Blog: Issue Brief: Provider Data Accuracy
I recently relocated to the lovely city of Springfield, IL — the Illinois state capital and the home of Lincoln. For me, that meant finding a new primary care doctor to see for prescription refills, annual check-ups, and referrals to specialist care, among the countless other things a primary care physician will do for their patients.
I used my insurance’s provider directory website to locate a primary care clinic that was nearby, highly-rated, and most importantly, in-network, and made my appointment. My initial impressions of the doctor were wonderful. He provided a physical and some routine blood work, and I walked out the door with refills on my prescriptions, referrals to some local specialists, and a sense of relief that I now had a doctor to see in my new home.
All was well until months later when I received a letter from a collections agency about an unpaid, several-hundred-dollar bill from this same primary care clinic. As it turns out, they weren’t actually in my insurance’s network.
They had ceased accepting this insurance several months ago; the insurance website simply had the wrong data.
The Provider Data Problem — Widespread Inaccuracy and a Harsh Regulatory Landscape
This, unfortunately, is a pretty common occurrence in our healthcare system today. As it stands, on average, nearly half of all provider directory data is inaccurate. That’s not a typo. In a recent study, the Centers for Medicare and Medicaid Services (CMS) found an average 44.97% level of inaccuracy for provider data across the Medicare Advantage Organizations they sampled.
This level of inaccuracy has been corroborated by numerous other studies, across geographies and specialties (see the Resneck 2014 study on dermatologist data, for example). Independently, we at Orderly Health have seen similar numbers in our own engagements with payers.
Part of what makes accuracy so hard to maintain is that provider data changes 2–3% every month by most estimates, with providers being inherently difficult to reach.
In 2015, two major California insurers — Anthem Blue Cross, Blue Shield of California — were fined $650,000 based on statewide surveys that found approximately 25% of their provider data records to be inaccurate.
These fines are mere pocket change compared to the tens of millions both insurers were court-mandated to pay back to affected patients. That year, Blue Shield of California had already been forced to pay more than $38 million in refunds by the time of the report’s release.
It’s important to note here that Blue Shield’s data had HALF the level of inaccuracies compared to the national average (~50%), meaning that a payer with an average level of inaccuracy could theoretically be forced to pay double in fines and refunds.
Solving the Problem — Results of Leading Market Solutions
So, how has the market addressed this issue so far?
The current industry standard requires direct provider attestation in order to ensure the data is accurate . In other words, the highest tech solution available to payers is to call, fax — yes, using actual fax machines — or mail out the existing information to payers, and ask them to correct it if any information is out of date (ironic, in the case of mailing, as often the address is wrong).
All of this manual effort costs an average of $4 per provider record, and takes a full 6 to 8 weeks to return results. Clearly, there needs to be a better way.
In March of 2017, America’s Health Insurance Plans (AHIP) brought together 13 of its national member plans — Anthem, AvMed, Blue Shield of CA, CareMore, Cigna, Florida Blue, HealthNet, Humana, L.A. Care Health Plan, Molina Healthcare of CA, SCAN Health Plan, Wellcare, Western Health Advantage — to pilot two of these market-leading vendor solutions.
These results were… underwhelming.
In the majority of cases, fewer than 20% of providers completed the vendors’ provider verification processes, and only after multiple contact attempts.
According to these surveys, clinicians and their staff:
- “Expressed a general lack of awareness regarding the need to proactively alert plans of changes to their information”
- “Did not understand the purpose of, or need for, responding to plan requests to validate or update their information”
- “Felt overwhelmed with responsibility and therefore prioritized activities that were required of them by regulation or to secure payment for the provider”
To combat the prevailing issues outlined above, Orderly takes a markedly different approach.
We’ve built OrderlyData, a custom machine learning engine, which sources data from 30+ public and proprietary data sources, to ingest, clean, and update provider directory data automatically, without providers needing to submit the data themselves.
Then as an additional layer of validation, we take a small, yet statistically significant sample of the providers and validate via the standard call center method, and compare the results to OrderlyData’s output.
By relying primarily on machine learning instead of call centers, email, and fax, OrderlyData is able to drastically reduce cost, increase efficiency, and improve overall data accuracy.
Early work with innovative payers has shown Orderly’s solution to deliver accuracy of over 90% for data fields such as provider phone number and address. Furthermore, OrderlyData is robust and flexible enough to validate myriad data fields and integrate seamlessly into existing workflows and solutions. If this sounds interesting, let’s talk.
An Eye Towards the Future — How The Provider Data Problem May Evolve, and How Your Organization Can Prepare
As consumers, providers, technologists, and politicians alike become more aware of the challenges inaccurate data presents, new legislation is underway at federal and state levels.
For clues as to how regulations may change over time, we can look to the history of how these regulations has evolved to date. The past decade has seen steadily increased regulation at all levels, with new states passing legislation for provider directory data every year — some establishing a minimum frequency for directory updates, with others establishing minimum accuracy threshold.
The size of federal fines is increasing at a commensurate rate, as is the frequency of federal blocks placed on new enrollment (i.e. ceasing new enrollments until data accuracy rises above a specified threshold) and mandated reimbursements to affected patients for plans with insufficient provider data accuracy.
Recalling our previous example of Blue Shield of California, they were forced to pay out nearly $40 million in mandated reimbursements alone, and their data was 25% more accurate than today’s national average. In other words, an insurer of similar size with an average level of inaccuracy could be paying nearly $80 million in reimbursements to affected patients alone, not to mention millions more in federal and state fines.
The case of regulation of EMRs/EHRs can also provide some insight into legal changes to better regulate provider data. It’s possible that legislation similar to HITECH may be passed in the near future, with insurers mandated to use certified tools to manage their provider directory data and additional associated fines for non-compliance.
While no one can be certain of the specifics of pending legislation, the current trajectory of the regulatory landscape makes it clear that it is absolutely critical, now more than ever, for your organization to be prepared and have the proper protocols and tools in place to ensure the accuracy of your provider directory data.
If your organization could use a helping hand, please don’t hesitate to reach out.