Many in the healthcare community are concerned with the unprecedented government release of Medicare payment data. Secretary Sebelius made it clear earlier in June that this won’t be the last of it either. Is it too much? Is it not enough? What if people misuse the data? These are just some of the concerns various stakeholders across the healthcare spectrum have voiced. Tom Wadsworth, who leads product strategy at MedeAnalytics, shared his views on the topic with Katie Bo Williams of Healthcare Dive earlier this month. We had the chance to catch up with Tom and asked him to expand on his thoughts:
Q: Given the legitimate limitations to Medicare data, what can providers glean from this data?
A: Providers can learn a lot from the data release but the AMA’s concerns are understandable. However, I don’t agree with all of their recommendations. Data when it’s first distributed is never 100% right. Back in the early 90s, we started releasing information to drive performance improvement. That data had flaws but if we did not start somewhere, we would never improve. Providers need to take the data and learn to expand their care based on what they can glean. The more they review the data, the more they will be able to know how to understand it.
Q: Since the data is only a slice of the physicians entire patient group, is this data a really good gauge for an entire patient population?
A: Since the data is only surrounding Medicare patients, physicians can only apply insights gleaned to Medicaid patient populations. Expanding the data to build assumptions around the entire patient population wouldn’t work.
Q: Can health payers benefit from this data?
A: It’s a data source for a couple of things: fraud is one. For example, in E&M charge codes (evaluation management, 1-5, 5 being the most intense). Find a physician who is constantly billing for level 5 care, it’s probably not accurate. That could lead to a fraud investigation to explain the issue. When everything is transparent natural errors and blatant fraud can be prevented.
The data can also help improve a value-based system. Payers can look at the cost of prescription medication being recommended and determine whether a more expensive drug is being encouraged when an affordable version is available. This can help reduce costs and still provide high quality care.
Q: Should employers and hospitals look at it from this angle as well?
A: If I’m an employer and I have a self-insured health plan, all of a sudden I have the same concerns as a regular health plan. You will be able to see if you have providers in-network that are resulting in higher costs. Hospitals will have options in their benefit design to say in-network MDs will have lower pay than out-of-network. This can be a way of encouraging employees to go to high quality, low cost providers.
Q: Stepping back to data at large, what is the provider standpoint? Are there other things the provider can acquire out of the data?
A: From a provider perspective, those would be the major items. They might be able to surmise what high-volume competition is charging for competing programs. That is a realistic possibility.
Q: Might this theoretically have an impact on M&A decisions?
Q: Do you think next year’s data is going to be a little more sophisticated or will we get better at analyzing?
A: That could go both ways. Now that MDs know this data is going to be published, this will probably influence MD behaviors and limit fraud. To a degree, AMA’s points are valid and can be corrected in reporting of data and I think CMS will do that. I don’t agree with AMA proposing we postpone the release of data based on the potential for misinterpretation; there’s always room for misinterpretation. I think CMS needs to release more data, not less.
Q: Is there any other data that the government can release?
Yes, they can expand their release to include Part A with Part B. For example, if Medicare denies a hospital payment for an admission in Part A. On the Part B side, the MD will be getting paid for that. The hospital gets that payment but the MD is getting paid. This would be a good opportunity to link Part A & Part B to provide a full picture. If MDs were held accountable for an admitting patient that didn’t meet criteria, you could match up Part A & Part B better.
Q: What about allowing a MD to edit data prior to release?
A: When I was at the Cleveland Clinic, everyone could see the data before it was released to the chairman and fix it. The problem is the data set is huge – how much time do you need to correct that? What would they correct? There is no easy answer. It might confound the project so much that the data wouldn’t be released.