Lots of systemic problems in healthcare carried over into the new year. A shortage of doctors, spiraling drug costs, escalating market frictions, cybersecurity, mergers and acquisitions to name a few. With a big election prize in November, it’s shaping up to be a bobby dazzler of a year.
Healthcare IT is poised to help with many of the technical challenges, of course, but much of what’s transpired to date has been a kind of tactical overlay to a largely analog bureaucracy. With a little federal goosing, the business side of healthcare became priority one, and big billing engines easily trumped the clinical side for software solutions at scale.
In that sense, much of what we’ve seen so far in the way of “wiring healthcare” (a great phrase coined by Dr. Bob Wachter) falls under the rubric of the old Sun Tzu quote:
Strategy without tactics is the long road to victory, but tactics without strategy is the noise before defeat.
Seen through this lens, healthcare IT has been heavily tilted toward tactics because the billing side marches to a quarterly drumbeat for both technology solutions and the institutions they serve.
Some of that tactical defeat was evident in a surprise announcement by Acting CMS Administrator Andy Slavitt at the JPM Healthcare conference earlier this month.
In 2016, MU as it has existed– with MACRA– will now be effectively over and replaced with something better #JPM16
— Andy Slavitt (@ASlavitt) January 12, 2016
Lot’s to mull over with that comment (including the subsequent “backpedaling” and “clarification”), but the JPM event itself – predominantly for the investor communities – was an odd venue for such an opaque and tantalizing announcement. Meaningful Use was purely tactical from the start – as a way to prime the proverbial software pump. After all, rules and regulations had to be in place to account for those generous federal subsidies paid directly to providers. A follow-on tweet by Mr. Slavitt was a little more specific – albeit in a related but different direction.
Everyone’s much relieved to hear that, of course, even if the concept of active “data blocking” has a legal definition not entirely unlike porn. I’ll know it when I see it. Good luck with that prosecution.
The first half of that tweet is encouraging, however, especially with the all important qualifier of being deadly serious. I worry about the use of that one a little in healthcare – but we get the gist.
That seriousness will absolutely be tested because Congress holds the legislative pen and at least so far has done absolutely nothing to address the largest single failing in all of healthcare “interoperability” – patient identification. They could fix it. They wrote the fix into legislation back in 1996 (HIPAA) then quickly backed out the funding and banned further research from the legislative side.
All of which brings us to one of the more exciting announcement so far this year, the $1 million prize that’s being offered by the College of Healthcare Information Management Executives – or CHIME.
The announcement was actually anticipated last fall – but now it’s official – and here’s the website.
According to inside sources, there are already over 60 formal registrations, so that’s a reasonably good sign of support from the tech community. The size and scale of the problem cannot be overstated.
The “best” error rate for patient identification (estimated by ONC in 2014) is 7% – and it’s likely closer to 10-20%. That number only gets dramatically worse – quickly.
“For example, Kaiser Permanente (which has 17 instances of Epic across its regions) reported a match rate of greater than 90 percent within each instance; that rate fell to around 50 percent to 60 percent when sharing between regions using a separate instance of Epic or with outside Epic partners.” Patient Identification and Matching – ONC Final Report (February 7, 2014)
Marc Probst – the CIO of Intermountain Healthcare – estimates that they spend about $4 to $5 million per year on “technology and processes to ensure proper patient identification.” While not remotely scientific, if we take the middle number of that range and divide it by the number of beds for Intermountain (about 450), it translates to about $10,000 per bed – per year – just on patient matching.
Multiply that $10,000 by the total number of staffed beds in all U.S. registered hospitals (about 900,000) and the current annual spend is about $9 billion. Now granted, that’s total speculation and entirely unscientifc, but if you add outpatient clinics and ancillary providers who have a similar patient matching problem, I suspect the number is a lot higher – and that’s just the economic toll for a flawed solution. The patient safety component is really more of an ethical calculation – and long overdue.
Having all the API’s in the world (including the latest exciting one – FHIR) does not solve patient identification. I asked the principal architect and lead developer of FHIR, Grahame Grieve, about this specifically. As a part of every healthcare facility and EHR installation, a Master Patient Index (or MPI) has to be configured and implemented. APIs (like FHIR) can easily exchange MPIs, but they don’t generate them.
“Yup, MPI is unavoidable.” Grahame Grieve
There is simply no other single issue that has such broad implications for the wiring of healthcare as patient identification. Personalized medicine, ACOs, population health, patient engagement and, of course, patient safety – all have a core dependency on a secure and industrial-strength system of patient identification.
The HeroX Challenge is designed to excite policy makers and the industry with what’s possible technically as a way to abandon the antiquated thinking that started with Social Security and has largely remained in the world of dumb numbering. The technology is there and CHIME should be commended for funding the challenge to surface specific solutions to this intractable and systemic problem.
Of course even with a winning technical solution, there will be a lot more work to propose and pass the legislation necessary to mandate it’s use, but it should definitely help to see the exciting technical possibilities. The winner of the Patient ID Challenge is scheduled to be announced in February of 2017. None too soon and worth watching for a solution that’s at the very heart of a lot of healthcare – especially patient safety.
Latest posts by Dan Munro (see all)
- Obamacare in Gartner’s ‘trough of disillusionment’ - September 13, 2016
- Why Ireland has a National Patient Identifier (and we don’t) - August 31, 2016
- Digital health lessons from BART - April 26, 2016