This post originally appeared on
A guest column by the American College of Physicians, exclusive to KevinMD.com.
This will not be another article focused on the Medicare Access and CHIP Reauthorization Act (MACRA) and its Merit-Based Incentive Payment System (MIPS). Everyone else is writing about that. (But I can't promise that I won't mention MIPS at all.)
Hopefully, by now, most of you know what MACRA and MIPS are, but how many of you have heard of an HCC? HCCs are not new. They already play a role in determining how some of us get paid. They will play a greater role under MIPS and alternative payment models such as accountable care organizations (ACOs).
"HCC" stands for hierarchical condition categories. It is the methodology that CMS uses for risk adjustment.
As a fee-for-service private practice internist for 26 years, I have paid attention to diagnosis codes as much as I needed to in order to get paid for my evaluation and management (E/M) services or procedures, or to avoid a call from a lab or patient when a test was not covered because of "medical necessity." When Medicare required greater specificity in coding, such as reporting whether a patient had type 1 or type 2 diabetes instead of just "diabetes," I coded to the level of specificity necessary but not beyond. It really didn't make a difference if I also coded for nephropathy or retinopathy if either or both were present. Payment for a level 3 or 4 visit was based on using diagnosis codes that documented "medical necessity," not on the level of detail of the ICD codes or how many were submitted.
In the past, while the specificity and completeness of my ICD coding didn't affect my bottom line, it affected others'. Medicare Advantage (MA) plans requested records or sent nurses to my office to review charts to "mine" for diagnoses that weren't captured in claims. The reason, I was told, was that by capturing these additional codes, the MA plan could get paid a higher capitated rate. But it didn't affect what I got paid for a 99213 or 99214 visit. Or so I thought.
Perhaps it did affect my fee-for-service payment, albeit indirectly and imperceptibly. While we were paid fee-for-service by the MA plans, their financial health could affect their ability to adjust their payment schedule. Call it "trickle-down economics" for the fee-for-service physicians, which like its namesake was more of a theory than a reality.
Fast forward to the present. Now, I'm in a few "shared savings" contracts with MA plans as well as CMS, as a member of an accountable care organization. While my E/M payments are not adjusted based on ICD-10 diagnoses, my choice of codes makes a big difference in how big the shared savings pool is, since it is based on risk adjustment using HCC.
For example, rightly or wrongly, one might be in the habit of using the ICD-10 code 11.9 for all their diabetic patients, and that might be enough to get you paid for your E/M visits. But if you're in a shared savings or capitated program, based on that code, CMS allocates as much payment to the MA plan or ACO for a diabetic with nephropathy as it would to an otherwise healthy type 2 diabetic who is diet-controlled. So instead, you should use the more specific codes, such as E11.21 or E11.22 and include in your claim the ICD-10 codes for the type of nephropathy (proteinuria, chronic kidney disease stage, etc.). This moves the patient into a higher-risk HCC.
What if you're not in a risk-sharing payment structure? Does any of this matter? Absolutely. One "feature" of MIPS is a "value-based" payment adjustment that is calculated by comparing your cost for each beneficiary to the expected per-capita cost determined by the risk of your patients. HCC will be the basis for determining that risk adjustment and how you code your claim will determine the HCC for your patients.
Using risk adjustment in fee-for-service Medicare is not a new phenomenon. It's being used now, in the "value based payment modifier" (VBPM) that was established in 2013 and is being rolled out gradually (and will become part of MIPS). The VBPM is an adjustment to traditional Medicare payments based on the risk-adjusted cost of each physician's care.
If that doesn't seem bad enough, the risk adjustment is based on claims submitted 2 years earlier, in most cases. In other words, we should have been coding with HCC in mind a while ago. But it's not too late.
The CMS website has information on risk adjustment using HCC. Start submitting all the appropriate codes applicable to your patients whose conditions are subject to HCC risk adjustment, as in the example above. If they're on insulin, include Z79.4 as well. You're not limited to four diagnoses per claim any longer -- you can submit up to 12 in an electronic claim.
Until you get into the habit of thinking about risk adjustment, it seems like another hassle. In my ideal world, supercomputers would use "big data" to figure out how sick a patient is from claims, test results, and EHR information, but I don't see that happening anytime soon. Therefore, we are stuck with using ICD-10 codes in claims to risk adjust. However, if these new methods of paying physicians deliver on their promises, they should be a step forward in addressing the perennial complaint that the system pays no more for treating complex patients than it does for healthier patients.
Yul Ejnes is an internal medicine physician and a past chair, board of regents, American College of Physicians. His statements do not necessarily reflect official policies of ACP. This post appeared on KevinMD.com.