What happened to all that medical price data? 

It’s been almost a year since group health plan sponsors and issuers, in order to comply with the Transparency in Coverage final rule, posted machine-readable files (MRFs) that contain in-network negotiated charges for every medical service with every provider in their networks. This data had previously been considered by insurers as proprietary and confidential, but the government saw the need to put more data about the cost of healthcare into the hands of consumers. There have been eye-opening reports in the past revealing wide discrepancies in the price of medical services from one healthcare facility to the next; the hope was that making price data freely available would make these kinds of studies much easier to do. Given that the data has been available to the public since July 2022, why haven’t we seen more stories on the cost of healthcare in the United States? What happened to all that medical price data?

Well, it’s certainly out there. By some estimates, over 700 terabytes of MRF data have been published since last summer. This number will only grow larger as updates are made to these files to reflect changes made during insurer negotiations with providers and health systems, essentially a continuous process. It’s hard to grasp just how much information this really is, so for context we’ll put it this way: this amount of data capacity would allow you to store about 70 million photos on your smartphone.

Which brings us to the first reason why we haven’t heard much about this information – the sheer amount of it. This new ocean of data was dropped on the internet all at once, but not all in one place. Collecting it, analyzing it, validating it against real claims charges, and summarizing it for use are each major tasks that take time to do properly.

Further, some of these tasks require more than healthcare knowledge, which is another roadblock. To unpack the MRFs at any sort of scale requires meandering through nested json files (a file type many had never even heard of before now), scouring the internet for all the nooks and crannies where plans have squirreled this information away, and bringing it all under one roof. This requires some pretty deep background in data science techniques, which makes this a “don’t try this at home” situation for many in the healthcare field, at least for the time being.

There is even a third, potentially more important, reason that explains why we haven’t heard much about this information: it lacks any utilization data to go along with it. While we can see the sticker price for an MRI for two different providers, unless we also know how many MRIs are scheduled and billed between the two providers, we can’t calculate the average price of an MRI. This would be necessary to allow us to make general statements about the average cost of care in any particular network or geography.

While this may sound less than encouraging, we believe the data will become more useful over time. We have been taking steps to explore the art of the possible here, such as partnering with clients to explore their own MRF data and exploring our parent company’s MRF data. We are also evaluating the landscape of companies who are in the midst of summarizing and validating this new data; what we’ve learned is that not every company with expertise in data science has the same level of expertise in healthcare, although there are some who have both.

While we’re excited about the possibilities for this price data, we’ll need to be patient as the market moves to put more of it to use. Importantly, this data also underpins the final leg of the transparency regulations, which requires real-time benefit cost estimators be made available to the public for 500 shoppable services this year and for everything else by 2024. So if you haven’t already, make sure your data partners are ready for this next step of compliance.

While we are still a long way from achieving price transparency in health care, this data is a crucial starting point.