HDG #024: Two Data Groupers Healthcare Analysts Shouldn’t Be Without

Read time: 5 minutes

Have you ever had to work with diagnoses and CPT/HCPCS procedure codes — two hallmarks at the core of healthcare analytics? Healthcare analysts will often get an oddball request like “how many diabetics [diagnosis] living in this county and had a hernia repair [procedure] this year?”

We might dive right in and start combing through lists for diagnoses that we’ll need to pull, plucking out the ones that look or sound like they’re related to the things in question, searching for diagnosis descriptions that contain “diabetes, diabet, diab” etc.

If we’re lucky, we’ll already have a vetted codeset available to us that identifies the codes in this grouped fashion, have a medical biller around to consult, or perhaps a clinician champion available to validate our assumptions.

But in my career, having a reliable crosswalk laying around has only been the case about 50% of the time, so I’ve had to be industrious.

I’ve found 2 datasets to be extremely helpful over the last decade as a healthcare analyst, which I’m sharing here.

Diagnosis Crosswalk:

Ahh, the good ol’ diagnosis: asthma, diabetes, headache, pecked by a turkey, hypertension. Loosely speaking, a primary diagnoses code represents the reason that your doctor documents for your visit, or the condition that they have determined you have, like diabetes. Your doctor can also capture additional (secondary or tertiary) diagnoses that impact the primary or factors that were also presented, like a comorbidity of asthma, or having been pecked by a turkey and rushed to the emergency department (ED).

Sounds simple enough — I mean, how many disease states are there? Well, in October of 2015, when we switched from ICD-9 to ICD-10 diagnoses, we went from having about 14,000 different diagnosis codes to nearly 70,000 more robust, descriptive, and nuanced diagnosis codes. For example, under ICD-9, there were almost 70 different codes for diabetes mellitus with or without complications. Under ICD-10, diabetes mellitus has over 475 different codes, all indicating something slightly different, but still under the diabetes mellitus with or without complications.

So what is an analyst to do when they need to pull any patient with a history of diabetes?

Enter the AHRQ’s CCS for Diagnosis Grouper.

You can read about what the CCS is and all of its nuances here, but in a nutshell, this is a robust diagnosis to clinical condition or disease crosswalk that groups all like-diagnosis codes into higher level categories that represent diseases as us layfolk think about them. It is also good to use as a simple diagnosis code to description crosswalk.

Check out a sample of the crosswalk for diabetes mellitus with complications. The ICD-10-CM code is the diagnosis code (with no “.”), the ICD-10-CM Code Description is the formal meaning of that diagnosis code, and the CCSR Category Description is the field that groups them all into the same “Diabetes mellitus with complication” group:

As with everything, some exclusions or special codes may need to further honed down based on your needs. You can see that this includes just about every diabetes diagnosis under the sun, by design. As always, we still need to understand the request, the reason for the request, the underlying data, and what we are looking at. All of that said, I have found this crosswalk to be an invaluable place to start.

Download the ICD-10 diagnoses crosswalk Excel list here.

. . .

CPT/HCPCs Grouper:

If the difference between CPTs and HCPCS still confuses you, check out my Healthcare Glossary & Definitions.

If you’re trying to identify all of the Inguinal and femoral hernia repairs that happened, check out the AHRQ’s CCS for Services and Procedures Tool. This crosswalk is still useful, but requires a bit of cleanup, and doesn’t get as granular (in my opinion) as the diagnosis crosswalk, but it’s still a good start for some things.

I have often hoped this crosswalk would be improved upon by adding another layer of grouping beyond the CCS procedure group, such as the more specific type of procedure. As an example, Inguinal and femoral hernia repairs could further be broken out into laparoscopic hernia repair vs. open hernia repair.

Download the 2020 procedure code crosswalk zip file (which includes the CSV that you want called CCS_services_procedures_v2020–1.csv) here.

I cannot stress enough that this is a very high-level grouper. Many companies charge a lot of money for good groupers because of their value, complexity, and how often coding guidelines change. No one CPT/HCPCS grouper or crosswalk may be the same as another, though general categories tend to be generally accepted (like evaluation and management or E&M).

If you’re going deeper into service line or procedure-level analysis, a next step would be going beyond the CCS grouper above and into more specific services. Consider the chart below from the AAPC. It is still very high-level, and some of these categories don’t mirror the AHRQ grouper, even for some of the same HCPCS codes. And some do:

There is no one-size-fits-all in healthcare.

Even the tools above have their shortcomings and limitations. Still, I have found them to be useful on numerous occasions, updated regularly (they just added a new CCS diagnosis group for COVID-19) and generally accepted/used somewhat commonly.

As you begin working with organizations that may purchase more robust or even different service groupers from commercial companies like Milliman or Optum, or who have developed their own internal crosswalks, these crosswalks may become obsolete for you (or potentially even incongruent to how your organization prefers to analyze diagnoses and service procedures). If that’s the case — count your blessings that you have a reliable organization-wide definition to lean on.

. . .

Actionable Idea of the Week:

Simplicity rules this week: familiarize with these groupers, keep them in your back pocket, and analyze on!

. . .

See you next week!

-Stefany

2 more ways I can help you:

Like my content?

If you want to learn more about health data quickly so you can market yourself, your company, or just plain level up your health data game, I recommend subscribing to this newsletter and checking out my free Guides. Courses and more resources are coming soon, so check back often!

Want to work together?

I work with healthtech startups, investors, and other health organizations who want to transform healthcare and achieve more tangible, equitable outcomes by using data in new ways. Book some time with me to talk health data, advice for healthtech startup and investment, team training+workshops, event speaking, or Fractional Analytics Officer support + analytics advisory.

Follow me on LinkedIn, Twitter, and Medium to stay up-to-date on free resources and announcements!