Heart failure is a chronic condition characterized by inability of the heart to maintain adequate blood flow to body tissues.1 In heart failure, the heart muscle weakens slowly over many years—a pattern of disease progression that contributes to high long-term healthcare resource utilization for patients with this condition. Although the medical and economic burden of heart failure in the United States is already substantial, it will likely grow as the population ages and life expectancy increases. Not surprisingly, most of the heart failure burden is borne by individuals aged ≥65 years, many of whom are Medicare beneficiaries.
At present, approximately 870,000 new cases of heart failure are diagnosed each year in the United States, and the incidence of heart failure is approaching 10 per 1000 individuals aged >65 years.2 Approximately 5.7 million individuals in the United States are currently living with heart failure,1,2 and estimates suggest that the prevalence will increase by approximately 46% in the next 15 years.3 If true, more than 8 million Americans will be living with heart failure in 2030.3
Patients with heart failure have a high rate of hospital admission and readmission and generate substantial treatment-related costs. For example, a 2011 study showed that approximately 80% of Medicare beneficiaries with heart failure were hospitalized in their last 6 months of life.4 Heart failure is also the most common hospital discharge diagnosis for patients aged >65 years, and the number of patients discharged with this diagnosis remained nearly unchanged from 2000 to 2010 (1,008,000 vs 1,023,000, respectively).2 Among Medicare beneficiaries, admissions for heart failure have the highest rate of hospital readmission within 30 days of discharge.5 In 2012, the total direct medical cost for heart failure in the United States was $20.9 billion, a figure that is expected to increase to $53.1 billion by 2030.3 Overall, 80% of the costs attributed to heart failure are associated with hospitalization.3
Although the aggregate current and projected costs of heart failure have been reported, there is a lack of claims-based studies quantifying the cost of heart failure in the Medicare population. For payers and providers managing healthcare costs for the Medicare population, the cost contributed by patients with heart failure, and the resource utilization of inpatient admissions and admissions to skilled nursing facilities are not well understood.
The purpose of the current study was to analyze the utilization and costs associated with inpatient admissions, readmissions, and admissions to skilled nursing facilities among Medicare fee-for-service (FFS) beneficiaries with heart failure.
Study Design and Patient Population
We used the 2011-2012 Medicare 5% sample (limited data set) to perform a retrospective claims data analysis of the utilization and cost of inpatient admissions and readmissions for Medicare FFS beneficiaries with heart failure. The year 2011 was used as a look-back year, and 2012 was the analysis year for this study. This data set is made available to the public for research purposes by the Centers for Medicare & Medicaid Services (CMS), and it contains all Medicare FFS Part A and Part B paid claims generated by a statistically balanced sample of Medicare FFS beneficiaries. Prescription drug data were not included in this analysis. The claims include various codes (diagnosis, procedure, revenue, and diagnosis-related group), as well as information on the site of service, beneficiary age, monthly eligibility status, and monthly status of HMO or Medicare Advantage enrollment.
The study population was selected to have Medicare eligibility in all months of 2011 (to perform a look-back analysis), eligibility for ≥1 months in 2012, Part A and Part B eligibility for all months of eligibility, and no HMO or Medicare Advantage enrollment at any time during the study period. Among all Medicare FFS beneficiaries meeting these criteria, heart failure beneficiaries were identified as beneficiaries with ≥1 claims of the Health Effectiveness Data and Information Set (HEDIS)specified types containing an International Classification of Diseases, Ninth Revision (ICD-9) code for heart failure (ie, 398.91, 402.01, 402.11, 402.91, 404.01, 404.03, 404.11, 404.13, 404.91, 404.93, 428.xx) in any position of the claim during 2012. Qualifying claim types included acute inpatient, emergency department, nonacute inpatient, and outpatient care claims, as identified by HEDIS-specified Current Procedural Terminology and revenue codes.
These Medicare FFS beneficiaries with heart failure were stratified into 5 status categories—aged/dual eligible, aged/not dual eligible, disabled, end-stage renal disease (ESRD), and institutionalized (non-ESRD). Death dates were documented in the eligibility file for beneficiaries who died in 2012, and mortality rates were calculated from this information.
We report utilization of inpatient admissions and admissions to skilled nursing facilities per 1000 beneficiaries for the Medicare FFS heart failure population and the total Medicare FFS population. Inpatient admissions included medical, surgical, psychiatric, substance abuse, acute rehabilitation, and maternity admissions, unless otherwise specified.
Inpatient and skilled nursing facility admission claims were identified from the Medicare files. Professional and other claims incurred and paid during an inpatient or a skilled nursing facility stay were identified as claims with dates of service spanning the admission through discharge dates. For claims with service dates corresponding to the actual date of admission or discharge, the claim was required to have a place-of-service code 21 or code 61 (ie, acute rehabilitation) for inpatient admission and code 31 for a skilled nursing facility.
All-cause and heart failure–related 30-day readmission rates were calculated using a modified version of the methodology developed by the Agency for Healthcare Research and Quality.6 We included all inpatient discharges in 2012 and followed them for 30 days after discharge to determine whether readmission had occurred (December discharges were followed only through December 31).
Inpatient rehabilitation and transfers to other acute-care facilities did not qualify as readmissions. In the readmission analysis, we also included all patients discharged in December 2011 and followed them for 30 days after discharge (into January 2012) to determine which January 2012 cases should be considered readmissions. We identified admissions related to heart failure as those assigned to Medicare severity diagnosis-related groups (MS-DRGs) 291, 292, or 293.
The costs for key metrics included all allowed payments (Medicare payments plus beneficiary cost-sharing) in 2012 US dollars. Per-member per-month (PMPM) costs were calculated using the total number of eligible member months for the indicated population.
We report findings for key cost and utilization metrics for the subset of the Medicare FFS population with heart failure and the total Medicare FFS population. We did not attempt to normalize the heart failure population to the total population through demographic or risk adjustment.
The total Medicare FFS population consisted of 1,461,935 patients (1,301,545 without heart failure; 160,390 with heart failure). The prevalence of heart failure among the total Medicare FFS population was 11%, and varied by Medicare status, ranging from 5.7% for the disabled category to 30.1% for the institutionalized (non-ESRD) category. The average age of the heart failure subset population was 77.5 years, which is substantially higher than that of the total Medicare FFS population (71.7 years). The mortality rate for the Medicare FFS population with heart failure (21.5%) was 3.5 times higher than that of the total Medicare FFS population (6.0%).
PMPM-allowed costs in the heart failure population subset and in the total Medicare FFS population were $3395 and $1045, respectively. The costs incurred by the heart failure population subset accounted for 34% of the total annual Medicare FFS population–allowed costs (Table 1 and Table 2).
The annual number of inpatient admissions per 1000 beneficiaries was 1307 in the heart failure population compared with 345 in the total Medicare FFS population. Inpatient admissions for the heart failure population among Medicare FFS beneficiaries represented 41.5% of all inpatient admissions for the total Medicare FFS population. The costs of inpatient admissions (including facility payments and other Medicare FFS expenditures during those admissions) contributed a larger portion to the total allowed costs of the population with heart failure (51.6% of allowed cost) than that contributed by inpatient admissions in the total Medicare FFS population (39.5% of allowed cost). Inpatient costs for the heart failure population contributed $182 PMPM (or 17.5%) to the total Medicare FFS population–allowed cost of $1045 PMPM (Table 3).
The annual all-cause 30-day inpatient readmission rate was 28% for the Medicare FFS population with heart failure and 21% for the total Medicare FFS population. Inpatient readmissions for the heart failure subset represented 55.3% of all inpatient readmissions in the total Medicare FFS population. Readmission costs (including facility payments and other Medicare FFS expenditures during readmission stays) contributed a larger portion to the total allowed costs in the heart failure population (16.4%) than readmission costs in the total Medicare FFS population (9.5%). Readmission costs for the heart failure population contributed $58 PMPM (or 5.6%) to the total Medicare FFS population–allowed cost of $1045 PMPM. The all-cause 30-day readmission rate after heart failure admission (MS-DRG codes 291, 292, and 293) was 31% (Table 4).
The annual number of admissions to skilled nursing facilities per 1000 beneficiaries was 356 in the heart failure population and 79 in the total Medicare FFS population. Admissions to skilled nursing facilities among Medicare FFS beneficiaries with heart failure represented 49.5% of skilled nursing facility admissions in the total Medicare FFS population. The costs associated with stays in skilled nursing facilities (including facility payments and other Medicare FFS expenditures during those skilled nursing facility stays) contributed a larger portion to the total allowed costs in the heart failure population (13.1%) than skilled nursing facility costs in the total Medicare FFS population (9.2%). Skilled nursing facility costs for the heart failure population contributed $46 PMPM (or 4.4%) to the total Medicare FFS population allowed cost of $1045 PMPM (Table 5).
Compared with the total Medicare FFS population, substantially more beneficiaries with heart failure had ≥1 inpatient admissions (19.8% vs 60.2%, respectively) annually. The majority of the heart failure population with ≥1 inpatient admissions had ≥2 admissions annually (52.4%) compared with 36.2% in the total Medicare FFS population (Table 6).
Nearly half (49.2%) of the heart failure admissions were in 2 major diagnostic categories involving (1) diseases and disorders of the circulatory system, and (2) diseases and disorders of the respiratory system (Table 7). By contrast, only 35.1% of admissions for the total Medicare FFS population were in these 2 categories. MS-DRG codes 291, 292, and 293, which indicate admissions specific to heart failure and are within the “diseases and disorders of the circulatory system” category, accounted for 11% of inpatient admissions among the heart failure population (Table 7).
Our findings support earlier studies identifying the high utilization and cost of inpatient admissions and readmissions associated with the Medicare heart failure population.3,7 Specifically, we found that although the prevalence of heart failure was only 11% in the total Medicare FFS population, the heart failure subset population contributed 34% of the total allowed costs, 41.5% of total inpatient admissions, 55.3% of total readmissions, and 49.5% of the total admissions to skilled nursing facilities. We found that 51.6% of spending on the heart failure population was for inpatient admissions and another 13.1% was for skilled nursing facility admissions, a finding supported by a previous study, which estimated that 80% of the costs attributable to heart failure were associated with inpatient admissions (ie, admissions, readmissions, and skilled nursing facilities stays).3
We did not analyze observation stays for heart failure, and we recognize that, with the increased focus on observation stays as a cost-mitigation measure for Medicare,8,9 it would be important to analyze potential changes in the distribution of inpatient versus observation services for the heart failure population in a subsequent study.
Given the substantial disease burden of heart failure, CMS recently developed relevant performance measures pertaining to the treatment of heart failure. For example, the consumer-oriented website Hospital Compare (created by CMS) and the Hospital Quality Alliance offer public access to information about the quality of care in more than 4000 Medicare-certified US hospitals.10 In 2005, CMS published its first set of 10 core process measures, including those related to heart failure, on the Hospital Compare website.10 These measures were expanded in 2008 to include hospital 30-day mortality rates for heart failure and were further expanded in 2010 to include 30-day all-cause readmission rates after admission for heart failure.10
Other CMS efforts represent a substantial investment in improving the management of patients with heart failure during admission and readmission. In 2012, CMS implemented the Hospital Readmissions Reduction Program (HRRP), an initiative under which hospitals are penalized financially for excessive readmissions; heart failure is 1 of the 5 applicable conditions included since inception of the program.11
In the first year of HRRP, the aggregate penalty was approximately $280 million in Medicare payments distributed across 2200 hospitals, and approximately 30% of hospitals received no penalty.12 For the second year of HRRP, CMS estimates that penalties will amount to approximately $227 million distributed across 2225 hospitals and that, again, approximately 30% of hospitals will receive no penalty.11 In year 3 of HRRP, more hospitals will be penalized (78% of all hospitals nationally), and the total penalties will amount to approximately $428 million.13
Since the program was introduced, the risk-standardized median unplanned readmission rates reported by CMS showed a decline between 2009 and 2012 (ie, decreasing from 23.3% in 2009-2010 to 22.5% in 2011-2012).14 Nonetheless, there was wide variation in hospital performance throughout the United States, with outcome rates for patients with heart failure (such as risk-standardized mortality following hospitalization for heart failure) being at least 4 absolute percentage points lower in the best-performing (10th percentile) hospitals than in the worst-performing (90th percentile) hospitals.14
Another CMS program, the Medicare Spending Per Beneficiary (MSPB) initiative, also has implications for heart failure readmissions. First implemented in fiscal year 2015, MSPB penalizes hospitals where the costs for discharged patients with Medicare incurred in the period from 3 days before admission through 30 days after discharge exceed an aggregate threshold target cost, which is assessed annually for all admissions.15 Because a substantial portion of Medicare admissions are for beneficiaries with heart failure, hospitals have an incentive to more aggressively manage the post–acute care of these beneficiaries, with a focus on reducing readmissions. Some hospitals are attempting to manage this post–acute care independently, whereas others are relying on outside care management firms.
Two additional CMS programs, the Bundled Payments for Care Improvement (BPCI) program and the Medicare Shared Savings Program (MSSP), have the potential to focus efforts on reducing admissions and readmissions for beneficiaries with heart failure. BPCI, which was launched in early 2013, provides a bundled payment for the care of patients with heart failure. The episode-of-care payment strategy under BPCI includes admission to a hospital for a heart failure MS-DRG, as well as services rendered in the period after discharge, extending for up to 90 days. Based on our unpublished analysis of Medicare BPCI claims data, we believe the opportunity to reduce Medicare expenditures in that period is largely derived from reducing hospital readmissions and admissions to skilled nursing facilities in the 30 days after discharge.
The voluntary MSSP initiative offers physician and hospital groups an opportunity to form accountable care organizations (ACOs) that can share in savings from managing their attributed Medicare population’s total cost of care. If an ACO is successful in delivering high-quality care and spending healthcare dollars more efficiently, it is eligible to receive some of the cost-savings. As of January 1, 2015, enrollment in the MSSP included 405 unique organizations, with approximately 15% of the total Medicare population attributed to an MSSP ACO.16
Three of the ACO quality metrics that determine whether savings will be shared with an organization are related to heart failure—ambulatory sensitive admissions for heart failure (ACO-10), beta-blocker therapy for left-ventricular dysfunction (ACO-31), and all-cause unplanned admissions for patients with heart failure (ACO-37).16
We acknowledge several study limitations. First, the identification of heart failure in claims data was based on the accuracy of claims coding. Because of this, our study did not capture patients with heart failure who have not had claims coded with a heart failure ICD-9 diagnosis code.
Second, it is possible that patients with low-severity heart failure did not have any claims with a heart failure ICD-9 code in the analysis year 2012. Therefore, it is possible that our definition of the heart failure population caused us to systematically miss individuals who had less severe disease or were subject to less medical intervention (and thus did not have the diagnosis coding required for study inclusion). We also did not include pharmacy claims data, and, as such, we may have missed patients who were effectively managed on heart failure medication. This could mean that our results are skewed toward the higher-cost patients with heart failure.
Third, our findings reflect 2012 experience only; studies of more recent periods may produce different results.
Fourth, our utilization and cost data are based on the Medicare FFS population and may not reflect the experience of Medicare Advantage populations. These results do not apply to commercial or Medicaid populations. However, our goal was to analyze detailed claims information for the Medicare FFS population and, as such, we believe that the results are useful, regardless of this limitation.
Fifth, we required all Medicare members in our analysis to have coverage eligibility in all months of 2011, which eliminated the Medicare members who enrolled in 2012 analysis year, thereby resulting in a slightly older population than the general Medicare population. Because Medicare eligibility begins at age 65 years for most individuals, the generalizability of our findings may be reduced for individuals aged <66 years.
Finally, the outcomes that are reported are national averages; however, the prevalence of heart failure varies regionally. Therefore, region-specific studies would yield different cost contributions to total Medicare population spending. Practice patterns and community infrastructure also vary by region and would be associated with different rates of inpatient admission, readmission, and admission to skilled nursing facilities. Therefore, our results should be interpreted with caution if applied to a particular region of the country.
The findings from our claims-based analysis support those of earlier studies showing that heart failure is associated with high inpatient utilization and costs. The medical and financial burden of Medicare beneficiaries with heart failure is far greater than may be expected from the 11% prevalence of heart failure; based on our study, patients with heart failure account for 41.5% of all inpatient admissions, 55.3% of all readmissions, and 49.5% of admissions to skilled nursing facilities of the total Medicare FFS population. In recent years, CMS implemented programs to improve the management of heart failure to reduce inpatient utilization and cost. Collectively, CMS efforts represent an important investment toward the delivery of high-quality care for patients with heart failure. Although the full value of these newly implemented programs is not yet known, it is clear that the management of Medicare beneficiaries with heart failure needs to be a continuing priority given the increase in prevalence, the minimal improvement in survival rates to date, and the current payment reform climate.
Much can be done in the short-term to further improve the management of heart failure, such as addressing the geographic variation in the utilization of medical services among Medicare beneficiaries with heart failure.17 Because a substantial portion of Medicare admissions are for beneficiaries with heart failure, it would be reasonable for hospitals to more aggressively manage the post–acute care of these beneficiaries, with a focus on reducing readmissions. The current findings highlight the high service utilization and cost among the heart failure population that point to potentially high-value targets for both payers and providers.
Funding for this study was provided by Novartis Pharmaceuticals.
Author Disclosure Statement
Ms Fitch, Ms Pelizzari, and Mr Pyenson received funding from Novartis for this study.
1. American Heart Association. What is heart failure? 2015. www.heart.org/idc/groups/heart-public/@wcm/@hcm/documents/downloadable/ucm_300315.pdf. Accessed March 6, 2015.
2. Mozaffarian D, Benjamin EJ, Go AS, et al; for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics—2015 update: a report from the American Heart Association. Circulation. 2015;131:e29-e322. Errata in: Circulation. 2015;131:e535; Circulation. 2016;133:e417.
3. Heidenreich PA, Albert NM, Allen LA, et al; for the American Heart Association Advocacy Coordinating Committee; Council on Arteriosclerosis, Thrombosis and Vascular Biology; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; Council on Epidemiology and Prevention; Stroke Council. Forecasting the impact of heart failure in the United States: a policy statement from the American Heart Association. Circ Heart Fail. 2013;6:606-619.
4. Unroe KT, Greiner MA, Hernandez AF, et al. Resource use in the last 6 months of life among Medicare beneficiaries with heart failure, 2000-2007. Arch Intern Med. 2011;171:196-203.
5. Hines AL, Barrett ML, Jiang HJ, Steiner CA. Conditions with the largest number of adult hospital readmissions by payer, 2011. Statistical brief #172. April 2014. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/reports/statbriefs/sb172-Conditions-Readmissions-Payer.pdf. Accessed March 22, 2016.
6. Barrett M, Raetzman S, Andrews R; for the US Agency for Healthcare Research and Quality. HCUP Methods Series Report: Overview of Key Readmission Measures and Methods. Report #2012-04; December 20, 2012. www.hcup-us.ahrq.gov/reports/methods/2012_04.pdf. Accessed March 22, 2016.
7. Wang G, Zhang Z, Ayala C, et al. Costs of heart failure–related hospitalizations in patients aged 18 to 64 years. Am J Manag Care. 2010;16:769-776.
8. Baugh CW, Venkatesh AK, Hilton JA, et al. Making greater use of dedicated hospital observation units for many short-stay patients could save $3.1 billion a year. Health Aff (Millwood). 2012;31:2314-2323.
9. Baugh CW, Schuur JD. Observation care—high-value care or a cost-shifting loophole? N Engl J Med. 2013;369:302-305. Erratum in: N Engl J Med. 2013;369:1476.
10. Centers for Medicare & Medicaid Services. Hospital Compare. www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/HospitalCompare.html. Accessed March 29, 2015.
11. Centers for Medicare & Medicaid Services. Readmissions Reduction Program (HRRP). www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Readmissions-Reduction-Program.html. Accessed March 28, 2015.
12. James J. Health policy brief: Medicare Hospital Readmissions Reduction Program. Health Affairs. November 12, 2013. http://healthaffairs.org/healthpolicybriefs/brief_pdfs/healthpolicybrief_102.pdf. Accessed March 29, 2015.
13. Boccuti C, Casillas G. Aiming for fewer hospital U-turns: The Medicare Hospital Readmission Reduction Program. http://kff.org/medicare/issue-brief/aiming-for-fewer-hospital-u-turns-the-medicare-hospital-readmission-reduction-program/. Kaiser Family Foundation. January 29, 2015. Accessed March 22, 2016.
14. Suter LG, Li S-X, Grady JN, et al. National patterns of risk-standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release. J Gen Intern Med. 2014;29:1333-1340.
15. Centers for Medicare & Medicaid Services. Hospital Value-Based Purchasing. www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/hospital-value-based-purchasing/index.html?redirect=/hospital-value-based-purchasing/. Accessed March 29, 2015.
16. Centers for Medicare & Medicaid Services. Shared Savings Program: news and updates. www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/News.html. Accessed March 29, 2015.
17. Sacks CA, Jarcho JA, Curfman GD. Paradigm shifts in heart-failure therapy—a timeline. N Engl J Med. 2014;371:989-991.