Impact of Care Coordination Based on Insurance and Zip Code

ABSTRACT

Objectives: To examine whether a care transitions program, Bridges, differentially reduced rehospitalizations among patients who underwent percutaneous coronary intervention (PCI) based on insurance status and zip code poverty level.

Study Design: Retrospective observational cohort.

Methods: We examined data from a single health system in Delaware, collected as part of a care transitions program for patients who underwent PCI from 2012 to 2015 compared with an unmatched historical control cohort from 2010 to 2011. Socioeconomic status was assessed by insurance status and zip code—level poverty data. Patients were divided into tertiles based on the proportion of their zip code of residence living under 100% of the federal poverty level. Rehospitalization rates were analyzed by negative binomial regression and included interaction terms to examine differential effects of Bridges by insurance and poverty level.

Results: There were 4638 patients representing 5710 hospitalizations: 3212 in the historical control and 2498 in the Bridges cohort. Among patients with Medicaid who received the Bridges intervention, those living in the wealthiest zip codes were 15.5% less likely to be rehospitalized than patients with Medicare and 9.4% less likely than patients with commercial insurance (P = .04). However, patients with Medicaid who lived in the poorest zip codes and those with dual Medicare/Medicaid status had higher rates of rehospitalization post intervention.

Conclusions: The Bridges intervention was associated with improved rehospitalization rates for Medicaid patients compared with those with Medicare or commercial insurance within Delaware’s wealthier communities. Care transitions programs may differentially affect Medicaid patients based on the wealth of the communities in which they reside.

Am J Manag Care. 2019;25(6):e173-e178Takeaway Points

Among patients who underwent percutaneous coronary intervention who received a care transitions intervention, patients with Medicaid who lived in the wealthiest zip codes were 15.5% less likely to be rehospitalized than their peers with Medicare and 9.4% less likely than their peers with commercial insurance (P = .04). Those who lived in the poorest areas had the highest rates of rehospitalization.

• Care transitions interventions may be more effective in reducing rehospitalizations if targeted toward patients with Medicaid who live in areas with greater resources.
• Alternately, rehospitalization rates may increase among patients with Medicaid who live in poorly resourced areas.

Since 2012, CMS has incentivized hospital systems to reduce 30-day rehospitalizations for patients with a history of acute myocardial infarction (AMI).¹ Patients experiencing an AMI are medically complex,² and their high level of medical acuteness is an important driver of high readmission rates. Furthermore, given the additional barriers that patients of low socioeconomic status (SES) face, they are at even greater risk for readmission^3,4 and other adverse outcomes.^5-8 In an effort to reduce unnecessary rehospitalizations, many health systems have implemented programs to help patients overcome barriers to care, including interventions to improve care transitions from hospital to home.^9-12 However, a systematic review of care transitions interventions for patients with AMI found that readmission rates did not differ between those who were enrolled in care transitions programs and those who were not.¹³ Although patients are often selected for care transitions programs based on variables related to high risk of readmission, variation in the effects of these interventions due to these characteristics is often not examined.¹⁴ For example, although patients with a history of AMI are at higher risk of readmission and mortality if they live in high-poverty areas or have Medicaid insurance,^4-7 few studies, if any, have examined whether implementation of a care transitions intervention can attenuate these risks. The objective of this study was to examine whether an intensive care transitions program was associated with a reduction in 30-day rehospitalization rates among patients who received percutaneous coronary intervention (PCI) based on individual and combined indicators of low SES.

METHODS

These data were collected as part of a quality improvement program that implemented an intensive care transitions intervention for patients who received PCI at a single regional health system in Delaware from 2012 to 2015. The intervention was called “Bridging the Divides” (“Bridges”) and was developed with a grant from the Center for Medicare and Medicaid Innovation. Details of the intervention have been described previously in the literature.¹⁵ Briefly, the intervention consisted of medication education for the patient from a clinical pharmacist and bedside nurse prior to discharge, biweekly phone calls during the first 2 weeks post discharge, and regularly scheduled calls from a care management nurse for up to 1 year if needed. As part of the program, called CareLink, the care management team helped to coordinate many aspects of postdischarge care, including making follow-up appointments, ensuring access to medications, and arranging transportation. During the postdischarge phone calls, the CareLink team assessed clinical issues (eg, clinical symptoms, medication adherence) and nonclinical issues (eg, barriers to obtaining follow-up care or medications). CareLink placed heavy emphasis on directing patients to appropriate sites for regular follow-up or urgent care and proactively identified issues that could lead to the need for emergency care.¹⁵ The program and approval to publish results of the study were obtained from the Christiana Care Institutional Review Board.

We collected data for consecutive patients who underwent PCI from 2012 to 2015 who were enrolled in the Bridges program (Bridges cohort); we compared this cohort with historical controls for patients who underwent PCI at the same institution from 2010 to 2011 (control cohort). Low SES was determined by insurance status and zip code—level poverty data obtained from the 2011 to 2015 estimates from the US Census Bureau.^4,8,16 Patients were divided into tertiles based on the proportion of individuals within their zip code of residence who live under 100% of the federal poverty level (FPL). The 3 tertiles included the “wealthy” zip code (defined as those living in zip codes where <9.1% of individuals live under the FPL), the “moderate wealth” zip code (where 9.2%-13.1% of individuals live under the FPL), and the “poor” zip code (where >13.1% of individuals live under the FPL) tertiles.¹⁷ Insurance status was defined as Medicare, Medicaid, dual eligible (Medicare/Medicaid), commercial, or uninsured/other. Thirty-day rehospitalization was defined as a composite of inpatient, observation, and emergency department (ED) visits to the index site of service. We used a combined metric of rehospitalization because the Bridges intervention sought to reduce the need for all types of emergency care.

We then examined the unadjusted association between poverty level, insurance status, and the Bridges intervention with our outcome of 30-day rehospitalization using Poisson regression or negative binomial regression if the distribution of counts was overdispersed. For Poisson distributions, the mean and variance are equal; overdispersion occurs when the variance is greater than the mean. Analogous to an odds ratio in logistic regression, Poisson regression gives an incident rate ratio (IRR)—that is, a ratio of incident rates between categorical variables or increase (or decrease) for units of a continuous variable. Counts were analyzed because patients could have more than 1 rehospitalization.

We constructed a multivariable count model with poverty tertile as the primary independent variable and 30-day rehospitalization as the dependent variable. We controlled for insurance status, Bridges intervention, age, and all variables that were significantly different (P <.05) between the historical control and Bridges cohorts.^3,18,19 To determine if there was a differential improvement in 30-day rehospitalization rates after the Bridges intervention among patients with combined indicators of low SES (Medicaid insurance/poor zip code), we tested the 3-way interaction (control vs Bridges, insurance status, poverty tertile) in the models. We then compared differences in predicted proportions of 30-day rehospitalization rates in the control and Bridges cohorts, stratified by poverty tertile and insurance status. All analyses were conducted in Stata version 14 (StataCorp; College Station, Texas).

RESULTS

Study Population

There were 4638 patients, with 5710 hospitalizations, who had valid zip code—level poverty data and who underwent PCI from 2010 to 2015. Patients in the “poor” zip code tertile were significantly different from those in the “wealthy” zip code tertile in that they were younger (mean age, 63.3 vs 65.2 years, respectively; P <.001); were more likely to be female (594/1593 [37.2%] vs 423/1470 [28.8%]; P <.001), be African American (432/1593 [27.1%] vs 131/1470 [8.9%]; P <.001), and have Medicaid (213/1944 [11.0%] vs 84/1817 [4.6%]; P <.001) or dual-eligible status (149/1944 [7.7%] vs 59/1817 [3.2%]; P <.001); and had a greater number of comorbidities (2.11 vs 1.72; P <.001), higher rates of multiple prior hospitalizations (14.5% vs 11.0%; P <.004), and longer length of stay (3.05 vs 2.64 days; P = .006) (Table 1).

Demographic differences were also examined in the control and Bridges cohorts. Overall, there were no statistically significant differences between the control and Bridges groups, except that in the Bridges cohort, there were more patients with Medicaid as their primary form of insurance compared with the control cohort (235/2498 [9.4%] vs 211/3212 [6.5%]; P = .038; data not shown).

In our unadjusted analysis, we found no significant association between zip code poverty tertile and 30-day rehospitalization rate. Similarly, the Bridges intervention was not associated with 30-day rehospitalization rate. However, insurance status was; patients with dual Medicare/Medicaid eligibility had the highest rates of rehospitalization (35.5%) whereas those with commercial insurance had the lowest rates (13.6%) (Table 2).

Poverty, Insurance, Bridges, and 30-Day Rehospitalization

In our first multivariable model, we found that poverty level was not independently associated with higher rates of rehospitalization after adjustment for insurance status, Bridges intervention, and age, but that insurance status was. Study participants with dual Medicare/Medicaid eligibility had higher rates of rehospitalization (IRR, 1.57; 95% CI, 1.18-2.09) and commercially insured participants had lower rehospitalization rates (IRR, 0.54; 95% CI, 0.42-0.70) compared with other insurance types (eAppendix [available at ajmc.com]). Participation in the Bridges intervention was not associated with lower rates of rehospitalization compared with historical controls.

In our subsequent models (models 2-4), we examined all 2-way interactions between zip code poverty tertile, insurance status, and Bridges intervention and found that only the effect of Medicaid status on rehospitalization significantly varied based on zip code poverty level (model 4; P = .034). Our final model (model 5) included all 2-way and 3-way interaction terms and found that there was a significant 3-way interaction among insurance status, zip code poverty tertile, and Bridges intervention (P = .04) (eAppendix Tables 1 and 2).

As shown in eAppendix Table 2, using predicted proportions of readmissions based on model 5 (all interaction terms), we examined whether, within zip code poverty tertiles, changes in rehospitalization rates were greater among Bridges versus historical control patients based on insurance status. This model represented 3212 hospital revisits in the historical control group, 6.5% (n = 211) of which were from Medicaid patients. Of these, 30.2% (n = 110) of revisits were from patients from poor zip codes, 26.5% (n = 55) were from moderate-wealth zip codes, and 17.6% (n = 46) were from wealthy zip codes. For the Bridges cohort, the model represented 2498 revisits, 9.4% (n = 235) of which were from Medicaid patients. Of these, 45.8% (n = 103) of revisits were from poor zip codes, 16.5% (n = 94) were from moderate-wealth zip codes, and 4.5% (n = 38) were from wealthy zip codes (eAppendix Table 2). Among patients with Medicaid who received the Bridges intervention, those who lived in the wealthiest and moderately wealthy zip codes had lower rehospitalization rates than historical controls (—12.5% and –10.6%, respectively), whereas those who lived in the poorest zip codes had a higher postintervention rehospitalization rate (16%) (Table 3). Patients dually insured with Medicare and Medicaid had higher rehospitalization rates post intervention regardless of zip code poverty level (Table 3).

We then compared adjusted 30-day rehospitalization rates for Medicaid patients with those for patients with Medicare and commercial insurance according to zip code poverty level. Among patients who received the Bridges intervention, patients with Medicaid who lived in the poorest zip codes were 24.5% more likely to be rehospitalized than patients with Medicare (46% vs 21.5%) and 36.5% more likely than those with commercial insurance (46% vs 9.5%) (Table 3). These rates were higher than those of historical controls. However, patients with Medicaid who received the Bridges intervention who lived in moderate wealth zip codes were 7% less likely to be rehospitalized than patients with Medicare (16.5% vs 23.5%) (Figure). Those with Medicaid living in the wealthiest zip codes were 15.5% less likely to be rehospitalized than patients with Medicare (4.7% vs 20.2%) and 9.4% less likely than patients with commercial insurance (4.7% vs 14.1%; P = .04) (Table 3; Figure).

DISCUSSION

In this secondary analysis of data of patients who underwent PCI, we found that after implementation of a care transitions program, 30-day rehospitalization rates decreased among patients with Medicaid who lived in relatively wealthier zip codes compared with other patients.

Similar to findings of prior studies that have examined the impact of care transitions programs on rehospitalization rates,^20-24 the 2017 overall analysis of the Bridges program found that the intervention did not reduce rehospitalization or 30-day ED utilization rates.¹⁵ However, the findings of our current study suggest that there may be merit in examining the impact of such longitudinal care management interventions based on patient- and community-level socioeconomic factors. Prior work has demonstrated that among patients with a history of myocardial ischemia, residence in a high-poverty area^5-7 and low income status²⁵ are associated with worse outcomes, whereas residence in wealthier neighborhood environments has a positive impact on outcomes.²⁴ In the general population, even as hospitalization rates decline, patients with Medicaid are hospitalized at higher rates and those from wealthier communities are hospitalized at lower rates.²⁶ Additionally, prior work has demonstrated that the positive health effects of wealthy communities can modify the negative health effects associated with low-income status.²⁵ Our study builds on this work in that it examines the interaction between individual- and community-level indicators of poverty and suggests that, among low-income patients, the wealth of their area of residence could potentially influence the effectiveness of a care management intervention.

In the general medicine literature, there are several examples of effective care transition programs that were created specifically for patients of low SES. The IMPaCT program, which targets patients with Medicaid or no insurance who live in low-income zip codes, has demonstrated a reduction in the rate of multiple rehospitalizations.¹⁶ The C-TraIn program also targets low-income patients and has demonstrated improvement in the quality of care transitions.¹² However, neither intervention demonstrated a reduction in 30-day rehospitalization rate, and neither stratified their results by neighborhood or zip code characteristics. This may be because most studies of this nature, including ours, are not powered for such subgroup analyses a priori. Nevertheless, it may be important to consider this for future studies.

Many health systems have invested millions of public and private dollars into care transitions programs with variable results.²⁷ To our knowledge, this is the first study that suggests a differential impact of such programs based on the interaction of insurance status and the community in which patients live. Our findings suggest that the effectiveness of such programs may be further accentuated among traditionally underserved groups who live in wealthier, presumably better-resourced environments. A deeper understanding of where these programs may have the greatest impact could allow us to continue to innovate new approaches to care and also could help to direct precious resources to communities and patients who may benefit the most.²⁸ It could also help to demonstrate the long-term value of community investment to policy makers and health systems.

We found that among certain subgroups, such as those with Medicaid who lived in poorer areas and those with dual-eligible status, revisit rates increased with the Bridges intervention. Among the Medicaid patients, this may reflect greater barriers to accessing ambulatory care that exist within poorer communities that could have resulted in higher rates of hospital use.¹¹ The high revisit rates seen among patients with dual Medicare/Medicaid status likely reflect their greater complexity as patients, including higher level of comorbidity and disability.²⁹ However, because a key element of the Bridges intervention was regular assessment for concerning symptoms such as chest pain, these increases in utilization may reflect appropriate, potentially life-saving interventions within populations of high risk and great need. Indeed, efforts to reduce 30-day readmissions among patients with heart failure have been associated with increases in 1-year mortality.³⁰

Limitations

It is important to interpret the findings of this study within the context of its limitations. This was a secondary, post hoc analysis of a single-center study using noncontemporaneous unmatched controls, which was not powered to examine differences in outcomes based on zip code of residence. Therefore, our findings should be interpreted with caution, and causality cannot be inferred from our results. The variables used to examine poverty status were not based on personal income. Rather, we used Medicaid status, which is a common surrogate in the health services literature, and US Census poverty data for zip code of residence, which has been used previously.^4,8 The demographic characteristics of our poor zip code tertile were similar to those in prior studies of patients with a history of myocardial ischemia that had used more granular Census block poverty data,⁶ which adds credibility to our findings. The Bridges cohort had a greater proportion of patients with Medicaid insurance than the historical control cohort, which may reflect changes related to Medicaid expansion under the Affordable Care Act. We could not account for unmeasured confounders, such as self-efficacy and social cohesion, that could potentially differentiate patients with Medicaid who live in areas of relative wealth versus those who do not and could potentially affect health behaviors. We also could not account for community characteristics that may have influenced the effect of the intervention.

CONCLUSIONS

In this large, single-center study of patients who have undergone PCI, a care transitions intervention appeared to reduce 30-day rehospitalization rates more effectively among patients with Medicaid who lived in areas of relative wealth. In contrast, those who lived in areas of relative poverty and those with dual Medicaid/Medicare eligibility status experienced a higher rate of rehospitalization. Transitions of care interventions may be more effective if targeted toward patients of high need who live in areas with greater resources. Alternately, for patients of lower SES living in poorer areas, alternate or more intensive interventions may be needed. As our findings were post hoc in nature with a relatively small sample size, we caution that further research is needed. Nonetheless, SES should be considered when conducting studies of transition of care programs.Author Affiliations: Department of Medicine, Christiana Care Health System (JNG, DJE, LSH), Newark, DE; Department of Pharmacy, University of Delaware (MS), Newark, DE; Heart & Vascular Institute, Georgetown University (PK, WSW), Washington, DC.

Source of Funding: Supported by grant number 1C1CMS331027 from HHS, CMS, and the Delaware INBRE program with a grant from the National Institute of General Medical Sciences (P20 GM103446) from the National Institutes of Health and the state of Delaware. Drs Weintraub and Kolm are supported by an Institutional Development Award from the National Institute of General Medical Sciences of the National Institutes of Health under grant number U54-GM104941 (PI: Binder-Macleod).

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (JNG, DJE, WSW, LSH); acquisition of data (JNG, MS, DJE, WSW); analysis and interpretation of data (JNG, MS, PK, DJE, WSW); drafting of the manuscript (JNG, MS, PK, LSH); critical revision of the manuscript for important intellectual content (JNG, PK, DJE, WSW, LSH); statistical analysis (JNG, MS, PK, LSH); provision of patients or study materials (DJE); obtaining funding (JNG, DJE, WSW); administrative, technical, or logistic support (JNG); and supervision (JNG).

Address Correspondence to: Jennifer N. Goldstein, MD, MSc, Christiana Hospital, 4755 Ogletown Stanton Rd, Ammon Education Bldg, Ste 2E70, Newark, DE 19713. Email: jgoldstein@christianacare.org.REFERENCES

1. Acute inpatient PPS. CMS website. cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps. Updated March 8, 2019. Accessed May 7, 2019.

2. Bernheim SM, Spertus JA, Reid KJ, et al. Socioeconomic disparities in outcomes after acute myocardial infarction. Am Heart J. 2007;153(2):313-319. doi: 10.1016/j.ahj.2006.10.037.

3. Khera R, Jain S, Pandey A, et al. Comparison of readmission rates after acute myocardial infarction in 3 patient age groups (18 to 44, 45 to 64, and ≥65 years) in the United States. Am J Cardiol. 2017;120(10):1761-1767. doi: 10.1016/J.AMJCARD.2017.07.081.

4. Kind AJ, Jencks S, Brock J, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Ann Intern Med. 2014;141(11):765-774. doi: 10.7326/M13-2946.

5. Foraker RE, Patel MD, Whitsel EA, Suchindran CM, Heiss G, Rose KM. Neighborhood socioeconomic disparities and 1-year case fatality after incident myocardial infarction: the Atherosclerosis Risk in Communities (ARIC) Community Surveillance (1992-2002). Am Heart J. 2013;165(1):102-107. doi: 10.1016/J.AHJ.2012.10.022.

6. 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/s11606-012-2147-9.

7. Gerber Y, Benyamini Y, Goldbourt U, Drory Y; Israel Study Group on First Acute Myocardial Infarction. Neighborhood socioeconomic context and long-term survival after myocardial infarction. Circulation. 2010;121(3):375-383. doi: 10.1161/CIRCULATIONAHA.109.882555.

8. Bucholz EM, Ma S, Normand SL, Krumholz HM. Race, socioeconomic status, and life expectancy after acute myocardial infarction. Circulation. 2015;132(14):1338-1346. doi: 10.1161/CIRCULATIONAHA.115.017009.

9. Verhaegh KJ, MacNeil-Vroomen JL, Eslami S, Geerlings SE, de Rooij SE, Buurman BM. Transitional care interventions prevent hospital readmissions for adults with chronic illnesses. Health Aff (Millwood). 2014;33(9):1531-1539. doi: 10.1377/hlthaff.2014.0160.

10. Burke RE, Guo R, Prochazka AV, Misky GJ. Identifying keys to success in reducing readmissions using the ideal transitions in care framework. BMC Health Serv Res. 2014;14:423. doi: 10.1186/1472-6963-14-423.

11. Kangovi S, Barg FK, Carter T, Long JA, Shannon R, Grande D. Understanding why patients of low socioeconomic status prefer hospitals over ambulatory care. Health Aff (Millwood). 2013;32(7):1196-1203. doi: 10.1377/hlthaff.2012.0825.

12. Englander H, Michaels L, Chan B, Kansagara D. The Care Transitions Innovation (C-TraIn) for socioeconomically disadvantaged adults: results of a cluster randomized controlled trial. J Gen Intern Med. 2014;29(11):1460-1467. doi: 10.1007/s11606-014-2903-0.

13. Prvu Bettger J, Alexander KP, Dolor RJ, et al. Transitional care after hospitalization for acute stroke or myocardial infarction: a systematic review. Ann Intern Med. 2012;157(6):407-416. doi: 10.7326/0003-4819-157-6-201209180-00004.

14. Kansagara D, Chiovaro JC, Kagen D, et al. So many options, where do we start? an overview of the care transitions literature. J Hosp Med. 2016;11(3):221-230. doi: 10.1002/jhm.2502.

15. Weintraub WS, Fanari Z, Elliott D, et al. The impact of care management information technology model on quality of care after percutaneous coronary intervention: “Bridging the Divides” [published online July 3, 2017]. Cardiovasc Revasc Med. doi: 10.1016/j.carrev.2017.06.014.

16. Kangovi S, Mitra N, Grande D, et al. Patient-centered community health worker intervention to improve posthospital outcomes: a randomized clinical trial. JAMA Intern Med. 2014;174(4):535-543. doi: 10.1001/jamainternmed.2013.14327.

17. Goldstein JN, Zhang Z, Schwartz JS, Hicks LS. Observation status, poverty, and high financial liability among Medicare beneficiaries. Am J Med. 2018;131(1):101.e9-101.e15. doi: 10.1016/j.amjmed.2017.07.013.

18. Kwok CS, Wong CW, Shufflebotham H, et al. Early readmissions after acute myocardial infarction. Am J Cardiol. 2017;120(5):723-728. doi: 10.1016/J.AMJCARD.2017.05.049.

19. Cafagna G, Seghieri C. Educational level and 30-day outcomes after hospitalization for acute myocardial infarction in Italy. BMC Health Serv Res. 2017;17(1):18. doi: 10.1186/s12913-016-1966-5.

20. Kotowycz MA, Cosman TL, Tartaglia C, Afzal R, Syal RP, Natarajan MK. Safety and feasibility of early hospital discharge in ST-segment elevation myocardial infarction—a prospective and randomized trial in low-risk primary percutaneous coronary intervention patients (the Safe-Depart Trial). Am Heart J. 2010;159(1):117.e1-e6. doi: 10.1016/j.ahj.2009.10.024.

21. Costa e Silva R, Pellanda L, Portal V, Maciel P, Furquim A, Schaan B. Transdisciplinary approach to the follow-up of patients after myocardial infarction. Clinics (Sao Paulo). 2008;63(4):489-496. doi: 10.1590/S1807-59322008000400013.

22. Sinclair AJ, Conroy SP, Davies M, Bayer AJ. Post-discharge home-based support for older cardiac patients: a randomised controlled trial. Age Ageing. 2005;34(4):338-343. doi: 10.1093/ageing/afi116.

23. Hanssen TA, Nordrehaug JE, Eide GE, Hanestad BR. Does a telephone follow-up intervention for patients discharged with acute myocardial infarction have long-term effects on health-related quality of life? a randomised controlled trial. J Clin Nurs. 2009;18(9):1334-1345. doi: 10.1111/j.1365-2702.2008.02654.x.

24. Unger E, Diez-Roux AV, Lloyd-Jones DM, et al. Association of neighborhood characteristics with cardiovascular health in the multi-ethnic study of atherosclerosis. Circ Cardiovasc Qual Outcomes. 2014;7(4):524-531. doi: 10.1161/CIRCOUTCOMES.113.000698.

25. Southern DA, McLaren L, Hawe P, Knudtson ML, Ghali WA; APPROACH Investigators. Individual-level and neighborhood-level income measures: agreement and association with outcomes in a cardiac disease cohort. Med Care. 2005;43(11):1116-1122. doi: 10.1097/01.mlr.0000182517.57235.6d.

26. McDermott KW, Elixhauser A, Sun R. Trends in hospital inpatient stays in the United States, 2005-2014. Agency for Healthcare Research and Quality website. hcup-us.ahrq.gov/reports/statbriefs/sb225-Inpatient-US-Stays-Trends.jsp. Updated June 23, 2017. Accessed May 7, 2019.

27. Community-based Care Transitions Program. CMS website. innovation.cms.gov/initiatives/CCTP/?itemid. Updated February 12, 2019. Accessed May 7, 2019.

28. Peikes D, Chen A, Schore J, Brown R. Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials. JAMA. 2009;301(6):603-618. doi: 10.1001/jama.2009.126.

29. Section 4: dual-eligible beneficiaries. Medicare Payment Advisory Commission website. medpac.gov/docs/default-source/data-book/jun17_databooksec4_sec.pdf. Published June 2017. Accessed May 7, 2019.

30. Gupta A, Allen LA, Bhatt DL, et al. Association of the Hospital Readmissions Reduction Program implementation with readmission and mortality outcomes in heart failure. JAMA Cardiol. 2018;3(1):44-53. doi: 10.1001/jamacardio.2017.4265.

Download PDF: Impact of Care Coordination Based on Insurance and Zip Code