Push for Simplicity Spurring New Approaches to Quality Measurement and Reporting

By Robert C. Saunders, PhD, Elizabeth Hamlett, BS, and Julia Sheriff, MAS
June 28, 2021

On April 30, 2021, CMS released its latest edition of the hospital star ratings on the Care Compare website. The 2021 ratings use CMS’s revamped scoring methodology, which is intended to simplify both calculation and interpretation of performance scores by hospital administrators and consumers. The new system uses averages of measure performance across five domains—Mortality, Safety of Care, Readmission, Patient Experience, and Timely & Effective Care—and compares hospitals to peers that report on a similar number of measure groups.

While simplifying these calculations is a step in the right direction, the overall ratings program is still complex. For example, the 2021 scoring system creates a composite score from 48 measures, many of which are themselves composites of other indicators. The volume of measures and use of domains are intended to capture the multidimensional nature of quality and provide confidence in the consistency of the results for any domain and overall. The tradeoff, however, is that the large number of measures makes it hard for hospitals to identify key drivers of performance to promote improvement.

These challenges are compounded by the multiplicity of value-based purchasing programs that CMS operates—the Hospital Readmissions Reduction Program (HRRP), Hospital Value-based Purchasing (VBP) Program, and Hospital-Acquired Conditions (HAC) Reduction Program. These programs reuse many of the same measures to produce program-specific ratings that influence payment, typically in the form of reductions for poor performance.

The push for further simplicity in terms of measurement and reporting programs is consistent with trends in CMS’s value-based payment innovation programs. For example, the Comprehensive End-Stage Renal Disease (ESRD) Care model, which creates accountable care organizations for ESRD patients on Medicare, cut back from 18 quality measures to 5 between 2018 and 2019. Concurrently with model changes like these, the Biden administration is moving to cull the number of innovation models that CMS operates and to control the multiplicity of measures through its “Meaningful Measures” initiative.

In January 2019, the Medicare Payment Advisory Commission (MedPAC) proposed a radical simplification of CMS’s hospital value-based purchasing programs into a consolidated Hospital Value Incentive Program. Hospitals would compete against peers on five equally weighted outcome, patient experience, and value measures.  The proposal supports simplicity in calculation by setting prospective thresholds for star rating scores and using a single set of equal measure weights.

The measures selected by MedPAC also support simplicity in terms of reporting, as the measures either are directly calculated by CMS (30-day Hospital-wide Readmissions, 30-day Mortality, and Medicare Spending Per Beneficiary) or leverage existing national reporting systems (Hospital Acquired Conditions through the Centers for Disease Control & Prevention’s National Healthcare Safety Network and patient experience via the Consumer Assessment of Healthcare Providers and Systems survey).

Whether MedPAC has hit on the “right” formula with these measures remains to be seen. Some will naturally question whether the quality measures selected sufficiently capture the performance of all hospitals. The proposal also introduces new features, such as defining peer groupings of hospitals based on how many patients were dually-eligible for Medicaid. It is also important to remember that simplicity in all its forms—number of measures, calculations, summarization, display—is but one desirable feature. The metrics used will still need to demonstrate validity, reliability, sensitivity to change, and other attributes to support the informational, accountability, and other goals of the reporting program.

Also, with any change in methodology, an additional question, at least for those being measured, is how different will the results be. In 2020 KNG Health compared the financial impact on hospitals of MedPAC’s proposal taking as given that the measures selected are appropriate. We looked at differences in incentive payments to hospitals by urban location, bed size, teaching status, profit status, and Census region. We set the amount hospitals were at-risk at 2% and 5% of IPPS payments withholding. Payments to hospitals under MedPAC’s proposal are budget neutral, so all withholdings are paid out to hospitals.

Peer grouping, in combination with budget neutrality, redistributes payments among different types of hospitals. Allowing hospitals to compete in peer groups on quality allows rural hospitals, for example, to receive an average of 0.13% more in IPPS payments (under a 2% withhold) compared to an average 0.76% reduction under CMS’s current programs; with a 5% withhold, the bonus payment increases to 0.31%. Similar shifts happen at the regional level. For example, hospitals in the West North Central census region currently face payment reductions of 1.09%; under the MedPAC proposal, the same hospitals would earn an average bonus of 0.13% under a 2% withhold and 0.31% under a 5% withhold.

As CMS considers changes to the design of its reporting programs, it will be important for CMS to continue to be transparent with the analyses that support the changes. In addition, ensuring the availability of data to support additional independent assessment of the methodology will be necessary to support the goals of the reporting program.

Services : Blog, Payment Policy & Delivery System Innovation