The Risk Adjustment Data Validation audits will be underway soon for CY 2007 payments, so 2006 dates of service. About 30 Medicare Advantage plans will be notified this month that they have been selected for audit. CMS plans to sample at least 200 patients per plan to make sure their HCC codes are supported by valid documentation. Plans should select the “one best medical record” that supports that HCC code.
The CMS risk model gives credit for any type of vertebral fracture, whether it be traumatic (ICD-9 805.x or 806.x) or simply a pathologic, compression fracture – 733.13 – due to an underlying condition such as osteoporosis. Any type of vertebral fracture maps to HCC 157.
CMS frowns upon capturing ICD-9 codes from the admission History and Physical. The diagnoses, especially those addressing the acute reason for admission, may be potentially erroneous and turn out later to not be valid. Be wary of diagnoses from the ER, for example, that may turn out to be “working diagnoses” that are not corroborated by further evidence during the stay. Well-documented progress notes, either from the attending or consultant, and the discharge summaries are the best places to look for validatable, well-documented ICD-9 codes in the HCC model. Nurses notes, radiology reports, and lab values alone cannot be used by an auditor to “infer” a diagnosis that has not been corroborated by the provider. If you decide to capture data from the H&P, then it must be coded according to the Outpatient Coding Guidelines.
January 31, 2009 is the absolute final deadline for 2007 dates of service (1/1/07-12/31/07).
March 6, 2009 aka “March Sweeps” is the initial submission deadline for 2008 dates of service (1/1/08-12/31/08).
The CMS-HCC model is not limited to chronic diseases. There are a number of traumatic conditions in the model, including skull fractures – HCC 154, and HCC 155. As well as third-degree burns, HCC 150. For a burn to be included, it must be at least 10% surface area AND 10% third-degree burn.
HCC 148, HCC 149 in the skin disease group is a good example of how it may be helpful to know the HCC hierarchy. In this case, Chronic ulcers HCC 149 is trumped by HCC 148 – pressure ulcers (changed in 2009 from “decubitus” ulcers). So, a patient with a well-documented pressure ulcer already has the most-valued HCC code in the Skin Disease Group hierarchy, so it would not be efficient to look for or code for HCC 149 as well as it would be canceled out by the HCC 148 during that date of service data collection period.
Addendum (4/21/09) HCC 150 burns can be co-reported with HCC 148 or HCC 149 ulcers.
A diagnosis of Acute or Unspecified Hepatitis B or Hepatitis C has no HCC value. Thus far, only chronic hepatitis B and C have been mapped to hierarchical condition categories in the current version of the CMS-HCC payment model. I keep seeing merely “Hep B”or “Hep C” in provider progress notes with no mention of whether acute or chronic. While, as an auditor with the chance to view the entire chart, I might be able to infer whether it is acute or chronic, a CMS validator with just the “one best record” to audit, will not be able to make that inference and thus will likely not be able to substantiate the ICD-9 code. Make sure your providers are well-versed on the criteria for diagnosing chronic hepatitis B or C and to document their findings appropriately.
Faxing in records for central review dramatically reduces travel time and costs for your auditors. The best benefit may well be the creation of a digital file as most fax services can convert incoming faxes automatically to digital files such as .pdf. One trick is to name the .pdfs with a standardized system so that files can be retreived in case of a CMS audit. Plans can name digital files by a unique plan member ID number or can tag .pdf files with searchable Text Boxes that might contain redundant searchable identifiers such as Member ID #, HIC, DOB, Name. Current Adobe Acrobat (9) software can quickly search through a folder of .pdfs that have been tagged with Text Boxes which will enable your plan to quickly find records for validation without having to bother the provider again.
One of the biggest lessons I have learned from doing Medicare risk-related chart reviews is that doctors and the people who create ICD-9 terms often do not use the same common terminology. Doctors so often use synonyms that you might only see in the “extra descriptions” of the ICD-9 codes. A good example is “Unstable Angina”, which we use all the time – and document as well, is buried in the extra description of ICD-9 code “411.1 Intermediate Coronary Syndrome” which is mapped to HCC 82 and RxHCC 92. A majority of doctors and providers are more likely to document “unstable angina”, so be sure that your coders doing chart audits are aware that many risk-related ICD-9 codes are at times better found by using the Volume 2 alphabetic index.
Here are the pages from the CMS Training Guide that demonstrate: CMS Audit Request that shows how CMS will ask you for charts to validate your reported HCCs. You’ll get a list a patients by their HIC number, a list of captured HCCs and all the reported ICD9s that map to each HCC. The plan’s task will be to find the one best medical record that supports each HCC. So, just finding one record that clearly validates any one of the matching reported ICD9s for a reported HCC will be sufficient to justify that HCC. Even better, even if an ICD9 has not been reported yet, but you come across it in your own audits, you can still use that valid, unreported ICD9 to substantiate that HCC.
Should your auditors find that erroneous ICD-9 data had been submitted to RAPS, your IT department will have to resubmit the same “diagnosis cluster” to RAPS this time with a “D” or delete code to remove it from the RAPS database. See deletions.
A diagnosis cluster for a patient is the ICD9, From Date, Through Date, Provider Type.
The RAPS database stores “unique” diagnosis clusters, i.e. those that are not duplicates. Duplicate diagnosis clusters trigger the 502 error “Already stored in RAPS”.
