April 18, 2007

Medicare Feedback Group
Centers for Medicare & Medicaid Services
Mail Stop C4-13-07
7500 Security Blvd.
Baltimore, MD 21244-1850

RE: Medicare Value-Based Purchasing Options Paper, March 23, 2007

Dear Members of the Medicare Value-Based Purchasing Work Group,

On behalf of Pennsylvania’s 225 hospitals and health care systems, the Hospital and Healthsystem Association of Pennsylvania (HAP) welcomes the opportunity to comment on the Centers for Medicare & Medicaid Services’ (CMS) Value-Based Purchasing Options Paper. HAP has taken several different opportunities to provide comments to CMS to help in the development of this initiative, beginning with comments submitted last year following the publication of the proposed hospital inpatient prospective payment system rule. HAP appreciates that CMS considered the recommendations offered by the hospital community in the development of the Options Paper. Medicare’s transition to a value-based purchasing program represents a significant change in how hospitals will receive a portion of their payment from Medicare and needs to reflect the value that Medicare places on the delivery of safe and high quality care to all patients who seek care and treatment in America’s hospitals.

HAP understands the time constraints that CMS faces in delivering a final Medicare value-based purchasing design plan to the U.S. Congress for its consideration. However, HAP would urge CMS to consider one final round of interested stakeholder input before presenting its final design concept to Congress. HAP has previously recommended that the development of the Medicare value-based purchasing plan be an iterative process where the hospital community has sufficient opportunity to comment on a proposal(s) and modifications to the proposal(s) based on comments received. HAP believes that an iterative process will need to be repeated several times to prepare the hospital community for the value-based purchasing program that Congress and CMS ultimately decide to implement.

In particular, there needs to be increased involvement of individuals responsible for hospital reimbursement and finance in discussions around the Medicare value-based purchasing program. HAP does not believe that there has been adequate time for hospital quality and finance staff to fully understand and digest the potential implications associated with the implementation of a Medicare value-based purchasing on a hospital’s overall payment. HAP believes that it is critical to try to involve those responsible for hospital finances more intimately in the plan’s development, and HAP recommends that CMS work directly with the American Hospital Association to develop a mechanism that assures that hospital finance staff have had an ability to more fully engage with CMS in discussion about the value-based purchasing plan.

HAP again recommends that CMS move cautiously in implementing a value-based purchasing incentive program in order to identify and potentially avoid any unintended consequences that may adversely affect hospitals and ultimately the patients they serve. This is particularly important as a number of major regulatory changes are expected in fiscal year (FY) 2009, including continued transition to cost-based weights, a possible new classification system to address patient severity, implementation of the Deficit Reduction Act provision on health care-acquired infections, and potentially significant wage index changes. The implementation of a value-based purchasing program, in and of itself, will be challenging and resource intensive for hospitals.

After a careful review of the Options Paper, discussion with Pennsylvania hospitals and health systems, and participation in the CMS listening session held earlier this month, HAP offers the following comments and recommendations with respect to the various components of the value-based purchasing plan as set forth in the CMS Options Paper.

Performance Assessment Model

Summary of Proposal in Options Paper:
CMS outlines a performance assessment model that seeks to reward hospitals based on achieving benchmark or attainment levels of performance. CMS has also designed the performance assessment model such that it can also reward a hospital for improving its performance on a particular measure. Specifically, every measure has a benchmark and an attainment threshold, which are determined from the distribution of national performance on that measure during the previous reporting period. These thresholds would be established using the most recent data available to CMS, and hospitals would receive a point value for each measure depending on their individual hospital score on that measure. In addition, hospitals can receive points for improving performance on a particular measure. If the point value for improvement is greater than the point value associated with attainment on that measure, the hospitals receive the higher point value for that measure, provided that the score for the measure is in the attainment range established by CMS.

A hospital’s overall value-based purchasing performance score would be based on all measures that count toward the financial incentive for which the hospital submitted data and for which the hospital had a sufficient number of cases. This means that the number of measures included in the value-based performance score for each hospital would vary depending on the services that the hospital provides and the volume of eligible patients it treats. For each applicable measure, a hospital will receive 0-10 points based on the higher of the hospital’s attainment or improvement score. The points earned would be summed to determine total points earned and then divided by the corresponding universe of total possible points that could be achieved based on the number of eligible measures for that hospital. This means that some hospitals might only have a handful of measures on which its overall value-based performance score would be calculated, while others might have their score based on all measures contained in the Medicare’s value-based purchasing program’s financial incentive plan.

HAP Comments and Recommendations:

• HAP generally supports the performance assessment model as proposed by CMS. HAP had recommended previously that any model developed by CMS should be one that rewards both attainment and improvement, and as outlined conceptually, this model attempts to do just that.
• Hospitals need to know in advance the target they must meet to receive an incentive payment. This predictability for purposes of budgeting and program planning is important in a prospective payment system and would help ensure that all hospitals that meet the threshold receive an incentive payment.
• HAP has concerns that pay-for-performance systems that provide incentives based on hospitals’ relative performance create competition among providers that impedes a collaborative environment focused on quality and patient safety improvement in which hospitals share best practices and lessons learned. Therefore, HAP is pleased to see that CMS has proposed absolute thresholds and benchmarks in the Options Paper. However, HAP is concerned that CMS’ proposal to base performance targets relative to hospitals’ prior year performance will not ensure the transparency and predictability that hospitals need. Information on the prior year’s performance under current data submission and processing deadlines would not be available until midway through the payment year. Even under CMS’ proposed expedited time frame, more than a third of the year will have passed before hospitals know the thresholds and benchmarks that they must meet. This makes budgeting, program planning, and implementation of strategies more difficult.
• CMS has indicated that each hospital’s overall value-based performance score will be based only on the measures applicable to that hospital, given its patient population and the services it provides. While HAP supports this policy, HAP recommends that CMS institute a formal process whereby hospitals can apply to opt out of measures that are not applicable to their institutions and receive formal, written confirmation of opt-out decisions from CMS prior to the measurement period. A formal opt-out process will assure that the hospital and the agency agree on the performance measures that will be used before the measurement period commences. CMS should strive to develop this formal process in a manner that does not place undue administrative burden on hospitals applying to opt out of certain measures, particularly because such hospitals are likely to be low-volume, smaller hospitals that are critical in assuring access to care in rural communities.
• There may be some inherent unfairness in the model in measuring hospitals that report and improve performance on the entire cadre of measures identified for inclusion in the value-based purchasing program incentive plan against those hospitals that might only treat or have enough patients to report on a subset of the measures in the incentive plan. In particular, those hospitals that are reporting on all measures may have an inherently more difficult job in terms of making improvements across multiple measures and conditions than those hospitals that can focus on one set of measures or condition. HAP recommends that CMS consider ways to adjust its model to account for these inherent differences among hospitals based on the number of measures each hospital reports.
• While HAP strongly supports a way to reward improvement, HAP believes that the consistency in looking at a performance on a measure from year to year may be seriously compromised by changes in the directions provided to hospitals on how to abstract the information that is used to calculate compliance with a particular measure Even though the measure at the highest level appears to be the same measure, it may be considerably different from year to year based on changes made in the measure’s specifications, documentation instructions, or inclusion or exclusion criteria. HAP was provided with numerous examples of how such changes in abstraction directions can impact overall calculated performance on a measure. One example is in relation to administration of aspirin at arrival. Abstractors were previously able to exclude patients from this measure based on information in the patient’s medical record that was indicative of a history of bleeding problems. However, there now must be explicit physician documentation of such bleeding problems in order to have the case excluded from the measure calculation for that hospital. This one change in abstraction instructions could end up including more patients who were not eligible for aspirin administration in the denominator, thereby decreasing a hospital’s overall performance score on that measure. HAP strongly recommends that CMS consider the impact on how changes in abstraction directions could impact a hospital’s opportunity to demonstrate improvement on a measure from year to year and that efforts be made to maintain stability in abstraction directions for any measure in the value-based purchasing program’s incentive payment plan.
• For many of the measures, only 10 percent of all hospitals will have the ability to score the maximum of 10 points in the model developed by CMS. HAP recommends that CMS consider revisiting how it will establish a benchmark score for those measures with a broad distribution of performance scores in the initial year(s) of the program such that more than 10 percent of all hospital have the ability to score the full 10 points. CMS should then implement changes in the benchmark calculation to raise the bar in expected hospital performance on selected measures over time.
• While supportive of the model conceived by CMS, HAP is concerned that the development of the thresholds in the value-based performance incentive plan may be somewhat arbitrary. The selection of the thresholds are based on actual hospital performance on the measures; however, there may be limited clinical significance in patient outcomes between hospitals that achieve a certain level of performance on measure and those that score slightly below that level of performance. In essence, CMS may be rewarding hospitals for attainment of performance when there may be little difference in patient outcomes along a continuum of performance.

Translation of Value-Based Performance Score into the Incentive Payment

Summary of Proposal in Options Paper:
CMS is seeking direction on how best to translate the calculated value-based performance score into the hospital’s incentive payment. CMS has suggested in the Options Paper that there might be a minimum performance level below which a hospital would receive none of the value-based incentive and that only those hospitals that reach a benchmark level of performance would receive a full incentive amount. In between these two extremes, CMS is suggesting that there would be some exchange rate or factor established that would translate the hospital’s value-based performance score into a percentage of value-based incentive payment earned by the hospital. The exchange rate could be a linear function or a non-linear function. In all the options outlined in the paper, there would be additional monies in the value-based incentive model that would not be distributed to hospitals, and CMS has asked for recommendations on what to do with this pool of residual dollars.


HAP Comments and Recommendations:

• Implementation of the CMS value-based purchasing incentive program is premised on the notion that the plan must be budget neutral, meaning that the dollars to create the incentive pool will come from existing hospital DRG payments and that pool of money will be redistributed across hospitals based on their performance on the measures included in the value-based purchasing program’s incentive payment plan. HAP is concerned that CMS’ proposal to make the value-based incentive program budget neutral will negatively impact some hospitals, including those that may be providing access to vulnerable populations. Incentive-based approaches to payment should use a system of rewards, not penalties, to help change behavior. As currently envisioned, each hospital’s initial base DRG rate would be decreased by the percent reserved for the incentive pool.
• Initially, the dollars set aside for the value-based incentive pool should be limited to no more than two percent of present hospital DRG payments. HAP understands that CMS has indicated a desire to increase the percent of the quality incentive at risk over time. Given other changes that CMS anticipates making to the DRG system at or about the same time as CMS plans on transitioning from Reporting Hospital Quality Data for Annual Payment Update (RHQDAPU) to the value-based purchasing program incentive payment plan, HAP recommends limiting the pool of dollars that are at risk in the value-based incentive program until some stability is reached in the implementation of other DRG system changes. Through the CMS Premier demonstration project, CMS has recognized that small levels of incentives can result in changes in hospital performance.
• HAP also recommends that CMS limit the amount of dollars at risk given the lack of convincing evidence on the impact of pay-for-performance systems. In a recent study, “Public Reporting and Pay for Performance in Hospital Quality Improvement,” published in the New England Journal of Medicine, researchers compared the quality performance of 207 hospitals participating in the CMS/Premier demonstration project with 406 hospitals participating only in a public reporting initiative. After adjusting for baseline performance and other hospital characteristics, pay-for-performance was associated with improvements ranging from 2.6 percent to 4.1 percent during the two-year period. The researchers tentatively concluded that financial incentives might modestly improve care quality. The researchers also recommended more research to determine whether different incentives would stimulate more improvement and whether the benefits of pay-for-performance programs outweigh their costs.
• HAP does not support the withholding of any dollars to hospitals in the initial years of the Medicare value-based purchasing program’s incentive payment plan as would be the case in the options outlined in the Medicare Options Paper. Without knowing the array of hospital performance scores, it is difficult to recommend a specific slope for the exchange rate; however, in order to award all dollars available to the program, the exchange rate slope needs to be greater or shifted to the left, and in both cases needs to extend beyond 100 percent.
• HAP recommends that CMS consider implementing incentive payments based on ranges of overall value-based performance scores attained by hospitals. For illustration purposes only, a potential payment structure could be as follows. The actual determination of what the cutoff points might be would need to be determined based on the overall performance of all hospitals included in the value-based purchasing program’s incentive payment plan. Additionally, the incentive amount might be set at different levels depending on the number of measures reported by hospitals with more money in the incentive pool for those hospitals that must collect and report more measures.
  • 150% of incentive reserved for hospitals with an overall value-based performance score of 90 percent or greater.
  • 125% of incentive reserved for hospitals with an overall value-based performance score between 85 percent and 89 percent.
  • 100% of incentive reserved for hospital with an overall value-based performance score between 75 percent and 84 percent.
  • 75% of incentive reserved for hospitals with an overall value-based performance score between 50 and 74 percent.
  • 50% of incentive reserved for hospitals with less than an overall value-based performance score less than 50 percent.
• HAP is concerned that a value-based purchasing program’s incentive payment plan, depending on how it is structured, could create competition among providers and impede a collaborative environment in which hospitals share successful approaches to improving the delivery of care and ways to improve overall performance on measures in the value-based performance program. To this end, HAP would recommend that CMS consider requiring each state’s Quality Improvement Organization (QIO) to work with high performing hospitals to ensure that their successes are widely shared as a condition of the next CMS Scope of Work contract with QIOs. Each QIO could determine how best to ensure the dissemination of that information. Additionally, there need to be requirements in the next Scope of Work contract that require QIOs to support overall performance improvement in those hospitals that are performing at the lowest levels in the Medicare value-based purchasing program’s incentive payment plan.
• CMS also needs to model the impact of the proposed value-based purchasing program’s incentive payment plan on various categories of hospitals, including small, mid-sized, and large hospitals; rural and urban hospitals; hospitals with disproportionate shares of Medicare and Medicaid patients; teaching and non-teaching hospitals; and combinations of the above hospital characteristics to fully understand the plan’s impact across different types of hospitals.
• HAP supports a phase-in transition from the current hospital pay-for-reporting program to a hospital value-based purchasing program incentive payment plan over several years. As HAP has already noted, there is insufficient evidence to unequivocally support pay-for-performance programs. Pay-for-performance programs have had mixed results, with sometimes unexpected or unintended effects on care delivery, and CMS should take a careful and deliberate approach in implementing value-based purchasing.

Structuring Incentive Payments

Summary of Proposal in Options Paper:
CMS does not specifically propose a method for the structuring of the incentive payments in the Options Paper. However, in the listening session, CMS indicated that it would apply an incentive-payment factor to the base DRG to determine hospital payment. This incentive-

payment factor would be based on the hospital’s overall performance and would be published as part of the annual hospital inpatient prospective payment system rule. Additionally, CMS indicated that the incentive-payment factor would apply across all DRGs and not just those against which the measures apply.

HAP Comments and Recommendations:

• HAP believes that the structure of the financial incentive should be as simple and straightforward as possible, especially since many of the aspects of the value-based performance incentive program will be very complex. As a starting point, HAP believes that only the base DRG payment, adjusted for geographic wage differences should serve, as the basis upon with the value-based incentive-payment factor should be applied for payment purposes.
• Given the other anticipated changes to the inpatient prospective payment system in fiscal year 2009, HAP believes that it would be extremely difficult for CMS to accurately predict the financial impact of including other components of hospital payment.
• As already described, the current Options Paper may put general acute care hospitals, which deliver a wide range of services to all types of patients at a disadvantage because they must perform well on a broad array of measures to qualify for a full incentive payment, whereas hospitals with a more limited scope of services might do quite well on a smaller set of measures. CMS needs to consider ways to deal with this inequity. CMS might consider funding the value-based performance incentive payment on a DRG-specific basis. Using such an approach, hospitals would see their DRG payments increased or decreased only for those DRGs related to measures on which they are reporting for payment incentive.

Value-Based Purchasing Measures

Summary of Proposal in Options Paper:
CMS is using the term value-based purchasing measures to represent those measures that will be publicly reported as well as those that will be used to calculate the value-based purchasing program’s incentive payment. CMS envisions that the universe of measures that will be publicly reported will always be greater in number than those included in the value-based purchasing program’s incentive payment. However, in order for a hospital to qualify for the incentive payment, the hospital would also need to be submitting data on other measures that will be publicly reported, but not necessarily included in the value-based purchasing performance incentive payment calculation. In addition, CMS provides a framework for how measures will be selected, then introduced into public reporting, and finally included for the value-based performance incentive payment.

HAP Comments and Recommendations:

• HAP supports CMS’ plan to build upon the existing set of RHQDAPU measures for the value-based purchasing program. The measures selected for the value-based purchasing program should be evidence-based, endorsed by the National Quality Forum (NQF), and adopted by the Hospital Quality Alliance. These processes have enabled hospital and clinical input and comment in selecting measures. HAP agrees that not all measures currently used in the RHQDAPU program – nor all measures developed in the future – may be appropriate for inclusion in the value-based purchasing program’s incentive payment plan. But, all measures suitable for Medicare’s value-based purchasing program’s incentive payment plan should be appropriate for public reporting.
• HAP agrees with CMS’ criteria to evaluate the suitability of measures for the value-based purchasing program based on their importance, scientific availability, feasibility, improvability, usability, controllability, potential for unintended consequences, and contribution to comprehensiveness. In particular, HAP strongly supports selecting those measures that are significantly under the control of the hospital. HAP understands the pressures to begin the public reporting of outcomes measures, which led to CMS’ adoption of 30-day mortality for heart failure and heart attack patients. As HAP has indicated previously, HAP does not believe that use of a 30-day risk-adjusted mortality for acute myocardial infarction and heart failure represents the best outcome measures that could have been selected by Medicare to reflect the quality of care delivered to patients in hospitals. Outcomes for the measures for 30-day mortality of heart attack and heart failure patients are not solely within hospitals’ control. Social, environmental and cultural factors all contribute to patients’ mortality in the community and cannot necessarily be managed by a hospital. The results of measures included in a Medicare value-based purchasing incentive program for hospitals should be directly attributable to the work of the hospital and its clinicians. This should apply to measures that are publicly reported, and more importantly to those measures identified for inclusion for pay-for-performance.
• HAP generally supports CMS’ plan that new measures be introduced into the value-based purchasing program using a staged approach as presented in the Options Paper. When new measures are developed, they should be thoroughly tested and undergo a dry run in the field before they are incorporated into public reporting. Performance measurement systems, hospitals, clinicians, and CMS must gain experience with the data collection, submission, validation and reporting of new measures before they are linked to any value-based purchasing performance incentive payment.
• HAP agrees with CMS’ proposed three-phased approach of a preliminary data submission period, a public reporting/baseline data period, and a final stage when measures are included for financial incentive, if appropriate. However, value-based purchasing measures that will be included as part of the calculation for the financial incentive should be selected during the public reporting period and not during the preliminary data submission period as CMS proposes. Selection of the measures during the preliminary data submission period is premature. HAP also encourages CMS to provide at least a one-year notice of the implementation of new measures so that systems and processes can be put in place, similar to CMS’ decision to publish fiscal year 2008 pay-for-reporting requirements in the fiscal year 2007 hospital outpatient prospective payment system rule.
• It is equally important to retain some stability in measures selected for inclusion in the value-based purchasing program’s incentive payment plan given the performance assessment model described by CMS. Otherwise hospitals will loose the ability to develop strategic and financial plans for improvement from year to year.
• HAP does not support CMS’ intent to include the clinical quality outcomes measures and the Hospital CAHPS (HCAHPS) measures into the value-based purchasing program’s financial incentive for fiscal year 2009. As already stated in this letter, HAP remains unconvinced that the 30-day mortality rates for acute myocardial infarction and heart failure patients should be incorporated into the value-based performance incentive payment plan. And while HAP agrees that patients’ perception of care provides important information to consumers and hospitals through public reporting, it is not clear yet whether the HCAHPS measures should be part of the value-based performance incentive payment plan calculation. HAP recommends not including the mortality measures and delaying inclusion of any HCAHPS measures into the value-based performance incentive payment until at least fiscal year 2010. CMS needs to carefully examine the HCAHPS data to develop the most suitable way to include HCAHPS information into a value-based purchasing program’s incentive payment plan.
• HAP has reviewed the measures that CMS is considering excluding from the value-based purchasing program’s incentive payment plan in fiscal year 2009. HAP disagrees with the exclusion of the left ventricular ejection fraction assessment for heart failure as a candidate measure for the value-based purchasing program’s financial incentive plan. Many Pennsylvania hospitals have indicated that they believe that this assessment measure is as important to include in the financial incentive plan as is prescribing of ACE inhibitors or angiotensin receptor blockers for left ventricular systolic dysfunction. HAP received many mixed clinical reviews about the exclusion of initial antibiotic received within 4 hours of hospital arrival for pneumonia. An almost equal portion of hospitals spoke in favor of excluding this measure, as did those in recommending its inclusion in the value-based purchasing program’s payment incentive calculation for FY 2009.
• Just as CMS has proposed exclusion of the prophylactic antibiotic selection measure for surgical patients in the fiscal year 2009 value-based performance incentive calculation because the availability of specific antibiotics has been unstable over time, CMS may need to consider suspending measures from public reporting and the value-based purchasing performance financial incentive calculation. This could happen based on major clinical studies or new emerging evidence that supports changes or suspension of a particular measure. Additionally, measures should be suspended when an event is outside the hospital community’s control, as has been the case with influenza vaccine shortages, which precludes hospitals’ ability to perform well on that measure. This is particularly the case with vaccines when there is tremendous geographic variability around the availability of the vaccine that could inappropriately advantage or disadvantage hospitals in certain regions even within a state on a measure outside of their control.
• HAP is concerned about the types of measures and time frame that CMS proposes to incorporate new measures into the value-based purchasing program. CMS has indicated in its Options Paper that it plans to “rapidly evolve the value-based purchasing measure set beyond the relatively limited set of conditions and services now covered” by fiscal year 2010 and 2011 through inclusion of efficiency, care coordination, and patient safety measures. Measures for these areas have not been developed, endorsed by NQF, or adopted by the Hospital Quality Alliance and are unlikely to be ready for inclusion in a value-based purchasing program by fiscal years 2010 or 2011.
• HAP does not believe that inclusion of outpatient care quality measures should be part of the inpatient value-based purchasing program. Specifically, HAP does not support tying the inpatient value-based performance incentive payment to the public reporting of outpatient quality care measures or inclusion of such measures into the overall calculation of hospital value-based performance score. Medicare should make these a condition of hospital outpatient payments, but not inpatient payment.

Participation by Hospitals with Limited Data

Summary of Proposal in Options Paper:
In the Options Paper, CMS specifically identifies some of the issues associated with the development of a value-based purchasing program incentive payment plan for those hospitals reporting a small number of cases in the measure denominator for one or more of measures proposed for use in the value-based performance incentive payment. Very low numbers of cases provide only limited approximations of the true underlying performance of the hospital. In the Options Paper, CMS outlined some different approaches to dealing with the problems of smaller numbers and asked for feedback on its approaches as well as other suggestions from the field.

HAP Comments and Recommendations:

• HAP appreciates CMS’ recognition that because some hospitals may have limited data available for reporting, their data may not be statistically stable or sufficiently indicative of their real performance to enable meaningful participation in a program that pays based on performance data.
• HAP recommends that hospitals with low volumes be required to report applicable data for their patient population in order to receive Medicare payment. With CMS’ agreement, such hospitals could continue in the public reporting portion of the value-based purchasing program but not be included in the pay-for-performance portion of the program. Any hospitals that are excluded from the pay-for-performance aspect of the value-based purchasing program should not see their payments reduced to fund the program.
• CMS needs to consider development and implementation of measures that will allow all hospitals to fully participate in the value-based performance incentive program. CMS should consider devoting resources to the identification of measures that cut across populations of patients, such as pain management measures.

Data Infrastructure and Validation

Summary of Proposal in Options Paper:
CMS has suggested revised timeframes for the submission of measure data to CMS; a time period for data resubmission; and a revamped process around data validation. CMS has invited comments on all of its suggested revisions, including comments on the need for a single hospital quality data repository.

HAP Comments and Recommendations:

• HAP does not support compressing the data submission timeline to 60 days. Pennsylvania hospitals indicated that a 60-day timeframe for the submission of measure data would be impossible to meet given the internal processes that need to be completed, including obtaining any final physician documentation in the record; coding of the information contained in the medical record for billing purposes; abstraction of the data; the submission of the data to a hospital performance measurement vendor for processing; and the time frame needed by vendors to ensure the integrity of the data before submission to other repositories on behalf of the hospital. Pennsylvania hospitals were adamant that it would be impossible to submit data to CMS in any shorter timeframe without compromising the collection of accurate data.
• As stated in the Options Paper, the current RHQDAPU process does not allow for data resubmission. HAP supports efforts by CMS to allow hospitals some time period after the close of each data submission period to allow hospitals to resubmit their data. This should allow for more accurate submission of information that is important for public reporting and carries with it additional importance since performance may determine hospital payment. Hospitals and their vendors should be offered an opportunity to correct identified errors prior to the “lock down” of the data that will be used to determine incentive payments.
• HAP appreciates CMS’ proposal to move to an annual audit of a random sample of hospitals, in addition to a targeted audit of outliers, for the purpose of validating data that has been submitted to CMS. HAP had recommended such an approach in previous comments submitted to CMS. Additionally, HAP recommends that CMS attempt to match the number of charts selected for audit with the hospital’s experience in providing care to those patient populations.
• HAP is concerned about the proposed validation method that was presented for the first time in the CMS listening session. This revised scheme was not explicitly outlined in the CMS Options Paper, thereby limiting the ability of the hospital community to fully discuss the proposed validation method process changes. HAP has several initial concerns about the revised methodology, especially since CMS proposes to determine appropriateness of case inclusions and exclusions for the different measures to calculate what will be called a measure match rate calculation. As previously noted, HAP has concerns about the constantly changing measure abstraction specifications, particularly with regard to criteria about what cases are included or excluded on the type of documentation in the medical record. HAP previously indicated that these changes could affect overall hospital performance on a measure. Additionally, such change could affect validation depending on the specifications being used by the vendor identified to conduct the chart audits. HAP believes that there needs to be greater consideration around options for conducting the audits as well as methods that CMS could use to validate the data. There needs to be continued dialogue with the field with the option discussed at the listening session being shared in greater detail with the field along with other options for consideration.
• HAP supports the proposal that the validation and appeals process be conducted post-payment to avoid delays in incentive payment adjustments and public reporting.
• HAP recommends that CMS reconsider its proposal to suppress hospital performance data on Hospital Compare for a full year should the hospital fail a validation audit. HAP would support suppression of performance data only if the accurate results are significantly different from the posted results and only for as long as necessary to gather and post accurate data.

Single Hospital Quality Data Repository

HAP Comments and Recommendations: HAP supports the concept of a single hospital quality data repository that minimizes the administrative burden on hospitals and that allows all users easy access to needed data, with appropriate protections for patient privacy. As more insurers and other organizations engage in public reporting or structure pay-for performance, it will

become more burdensome for hospitals to collect and report on different measures that use different parameters to be submitted to different data repositories. A single, comprehensive source of information to enable transparency for hospitals would help to resolve these problems as long as it:

• enables accrediting bodies, insurers, consumers, hospitals, health care providers, and others to have broad access to the quality data;
• ensures that the data are reliable and valid;
• enables collection of a broad array of patient data, regardless of age, gender, or insurer;
• provides a clear vehicle for updating measures when new scientific and clinical information affecting the measures emerges;
• minimizes the reporting burden on providers; and
• maximizes efficiency in collection and reporting for all stakeholders.

HAP encourages CMS to work with other stakeholders to ensure that the chosen repository will meet the needs and expectations of those who will be using it, including hospitals, accreditation organizations, insurers, consumers and other interested stakeholders, such as researchers.

Public Reporting of “Topped Out” Measures

Summary of Proposal in Options Paper:
CMS has indicated that it will continue to require hospitals to submit data for public reporting purposes on “topped out” measures, as defined by CMS, even if these measures are not included in the value-based performance incentive payment. At the same time, CMS has indicated an interest in rapidly expanding the measures it would include in the value-based purchasing program, which initially includes expanding the measures that must be publicly reported by hospitals.

HAP Comments and Recommendations:
HAP understands CMS’ interest in making data available to consumers for use in medical decision-making and in ensuring sustained performance on “topped out” measures. However, HAP is concerned about the potential burden placed on hospitals in having to abstract and report on an ever-expanding universe of measures without some process for retiring ongoing measure reporting for “topped out” measures. HAP recommends that CMS give consideration to a process that would allow for the retirement of measure submission once the hospital community has achieved a sustained and consistent high level of performance on that measure. Additionally, CMS should consider development of a process to monitor hospital performance on retired measures to ensure that performance on those measures is being sustained over time.

Conclusion

HAP appreciates the opportunity to provide comments and recommendations on behalf of Pennsylvania hospitals and health systems on the Medicare Options Paper. We look forward to having the opportunity to engage in further dialogue with CMS about the design of its value-based purchasing plan.


If you have any questions about HAP’s submitted comments, feel free to contact Lynn Leighton, vice president, professional and clinical services, HAP at (717) 561-5308 or lgleighton@haponline.org.

Sincerely,

PAULA A. BUSSARD
Senior Vice President, Policy and Regularoty Services