Structured Data + ICD-10: Readiness Is All

Tidal Wave of Change for Healthcare

"Crises Precipitate Change." - Deltron 3030

Healthcare workers have recently been hit by a tidal wave of changes to their occupations. From the Affordable Care Act (and its various ancillary changes), to constant discoveries in research, to the additions of technology - not the least is the adoption of electronic health records. While there are always new problems arising, the past few years have seen many facilities, physicians and staff members having to contend with a veritable sea change in how they do what they do. And yet another looms large ahead in the coming months - conversion to ICD-10.

For the uninitiated - diagnoses are coded using a very old system known as ICD. The United States currently is using ICD-9, while the rest of the world has been on ICD-10 for some time. This fall, after multiple delays for a multitude of reasons, the US will convert to ICD-10 for all of its diagnosis coding (and some of its procedural coding, but that's another kettle of fish). This seems fairly routine - just upgrading to the latest version and finally joining the rest of the industrialized world in cataloging various diagnoses, right? The issue comes in the fact that ICD-10 has many more codes than ICD-9 - and that the US version of ICD-10 (ICD-10-CM) has many many more codes than other countries' versions of ICD-10. How many more? Over ten times the number of codes, going from roughly 14,000 to 155,000. And this is all expected to happen overnight: on September 30, 2014 the hospital is using ICD-9, on October 1, they are using ICD-10.

And, because all of this wasn't problematic enough, there's yet another wrinkle - ICD-10 codes are structured differently than ICD-9 codes. What was once a 5-digit, numerical code is now an up to 7-digit code with letters and numbers. This means not only do the people have to learn a whole bevvy of new codes to enter into a health system for proper medication, billing, reimbursement, etc. but that the health systems themselves have to be reformatted to be able to take in this new structure. Think of it like the Y2K issue that plagued software companies in 1999 - suddenly changes have to be made to allow for more spaces and digits than before, along with different types of data being entered.

ICD-9 vs ICD-10 Structure

So with all of these changes that it requires - from billers and coders learning new codes, to changing computer programming to accept the new codes - it's not surprising that there have been a lot of delays for ICD-10 adoption. From October 1, 2011 to October 1, 2013 to its current deadline of October 1, 2014. Kicking the can down the road (pushing up against the 2015 beta testing debut of ICD-11 elsewhere in the world) was meant to provide enough time for everyone to get their collective ducks in a row. Large health systems would coordinate with vendors to ensure a proper rollout of the codes while coding staffs would have time to better understand the new code layout and its applications. The date was meant to be a hard stop to the constant pushbacks and everyone was supposed to be working together to make sure that October 1st, 2014 would not be a date that lived on in hospital administrator infamy.

At least...that was the plan. It's been recently reported multiple places - including at Healthcare IT News - that the prospect of meeting that deadline is slowly slipping away. That post includes the results of a survey that the Workgroup for Electronic Data Interchange conducted of healthcare providers found that a "whopping 80 percent of healthcare providers will fail to complete their business changes and begin testing before 2014." Furthermore, the survey revealed an interesting contradictory finding between what is impeding healthcare providers in being ready, and what is slowing healthcare vendors from being ready:

  • The top three barriers to providers' ICD-10 delays were cited to be: staffing, competing priorities and vendor readiness.
  • Vendors indicated their top three barriers were: customer readiness, competing priorities and other regulatory mandates.

Both sides say that the other isn't ready and therefore is slowing down the preparations needed to meet that rapidly approaching deadline. This is a bad situation where neither side feels like they are working together towards something, but instead at competing angles. They both acknowledge that there are a lot of issues that simultaneously require their time, attention, and resources ('competing priorities') - but does that not suggest they don't perceive the need to reprioritize ICD-10 conversion to a higher level? Perhaps, with all of the previous delayed start dates, both sides are hoping that yet another 11th hour delay will be instituted once more? Whatever the case, the picture is one of imminent failure unless there are changes in priorities, behavior and even thinking about how to best implement this change.

I've written before about this upcoming conversion and then, just as now, I was an advocate of using ICD-10's own structured design as the best way to ease conversion. As seen in that graphic above, each ICD-10's digits have a meaning - whether it's the area of the body, type of illness or injury, the severity, etc. So if each of those digits represent a question (where did it happen, how bad was it,), then supplying the answers can automatically populate that code. If physicians are presented with a synoptic report with clearly delineated, standardized fields, then they will produce structured data out of it. Synoptic pathology report with ICD-10 That is, itemized pieces of information that can then be used to derive a response.

At mTuitive, we've already incorporated this into our pathology product (though it obviously won't be live until that October 2014 date) as you can see on the right. By having users enter in information - important to note: information they are already capturing, albeit in unstructured manners like dictation or narrative reporting - by having these physicians fill out these fields and enter in this data, we are able to automatically derive however many codes we need to help them justify their procedures/work and to get reimbursement started faster. ICD-10 is already a piece of structured data - why not use that to our advantage? By encouraging more synoptic format reports and an increased use of initial structured data capture (as opposed to abstractors wading through long paragraphs), healthcare providers will be able to meet this new tsunami of change. That tidal wave is bearing down on every healthcare provider in the United States - the question is, when it hits - will you sink or will you swim?