Abstract
Background
The Hospital Readmission Reduction Program (HRRP) penalizes hospitals with “excess” readmissions up to 3% of Medicare reimbursement. Approximately 75% of eligible hospitals will receive penalties, worth an estimated $428 million, in fiscal year 2015.
Objective
To identify the demographic, socioeconomic and hospital characteristics that distinguish maximum penalty hospitals from geographically matched no-penalty hospitals.
Design
Case-control. Cases were hospitals to receive the maximum 3% penalty under the HRRP during fiscal year 2015. Controls were drawn from no-penalty hospitals, and matched 1:1 to cases by geographic proximity.
Setting
3,383 U.S. hospitals eligible for HRRP
Participants
Thirty-nine (39) case and 39 control hospitals from the HRRP cohort
Measurements
Socioeconomic status variables collected by the American Community Survey (ACS). Select measures of hospital quality and health system capacity collected by the Centers for Medicare and Medicaid Services and the Dartmouth Atlas of Health Care.
Results
39 hospitals received a maximum penalty. Low income, low education and low property values all increased the odds of a maximum penalty designation after adjusting for age, sex, race and English language. For every $1,000 increase in median income, the odds of a maximum penalty decreased by 15% (OR = 0.85, 95% CI = 0.74, 0.97).
Conclusion
Cases were more likely than controls to be located in counties with low socioeconomic status, raising questions about unintended consequences of national benchmarks for phenomena underpinned by environmental factors; specifically, whether maximum penalties under the HRRP are a consequence of underperforming hospitals or a manifestation of underserved communities.
Keywords: Hospital Readmission Reduction Program, readmission, health policy, socioeconomic status
Introduction
According to Centers for Medicare and Medicaid Services (CMS), approximately 1 in 5 patients discharged from a hospital will be readmitted within 30 days.1 The Hospital Readmission Reduction Program (HRRP) is designed to reduce readmission by withholding up to 3% of all Medicare reimbursement from hospitals with “excess” readmissions; however absent from the HRRP is adjustment for socioeconomic status (SES), which CMS holds may undermine incentives to reduce health disparities and institutionalize lower standards for hospitals serving disadvantaged populations.2
Lack of SES adjustment has been criticized by those who point to evidence highlighting post-discharge environment and patient socioeconomic status as drivers of readmission, as hospitals that serve low SES individuals will bear a disproportionate share of penalties.3–6 Single-center,3,7,8 regional,9,10 and nationwide6,11 studies highlight census tract level socioeconomic variables as predictive of readmission. Single-center studies, robust in controlling for confounders including staffing, training, electronic medical record utilization and transitional care processes, do not allow comparisons between hospitals, limiting utility in HRRP evaluation. Multi-center cohorts, on the other hand, allow for comparisons between high and low penalty hospitals, pioneered by Joynt and colleagues after the first round of HRRP penalties;12 yet this technique may not account for confounding caused by extensive demographic, socioeconomic and hospital characteristic heterogeneity inherent in a national cohort. Analysis of the 2015 HRRP penalty data by Sjoding et al revealed higher COPD readmission rates in the Mid-Atlantic, Midwest and South relative to other regions; however the magnitude of small-area variation, as well as its relationship to population SES has yet to be characterized.
Therefore, our analysis plan involved a matched case-control design, whereby each maximum penalty hospital was matched to the nearest geographically located non-penalty hospital. We used geographic matching to isolate SES factors predictive of readmission within specific geographies, an effort to control for regional population differences. We hypothesized that among localized hospital pairs, disparities in population socioeconomic status are the most significant predictors of a maximum penalty. Now in the third year of the HRRP, with approximately 75% of eligible hospitals to receive penalties worth an estimated $428 million in fiscal year 2015,13 we offer a small-area analysis of bipolar extremes to inform debate surrounding the HRRP with evidence regarding the causes and implications of readmission penalties.
Methods
Design & Study Sample
This study relies on a case-control design. Cases were defined as hospitals in the United States to receive the maximum, 3%, HRRP penalty in fiscal year 2015. Controls were drawn from the national cohort of hospitals potentially subject to HRRP penalties that received no readmission penalty in fiscal year 2015 with at least 1 index admission for any of the following conditions: heart failure (HF), acute myocardial infarction (AMI), pneumonia (PN), total knee or total hip arthroscopy (THA/TKA) or chronic obstructive pulmonary disease (COPD).
Data sources
Penalty data was drawn from the 2015 master penalty file,14 accessed via CMS.gov. County-level demographic and socioeconomic data was gathered from the 2013 American Community Survey (ACS), a subsidiary of US Census. Data on hospital characteristics and capacity as well as regional health care service utilization and supply was drawn from the 2010 Dartmouth Atlas15, the 2012 Medicare Cost Report16 as well as the 2014 Hospital Care Downloadable Database. See eTable 1 for a list of data sources, data sets and variables.
Hospital-level CMS data was linked to the master 2015 penalty file. Dartmouth Atlas data was subsequently linked to the file using the Dartmouth Atlas “Hospital to H.S.A./H.R.R. Crosswalk” file (accessed via DartmouthAtlas.org.) Each hospital was assigned the profile of the hospital service area (H.S.A.) and Hospital Referral Region (H.R.R) in which it is located. A hospital service area is a geographic region defined by hospital admissions; the majority, but not entirety, of residents of a given H.S.A. utilize the corresponding hospital. Similarly, an H.R.R. is a geographic region defined by referrals for major cardiovascular and neurosurgery procedures. County-level socioeconomic data was linked to the dataset by county name; thus, hospital socioeconomic profiles are based on the county in which it is located. Any missing hospital-level data is acknowledged in eTable 1.
Case-control matching
Hospitals were geocoded by zip code. Geographic Information Systems mapping software (ESRI ArcGIS, Redlands, CA) relied upon Euclidean allocation distance spatial analysis17,18 to match each maximum penalty hospital to the nearest non-penalty hospital. As shown in eTable 2, each case was matched to a distinct control; duplicate controls were replaced with the nearest unmatched non-penalty hospital.
Statistical analysis
Univariate analyses utilized paired t-test. Given the matched design, conditional logistic regression was used to calculate the main effect, defined as the odds ratio of several potential socioeconomic predictors associated with maximum penalty designation. Multivariate conditional regression modeling was used to adjust main effects for county level age, sex and race profiles of each hospital region. Self-reported race and ethnicity collected by the American Community Survey was used to identify demographic variation among hospitals. Statistical analyses were conducted using STATA (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX)
Results
Vector-Matching Maximum Penalty and Non-Penalty Hospitals
Of 3,383 hospitals eligible for the HRRP, 39 received the maximum penalty and 770 received no penalty. 39 control hospitals were identified using a GIS vector analysis algorithm, which matched each case to the nearest non-penalty hospital. The median distance between case and control was 42.5 miles (interquartile range: 25th percentile, 15.4 miles; 75th percentile, 98.4 miles). 17 case-control pairs (44%) were located in the same hospital referral region (HRR), 6 of which were in the same hospital service area (HSA) while an additional 10 hospital pairs (26%) were located in different HRRs within the same state. The remaining 12 hospital pairs (31%) were located in adjacent states. Seven pairs (18%) were located within the same county.
Hospital characteristics
Case and control profiles are presented in Table 1. Maximum penalty hospitals were significantly smaller than non-penalty hospitals in terms of total inpatient days (7,220 vs 20,142, p=0.005), number of staffed beds (81.7 vs 196.2, p=0.002) as well as case volume for each of the 5 conditions measured by the HRRP (PN, CHF, AMI, COPD THA/TKA; p<0.05 for each). Maximum penalty and non-penalty hospitals were located in hospital service areas with similar physician staffing profiles; however, cases had significantly more per capita beds than controls (3.4 vs 2.5 beds per 1000 residents, p=0.002).
Table 1.
Hospital characteristics of max. penalty and no penalty hospitals, matched by hospital characteristics
| No penalty | Max. penalty | p-value | |
|---|---|---|---|
| Matched Criteria | |||
| Hospital Owner (n)+ | 1.0 | ||
| Government | 4 | 4 | |
| For-profit | 15 | 15 | |
| Non-profit | 17 | 17 | |
| Quartile of beds+ | 1.0 | ||
| First | 16 | 16 | |
| Second | 16 | 16 | |
| Third | 5 | 5 | |
| Fourth | 1 | 1 | |
| Case mix index (above median, %)+ | 42.1 | 42.1 | 1.0 |
| Total HRRP eligible cases** (above median, %) | 84.2 | 84.2 | 1.0 |
| Patients with ambulatory care visit within 14 days of discharge (above median, %)‡ | 34.2 | 34.2 | 1.0 |
| Volume and service mix | |||
| Total HRRP eligible cases | 937 | 882 | 0.7 |
| Total inpatient days+ | 15,600 | 15,000 | 0.9 |
| Total discharges+ | 3,800 | 3,660 | 0.9 |
| Total beds+ | 89.0 | 82.3 | 0.7 |
| Emergency Services (%)῏ | 87 | 82 | 0.5 |
| Geography: rural (%) | 5 | 24 | 0.022 |
| Case mix index+ | 1.69 | 1.51 | 0.12 |
| Charity care charges ($ millions)+ | 3.4 | 4.9 | 0.3 |
| Mortality and Complications | |||
| 30-day AMI mortality (%)῏ | 15.3 (n=20) | 15.4 (n=20) | 0.8 |
| 30-day CHF mortality (%)῏ | 12.2 (n=29) | 11.6 (n=27) | 0.09 |
| 30-day PN mortality (%)῏ | 12.3 (n=29) | 12.2 (n=28) | 0.8 |
| Complications hip/knee (%)‡ | 3.17 | 3.64 | 0.0031 |
| Age sex race adjusted hospital-mortality (%)‡ | 4.9 | 5.3 | 0.009 |
| Primary care (Medicare enrollees)‡ | |||
| Mammogram within past year; ages 67-69 (%) | 63.6 | 57.9 | 0.0006 |
| Discharges for ACSC× conditions (per 1000 enrollees) | 61.3 | 108 | 0.0002 |
| Ambulatory care visit in past year (%) | 80.7 | 80.6 | 0.9 |
| Ambulatory care within 14 days of discharge | 63.4 | 61.6 | 0.3 |
| ED visit within 30 days of discharge | 18.4 | 20.8 | <0.0001 |
| Medicare diabetics with annual A1c test (%) | 85 | 83.3 | 0.07 |
| Medicare diabetics with annual eye exam (%) | 66.4 | 61.6 | 0.008 |
| Financiall¥ | |||
| Operating margin (%) | 6.9 | −0.1 | 0.033 |
| Total assets (millions) | 97.1 | 83.1 | 0.6 |
| Net patient revenue | 93.9 | 71.9 | 0.3 |
Medicare Cost Report (2012);
Dartmouth Atlas of Healthcare (2012);
CMS Hospital Compare (2014);
AHA Hospital Statistics Database (2012);
AMI, COPD, HF, PN, THA/TKA;
Ambulatory care sensitive conditions
Demographic Characteristics
Demographic indicators are presented in Table 2. Maximum penalty and non-penalty hospitals were located in counties with similar median age, sex and race profiles. Relative to non-penalty hospitals, maximum penalty hospitals were located in counties with a larger share of individuals who spoke only English at home (90.0% vs 87.2%, p=0.087) and were more likely to be located in rural environments (48.7% vs 28.2%, p=0.063); however neither difference was statistically significant.
Table 2.
Socioeconomic profile of maximum penalty and no penalty hospitals, matched by hospital characteristics
| No penalty | Max. penalty | p value | |
|---|---|---|---|
| Income and Employment | |||
| Per capita income ($) | 26,400 | 24,500 | 0.2 |
| Median household income ($) | 50,500 | 46,700 | 0.3 |
| Families below poverty line (%) | 10.9 | 14.7 | 0.004 |
| Individuals below poverty line (%) | 15.5 | 19.1 | 0.015 |
| Food stamps (%) | 12.7 | 16.8 | 0.005 |
| Labor force participation (%) | 63.6 | 57 | 0.0008 |
| Unemployment (%) | 4.5 | 4.7 | 0.4 |
| Health Insurance | |||
| Uninsured (%) | 12.6 | 13.2 | 0.6 |
| Private insurance (%) | 68.1 | 61.5 | 0.003 |
| Medicaid (%) | 12.2 | 15.8 | 0.005 |
| Demographics | |||
| Males per 100 females | 98.9 | 99.2 | 0.9 |
| Median age | 37.9 | 39.1 | 0.2 |
| 65 years old (%) | 14.8 | 15.3 | 0.6 |
| White (%) | 76.8 | 76.2 | 0.9 |
| Black (%) | 5.8 | 9.7 | 0.1 |
| Hispanic or Latino (%) | 13.5 | 8.9 | 0.2 |
| Education | |||
| High-school graduate (age >25, %) | 87.5 | 82.2 | 0.001 |
| Bachelors degree (age >25, %) | 25.4 | 22.5 | 0.2 |
| Age 5 and older who speak English less than “very well” (%) | 2.0 | 2.0 | 0.9 |
Source: American Community Survey (2015)
Socioeconomic characteristics
County-level socioeconomic indicators are presented in Table 2. The socioeconomic profiles of counties containing maximum penalty hospitals were lower than those in which non-penalty hospitals were located. Maximum penalty hospitals were located in counties with lower median household income ($45,500 vs $51,100, p=0.006) as well as greater prevalence of families below the poverty line. Median property values were lower in counties with maximum penalty hospitals compared to those with non-penalty hospitals ($138,000 vs $175,000, p<0.001). Federal poverty assistance programs were more common in counties with maximum penalty hospitals relative to non-penalty hospitals, in terms of percentage of the population to receive Medicaid (20% vs 16.5%, p=0.006) as well as SNAP and Supplemental Security Income (32.6% vs 27.0%, p=0.003). The proportion of the population with a bachelor’s degree was lower in counties with maximum penalty hospitals relative to non-penalty hospitals (6.6% vs 8.5%, p =0.003)
As shown in Table 3, counties with lower SES profiles had higher odds of receiving the maximum penalty. Low income, low education and low property values all increased the odds of a maximum penalty. Low income was the most significant predictor; for every $1,000 increase in median income, the odds of a maximum penalty decreased by 15% (OR = 0.85, 95% CI = 0.74, 0.97; p=0.019). All reported odds ratios are adjusted for county level profiles of age, sex, race and exclusive household use of English language.
Table 3.
Select socioeconomic and hospital characteristics associated with maximum-penalty hospitals and geographically paired no-penalty hospitals
| No penalty (n=39) |
Max. penalty (n=39) |
p value | |
|---|---|---|---|
| Socioeconomic and demographic profile‡ | |||
| Income, employment and education | |||
| Unemployed (%) | 4.7 | 4.7 | 0.9 |
| Labor force participation (%) | 63.1 | 56.7 | <0.001 |
| Receive SNAP (%) | 13.1 | 17.0 | 0.005 |
| Individuals below poverty line (%) | 15.6 | 19.2 | 0.007 |
| Median household income ($) | 52,000 | 46,400 | 0.004 |
| Medicaid (%) | 12.6 | 15.9 | 0.014 |
| High school graduate (age>25, %) | 86.4 | 82.1 | 0.005 |
| Bachelor’s or higher (age>25, %) | 28.1 | 22.3 | 0.002 |
| Demographics | |||
| White (%) | 73.2 | 76.3 | 0.3 |
| Black (%) | 9.7 | 9.4 | 0.9 |
| Hispanic or Latino (%) | 11.5 | 8.7 | 0.09 |
| Median Age (years ± SD) | 37.4 ± 7.6 | 39.1±4.2 | 0.2 |
| Greater 65 years old | 15.2 | 15.3 | 0.9 |
| Males: 100 females | 93.8 | 98.1 | 0.09 |
| Hospital and health system characteristics | |||
| Inpatient profile | |||
| For-profit (%)** | 20.5 | 15.4 | 0.082 |
| Emergency services (% Yes)** | 84.6 | 82.1 | 0.7 |
| Geographic region: urban (%) | 71.8 | 51.3 | 0.063 |
| Total beds+ | 196 | 81.7 | 0.002 |
| Total HRRP-eligible admissions | 1,580 | 880 | 0.03 |
| Case mix index+ | 1.56 | 1.51 | 0.5 |
| Charity care charges ($ millions)+ | 8.5 | 4.9 | 0.095 |
| Population health (Medicare enrollees)* | |||
| ASR adjusted mortality among Medicare enrollees (%) | 4.78 | 5.37 | 0.0001 |
| Ambulatory care visit within 14 days of discharge (%) | 62.8 | 61.1 | 0.24 |
| Had ED visit within 30 days of discharge (%) | 18.7 | 21.0 | 0.0004 |
| Discharges for ambulatory care sensitive conditions (per 1,000 Medicare enrollees) | 59.6 | 109.6 | 0.0001 |
| Beneficiaries with ambulatory visit in past year (%) | 80.6 | 80.6 | 1.0 |
| Females 67-69 with having at least 1 mammogram within past year (%) | 62.6 | 57.8 | 0.0028 |
| Diabetic pts (65-75 yo) with annual HA1c test (%) | 85.4 | 83.2 | 0.015 |
| Diabetic pts (65-75 yo) with annual eye exam (%) | 66.0 | 61.5 | 0.001 |
Source:
2015 American Community Survey
2012 Dartmouth Atlas,
2012 CMS Hospital Compare,
2012 CMS Cost Report
Discussion
Our analysis reveals county-level socioeconomic profiles to be predictors of maximum HRRP penalties. Specifically, location within counties with low income, low property values and low education levels all significantly increased the odds of maximum penalty, after adjusting for age, sex, race and ethnicity. In contrast, we observed no difference between cases and controls with regard to county-level race and ethnicity distributions.
Our study complements that of Joynt and colleagues,12 whose analysis of the first year of the HRRP revealed safety net hospitals (top quartile in disproportionate share index) had nearly double the odds of other hospitals to receive a high penalty (highest 50% of penalties). We add to current literature with evidence that local variation in readmission penalties is a function of income and education, but not race and ethnicity. Others have shown racial and ethnic disparities in readmission rates, even after adjusting for income and disease severity;19,20 leading the American Hospital Association to call for race and ethnicity adjustment of HRRP penalties.21 In contrast, we offer evidence that the difference between maximum penalty and non-penalty hospitals is not a function of race or ethnicity.
Hospital quality and population health
As shown in Table 1, maximum readmission penalties appear to be associated with lower hospital quality. Complication rates were slightly higher in maximum penalty hospitals, consistent with recent studies highlighting complications as important drivers of surgical readmissions.22,23 Moreover, hospital-wide mortality rates were significantly higher in maximum penalty hospitals relative to non-penalty hospitals (5.37 vs 4.78, p<0.001).
Using national data, Krumholz and colleagues found no correlation between rates of readmission and mortality for CHF, AMI, and PN24, a phenomenon acknowledged by MedPac in a 2013 report titled, “Refining the hospital readmission reduction program.”25 In large national cohort studies, others have shown low SES to be associated with elevated readmission but not mortality.10,11 In contrast we restricted our analysis to geographically matched pairs and are the first to present evidence of an association between readmission and hospital-wide mortality, adjusted for age, sex and race.
An alternative hypothesis is maximum penalties are less a function of hospital quality than a consequence of population health and public health capacity. Rates of ambulatory care sensitive condition (ACSC) discharges were twice as high in hospital service areas of maximum penalty hospitals relative to non-penalty hospitals (110 vs 60 per 1000 Medicare enrollees, p<0.001). ACSC discharges have been used to measure primary care quality for 30 years, the assumption being that admission for chronic conditions such as heart failure can be prevented with effective primary care.26,27 Moreover, patients discharged from maximum penalty hospitals were more likely to have an emergency room visit within 30 days of discharge (11.0 vs 18.7%, p<0.001). Higher rates of ACSCs and post-discharge ED visits suggest that outpatient resources in maximum penalty service areas struggle to effectively manage the disease burden of high risk populations. Geography may be a contributor; despite matching, maximum penalty hospitals were more likely to be located in rural settings than no penalty hospitals (48.7% vs 28.2%, p=0.058).
However, as shown in Table 1, benchmarks of primary care access and utilization do not explain the difference in readmissions, ACSC discharges or ED visits. Both maximum penalty and non-penalty hospitals were located in hospital service areas with similar concentrations of PCPs per 1000 residents. Moreover, neither the percentage of Medicare enrollees to receive an annual ambulatory care visit nor the proportion of patients to receive ambulatory care visits within 14 days of discharge was different between cases and controls. Our study is unable to explain root causes of disparate readmission rates with broad process measures reflecting primary care access and utilization, calling into question the utility of such measures as performance benchmarks.
Our findings suggest that hospitals providing care to vulnerable communities (defined by low income, low education, low property values and high rates of chronic disease) are disproportionately penalized. McHugh and colleagues reveal high nurse staffing levels to be protective against readmission penalties,28 yet high penalties to low-margin hospitals may encourage providers to reduce rather than increase staff. It may be better policy to direct resources rather than penalties to underserved communities and our findings echo others who raise concern about disproportionate penalties to hospitals serving low SES patients.2,5,6,29
Implications and Future Directions
In response to criticism surrounding the HRRP, the National Quality Forum endorsed the general concept of SES adjustment for hospital quality measures.30 Subsequently, in a briefing dated March 24th, 2015, the Medicare Payment Advisory Commission (MedPAC), a government agency which provides Medicare policy analysis to Congress, proposed an SES adjustment methodology of “dividing hospitals into peer groups based on their overall share of low-income Medicare patients, and then setting a benchmark readmissions target for each peer group;”31 in other words lower standards for hospitals that serve low-income populations. MedPAC’s proposal will reduce the penalties to “safety net” institutions - progress but not a permanent solution. Although the HRRP appears to be working, according to the US Department of Health and Human Services, all cause readmissions fell by 150,000 between January, 2012 and February, 2013,32 we suggest the program needs reform. Neither the current HRRP nor the MedPac revision proposal consider geographic and environmental components of readmission. The HRRP promotes national improvement in exchange for regional regression.
Fair quality measures are key to value-based reimbursement models; yet, implicit in penalties for “excess” readmissions is the assumed ability to calculate hospital performance targets. Benchmarks for hospital safety, timely and effective care and patient satisfaction can be uniform; rates of central line infections should not be influenced by “patient mix”. However, nine of the 39 maximum-penalty hospitals under the HRRP are located in rural Kentucky; the two non-penalty Kentucky hospitals are in urban settings. One could hypothesize many reasons why rural Kentucky is a hotbed for “excess” readmission, including dispersion of pharmacies, limited home care options, as well as regional affinity for tobacco and bourbon.
The fundamental question raised by our study is whether poor performance on quality measures is a function of underperforming hospitals or a manifestation of underserved communities. Moving forward, we encourage data systems and study designs that focus research on geo-spatial distribution of population health needs within the context of social and behavioral health determinants.33 Small-area studies into the factors that drive health outcomes will inform the rational alignment of health care policies and resources (including penalties and incentives) with the underlying needs of populations.
Strengths and Weaknesses of Study
Each hospital was assigned a county level socioeconomic profile, which may not precisely reflect hospitals patient populations. A more robust methodology would analyze patient level data; impractical given a cohort of 78 hospitals. Additionally, our control cohort was drawn from a sample of over 700 hospitals that received no penalty under the HRRP during FY 2015. It is possible that matching controlled for similar SES factors and skewed our results to the null, especially in terms of race and ethnicity. However, geographic matching adds strength to our assertion that specific SES factors such as income, property values and education drive the distinction between maximum penalty hospitals and non-penalty hospitals. Seven pairs, representing 18% of the cohort, were located in the same county; both maximum penalty and no-penalty hospitals were assigned the same socioeconomic profile, potentially skewing results towards null.
Thirty-nine hospitals received a maximum penalty in the third year of the Hospital Readmission Reduction Program. Relative to geographically matched non-penalty hospitals, maximum penalty hospitals were more likely to be rural and located in counties with low income, low property values, low rates of educational attainment and high rates of poorly controlled chronic disease. In contrast to nationwide studies, a geographically matched analysis plan revealed no difference between the cohorts in terms of race and ethnicity and also provided evidence that maximum penalty hospitals had higher rates of age, sex and race adjusted hospital-wide mortality.
Our results highlight potential consequences of nationally derived benchmarks for phenomena underpinned by social, behavioral and environmental factors and raise the question of whether maximum HRRP penalties are a consequence of underperforming hospitals or a manifestation of underserved communities. We are encouraged by MedPAC’s proposal to stratify HRRP benchmarks by SES, yet also recommend further small-area geographic analyses to better align quality measures, penalties and incentives with the resources and needs of distinct populations.
Supplementary Material
Figure 1.
Map of matched case-control pairs
Figure 1 maps the 39 maximum penalty hospitals (red dots) and 39 no penalty hospitals (blue dots) with dotted lines connecting the matched pairs. Median distance between pairs was 42.5 miles (interquartile range: 25th percentile, 15.4 miles; 75th percentile, 98.4 miles).
Table 4.
Sensitivity analysis excluding seven case-control geographically matched pairs within same county
| No penalty | Max penalty | p value | |
|---|---|---|---|
| Income, education and earnings | |||
| Individuals below poverty line (%) | 15.7 | 20.1 | 0.006 |
| Per capita income ($) | 27,500 | 23,600 | 0.001 |
| High school graduation (%) | 86.3 | 81.1 | 0.005 |
| SNAP (%) | 13.5 | 18.3 | 0.005 |
| Unemployment (%) | 4.7 | 4.7 | 0.9 |
| Labor force participation (%) | 62.6 | 54.8 | <0.0001 |
| Demographics | |||
| White(%) | 75.4 | 79.1 | 0.3 |
| Black (%) | 9.8 | 9.5 | 0.9 |
| Hispanic or Latino (%) | 9.6 | 6.3 | 0.09 |
| Median Age | 37.9 | 40.3 | 0.1 |
| Males per 100 females | 93.2 | 98.1 | 0.1 |
Source: 2015 American Community Survey
Acknowledgments
Funding: This project received no funding.
Footnotes
Original Manuscript: I certify that none of the material in this manuscript has been previously published and that none of this material is currently under consideration for publication elsewhere.
Disclosure: None of the authors on this manuscript have any commercial relationships to disclose in relation to this manuscript. All authors have reviewed and approved this manuscript and have contributed significantly to the design, conduct and/or analysis of the research.
Conflict of Interest:
No authors have any financial interests to disclose. No authors have any potential conflicts of interest to disclose. No authors have financial or personal relationships with any of the subject material presented in the manuscript.
References
- 1.Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. The New England journal of medicine. 2009;360(14):1418–1428. doi: 10.1056/NEJMsa0803563. [DOI] [PubMed] [Google Scholar]
- 2.Axon RN, Williams MV. Hospital readmission as an accountability measure. Jama. 2011;305(5):504–505. doi: 10.1001/jama.2011.72. [DOI] [PubMed] [Google Scholar]
- 3.Amarasingham R, Moore BJ, Tabak YP, et al. An automated model to identify heart failure patients at risk for 30-day readmission or death using electronic medical record data. Medical care. 2010;48(11):981–988. doi: 10.1097/MLR.0b013e3181ef60d9. [DOI] [PubMed] [Google Scholar]
- 4.Calvillo-King L, Arnold D, Eubank KJ, et al. Impact of social factors on risk of readmission or mortality in pneumonia and heart failure: systematic review. Journal of general internal medicine. 2013;28(2):269–282. doi: 10.1007/s11606-012-2235-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Feemster LC, Au DH. Penalizing hospitals for chronic obstructive pulmonary disease readmissions. American journal of respiratory and critical care medicine. 2014;189(6):634–639. doi: 10.1164/rccm.201308-1541PP. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sjoding MW, Cooke CR. Readmission penalties for chronic obstructive pulmonary disease will further stress hospitals caring for vulnerable patient populations. American journal of respiratory and critical care medicine. 2014;190(9):1072–1074. doi: 10.1164/rccm.201407-1345LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hu J, Gonsahn MD, Nerenz DR. Socioeconomic status and readmissions: evidence from an urban teaching hospital. Health affairs (Project Hope) 2014;33(5):778–785. doi: 10.1377/hlthaff.2013.0816. [DOI] [PubMed] [Google Scholar]
- 8.Mather JF, Fortunato GJ, Ash JL, Davis MJ, Kumar A. Prediction of pneumonia 30-day readmissions: a single-center attempt to increase model performance. Respiratory care. 2014;59(2):199–208. doi: 10.4187/respcare.02563. [DOI] [PubMed] [Google Scholar]
- 9.Philbin EF, Dec GW, Jenkins PL, DiSalvo TG. Socioeconomic status as an independent risk factor for hospital readmission for heart failure. The American journal of cardiology. 2001;87(12):1367–1371. doi: 10.1016/s0002-9149(01)01554-5. [DOI] [PubMed] [Google Scholar]
- 10.Bikdeli B, Wayda B, Bao H, et al. Place of residence and outcomes of patients with heart failure: analysis from the telemonitoring to improve heart failure outcomes trial. Circulation Cardiovascular quality and outcomes. 2014;7(5):749–756. doi: 10.1161/CIRCOUTCOMES.113.000911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lindenauer PK, Lagu T, Rothberg MB, et al. Income inequality and 30 day outcomes after acute myocardial infarction, heart failure, and pneumonia: retrospective cohort study. BMJ (Clinical research ed) 2013;346:f521. doi: 10.1136/bmj.f521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. Jama. 2013;309(4):342–343. doi: 10.1001/jama.2012.94856. [DOI] [PubMed] [Google Scholar]
- 13.Medicare Fines 2,610 Hospitals in Third Round of Readmission Penalties. Kaiser Health News. 2014 Oct 2; 2014. [Google Scholar]
- 14.Centers for Medicare and Medicaid Services. Fiscal Year 2015 IPPS Hospital Readmission Reduction Program Supplemental Data File. http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Readmissions-Reduction-Program.html2014.
- 15.Atlas D. Practice TDIfHPaC, editor. “Hospital and Post-Acute Care” and “Selected Hospital and Physician Capacity Measures”. http://www.dartmouthatlas.org/tools/downloads.aspx2010.
- 16.Services CfMaM. Cost Reports by Year: 2014. http://www.cms.gov/Research-Statistics-Data-and-Systems/Downloadable-Public-Use-Files/Cost-Reports/Cost-Reports-by-Fiscal-Year.html2014.
- 17.Walsh SJ, Page PH, Gesler WM. Normative models and healthcare planning: network-based simulations within a geographic information system environment. Health services research. 1997;32(2):243–260. [PMC free article] [PubMed] [Google Scholar]
- 18.Emch M, Ali M, Root ED, Yunus M. Spatial and environmental connectivity analysis in a cholera vaccine trial. Social science & medicine (1982) 2009;68(4):631–637. doi: 10.1016/j.socscimed.2008.11.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Joynt KE, Orav EJ, Jha AK. Thirty-day readmission rates for Medicare beneficiaries by race and site of care. Jama. 2011;305(7):675–681. doi: 10.1001/jama.2011.123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Vivo RP, Krim SR, Liang L, et al. Short- and long-term rehospitalization and mortality for heart failure in 4 racial/ethnic populations. Journal of the American Heart Association. 2014;3(5):e001134. doi: 10.1161/JAHA.114.001134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Detailed comments on the Inpatient Prospective Payment System Proposed Rule for FY 2013 [press release] 2012 http://www.aha.org/advocacy-issues/letter/2012/120619-cl-ipps.pdf., June 19, 2012.
- 22.Dailey EA, Cizik A, Kasten J, Chapman JR, Lee MJ. Risk factors for readmission of orthopaedic surgical patients. The Journal of bone and joint surgery American volume. 2013;95(11):1012–1019. doi: 10.2106/JBJS.K.01569. [DOI] [PubMed] [Google Scholar]
- 23.Tsai TC, Orav EJ, Joynt KE. Disparities in surgical 30-day readmission rates for Medicare beneficiaries by race and site of care. Annals of surgery. 2014;259(6):1086–1090. doi: 10.1097/SLA.0000000000000326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Krumholz HM, Lin Z, Keenan PS, et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. Jama. 2013;309(6):587–593. doi: 10.1001/jama.2013.333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Committee MPA. Chapter 4: Refining the hospital readmissions reduction program. Report to the Congress: Medicare and the Health Care Delivery System. http://www.medpac.gov/documents/reports/jun13_ch04.pdf?sfvrsn=02013.
- 26.Rutstein DD, Berenberg W, Chalmers TC, Child CG, 3rd, Fishman AP, Perrin EB. Measuring the quality of medical care. A clinical method. The New England journal of medicine. 1976;294(11):582–588. doi: 10.1056/NEJM197603112941104. [DOI] [PubMed] [Google Scholar]
- 27.Purdy S, Griffin T, Salisbury C, Sharp D. Ambulatory care sensitive conditions: terminology and disease coding need to be more specific to aid policy makers and clinicians. Public health. 2009;123(2):169–173. doi: 10.1016/j.puhe.2008.11.001. [DOI] [PubMed] [Google Scholar]
- 28.McHugh MD, Berez J, Small DS. Hospitals with higher nurse staffing had lower odds of readmissions penalties than hospitals with lower staffing. Health affairs (Project Hope) 2013;32(10):1740–1747. doi: 10.1377/hlthaff.2013.0613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Joynt KE, Jha AK. Thirty-day readmissions–truth and consequences. The New England journal of medicine. 2012;366(15):1366–1369. doi: 10.1056/NEJMp1201598. [DOI] [PubMed] [Google Scholar]
- 30.Forum NQ. Risk Adjustment for Socioeconomic Status or Other Sociodemographic Factors: a Technical Report. 2014 [Google Scholar]
- 31.Council MPA. The hospital readmission penalty: How well is it working? http://www.medpac.gov/blog/march-2015/march-2015/2015/03/24/the-hospital-readmission-penalty-how-well-is-it-working-2015.
- 32.Services CfMaM. Services USDoHaH, editor. New HHS Data Shows Major Strides Made in Patient Safety, Leading to Improved Care and Savings. http://innovation.cms.gov/Files/reports/patient-safety-results.pdf2014.
- 33.Harrison KM, Dean HD. Use of data systems to address social determinants of health: a need to do more. Public health reports (Washington, DC: 1974) 2011;126(Suppl 3):1–5. doi: 10.1177/00333549111260S301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Press MJ, Scanlon DP, Ryan AM, et al. Limits of readmission rates in measuring hospital quality suggest the need for added metrics. Health affairs (Project Hope) 2013;32(6):1083–1091. doi: 10.1377/hlthaff.2012.0518. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.