Measuring the Financial Impact of ICD-10 Conversion

When healthcare providers switch to ICD-10 in October 2014, how much will the conversion have cost them and how much will it cost them in the future? The time to find that out is now.

That’s the advice of Terri McCubbin, director of acute care consulting services for 3M Health Information Systems, who presented a session at this summer’s  HFMA ANI conference on “Assessing the Financial Impact of ICD-10.”

All healthcare providers will be required to migrate to International Classification of Diseases (ICD), Tenth Revision, on Oct. 1, 2014. McCubbin compares the migration to ICD-10 to when providers had to switch from CMS-DRGs to MS-DRGs and how they had been told the migration would be “budget neutral.” “We realized it was ‘budget neutral’ for CMS,” she says drily. “For individual providers, not so much,” provoking a chuckle from the audience.

This is an “unfunded regulatory event,” McCubbin says, and before beginning any ICD-10 project, a healthcare provider should determine how much the conversion will cost and how it will affect future revenues.

She recommends a two-phase process to determine the impact: Analyze how ICD-10 will affect the current claims process, and then identify the risks and opportunities resulting from that analysis.

How big is ICD-10? It’s big. Really big. There are 8 times as many individual codes as its predecessor, ICD-9, she says, but the first step to measuring the impact of ICD-10 is to determine how many codes a healthcare provider actually will have to use once the conversion is complete.

Finding the ‘One to Many’

What was once covered by one code in ICD-9 could now be two or more codes (in some cases, hundreds of individual codes) in ICD-10, McCubbin says. A healthcare provider must determine which frequently used codes will become multiple codes in ICD-10, and what documentation will be required to support those individual code sets.

For a real-world example, McCubbin presented data from the New York University Langone Medical Center, a 1,000 bed facility in New York City (NYU Langone’s vice president of finance and revenue cycle operations, Wesley Smith, had been scheduled to co-present with McCubbin, but he had been called for grand jury duty).

The NYU Langone team examined current inpatient claims to identify which commonly used ICD-9 codes translated into multiple ICD-10 codes and what documentation would be required to support the new code set. They then filtered the results as follows:

What percentage of claims that required conversion of 1 ICD-9 code to multiple ICD-10 codes were DRG (diagnosis-related group) codes, thereby directly related to reimbursements (NYU Langone found that almost 75 percent fell into this group);

Of the DRG-related code claims, which were claims with diagnosis codes (approximately 25 percent), which were claims with procedure codes (19 percent), which had both Dx and Proc codes (more than half);

Of the DRG-related claims, how did they break down by specialty. As NYU Langone specializes in orthopedics, it was not surprising to find that the majority of claims came from that group.

The next step is to determine whether the medical records behind those multiple-code claims had enough documentation to support ICD-10. “Do you have the documentation today that allowed you to select the highest granular code in ICD-10 or is that documentation missing,” says McCubbin. “You will find that some of that documentation needed for ICD-10 is already in that medical record but in ICD-9 there is no code to code that, so that is lost information.”

The NYU Langone team found that depending on specialty, anywhere between 13 to 21 percent of any given medical record required more documentation.  McCubbin’s firm 3M Health Information Systems, has conducted more than 200 reviews of healthcare providers and found that, on average, anywhere from 15 percent to 25 percent of inpatient claims were missing documentation for ICD-10; for outpatient, the range increases to 25 percent to 35 percent; and for the professional side of the provider mix, the range increases to 30 percent to 40 percent.

Once a provider has the data to project how many claims will be affected by the migration to ICD-10, the real work can begin — how much will it cost?

In Part 2 of our ICD-10 series, McCubbin helps NYU Langone quantify the risks and opportunities that arose during the ICD-10 conversion analysis.