Innovation Pulse: ICD-10 - Pick your own adventure

Charting a proper course might lead to a veritable treasure trove of data. Or providers could lose their way and smash into sunken productivity.
By Tom Sullivan
07:51 AM
Share
ICD-10 Treasure map

Even though it has felt, perhaps, as if the opposite was true for several years, hospitals and medical practices are captains of their own ICD-10 ships -- a fact that's more apparent now, literally days from shore, than ever before.

A caveat: If Congress has somehow sunk the ICD-10 boat with a piece of last-minute legislation between the time I scribed this and the moment you're gazing upon it, well, then either bookmark it for future reference or tear this page out and peg it to your office wall.

But as of now, as masters at the wheel, it's imperative to navigate past the cliffs of productivity, avoid floating in irons otherwise known as minimum compliance, and chase that chest – the one containing riches in the form of more granular patient data.

Ready to hoist anchor and begin the journey?

Hospital leaders already educating staff and training coders might as well look beyond the cliffs and skip to Chapter 2.

Data-savvy CIOs and their IT shops, meanwhile – the ones that have remediated systems, mapped out analytics tactics and plotted ways to put ICD-10 data to work – can bypass both chapters and head straight to the pot of population health gold.

Chapter 1: About those cliffs
...
Much the way explorers once believed the world to be flat – that if one sailed far enough into the horizon boat and crew would ultimately fall off the world's edge and meet their makers – the productivity of medical coders is widely expected to tumble over a cliff come Oct. 1, 2015.

ICD-10-CM and its approximately 69,000 diagnosis codes (ICD-9 has about 14,000), the dissenting argument goes, will require more effort to locate codes – the idea being that clinicians previously selecting from a handful of options to find a patient's condition will now have to pinpoint the best one among a greater number.

The ultimate upside to that granularity may one day be discovered among presently buried treasure, but in the near-term it's more likely to complicate matters.  

A prominent example of this: When Canada set sail to ICD-10, it experienced a dip in coder productivity that estimates project was somewhere between 25 percent and 50 percent and still hasn't fully come around. I'd imagine it's safe to say that varied from one coder to the next but, obviously, their experience shows that the new classification system will hurt at first.

Whether or not proficiency ever returns to ICD-9 levels here in America is anyone's guess. The harsh reality is that to maintain current productivity rates, in addition to educating and training existing staff, you'll grapple with the need to hire more coders.

There goes some of that found treasure. 

Provider and IT leaders interested in understanding degrees of compliance, keep reading. Those who already know they're gunning for more can skip ahead to this column's final chapter.


Chapter 2: Compliance
Whereas it's far too early to define any notion of maximum compliance, the phrase "minimum compliance" has been bandied about for years now, and the Centers for Medicare & Medicaid Services' surprising recent treaty with the American Medical Association inevitably wrinkled the sails of what, exactly, it will mean to meet the ICD-10 mandate.

So-called minimum compliance could look an awful lot like this: Providers will remediate clinical software, EHR, revenue cycle and any other systems that ICD-10 touches – and essentially spend the year learning ICD-10 one necessary code at a time, falling back on "unspecified" when apropos, knowing that CMS will reimburse those claims so long as they're submitted in ICD-10 and not ICD-9.

One potential pitfall within this approach is that among small and even some mid-sized providers, the temptation to defer to "unspecified" is fierce, which will only inhibit the nation's trek to a land with more detailed information for the greater good. The way things stand today, however, that will only last for one year, after which healthcare organizations of all sizes must use ICD-10 with accuracy to collect payments from CMS.

You've made it this far. Now it's time to hunt for that treasure chest.


Chapter 3: Population health pot o' gold
Congratulations! You've arrived at the unknown. Treasure maps tease of accurate and dense data sets that everyone from academics, clinicians, informaticists, researchers and scientists, in some cases patients themselves, can tap into and discover a deeper understanding of chronic conditions, disease states – even genomics, to a certain extent – to tailor precise and personal treatments that bolster outcomes for individuals and populations.

Indeed, population health and, beyond that, precision medicine, represent a new world of sorts. Like all uncharted territory, the only way to know whether it is really full of fertile soil, or arid and craggy, is to actually go there.

ICD-10 is a key spot on those geographies not just for its greater specificity but also because, at this point, ICD-9, an antiquated 30-year old classification system, is dried up and not expandable to include new codes that keep pace with so many exciting revelations emerging out of the medical community.

Here's the rub, though: No precedent exists for harnessing big iron and impactful analytics software to dig out information from ICD-10 data. Yes, it's true that several other developed nations made the switch to ICD-10 years ago, that much we know. We just don't have an "X" marking the spot so the U.S. or anyone else can see where ICD-10 triggered substantive improvements in patient care and illustrating exactly how to land there.

But does that mean it's impossible, you ask? Of course not.

I'd argue that no country – nor healthcare entity, for that matter – has yet been in a position to make use of ICD-10 data for population health or precision medicine. The technology simply wasn't there at the onset.

When our neighbors to the north transitioned to ICD-10 between 2001 and 2004, for instance, the armada of advanced analytics tools for healthcare was smallish, big data and cloud computing were not household buzzwords and, since that pre-dated the HITECH Act, neither were electronic health records.

Today's providers, rather, have reasonable access to those products as well as ICD-10-centric tools including clinical documentation improvement and computer-assisted coding software, natural language processing, a burgeoning cadre of apps, and many online nuggets, like free code conversion sites.

None of those, however, are guarantees of success. There's simply no way to tell yet whether ICD-10 will yield untold riches or merely be a bucket of fool's gold. 

And this is where I surrender the captain's wheel to you, dear reader, to steer toward your own fate: Will you comply minimally, just enough to get paid? Or implement robust technologies to uncover buried treasure within this new data chest? There's inevitably plenty of bounty to be had for providers who can manage the latter.

Time to write your own ending.