Value Maximizer

The data in Value Maximizer is updated quarterly by Medisolv. This occurs in January, April, July and October. Note that only measures for which CMS has published new data on Care Compare will be updated. Each quarter, Medisolv will provide clients with a report called the Quarterly Measure Update Detail Report, which will show which measures have been refreshed with new data from the previous quarter.  

Medisolv obtains measure performance data from provider files published by CMS on Care Compare. These include files for the HIQR program, as well as a variety of files posted for each specific program e.g., Hospital Acquired Infections, a variety of Hospital Value based purchasing files organized by domain, and files for Hospital Readmission Reduction Program.  Note that none of this data has patient level detail in it.  It is only summarized data by facility and includes any risk adjustment performed by CMS in the calculation of performance values.  Additional data needed for the calculation of performance results and financial penalties, including benchmarks, achievement thresholds and floor values are retrieved from the QualityNet Website. Hospital base operating rates and base operating payments, slope values and readmission supplemental data files are obtained from the most recent final IPPS rule published in the Federal Registry in August of each year.

While CMS provides hospitals with an “Excess Readmission Ratio” in their Hospital Specific Reports, which is the number of predicted readmissions, divided by the number of expected readmissions, Medisolv provides clients with the number of readmissions that would have to be avoided to avoid a financial penalty. This proprietary calculation is based on the number of observed unplanned 30-day readmissions minus expected readmissions. Estimated Excess Readmissions are calculated both prospectively for future payment years, and retrospectively for final-result years.

The prediction of unplanned 30-day readmissions is based on predictive modeling techniques developed by Medisolv, which consider an individual hospital’s historical index case volume, expected, and predicted readmissions. Each quarter, every hospital’s performance is recomputed, using one of thousands of machine learning models until a best-fit is achieved for a given measure and for a given hospital.

The Value Maximizer is a cloud-based solution requiring a web browser to access, however for those clients with a current license for Encor D, on-premises measures are available for selected measures included in the pay for performance program. These measures will re-summarize nightly and may be used to track performance concurrently, however the data from these measures will not be used in the forecasts in the Value Maximizer.

Because the three pay for performance programs are budget neutral, hospitals are essentially competing with other U.S. hospitals for higher incentive payments and lower penalties. For example, 25% of all U.S. hospitals will be penalized in the Hospital Acquired Conditions (HAC) Reduction Program. Likewise, in the Hospital Value Based Purchasing Program, the monies available for incentive payments are essentially coming from those hospitals with penalties. Therefore, to accurately forecast penalties or incentive payments for a given hospital, it is necessary to understand that hospital’s performance relative to all other U.S. hospitals that participate in each program. Because Medisolv does not have a contractual relationship with all U.S. hospitals, historical data published on Care Compare, which contains data for past applicable periods is used. As more recent data becomes available in the public domain, Medisolv will incorporate this data into your performance trend, adjusting future forecasts as needed to anticipate your hospital’s penalties or incentive payments relative to other U.S. hospitals.

Note that clients with a current license for Encor D will have access to a dashboard containing many measures included in the pay for reporting programs. These measures update nightly and may be used for concurrent tracking of performance; however, this data will not be used by Medisolv in the Value Maximizer forecasts.

All U.S. short-term acute care hospitals that participate in two or more of the three pay for performance programs, and for which there is data published by CMS in Care Compare are included in the Medisolv Value Maximizer forecasts. This excludes critical access hospitals and VA Hospitals.

Following a fully executed agreement with Medisolv for the Value Maximizer, clients may be up and running in as little as two weeks. Because there are no interfaces, data harvests or mapping required (see FAQ on Where Does Medisolv Get the Data), and because the Value Maximizer is a cloud-based solution requiring only a web browser, assistance from your IT department is minimal. Medisolv will request the names and email addresses of those users who will be provided access to the Value Maximizer. For healthcare systems that license the Value Maximizer for two or more hospitals, Corporate Reports will also be made available to designated users at your facility. Finally, for those hospitals with a current license for ENCOR Dashboards, a dashboard will be created containing some of the measures included in the pay for performance program. Once accounts are established by Medisolv, a URL link to access the Value Maximizer will be provided and a 90-minute on-line training session will be scheduled. During this session, your team will be trained using your hospital’s data. Following the initial training session, additional follow up sessions may be scheduled, as necessary.

Each quarterly update in the Value Maximizer, a model is fit for every measure and for every provider. Depending upon the amount of data and the fit, the selected model may be a time series, simple linear regression, mean, or simply carrying forward the most recent value. Note that models may change from one quarter to the next for a given provider and vary across all providers for a given measure. While confidence limits range for each specific model, provider and data set, the overall accuracy of predicted financial penalties for each program is measured by mean absolute error (MAE) each year. MAE rates are calculated annually by Medisolv and are available upon request.

Medisolv’s algorithms provide recommendations to enable the most efficient measure improvement pathway towards achieving your financial goals.  In some cases, it may be possible to achieve your financial goal by improving significantly in a single measure.  Recommendations in the Value Maximizer simulator try to balance between providing the most “efficient” pathway towards achieving your financial goals while suggesting a minimum number of measures with a realistic jump in improvement for each measure.   For measures in the Hospital Value Based Purchasing Program, Medisolv’s “efficiency index” describes the amount of effort it takes to achieve a set financial target for each point option. Efficiency is based on “net gain”, meaning the financial gain achieved with each measure score “leap” (0-10 points), divided by “effort”, which is the change in performance divided by the range in performance across other U.S. hospitals. For measures in the Hospital Acquired Conditions Reduction Program, Medisolv first computes observed complications, and then computes the financial impact of reducing each complication for each measure and provides the most efficient path to removing the penalty with the fewest measures. For measures in the Hospital Readmission Reduction Program, Medisolv computes excess readmissions and then computes the impact of reducing each excess readmission for each cohort; providing the most least number of reduced readmissions to meet a targeted financial goal.