Before joining Ribbon Health, I saw first-hand the challenges of health data and data exchange in the contexts of pharma, chronic disease management, value-based care, and patient record sharing. Through that experience, I’ve seen blockers to successful data exchange coalesce along similar themes:
- Missing incentives: Organizations have made costly investments into tech stacks that work for certain use cases internal to the org. Without incentives to build new infrastructure for data sharing, change management is not worth the cost and effort. We’ve seen this in previous attempts at co-ops and data models where the work eclipsed what systems were able to reap in rewards.
- Security risks: Change management is hard enough, but security risks add complexity to the ability to share clinical and personal health information, leading organizations to stick with tried and trusted infrastructure even if it’s not the most accurate or efficient.
- No common key: With different systems, data rules, business logic, and use cases, there is a significant challenge in the fundamentals of sharing and understanding information.
As they say, identifying you have a problem is the first step to healing. There has been a lot of movement through policy and technology to address the above and move us toward more free-flowing information.
When I look at provider data, it is a different beast. A 2018 CMS survey shows that over 50% of the information in health plans’ provider directories is inaccurate. Patients suffer from surprise bills even when they’ve gone through the trouble to look up a provider on their own health insurer’s directory page. I’ve heard people across the industry bemoan just how hard it is to get provider data right. And it doesn’t hold up against the above common root causes:
- We have a common identifier through the NPI (or TIN for facility-level)
- Provider information is publicly available on both provider and health plan’s sites
- Providers and health plans alike should have a vested interest in keeping their information and directories up to date to render needed services
So what is the problem with provider data?
To start, here’s a high-level view of three primary issues:
Provider data changes frequently
Provider data is complex, and “provider data” can mean a lot of different things. If you just look at the phone number and address at which a provider practices, this information changes 30% year over year. At Ribbon, we’ve seen 2 – 3% churn every month in the correctness of this information. Adding in dimensions like affiliations, insurance networks, and specialties lead to much higher rates of change.
There’s no single source of truth (or single use case for the truth)
Provider information is disjointed and stored all over the place; providers will use one vendor for credentialing, another for billing, and yet another for marketing and patient satisfaction. The use cases for this information are also disjointed: patients will want a certain view, regulators another, network design another yet. So what ends up happening is a lattice of redundant and manual work. Users who have workflows where provider data is one input end up searching across different platforms and screens to try to find the best guess of the necessary information in provider directories, public search engines, and CRMs.
Where there is no single source, there is also no single centralized place to update and maintain data changes. With all of the different users of provider information, we have many opportunities to confirm data points (calling the provider to book an appointment, seeing a provider for the wrong specialty), but there is no easy way to edit and share those confirmed updates. The errors just live on.
Confusion over “accuracy”
When my dad needed major surgery, we wanted to know: of the surgeons that took his insurance nearby, who had the most success with this particular surgery, and who could operate in the next few weeks? That type of information is far beyond the current accuracy evaluation criteria of whether the location and phone number in a directory are correct. While calling the correct phone number was an important way to get in touch with said surgeon, I wanted to make sure I had the right surgeon in the first place.
This is the crux of how we should approach a solution to this thorny provider data: we need to think about information sources and the accuracy of that information in a way that meets what patients and referring providers need in order to make informed decisions, while also meeting regulatory standards on accuracy for compliance and daily business use.
Read Part 2 to explore how locations factor into the provider data problem.