acute ischemic stroke: findings from Get With the Guidelines–Stroke.  J Am Heart Assoc. 2016;5(11):e004282. doi:10.1161/JAHA.116.004282PubMedGoogle ScholarCrossref
    6.
    James  ML, Grau-Sepulveda  MV, Olson  DM,  et al.  Insurance status and outcome after intracerebral hemorrhage: findings from Get With the Guidelines–Stroke.  J Stroke Cerebrovasc Dis. 2014;23(2):283-292. doi:10.1016/j.jstrokecerebrovasdis.2013.02.016PubMedGoogle ScholarCrossref
    7.
    Kapoor  JR, Kapoor  R, Hellkamp  AS, Hernandez  AF, Heidenreich  PA, Fonarow  GC.  Payment source, quality of care, and outcomes in patients hospitalized with heart failure.  J Am Coll Cardiol. 2011;58(14):1465-1471. doi:10.1016/j.jacc.2011.06.034PubMedGoogle ScholarCrossref
    8.
    Mathews  R, Wang  TY, Honeycutt  E,  et al; TRANSLATE-ACS Study Investigators.  Persistence with secondary prevention medications after acute myocardial infarction: insights from the TRANSLATE-ACS study.  Am Heart J. 2015;170(1):62-69. doi:10.1016/j.ahj.2015.03.019PubMedGoogle ScholarCrossref
    9.
    Ng  DK, Brotman  DJ, Lau  B, Young  JH.  Insurance status, not race, is associated with mortality after an acute cardiovascular event in Maryland.  J Gen Intern Med. 2012;27(10):1368-1376. doi:10.1007/s11606-012-2147-9PubMedGoogle ScholarCrossref
    10.
    Carman  KG, Eibner  C, Paddock  SM.  Trends in health insurance enrollment, 2013-15.  Health Aff (Millwood). 2015;34(6):1044-1048. doi:10.1377/hlthaff.2015.0266PubMedGoogle ScholarCrossref
    11.
    Uberoi  N, Finegold  K, Gee  E; Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. ASPE Issue Brief: health insurance coverage and the Affordable Care Act, 2010-2016. https://aspe.hhs.gov/pdf-report/health-insurance-coverage-and-affordable-care-act-2010-2016. Published March 2016. Accessed May 8, 2017.
    12.
    Kaiser Family Foundation. Status of state action on the Medicaid expansion decision as of January 1, 2017. http://www.kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/. Accessed May 8, 2017.
    13.
    Sommers  BD, Blendon  RJ, Orav  EJ, Epstein  AM.  Changes in utilization and health among low-income adults after Medicaid expansion or expanded private insurance.  JAMA Intern Med. 2016;176(10):1501-1509. doi:10.1001/jamainternmed.2016.4419PubMedGoogle ScholarCrossref
    14.
    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:10.1001/jama.2015.8421PubMedGoogle ScholarCrossref
    15.
    Martinez ME, Ward BW. Health care access and utilization among adults aged 18–64, by poverty level: United States, 2013–2015. https://www.cdc.gov/nchs/products/databriefs/db262.htm. Accessed May 8, 2017.
    16.
    Okoro  CA, Zhao  G, Fox  JB, Eke  PI, Greenlund  KJ, Town  M.  Surveillance for health care access and health services use, adults aged 18-64 years—Behavioral Risk Factor Surveillance System, United States, 2014.  MMWR Surveill Summ. 2017;66(7):1-42. doi:10.15585/mmwr.ss6607a1PubMedGoogle ScholarCrossref
    17.
    Miller  S, Wherry  LR.  Health and access to care during the first 2 years of the ACA Medicaid expansions.  N Engl J Med. 2017;376(10):947-956. doi:10.1056/NEJMsa1612890PubMedGoogle ScholarCrossref
    18.
    Nikpay  S, Buchmueller  T, Levy  HG.  Affordable Care Act Medicaid expansion reduced uninsured hospital stays in 2014.  Health Aff (Millwood). 2016;35(1):106-110. doi:10.1377/hlthaff.2015.1144PubMedGoogle ScholarCrossref
    19.
    DeLiere  T, Joynt  K, McDonald  R; Department of Health and Human Services, Office of Assistant Secretary for Planning and Evaluation. ASPE Issue Brief: impact of insurance expansion on hospital uncompensated care costs in 2014. https://aspe.hhs.gov/pdf-report/impact-insurance-expansion-hospital-uncompensated-care-costs-2014. Published September 24, 2014. Accessed May 8, 2017.
    20.
    Freedman  S, Nikpay  S, Carroll  A, Simon  K.  Changes in inpatient payer-mix and hospitalizations following Medicaid expansion: evidence from all-capture hospital discharge data.  PLoS One. 2017;12(9):e0183616. doi:10.1371/journal.pone.0183616PubMedGoogle ScholarCrossref
    21.
    Xu  J, Murphy  SL, Kochanek  KD, Arias  E.  Mortality in the United States, 2015.  NCHS Data Brief. 2016;(267):1-8. PubMedGoogle Scholar
    22.
    Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP). Introduction to the HCUP State Inpatient Databases (SID). https://www.hcup-us.ahrq.gov/db/state/siddist/SID_Introduction.jsp. Published February 2018. Accessed July 13, 2018.
    23.
    Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP). Why the NIS should not be used to make State-level estimates. https://www.hcup-us.ahrq.gov/db/nation/nis/nis_statelevelestimates.jsp. Published January 2016. Accessed July 13, 2018.
    24.
    US Department of Health and Human Services, Agency for Healthcare Research and Quality. HCUPnet. Healthcare Cost and Utilization Project. https://hcupnet.ahrq.gov/. Accessed July 17, 2017.
    25.
    Yeh  RW, Sidney  S, Chandra  M, Sorel  M, Selby  JV, Go  AS.  Population trends in the incidence and outcomes of acute myocardial infarction.  N Engl J Med. 2010;362(23):2155-2165. doi:10.1056/NEJMoa0908610PubMedGoogle ScholarCrossref
    26.
    Kokotailo  RA, Hill  MD.  Coding of stroke and stroke risk factors using International Classification of Diseases, revisions 9 and 10.  Stroke. 2005;36(8):1776-1781. doi:10.1161/01.STR.0000174293.17959.a1PubMedGoogle ScholarCrossref
    27.
    Roumie  CL, Mitchel  E, Gideon  PS, Varas-Lorenzo  C, Castellsague  J, Griffin  MR.  Validation of ICD-9 codes with a high positive predictive value for incident strokes resulting in hospitalization using Medicaid health data.  Pharmacoepidemiol Drug Saf. 2008;17(1):20-26. doi:10.1002/pds.1518PubMedGoogle ScholarCrossref
    28.
    Psaty  BM, Delaney  JA, Arnold  AM,  et al.  Study of cardiovascular health outcomes in the era of claims data: the Cardiovascular Health Study.  Circulation. 2016;133(2):156-164. doi:10.1161/CIRCULATIONAHA.115.018610PubMedGoogle ScholarCrossref
    29.
    Gunasekara  FI, Richardson  K, Carter  K, Blakely  T.  Fixed effects analysis of repeated measures data.  Int J Epidemiol. 2014;43(1):264-269. doi:10.1093/ije/dyt221PubMedGoogle ScholarCrossref
    30.
    Davis  MM, Gebremariam  A, Ayanian  JZ.  Changes in insurance coverage among hospitalized nonelderly adults after Medicaid expansion in Michigan.  JAMA. 2016;315(23):2617-2618. doi:10.1001/jama.2016.6303PubMedGoogle ScholarCrossref
    31.
    US Census Bureau. Annual estimates of the resident population: April 1, 2010, to July 1, 2017: 2017 population estimates. https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=PEP_2017_PEPANNRES&src=pt. Accessed July 13, 2018.
    32.
    Gebreab  SY, Davis  SK, Symanzik  J, Mensah  GA, Gibbons  GH, Diez-Roux  AV.  Geographic variations in cardiovascular health in the United States: contributions of state- and individual-level factors.  J Am Heart Assoc. 2015;4(6):e001673. doi:10.1161/JAHA.114.001673PubMedGoogle ScholarCrossref
    33.
    Blavin  F.  Association between the 2014 Medicaid expansion and US hospital finances.  JAMA. 2016;316(14):1475-1483. doi:10.1001/jama.2016.14765PubMedGoogle ScholarCrossref
    34.
    Dranove  D, Garthwaite  C, Ody  C.  Uncompensated care decreased at hospitals in Medicaid expansion states but not at hospitals in nonexpansion states.  Health Aff (Millwood). 2016;35(8):1471-1479. doi:10.1377/hlthaff.2015.1344PubMedGoogle ScholarCrossref
    35.
    Chokshi  DA, Chang  JE, Wilson  RM.  Health reform and the changing safety net in the United States.  N Engl J Med. 2016;375(18):1790-1796. doi:10.1056/NEJMhpr1608578PubMedGoogle ScholarCrossref
    36.
    Agarwal  SK, Wruck  L, Quibrera  M,  et al.  Temporal trends in hospitalization for acute decompensated heart failure in the United States, 1998-2011.  Am J Epidemiol. 2016;183(5):462-470. doi:10.1093/aje/kwv455PubMedGoogle ScholarCrossref
    ×