I'd never heard of Healwell before and took a look over their offerings. Has anyone used the products? Beyond the…
Readers Write: Inside the Most Challenging Data Problem in Healthcare
Inside the Most Challenging Data Problem in Healthcare
By Navdeep Alam
Navdeep Alam, MS is CTO of Abacus Insights of Boston, MA.
Payer data is one of the most exciting assets in healthcare, holding the most promise for dynamic, meaningful change to the way care is delivered and paid for. It’s also the most challenging data problem to exist in the industry.
We may not always realize it, but payer data is often the center of discussion around dinner tables across the country. Whether we’re talking about which treatments are covered by our health plan, how much our prescriptions cost, or how our specialists and primary care physicians interact — or fail to interact — we’re actually talking about payer data. When we have questions about our care, the first call we make is to our health plan, and we hope they have the wisdom and expertise to point us in the right direction.
Payers are the hub of healthcare. Our health plan is where we as consumers begin when we’re trying to navigate the healthcare landscape. Where do we go to receive care? What treatments are best for us as individuals? Which pharmacies can fill our prescriptions?
As we interact with the healthcare system over our lifetimes, our experiences are eventually filed as insurance claims. Health plans hold treasure troves of rich, complex data about the patient journey, information that is critical to understanding how we as individuals experience healthcare.
This is why healthcare can be so complicated: We have barely scratched the surface in realizing the potential of payer data.
Healthcare data does not come from a single source. It comes from electronic health records, primarily used for documenting clinical data. It comes from pharmacy records, which were designed primarily for inventory management. And it comes from insurance claims, which lack detail about the patient journey but are necessary for tracking our experiences across the healthcare system. All of this data is growing at an exponential rate. Over 1.2 billion clinical documents are produced annually in the United States, and that figure is growing at a rate of 48% per year — and it’s all held by health plans.
Health plans are ingesting millions of data points every day, and all of it is necessary to ensure that we, as patients, are receiving the right care at the right cost. If this data were clean and structured in the same format, it could paint a beautifully elaborate picture of how we experience healthcare every day. But it is not: 80% of medical data, for example, is unstructured and therefore disconnected from the wider healthcare system.
Ultimately, all data challenges across the healthcare system become payer data challenges. Our health plans are burdened with the responsibility of mediating these challenges and piecing together all of the fragments of our healthcare experiences. When we switch plans, those challenges are exacerbated: all of our information is siloed within our old plans, and our new plans are barred from seeing a full picture of our medical histories. This disjointedness within the system, coupled with prohibitive privacy regulations, is how we end up with multiple sources of “truth” for every patient. The result is wide variations in the quality and cost of care we ultimately receive.
The healthcare industry has been attempting to confront these issues for decades, despite spending nearly $2.1 billion annually to try to resolve them. These challenges can be boiled down to three major roadblocks.
- Data capture. There is a longstanding inability among health plans to capture clean, and complete data in a timely manner. This is largely due to legacy systems and the continuation of highly manual data processes as best practice, all of which lead to a bevy of downstream issues.
- Data cleanliness. Payers receive and ingest millions of messy, mis-formatted data points from different providers, data suppliers, and vendors every day. The lack of standardization of these data creates inaccuracies and inconsistencies. Fragmented data often remains siloed within health plans, non-interoperable and underused.
- Data sharing. The lack of standardization of data within health plans makes data sharing impossible. CMS’s interoperability mandate is a much-needed first step toward addressing this issue and will certainly be a driver toward more efficient data sharing practices, but it is exactly that, a first step.
Overcoming these challenges is not impossible, but it requires the best tools and immediate action. According to a recent PwC survey, only 53% of payers have mapped out their data to see what will be impacted by the CMS interoperability mandate. Even more concerning: only 24% of healthcare executives said they see the mandate as a strategic opportunity. Here’s how payers can act now to unlock and realize the full potential of their data.
- Clean up data. Now more than ever, patients have greater control over their data — data which, at the moment, is largely fragmented and incomplete. Payers should strive to achieve a single source of truth for each member. Doing so will allow plans to develop personalized member benefits and give providers a complete view of each patient, allowing them to make more informed clinical decisions and empowering patients to stay healthier, longer.
- Advocate for and improve interoperability. The CMS mandate is a necessary first step, but simply following the mandate is a missed opportunity to innovate and create real change in the ways in which we experience healthcare. It is imperative that payers take up the mantle and champion interoperability from this point forward. We cannot wait for the next iteration of interoperability regulations to improve the quality of healthcare data. Payers should be contemplating ways to ingest quality data, generate new insights, and work with one another to meaningfully engage patients as we experience healthcare.
- Democratize the use of payer data. Payers cannot drive innovation themselves. The promise of interoperability will only be realized once payer data is shared widely, allowing others to drive innovation, improve connectivity, and enhance our interactions with the healthcare system. By giving patients control over their healthcare data, we’re opening a new realm of possibilities. It is upon payers to lead the charge as we step into the future of healthcare.
With the CMS interoperability mandate, health plans have the chance to change the narrative: what has historically been an industry data challenge is now a business opportunity. By taking action to unlock healthcare data today, health plans can drive efficiency within the industry and innovate to build a more seamless, engaging, and dynamic healthcare system.
Even easier solution: single payer. Think of the potential savings with disintermediation that will be resolutely opposed while everyone is cashing in on the confusion of a health care system that is pure oxymoron.
I really like the product idea here. My simplification of it is insurance companies need a single or consolidated view for a patient. I’m not sure what exactly it enables payers to do and this article doesn’t say either. But payers definitely spend a ton of money on labor related to this and don’t have the in house process guts or technical expertise to do it.
It’s strange that we are seeing so many companies pop up that are a cross between a software product and a consulting company. Maybe they are encouraged by the success of Health Catalyst and Palantir? Integration consulting doesn’t seem like it could hit the 30% margin that Health Catalyst is aiming for with its data science consulting. The most interesting thing to know would be what non-consulting method does abacus insight use to make money. Usage based pricing of some shared product sold to insurance companies? Software sass style reoccurring licensing? Support and maintenance of custom built product for each major insurance company? Hookup and support fee for each “connected” data source?