January 30, 2012

Marilyn Tavenner

Acting Administrator
Centers for Medicare & Medicaid Services
Department of Health and Human Services
Hubert H. Humphrey Building
200 Independence Avenue, SW, Room 310G
Washington, DC 20201

Attn: CMS 3239-P

Re: Hospital Readmissions Reduction Program

Dear Ms. Tavenner:

The Healthcare Financial Management Association (HFMA) strongly supports CMS's efforts to implement a hospital readmissions reduction program. HFMA is committed to helping its members improve the value of the care they deliver and as such believes that this program is an important step in aligning financial incentives to encourage providers to improve quality in a manner that will lower the overall cost of care to purchasers.

HFMA appreciates the thoughtful approach CMS took towards the readmissions reduction program and is encouraged by some of the comments in the 2012 IPPS final rule.However we are still very concerned that unless CMS takes steps to provide hospitals with data related to readmissions, aligns financial incentives across the care continuum, and changes how excess readmissions are defined and calculated, hospitals will be inappropriately penalized for circumstances beyond their control. Given the severity of the penalties associated with the readmissions program, HFMA believes that inappropriately penalized hospitals could be irreparably harmed, jeopardizing care not only for Medicare beneficiaries but for the entire community served by these providers. In an effort to avoid what we believe CMS would agree is both an undesirable and unacceptable situation, HFMA respectfully submits the following comments, which were developed in consultation with our local chapters through our Local Information Network committees (LINKS) and our Healthcare Reform Advisory Committee (HRAC).

Data for Coordinating and Improving Care Delivery
We strongly believe that access to current data related to readmitted patients is necessary if hospitals are to be held responsible for readmissions. Given that 20 to 40 percent of rehospitalized patients are readmitted to a different facility from where the index admission occurred, hospitals have a significant knowledge gap to even understand the extent of the problem⁽¹⁾. In order to accurately diagnose the problems contributing to readmissions, hospitals will need cross-continuum claims data for all readmitted Medicare beneficiaries. This will allow providers to accurately identify readmission drivers, apply scarce resources to mitigating these causes, and improve the quality of care provided to Medicare beneficiaries.

HFMA believes that CMS should provide hospitals with all patient level Parts A, B, and D claims data on a monthly basis for readmitted patients attributed to the hospital. HFMA appreciates the concerns expressed in the final rule about the administrative burden this would pose to CMS. However, we believe CMS may be overstating the burden. The data set requested is no different from what CMS will provide to participants in the Shared Savings Program on a monthly basis and is for a limited set of Medicare beneficiaries.

Further, HFMA believes that CMS should provide hospitals with access to real-time Part D claims data for Medicare beneficiaries to assist with medication reconciliation. Proper medication management is a significant factor in preventing unnecessary readmissions. Our members' experience has shown that, despite best efforts, medication reconciliation performed at discharge with patients and family members varies significantly from reconciliation performed subsequently in the patients' homes by home care nurses. Having a nurse visit every discharged patient's home may not be the best way to allocate scarce nursing resources when there is a more cost-effective way to improve quality. Access to real-time Part D claims data would provide the care team with a snapshot of the patient's medications to perform an accurate reconciliation and could serve as an indication of uncoordinated care based on the number of prescriptions prescribed.

Misaligned Incentives Contributing to Readmissions
Across the care continuum, there are misaligned financial incentives that pose barriers or result in missed opportunities to coordinate care, reduce readmissions, and improve patients' outcomes. Examples of barriers HFMA's members have identified as most pressing, as well as recommended solutions, are outlined below.

Physicians: Research shows patients who receive timely physician follow-up care post discharge are significantly less likely to be readmitted.^(2,3) However, in many instances, beneficiaries do not receive follow-up care in the recommended time frame. There are many reasons for this, which include-but are not limited to-socioeconomic issues (addressed later), appointment availability, and patient compliance.

While hospitals and the physicians they employ are working diligently to address the issues listed above, these efforts are limited in their effect by a lack of aligned incentives with community physicians. We believe altruism motivates many community physicians to participate in efforts to reduce readmissions where they can. However, given the intractable nature of the issues that contribute to readmissions, we believe altruism is insufficient to create the urgency necessary.HFMA recommends that CMS use "a carrot and stick" approach to align physician and hospital incentives to reduce readmissions:

• Expedite implementation of a physician value modifier and include a component that rewards physicians with low volumes of risk-adjusted readmissions (RARs) using funds withheld from physicians with high volumes of RARs.
• Work with Congress to create a bonus payment for physicians who see patients within 72 hours of discharge. Based on discussion with our members, payment comparable to a level 4 new patient visit would be sufficient to ensure prompt follow-up. Additionally, for patients unable to come to the office within 72 hours, there needs to be adequate reimbursement for a home visit, even if the patient doesn't qualify for home health services.
• Work with the OIG to create waivers from Stark, Civil Monetary Penalties, and other legal barriers to hospital gain-sharing arrangements, allowing hospitals to pay bonuses to community physicians who have successfully collaborated with the hospital to lower readmission rates.

Skilled Nursing Facilities (SNFs): MedPAC estimates that 25 percent of Medicare SNF residents are readmitted to the hospital.⁽⁴⁾ Although hospitalizations are often medically necessary, expert evaluation suggests that 28 to 40 percent of such admissions might be avoided with high-quality SNF care.⁽⁵⁾ In most instances, the hospital where the index admission occurred does not have an ownership interest or financial relationship with the SNF that is the source of the readmission. Not only are the financial incentives between hospitals and SNFs misaligned but in some cases CMS's SNF reimbursement policy makes it financially advantageous to the SNF for patients to be readmitted. As a case in point, CMS's policy covering only the first 100 days of SNF care post-discharge creates an incentive for SNF patients to be readmitted. HFMA recommends that CMS:

• Expedite the implementation of a SNF value-based purchasing adjustment that includes performance measures related to conditions such as electrolyte imbalance, congestive heart failure, respiratory infection, urinary tract infection, and sepsis that are common causes of hospital readmissions
• Develop a corresponding penalty mechanism for SNFs with high rates of hospital readmissions
• Correct the SNF reimbursement system so that it is no longer financially advantageous for beneficiaries to be readmitted
• Work with the provider community and OIG to identify legal barriers that prevent hospitals and SNFs from collaborating and create sufficient exemptions that will further efforts to reduce preventable readmissions.

Telemonitoring and Other Technological Care Extenders: There is a growing body of evidence that proves the efficacy of telemonitoring as a means to coordinate care, improve patient outcomes, and reduce readmissions and overall episode cost. Despite these results, few hospitals have implemented it because the technology is expensive and currently not supported by the Medicare reimbursement system beyond limited circumstances.HFMA encourages CMS to pay Medicare providers for effective use of telemonitoring technologies for at-risk individuals. We propose to define at risk as a beneficiary who is burdened with one or more chronic diseases. This payment should be available for telemonitoring services provided in settings such as the patient home, SNF or group home settings. Given the technology's potential to improve outcomes and the readmission penalties, even a nominal level of payment would likely encourage wider adoption among fee-for-service providers and significantly improve beneficiary outcomes.

Patients: Socioeconomic issues play a significant role in hospital readmissions. For example, when an indigent patient cannot afford the necessary medication to manage a chronic condition or a patient with a limited support network cannot keep a follow-up appointment with a community healthcare provider, the odds of a potentially preventable readmission are significantly increased. While the scope of these problems is beyond what hospitals alone can solve, there are significant legal barriers related to beneficiary inducement that prevent hospitals from supporting vulnerable beneficiaries.HFMA recommends that CMS and the OIG create waivers, similar to those afforded participants in the Shared Savings Program, allowing hospitals to provide vulnerable Medicare beneficiaries with medically indicated services aimed at preventing readmissions. Following the examples discussed above, items covered by the waiver would run the gamut from providing low/no-cost prescriptions to rides to follow-up appointments.

HFMA realizes that many of these recommendations will require congressional action. We strongly recommend that while CMS pursues the necessary legislation to implement the recommendations outlined above, it also uses the Center for Medicare and Medicaid Innovation to launch pilot programs to test each recommendation and better align incentives.

Definition and Calculation of Excess Readmissions
HFMA's members continue to be deeply concerned with the manner in which a potentially preventable readmission for each of the selected conditions is defined and calculated. Specifically, HFMA believes:   

• There are insufficient exclusions for readmissions that are not related to the index admission   
• The risk adjustment mechanism fails to account for socioeconomic factors   
• The minimum number of conditions required for participation is insufficient to offer statistical reliability.   

HFMA first shared the concerns of its membership with CMS regarding these issues in a comment letter on the proposed rule. Unfortunately, CMS in the IPPS final rule published on August 18, 2001 accepted the language in the proposed rule unmodified. Given the severity of the financial penalty borne by providers who are inappropriately sanctioned by an ill-constructed readmissions measure, HFMA strongly encourages CMS to reconsider the following issues.

Insufficient Exclusions for Unrelated Readmissions: The final rule identifies a limited number of excluded "planned readmissions," as defined by the NQF, from the 30-day risk-standardized AMI measure. As a result, Percutaneous Transluminal Coronary Angioplasty (PTCA) and Coronary Artery Bypass Graft (CABG) are excluded from the AMI 30-day risk-standardized readmissions measure unless the principal discharge diagnosis for the readmission is one of the following diagnoses that are not consistent with a scheduled readmission: heart failure, acute myocardial infarction, unstable angina, arrhythmia, and cardiac arrest. No such exclusions exist for the 30-day risk-standardized measures for HF and PN.

CMS, in the final rule, declined to expand the number of exclusions for planned readmissions, citing Sec. 1886 [42 U.S.C. 1395ww] (q)(5)(A)(i)(I) of the Social Security Act requiring that "measures of such readmissions - have been endorsed by the entity with a contract under section 1890(a) (the NQF)." The final rule articulated CMS's concern that if the measures were modified, they would no longer be NQF-sanctioned and therefore ineligible for usage in the program. However, in this analysis HFMA believes that CMS overlooked the second subpart ([42 U.S.C. 1395ww] (q)(5)(A)(i)(II)), which also requires CMS to ensure that "such endorsed measures have exclusions for readmissions that are unrelated to the prior discharge (such as a planned readmission or transfer to another hospital)." CMS argues that the statute "does not state that the measures must account for all possible unrelated readmissions."⁽⁶⁾ Respectfully, however, the statute does not qualify its blanket requirement that the measures must have "exclusions for readmissions that are unrelated to the prior discharge." The NQF measures adopted by CMS do not satisfy this statutory requirement. Rather, as CMS has acknowledged, they are "'all-cause' readmission measures (that is, they count readmission regardless of the reason for readmission) . . . ."⁽⁷⁾ Thus, by definition, the NQF measures do not exclude unrelated readmissions.

By including the second subpart, Congress intended for CMS to modify the readmissions measures to account more accurately for unrelated admissions than is currently reflected in the NQF-endorsed measures. HFMA strongly encourages CMS to reconsider its position and recommends at a minimum that CMS develop modifiers to identify and exclude the following from the hospital specific readmissions count:

• Planned readmissions. We believe that this is necessary given that studies show 10 percent of readmissions are planned.⁽⁸⁾
• Readmissions related to random occurrences (i.e. motor-vehicle accidents). We appreciate CMS's perspective that "readmissions that are truly unrelated to the hospitalization should not affect some hospitals more than others, because these readmissions should have the same probability of occurring for similarly situated patients, regardless of where the patient was initially hospitalized." While we believe that statistically this is true in the long run, we remain concerned that in any given year the averaging process will inadvertently negatively impact some providers. Given the severe financial penalties associated with CMS's readmissions policy, we believe that CMS should take all available steps to eliminate any chance that a hospital is inaccurately identified as having higher than expected readmissions.

Further, HFMA recommends that CMS follow Congress's intent in the second subpart and create an exhaustive list of MS-DRGs for which a readmission could not be related to one of the conditions included in the readmissions policy. We strongly encourage CMS to reconsider a methodology similar to 3M's Potentially Preventable Readmissions (PPR) model to identify other conditions for which readmissions should be excluded. We believe it is appropriate to follow the review process used by 3M when defining clinically related admissions and use those criteria to review the relationship between the MS-DRGs included in CMS's readmissions program and all other MS-DRGs. MS-DRGs that are identified as either not clinically related or preventable should be excluded from the calculation of the risk-adjusted readmissions rate. This would align the readmissions reduction program with similar efforts occurring in the private sector. Commercial contracts that include readmissions programs will typically exclude readmissions related to cancer treatment, kidney failure, rehabilitation services, and mental health/substance abuse. Included in Attachment I is an overview of 3M's PPR methodology.

Insufficient Risk Adjustment: In the final rule, CMS declined to expand the risk adjustment factors considered to include socioeconomic factors. CMS stated that:

The proposed readmission measures are risk standardized readmission measures that adjust for case-mix differences based on the clinical status of the patient at the time of admission to the hospital. That is, they are risk-adjusted for certain key variables (for example, age, sex, comorbid diseases and indicators of patient frailty) that are clinically relevant and/or have been found to have strong relationships with the outcome. To the extent that race or SES results in certain patient groups having a greater disease burden, those factors are accounted for in the measure.

However, these measures are not adjusted for other factors such as race, English language proficiency or SES. We believe such additional adjustments are not appropriate because the association between such patient factors and health outcomes can be due, in part, to differences in the quality of health care received by groups of patients with varying race/language/SES. Differences in the quality of health care received by certain racial and ethnic groups may be obscured if the measures risk-adjust for race and ethnicity. Additionally, risk-adjusting for patient race, for instance, may suggest that hospitals with a high proportion of minority patients are held to different standards of quality than hospitals treating fewer minority patients.

We agree with CMS that race or ethnicity should not be included in the risk adjustment mechanism. However, given the predictive power that variables such as the presence of Supplemental Social Security Income (SSI) have related to readmissions⁽⁹⁾ which are not included, HFMA recommends that CMS include SSI and other similar socioeconomic indicators (e.g., presence of Medicaid as a secondary payer) to improve risk adjustment. If CMS does not believe that it can improve the risk adjustment mechanism without sanction from the NQF, HFMA recommends that CMS delay the implementation of the readmissions reduction program until NQF develops readmissions measures that fully account for socioeconomic drivers.

Recent analysis ⁽¹⁰⁾ has shown that safety net hospitals are more likely to have higher readmission rates for the conditions included in CMS's readmissions policy. We continue to believe that refining the risk adjustment mechanism is necessary to ensure a level playing field for all providers while protecting safety net hospitals and their communities from the unintended and counter-productive consequences of an incomplete risk-adjustment mechanism. For many of these facilities inpatient Medicare payments are a larger than average component of their revenue. Any reduction in Medicare payment related to an incomplete risk adjustment will have both direct and indirect consequences. As a direct consequence, it will limit providers' ability to invest in programs to reduce unnecessary readmissions, and the socio-economic factors that cause them, further harming Medicare beneficiaries. Indirectly, it will reduce employment and increase the ranks of uninsured in these communities as safety net hospitals will likely respond to additional financial pressure by reducing staffing levels.

In the future as more providers adopt EMRs, HFMA also recommends that CMS explore ways to use the embedded data to refine the risk adjustment mechanism and improve the accuracy of readmissions models. Examples of variables for consideration in the future should include housing discontinuities as measured by address changes, census tract, history of drug use, and marital status.⁽¹¹⁾

Minimum Discharges for Inclusion: The final rule sets 25 discharges for each condition as the minimum number needed to be included in the Hospital Readmissions Reduction Program. CMS chose this number as it is the current minimum required for reporting in the Inpatient Quality Reporting Program (IQR). The rule further states that CMS is "currently conducting additional analyses to further evaluate the appropriate minimum number of discharges needed to yield reliable excess readmission ratios for the three proposed measures."

HFMA continues to strongly believe that in the absence of complete certainty around a minimum threshold to "provide sufficiently reliable information on hospital performance," CMS needs to increase the minimum threshold to a sufficient size to guarantee data reliability with a 99 percent confidence level. While unreliable data is an undesired outcome in a reporting program, it is completely unacceptable in a program that could potentially reduce a hospital's inpatient Medicare revenue by up to three percent. The national average hospital operating margin in 2009 was 1.98 percent.⁽¹²⁾ Given that Medicare inpatient payments are 18 percent⁽¹³⁾ of total hospital payments, any reduction due to unreliable data will irreparably harm hospital finances and threaten hospitals' ability to serve the communities that depend on them. Further, a low (and therefore unreliable) minimum threshold will disadvantage smaller hospitals as not only will they have lower volumes for each condition, but also will tend to have smaller margins and a higher reliance on Medicare revenue to sustain operations.

HFMA looks forward to any opportunity to provide assistance or comments to support CMS's effort to create a preventable readmissions reduction program that improves the quality of care for all patients. As an organization, we take pride in our long history of providing balanced, objective financial technical expertise to Congress, CMS, and advisory groups.

We are at your service to help CMS gain a balanced perspective on this complex issue. If you have additional questions, you may reach me, or Richard Gundling, Vice President of HFMA's Washington, DC, office, at (202) 296-2920. The Association and I look forward to working with you.

Sincerely,

Richard L. Clarke, DHA, FHFMA
President and Chief Executive Officer
Healthcare Financial Management Association

Cc: Shaheen Halim

ENDNOTES  

1. Jencks, S. F., M. Williams, E. Coleman. "Rehospitalizations among Patients in the Medicare Fee-for-Service Program." New England Journal of Medicine, April 2, 2009
2. Hernandez, A., M. Greiner, G. Fonarow, B. Hammill, P. Heidenreich, C. Yancy et al. "Relationship Between Early Physician Follow-up and 30-day Readmission among Medicare Beneficiaries Hospitalized for Heart Failure." JAMA 303(17) (2010):1716-1722
3. Misky, G., H. Wald, E. Coleman. "Post-hospitalization Transitions: Examining the Effects of Timing on Primary Care Provider Follow-up." Journal of Hospital Medicine 2010; DOI 10.1002/jhm.666.
4. Medicare Payment Advisory Commission. A Data Book: Healthcare Spending and the Medicare Program (2006)
5. Saliba, D., R. Kington, J. Buchanan et al.. "Appropriateness of the decision to transfer nursing facility residents to the hospital" Journal of the American Geriatrics Society 48 (2) (2000):154-63.
6. 76 Fed. Reg. at 51667
7. 76 Fed. Reg. at 51668
8. S. Jencks et al., "Rehospitalizations"
9. S. Jencks et al., "Rehospitalizations"
10. http://www.kaiserhealthnews.org/Stories/2011/December/20/Readmissions-Methodology.aspx
11. Kansagara, D., Englander, H., Salanitro, A., Kagen, D., Theobald, C., Freeman, M., Kripalani, S. "Risk Prediction Models for Hospital Readmission: A Systematic Review." JAMA, October 19, 2011
12. Ingenix, Almanac of Hospital Financial and Operating Indicators, 2011
13. 2009 CMS HCRIS Database

About HFMA
HFMA is the nation's leading membership organization for more than 37,000 healthcare financial management professionals. Our members are widely diverse, employed by hospitals, integrated delivery systems, managed care organizations, ambulatory and long-term care facilities, physician practices, accounting and consulting firms, and insurance companies. Members' positions include chief executive officer, chief financial officer, controller, patient accounts manager, accountant, and consultant.

HFMA is a nonpartisan professional practice organization. As part of its education, information, and professional development services, HFMA develops and promotes ethical, high-quality healthcare finance practices. HFMA works with a broad cross-section of stakeholders to improve the healthcare industry by identifying and bridging gaps in knowledge, best practices, and standards.

Appendix I: Overview of 3M's Potential Preventable Readmissions Methodology
 

Publication Date: Tuesday, January 31, 2012