Medicare Value-Based Purchasing (VBP) Score Tracker
Report Description
As part of the Patient Protection and Affordable Care Act (PPACA) Congress directed the Secretary of Health & Human Services to establish a hospital value-based purchasing program to become effective in FY 2013. Hospitals meeting defined performance standards will receive incentive payments. Current speculations are that 2% of a hospitals annual marketbasket increase will be set aside in the VBP program and hospitals will be allowed to earn back their inflationary adjustment based on their overall score derived from annual data reporting.
According to PPACA, the Program mustl begin in FY 2013 and will impact payments for discharges on and after October 1, 2012. For FY 2013, the program will include measures that relate to at least the following five specific conditions or procedures: (1) acute myocardial infarction (AMI); (2) heart failure; (3) pneumonia; (4) surgeries; and (5) health care-associated infections. Beginning in FY 2014 the program must be expanded to include efficiency measures, in particular measures of Medicare spending per beneficiary.
In its 2007 Report to Congress, CMS proposes to derive VBP scores in three quality domains; Process of Care (RHQDAPU), Patients’ Perspectives of Care (HCAHPS Survey), and Outcomes (30-day mortality measures for AMI and heart failure). Most recently, the health care reform law now mandates that CMS implement a VBP program that incorporates these measures, at least at the start. The outcomes domain does not yet have enough publicly available data for score calculations; hence, this model focuses only on the RHQDAPU and HCAHPS indicators which are expected to constitute 70% of the combined score.
These reports were designed to provide hospitals with a first look atthe Centers for Medicare and Medicaid Services’ (CMS’) proposal for deriving one composite quality score from the data reported under the Medicare Reporting Hospital Quality Data for Annual Payment Update (RHQDAPU) program. These worksheets are designed to allow hospitals access to estimated impact of quality results on a real-time basis while the data is still actionable and corrective adjustments can be made to internal policies, procedures and practices to avoid financial repercussions. These reports ARE NOT intended to provide actual results as most variables in the final models have not yet been determined by CMS. THESE REPORTS ARE INTENDED ONLY TO ASSIST HOSPITALS IN PREDICTING POTENTIAL CASH IMPACT BASED ON THE BEST AVAILABLE DATA AT THE TIME OF THE REPORT OR THE MOST RECENT UPDATED MODEL.
In addition to providing details on how the RHQDAPU and HCAHPS scores are calculated, the Medicare VBP Score Analysis reports show two proposed payment scenarios for the VBP program, the first scenario reflects a prior recommendation from the Senate Finance Committee’s Policy Option paper for a linear payment function, the second follows the proposalfor a curvilinear payment function as outlined in CMS’ November 2007 Report to Congress. In both payment scenarios, the VBP pool contributions are modeled according to health reform law.
The scoring composite methodology is as presented in CMS’ Options Paper and Report to Congress. The formula for the curvilinear payment translation was derived to comport with the graphical depiction in CMS’ Report to Congress. The formula for the linear payment function was derived based upon the description in the Senate Finance Committee’s Policy Options Paper – hospitals with composite scores in the bottom quartile nationally will receive no payment from the VBP pool; hospitals with scores in the top quartile will receive 100% of their contribution into the pool; the middle 50% of hospitals will be paid according to a linear function that goes from 0% to 100%. Financial impact estimations
Worksheets have also been included that included summary historical data and graphed results to assist hospitals in reporting and evaluating their data internally.
Sources: The CMS historical quality comparative data used in this analysis is available to the public at the CMS website: Quality data for the most current period is available to the public at the CMS Hospital Compare website:
Medicare inpatient payment dollars and impacts are derived using CMS’Final FFY 2010Inpatient PPS Impact File for cases, case-mix, disproportionate share, and Indirect Medical Education adjustments, which are applied to the FFY 2010 payment rates and trended forward to 2013 through 2017.
Assumptions: The CMS Options Paper and Report to Congress discuss the theory and concepts behind the proposed VBP scores and payment options and provide examples of score calculations. Wherever possible, we have followed the methodology laid out in the Report to Congress.
The CMS Report to Congress did not address certain issues necessary to complete this analysis; therefore, the following assumptions were necessary, but may not reflect CMS’ intent:
Insufficient Data and/or Small Case Sizes: Hospitals with fewer than 10 reported cases for any given quality measure were considered to have insufficient data for that measure. These exclusions apply to the hospital’s composite score and to the calculation of the national benchmarks and thresholds. The health reform law exempts certain small hospitals (those with small case counts for individual measures and/or those with only a few useable quality measures). All hospitals with at least one useable measure have been included in this analysis.
Critical Access Hospitals and Small Rural Hospitals: These hospitals are exempt from the VBP program under health reform, but the Secretary of Health and Human Services is mandated to establish special demonstration projects for these two groups of providers. CAHs and small rural hospitals are included in these reports.
Score Calculation: Hospitals must have sufficient data for calculating both an Attainment Score and an Improvement Score in order for an individual measure’s score to be included in the composite. The CMS papers are silent as to what would be done when a hospital does not have sufficient base year data to calculate an improvement score; we have assumed that this would disqualify the measure from inclusion in the composite.
Redistribution of Residual Funds: This analysis assumes no distribution/allocation of residual VBP pool dollars. The health reform law requires the VBP program to be budget neutral, meaning that all monies created for the VBP pool must be paid out. By assuming no allocation of the residual amounts, the model may be overstating individual hospitals’ and states’ losses from VBP, but it does illustrate the amount of money that is at risk for redistribution.
Topped Out Measures: The CMS Report to Congress discusses a proposed methodology for determining when a particular quality measure should be deemed “topped out” – (if the top 25% of scores are statistically identical). Rather than calculate statistical tests on each indicator, this analysis assumes that any indicator with a calculated benchmark score of 100% must be topped out. Topped out measures are assigned a benchmark of 90% and a threshold of 60%
Inclusion of New Measures: Each measure must be reported in both the base year and the scoring year to be included in the VBP calculation. When new measures are adopted by CMS, they cannot be incorporated into this analysis until there are data for both periods.
Retirement of Old Measures: Measures that are no longer reported will not be reflected in the composite scores after the date of discontinuance. The CMS proposed methodology also allows for the Secretary to retire indicators that become topped out. This analysis does not retire any quality indicators except those that are no longer reported.
Report Description
Process Score Details: This report shows all of the components considered in the calculation of theProcess Measures composite score for the July 1, 2008 – June 30, 2009 measurement period. The first two columns of the report provide the national benchmark and threshold for each quality measure. The next four columns show the hospital’s own reported data for the base year and the reporting/measurement year. The next two columns show the hospital’s attainment and improvement scores, where applicable. In instances where a hospital receives an attainment score of 10, no improvement score is calculated because the maximum score has already been reached. The last column shows the final score for each process measure.
The measures that are shaded in grey were not included in the composite score calculation either because the hospital did not report sufficient data for inclusion or because the quality measure did not have enough data to calculate national benchmarks and thresholds.
The overall score at the bottom of the report is the score that is used to estimate the hospital’s VBP payments on the next page/report.
HCAHPS Score Details: This report shows all of the components considered in the calculation of the HCAHPS composite score for the July 1, 2008 – June 30, 2009 measurement period. The first two columns of the report show the national benchmark and threshold for each HCAHPS measure – 95% and 50% respectively. For each HCAHPS measure/question, the hospital must have a score that is above the 50th percentile (aka the median) in order to receive any performance points; HCAHPS scores that are at or above the 95th percentile rank yield 10 performance points and do not require calculation of an improvement score. The next four columns show the hospital’s data for the base year and the reporting/measurement year – the score and its percentile ranking. The next two columns show the hospital’s attainment and improvement scores, where applicable. In instances where a hospital receives an attainment score of 10, no improvement score is calculated because the maximum score has already been reached. The last column shows the final score for each HCAHPS measure.
The overall score at the bottom of the report is the score that is used to estimate the hospital’s VBP payments on the next page/report.
Linear Payment Analysis: This report shows the hospital’s composite score (70% process score and 30% HCAHPS, except when no HCAHPS score is available) and how it would translate into VBP payments using the linear payout function suggested by the Senate Finance Committee. The VBP pool would be funded with 1% of hospitals’ Medicare IPPS dollars in FFY 2013, 1.25% in 2014, 1.5% in 2015, 1.75% in 2016, and 2% in 2017 and thereafter. Hospitals with composite scores in the bottom quartile nationally would receive no payment back from the VBP pool; hospitals with scores in the top quartile will receive 100% of their contribution into the pool; the middle 50% of hospitals will be paid according to a linear function that goes from 0% to 100% The graph shows how the hospital’s composite score would “map” to a payment percentage – the dotted red line. The table on top shows the annual contributions the hospital would make to the pool and its expected pay-out. The table also provides figures for the state. This model does not allow hospitals to receive more than 100% of their original pool contribution, which may overstate individual losses due to VBP, but does illustrate the amount of money that is at risk for redistribution.
CurvilinearPayment Analysis: This report shows the hospital’s composite score (70% process score and 30% HCAHPS, except when no HCAHPS score is available) and how it would translate into VBP payments under a curvilinear payment function similar to the one proposed by CMS in its Report to Congress. The curvilinear approach uses a polynomial function to create a payment curve that begins paying for composite scores above 0% and pays the full incentive amount for scores above 85% - as illustrated in the Report to Congress.
The hospital’s score is denoted on the graph in red.
The Table above the graph shows the hospital’s score, payment percentage, expected dollars contributed to the VBP program, expected dollars paid out from the program, and amount left in the residual pool. The Table also provides information on the State-wide impact of this payment scenario. Again, this model does not allow hospitals to receive more than 100% of their original pool contribution, which may overstate individual losses due to VBP, but does illustrate the amount of money that is at risk for redistribution.
These two payment analyses are important for hospitals to understand as the debate continues regarding how VBP scores should be translated into payments and how the residual dollars should be distributed.
1