Rethinking the Hospital Readmissions Reduction Program
Reducing readmissions is an important way to improve quality and lower health care spending, and hospitals are making significant progress. The Centers for Medicare & Medicaid Services reported that the national readmission rate (i.e., instances when patients return to the same or different hospital within 30 days of discharge) fell to 17.5 percent in 2013, after holding steady at 19 to 19.5 percent for many years.
However, reducing readmissions is a complex undertaking because not all readmissions can or should be prevented; indeed, some are planned as part of sound clinical care. Furthermore, while hospitals are working to reduce readmissions caused by clinical care practices, there are many other factors beyond their control — sociodemographic factors like poverty and lack of access to supportive services in the community that aid post-hospitalization recovery — that increase the risk of readmission. In measuring hospital performance, policies must account for many factors beyond hospitals’ control to achieve accurate comparisons of performance. Public policy efforts intended to reduce hospital readmissions should target the reduction of only avoidable readmissions.
To encourage efforts to reduce readmissions, Congress included in the Affordable Care Act the Hospital Readmissions Reduction Program, or HRRP, which instructs CMS to penalize hospitals that have higher-than-expected readmissions for specific clinical conditions like heart attack, pneumonia and heart failure. The HRRP payment penalties, which took effect in fiscal 2013, can amount to as much as 3 percent of a hospital’s Medicare payments. While hospital readmissions are declining, there are serious questions about how the HRRP assesses penalties that affect the fairness and long-term sustainability of the program. Specifically, hospitals and other stakeholders have raised concerns about:
1. The lack of risk adjustment for key sociodemographic factors, usually outside of a hospital’s control, that influence the likelihood of readmission;
2. The inclusion of readmissions unrelated to the initial admission in determining HRRP penalties.
As CMS increases both the reimbursement at risk and adds conditions to the program, more hospitals will face penalties, further highlighting the urgency of addressing the program’s shortcomings. The HRRP’s approach to calculating hospital penalties needs refinement to reduce readmissions without unfairly penalizing hospitals.
Providers may be able to prevent certain readmissions if they ensure that their patients receive the right care at the right time, both in the hospital and in subsequent care settings. However, many readmissions may be unavoidable due to the natural progression of disease, accepted treatment protocol or a patient’s preferences. Recognizing the potential for confusion when evaluating readmissions, the American Hospital Association consulted with clinicians to create the following framework for types of readmissions:
1. A planned readmission related to the initial admission, such as placement of a ventricular assist device following a heart attack.
2. A planned readmission unrelated to the initial admission, such as removal of a lung tumor discovered during an admission for a heart attack.
3. An unplanned readmission unrelated to the initial admission, such as a fracture from a car accident following an initial stay for pneumonia.
4. An unplanned readmission related to the initial admission, such as a surgical-site infection.
Planned readmissions are typically part of clinically appropriate care. For example, during an acute care admission, clinicians may identify the need for a hysterectomy or hernia repair and plan these procedures within 30 days of the original hospital admission. CMS developed an algorithm to omit planned readmissions from the HRRP penalty calculation. Accordingly, many planned readmissions within 30 days of discharge no longer count as a readmission for the HRRP.
Similarly, CMS should not hold hospitals accountable for unplanned, unrelated admissions because they are unpredictable and not typically preventable. However, these readmissions are currently included in the HRRP penalty calculation, even though they are not associated with care delivered by the hospital.
Factors Outside of a Hospital’s Control
Hospitals are focused intently on reducing avoidable readmissions using a number of strategies. Nevertheless, the likelihood of patients being readmitted to the hospital is affected not only by the steps hospitals take to improve care, but also by a variety of clinical and nonclinical factors beyond providers’ control. For example, Medicare beneficiaries with six or more chronic conditions have a readmission rate of 25 percent, compared with 9 percent for those having only one or no chronic condition. Similarly, patients whose illnesses are more severe, or who have other comorbid conditions face greater challenges in recovering from illness and are more likely to be readmitted. While not all hospitals treat the same proportion of these types of patients, the HRRP requires a comparison of the performance of all hospitals. Therefore, it is important to ensure that hospitals do not incur greater penalties simply because they treat patients who have more complex conditions.
Risk adjustment is a widely accepted statistical technique that accounts for some of the factors outside the control of providers when one is seeking to isolate and compare the quality of care. It is intended to create a level playing field that allows fairer comparisons of whether providers are doing all they can to ensure the quality of care. The readmission measures used in the HRRP adjust risk for several clinical factors, including age, gender, comorbidities and patient frailty. CMS recognizes that comorbidities and frailty contribute to the cause and outcome of the admission, and acknowledges that hospitals have limited tools to cure or manage them during a single inpatient stay.
However, CMS does not apply a similar risk adjustment to account for sociodemographic factors within a hospital’s service area. Research shows that economically disadvantaged patients often have limited access to services and resources that can help to support their recovery post-hospitalization and, therefore, reduce the likelihood of readmission. Such supports include public transportation to get to follow-up appointments, grocery stores to support special dietary needs and social support. A study at Henry Ford Hospital in Detroit found that patients living in high-poverty neighborhoods were 24 percent more likely to have a readmission when compared with their peers in higher-income neighborhoods.
CMS has resisted adjusting for additional risk factors, including sociodemographic status, suggesting that doing so would “mask disparities in quality of care provided.” However, excluding important sociodemographic factors, such as income, education, occupation and primary language, creates an inherent disadvantage for hospitals that treat patient populations at higher risk for readmission. As a result, many researchers have demonstrated that hospitals that care for the neediest patients are much more likely to incur a penalty under the HRRP.
Policymakers recommend Risk Adjustment
Recognizing the disproportionate readmissions risk for hospitals that serve low-income patients, the Medicare Payment Advisory Commission, known as MedPAC, recommended in June 2013 that CMS account for sociodemographic factors in calculating HRRP penalties. Specifically, MedPAC recommended that hospitals continue to report unadjusted readmission rates so that data on potential disparities would be available. However, to calculate readmissions penalties, MedPAC recommended that CMS compare hospitals’ readmission rates against those of peer groups with similar proportions of low-income patients, rather than evaluating their performance based on national levels. Each hospital would have a fixed target readmission rate based on the percentage of its patients who receive Supplemental Security Income benefits. Hospitals that exceed the target would incur a penalty, while those below the target would not. MedPAC reiterated its recommendations in its March 2014 report, while other stakeholders also have called on CMS to incorporate sociodemographic factors:
• The National Quality Forum’s existing evaluation criteria prohibit using sociodemographic factors in risk-adjustment models. Similar to CMS, the NQF believed such an adjustment could mask disparities. However, in 2014, an NQF-endorsed expert panel published a report recommending that policymakers include sociodemographic factors in measures because “patient characteristics that are present before care begins can influence patient outcomes” and lead to incorrect conclusions about care quality.
The NQF is a nonprofit, consensus-standards organization that endorses quality measures. NQF-endorsed measures are used extensively in CMS’s quality reporting and pay-for-performance programs. In the case of the HRRP, CMS is required to use NQF-endorsed measures for the first three conditions in the program (heart attack, heart failure and pneumonia) and is expected to seek NQF endorsement of any measures added to the program.
• Legislators also have turned their attention to the issue of sociodemographic adjustment. In 2014, legislators in the 113th Congress demonstrated bipartisan support for addressing this issue by introducing two bills supported by the AHA that would have required CMS to include sociodemographic factors in the HRRP’s risk-adjustment methodology. Beginning in fiscal 2017, CMS could use an alternative method, such as the peer groups recommended by MedPAC.
These two bills, and the MedPAC and NQF expert panel recommendations, all acknowledged the need for the HRRP’s risk-adjustment methodology to account for additional sociodemographic factors to help ensure that CMS compares hospital performance fairly, while maintaining an incentive for all hospitals to prevent avoidable readmissions.
Modifications to the Penalty Formula
MedPAC also raised concerns about the penalty formula’s “multiplier effect” that results in (1) readmissions penalties that far exceed the cost of excess readmissions, and (2) an inverse relationship between national readmissions rates and hospital penalties.
Others have highlighted limitations with the statutory requirement defining how CMS must measure excess readmissions because rates for hospitals with less volume are more heavily weighted toward the national average.
Comparison against national averages also may inflate the number of penalized hospitals. Although readmission rates have declined nationally, as many as three-fourths of hospitals have incurred a penalty during each year of the HRRP.
MedPAC recommended to Congress an alternative method for assessing hospital performance that respects the intent of the program while recognizing improvements. The proposal would set a risk-adjusted readmission rate target based on historical national performance and a hospital’s share of low-income beneficiaries. Such a target would create a defined benchmark for hospitals to work toward during the performance year and, unlike the current system, every hospital could avoid a penalty by achieving fewer readmissions than the target. Ultimately, the program would continue to reduce expenditures related to readmissions, in addition to any penalties imposed on hospitals that fail to reduce readmissions below the predetermined target.
Hospitals are committed to reducing avoidable readmissions. The Health Resource & Educational Trust, an educational affiliate of the AHA, joined CMS’ Partnership for Patients as a Hospital Engagement Network that included 1,500 hospitals and 31 state hospital associations. The HENs have helped to advance the Partnership’s goal of reducing readmissions by identifying and sharing best practices. Early results are promising, as HRET’s HEN hospitals have decreased heart failure readmission rates by an average of 13 percent.
In 2008, the University of California San Francisco Medical Center started a team-based intervention to prevent readmissions for elderly patients with heart failure. Its multidisciplinary team includes two nurse program coordinators, geriatricians, hospitalists, cardiologists, clinical nurse specialists, case managers, social workers, pharmacists, dietitians and post-acute care providers. When the patient is first admitted to the hospital, the team alerts external providers who are responsible for the patient, such as primary care physicians, and engages others, like home health care providers, during the course of treatment and discharge.
During the inpatient stay, providers educate patients about their conditions using the teach-back method, which ensures that each patient understands and repeats back his or her care plan before returning home. After discharge, patients at the highest risk of readmission receive home visits from geriatricians to help manage cognitive conditions and improve medication adherence. The program has reduced 30-day heart failure readmission rates by 45 percent.
Swedish Covenant Hospital, a safety net provider in Chicago, implemented a care transitions program focused on reducing readmissions for patients with chronic conditions who were discharged home and uninsured or ineligible for home health services. The program features disease management coaching by registered nurses, a home visit to reconcile medications and conduct nutrition screening, and ongoing telemonitoring, telephone calls and home visits as needed for three months.
After implementing the program, Swedish reduced its Medicare fee-for-service readmission rate from 16 percent in 2012 to 14.25 percent in 2013. Rather than focusing on conditions included in the HRRP, Swedish focused on all types of patients with chronic conditions.
Not all readmissions are the same; they can be planned or unplanned, and related or unrelated to the initial admission. Improvements to the HRRP could focus the penalty on admissions that are avoidable and related to the initial admission. An adjustment for sociodemographic factors will ensure that hospitals that serve higher-risk populations do not incur disproportionate penalties. Critical changes to the HRRP evaluation time frame and performance rate calculation would promote continued innovation without unfairly penalizing hospitals that focus on providing care for their community populations.
This is an excerpt from “Rethinking the Hospital Readmissions Reduction Program” from the American Hospital Association. Since the report’s publication, two AHA-supported companion bills have been introduced in Congress. The AHA encourages hospitals to support the Establishing Beneficiary Equity in the Hospital Readmission Program Act of 2015 to address concerns with the HRRP. For more information and the complete report, go to www.aha.org.
Akin Demehin (email@example.com) and Michael Ward (firstname.lastname@example.org) are senior associate directors of policy, American Hospital Association, Washington, D.C.
Overview of the Hospital Readmissions Reduction Program
Mandated by the Affordable Care Act, the Hospital Readmissions Reduction Program is a payment penalty program designed to reduce Medicare fee-for-service hospital readmission rates for conditions that account for expensive, high-volume admissions and frequent readmissions. As of federal fiscal 2013, the Centers for Medicare & Medicaid Services reduces a hospital’s Medicare payments for all patients if it has a higher-than-expected 30-day readmission rate for patients with specific clinical conditions. By law, only those hospitals paid under the Inpatient Prospective Payment System are eligible for HRRP penalties.
CMS uses an ACA-mandated formula to determine each eligible hospital’s readmissions performance. The formula calculates an “excess readmission ratio” for each hospital using readmission measures for the clinical conditions in the program, and then translates that ratio into a financial penalty. As of fiscal 2015, the HRRP includes readmission measures for heart attack, heart failure, pneumonia, chronic obstructive pulmonary disease, and total hip and knee replacements. The excess readmission ratio calculated by the measures determines whether a hospital has a higher number of readmissions than the national average for other hospitals that treat a similar clinical mix of patients (i.e., patients with a similar mix of age and clinical risk factors for readmissions). CMS applies a risk adjustment to account for the differences in clinical mix across hospitals before making comparisons. Hospitals with excess readmissions when compared with the expected level incur a penalty — the higher the number of excess readmissions, the higher the penalty, up to a cap of three percent of Medicare inpatient payments. When calculating each hospital’s readmissions performance, CMS excludes patients who had certain planned readmissions, transferred to other hospitals or left against medical advice. However, CMS also includes patients readmitted for reasons unrelated to the initial hospital stay.