Close Server: KOPWWW05 | Not logged in


ICD-10 Testing

Do the heavy lifting before ICD-10 end-to-end testing

In April 2013, The Centers for Medicare and Medicaid Services (CMS) made it clear on a national provider call that they would not be performing end-to-end testing; meaning that the agency won't be testing claims from providers. This comes as no surprise for two reasons.

One, testing claims under ICD-10 is difficult, challenging work. This is where the rubber meets the road, and who could blame CMS for wanting to avoid the really hard part of the initiative?

Second, CMS has no real incentive to test claims from providers. And I would argue, not testing is great for CMS. Why? Claims submitted with unspecified codes, or NOS, may not qualify as a CC or MCC, which means they pay less in reimbursement.

The onus is now on providers, payers, clearinghouses and other involved entities to use a testing model to validate accurate claims creation, transmission, adjudication, and payment.

Many might be tempted to take the path of least resistance. That is, use a GEMs-based predictive modeling or financial modeling tool to identify potential coding problems and their associated financial risk, and then submit those claims for testing.

Unfortunately, the old adage - no pain, no gain - really does apply here. If providers simply use predictive or financial modeling, a significant set of 'at-risk' claims will be overlooked and not tested before the transition.

The Cold, Hard Truth - Pull the Charts

After a series of coding projects across the country where hospitals pulled charts and our coding experts reviewed and analyzed thousands of charts, we've found a 30 percent coding error rate. The coding errors, more than likely, originate further upstream at the point of documentation.

Did the physician document properly and in line with new ICD-10 requirements, or does coding need to be realigned?

Either way, I often have to tell many a HIM executive and director the words they would rather not hear: You have to pull the chart, hand it to the coder and code it natively in ICD-10.

This is the only way a provider can be assured that the testing will be valid, and perhaps most importantly, that the HIM department is uncovering DRGs that will cause the most heartburn further down the road-when they don't get paid correctly.

Why can't a provider just rely on GEMs based predictive modeling or financial modeling tools? Consider that Medicare's goal in creating the GEMs grouper was to make the transition to ICD-10 revenue neutral for providers. Much time and effort was placed on the top DRGs to assure payment neutrality. However, during hospital projects where we've pulled charts and natively ten-coded thousands of claims, we have found that a DRG shift more often occurs in the less- utilized DRGs.

Only looking at the top 20 DRGs through GEMs type predictors ignores these less-utilized DRGs and puts hospital revenue at risk.

Best Practices for Native Ten Coding

Pulling a percentage of charts won't be easy, but it will ensure that a provider is discovering and rectifying its own at-risk charts, not simply verifying what CMS considers the DRGs most likely to be 'revenue neutral.'

We recommend that providers plan on the following process to successfully 'audit' their own financial risk and make the appropriate coding and documentation adjustments:

• Natively re-code a minimum of ten percent of inpatient annual discharges, if there is a focus on specific DRGs and physicians. If there isn't a focus, pull and natively ten code 20 percent of inpatient annual discharges.

• Make sure you have AHIMA-Approved ICD-10-CM/PCS Trainers natively re-code charts. We've found that staff without this level of ICD-10 expertise very likely won't be experienced enough to identify if the problem is documentation or the coding itself.

• Are you prepared for the time to natively re-code? On average, coders complete two to three charts per hour in ICD-9. Even our best coders need on average a solid hour to natively re-code one chart in ICD-10. Do you have the staff to focus on a project of this scope?

Now, once you've pulled ten or 20 percent of your charts, and have a dedicated, certified coding team ready to natively ten code, you are ready to categorize any and all factors that result in variances between the ICD-9 coded claim and the ICD-10 coded claim.

First, make sure the coding team has the original claim data with ICD-9 codes and resulting MS-DRG assignment along with complete medical records.

The coding team should then review and validate the original ICD-9 coding and resulting DRG assignment. This step is critical so that risk mitigation opportunities can be properly categorized as ICD-10-related versus ICD-9-related. For example, if a case was not completely and accurately coded in ICD-9 and the ten-coding process resulted in a different DRG assignment, a presumption that the DRG shift was directly related to ICD-10 may not be correct.

This distinction is important in terms of identifying CDI opportunities having both current and future financial implications versus those with only future implications.

Discovering - and Fixing - Variances

Individual cases and charts normally fall into two categories:
• A diagnosis or procedure code in ICD-10 could not be assigned at all because documentation was incomplete, imprecise, inconsistent, unclear, or illegible; or
• A diagnosis or procedure code in ICD-10 could not be assigned to its highest level of available specificity because documentation was incomplete, imprecise, inconsistent, unclear, or illegible.

In addition, you want any analysis to included a statistical assessment of overall assignment of non-specific codes, also known as NOS or "not otherwise specified" codes. Drilling deeper into these statistics, a provider is able to identify its top-10 conditions where documentation improvement could result in code assignment that is more specific (the primary reason for the transition to ICD-10) and, in some instances, grouper-significant and financially-significant.

Armed with case-specific and provider-specific examples of clinical documentation improvement opportunities and baseline performance benchmarks, the provider is then able to develop a documentation roadmap for consideration and approval by executive leadership.

Related Content

Once and Done

How technology has impacted medical transcription.

Now, Comes the End-to-End Testing

These natively ten-coded claims are now ready for testing with payers. Additionally, by sampling a percentage of charts and natively re-coding, it becomes crystal clear what gaps need to be filled in coder education and physician documentation.

Using predictive or financial modeling based on a GEMs crosswalk might seem like the path of least resistance, but ICD-10 is not a financial exercise - it's a documentation and coding exercise. Pull the charts and natively ten code them as soon as you can. A little pain now will mean much gain later.

Kerry Martin is CEO of VitalWare. Martin is an adept leader and entrepreneur who understands the complex realities of healthcare's overwhelming challenges in transitioning to ICD-10, including the need for improved clinical documentation practices to ensure accurate and complete claims submission.

Articles Archives


Email: *

Email, first name, comment and security code are required fields; all other fields are optional. With the exception of email, any information you provide will be displayed with your comment.

First * Last
Title Field Facility
City State

Comments: *
To prevent comment spam, please type the code you see below into the code field before submitting your comment. If you cannot read the numbers in the below image, reload the page to generate a new one.

Enter the security code below: *

Fields marked with an * are required.

View New Jobs, Events and More


Back to Top

© 2017 ADVANCE Healthcare, an Elite CE company