Getting Started With Your Own Pilot: Conducting Initial Data Analysis

    • February 27, 2011

Why This Step is Essential

Before implementing PROMETHEUS Payment, payers will need to perform an initial data analysis of the health plan’s claims data. This is an essential step in determining the areas that offer the best opportunity to improve care quality and reduce costs. Before you dive in, this section will review a couple of points to consider. While it should not be difficult to conduct this analysis, it does require a highly competent SAS programmer, and it may occupy a significant portion of the programmer’s time. In fact, one may not be enough; some pilot sites report the need for an additional SAS programmer/data analyst.

In addition, payers must have “clean” and accurate claims data, formatted according to HCI3’s specifications. To help with compliance, HCI3 provides a checklist of specs, and an instruction manual on how to format data. Be sure to review these documents and check each data source before moving forward. If the data does not conform, it will produce erroneous results or no results at all. A simple example of “unclean” data would be a member ID that is not unique; in other words, it includes more than one person. If individual member identifiers or other information are missing from the data, the outputs will be certainly be compromised.

You Can Conduct the Initial Data Analysis Yourself

If you have a highly competent SAS programmer who can devote substantial time to this effort, and a solid analytics team to interpret results, you can conduct your own analysis through the ECR Analysis Tool on the HCI3 website. Once you set up your profile and review the user agreement, you can download the freeware and run the application. If you choose to run your claims data yourself, the data must first be mapped to a standard format. You will receive detailed instructions on how to execute the necessary steps.

The tool allows you to input your organization type, population type and number of lives covered. It then analyzes your claims data to create a snapshot of your current health care spending across multiple conditions, so you can examine how efficiently health resources are being used relative to benchmarks.

By comparing your PAC rates against national and regional averages, you can calculate the overall impact PROMETHEUS can have on your organization. Potential pilot partners can review the initial outputs to determine the percentage of dollars spent on PACs across any of the 21 ECR conditions and procedures. You will also receive an Opportunity Report that highlights the savings you can achieve by reducing those PACs. This report is essential to helping you identify the ECRs that offer the greatest quality improvement and cost saving opportunities.

Caution Point If you run the data in-house, you will have to license the full SAS ECR Analytic package by signing a license agreement. Again, a skilled SAS programmer will be needed for this. In addition, certain characteristics are required before you run your data. For example:

  • The file must consist of two years worth of incurred claims with a run-out period of 3-6 months.
  • All costs must be reported as “Allowed Amounts” and reconciled to adjust for interim bills and negative dollars.
  • A single “final” bill must be submitted for each claim.
  • Inpatient Stay Claims must be rolled up to one claim per admission/discharge.
  • Pharmacy claims must be reported with prescription dates, one NDC code per claim and associated costs.
  • Professional, Outpatient Facility, Ancillary and Other Claims have NOT been rolled up, and have remained at the claim line level.

Of course, the PROMETHEUS team is available for consultation to help with these and other issues.

Or Contract with Partner Data Analysis Firms

For a more in-depth report, you can have your initial data analysis performed by PROMETHEUS data partners such as Masspro or MedAssets. You will need to sign a Data Use Agreement and Business Associates Agreement before any data transfer can take place.

You can use the PROMETHEUS templates, or use your own standard agreements to expedite the agreement process. PROMETHEUS and its data partners are available to discuss the contents and format of the files prior to their submission for data analysis. There is an advantage to running your data this way: the ECR analytic outputs are more detailed.

In addition to a breakdown of the dollars spent on the chosen ECRs, you will receive provider-level outputs that can help you understand where the sources of variation are, in terms of both cost and quality. You will also receive a PAC drill-down report so you can see which specific PACs are occurring most often and are the most costly. Download a sample report that illustrates the level of in-depth information you can obtain if you choose this path.

Caution Point If you have a well-resourced IT and Quality Management team, you are in a good position to implement PROMETHEUS independently. But it’s important to be patient—examining data, running it, exploring it and executing legal agreements will take a while. If you choose to have your data run by Masspro or MedAssets, you should begin that process as early as possible; the timeframe for execution of legal agreements can vary widely.

What the Results Will Show

No matter how you choose to run your data, the results will show your PAC rates and identify the exact source of those PAC dollars. From there, you can determine the areas of your organization that offer the greatest potential to improve quality and reduce costs.

With these findings in hand, you can make an informed decision about whether, and how, to move forward with PROMETHEUS. Of course, the potential ROI and cost savings identified in your data analysis are not the only things you’ll need to consider. Any decision to implement PROMETHEUS must take many factors into account. The organization must be truly motivated to improve health care quality and move away from fee for service payment.

Providers and payers must make the pilot a priority, and be eager to work together for mutual success. On a purely practical level, participants must have the time and resources needed for implementation. And providers must have EMRs, and the ability to improve quality once the results are in (see Engaging Providers and Payers.) It’s important to understand this may not be easy. But of course, few things worth doing ever are.