How Technology Can Help Ensure the Utmost Compliance in Your Risk Adjustment

Recently, there has been a lot of buzz around compliance in risk adjustment, stemming from the decision by the US Department of Justice (DOJ) to join the False Claims Act lawsuit against UnitedHealth Group (UHG) as well as a handful of other organizations.

The complaint filed by the DOJ alleges that the defendants knowingly conducted activities to inflate risk adjustment payments from Medicare. More specifically, the allegations stress the lack of a proactive redaction process – the identification and removal of codes previously submitted on claims that should not have been present.

Below are some highlights of the allegations in the lawsuit:

  • Upcoding, or submission of risk adjustment codes that the member does not have
  • Conducting one-way only chart reviews
  • Failure to act on correcting inaccurate codes
  • UHG executive compensation plans tied to increasing risk adjustment payments
  • Vendor contracts that incentivize higher risk capture
  • Use of Patient Assessment Forms (PAFs) only for members who are subject to risk adjustment, focusing only on revenue increase

If you are like most risk adjustment professionals I know, you are not actively upcoding or blatantly ignoring deletes that should be submitted. However, many will agree that they simply do not have the resources to conduct a two-way review for all members, encounters, and diagnoses that are submitted by providers. Some choose to do blind coding – reviewing member records without knowledge of any previously submitted codes – but when coding results are compared with claims data and differences are found, further investigation is necessary to reconcile the data prior to submission, which can deplete more resources.

The good news is that there are technology-enabled solutions that can help to achieve utmost compliance in an efficient and effective manner.

NLP-enabled coding – not just for upside

Contrary to the belief that coding platforms using natural language processing (NLP) help only to “find more,” the technology is just as useful for finding inaccurate codes. Rather than spending time to recode each individual member, NLP can enable a much more targeted and streamlined downside review. By processing medical records along with claims data, NLP can identify instances where codes found on claims lack substantiating evidence in the records. The system can flag these for coders to review, enabling the creation of deletes as necessary.

Getting to the source – provider education

There is also an opportunity to better engage and educate providers for improved documentation and coding. Technology can help to systematically identify areas where providers are unknowingly submitting non-compliant codes. When organizations use Health Fidelity’s coding platform, coders can flag instances where diagnoses are not sufficiently documented in the chart but where the patient likely has the condition (i.e., under-documented conditions). A report that contains the details of these instances can then be generated and sent to the providers for review. Making use of this type of coder feedback helps to ensure that providers are continually improving their documentation and coding. Further, these under-documented conditions can be confidently folded into a prospective process where they can be addressed proactively for the next payment year.

Using analytics to proactively address gaps

One of the many benefits of conducting chart reviews on a coding platform is the data that can be generated for analysis. Running analytics on deleted and under-documented codes helps to identify specific providers and/or sets of conditions that are commonly miscoded. This enables organizations to better target and prioritize education efforts to improve compliant documentation and coding.

It may be quite some time before we understand which parts of the complaint the courts will address, but many speculate that the government’s intervention is indicative of its intent to implement more stringent programs for ensuring compliant risk adjustment submissions. Investing in an NLP-enabled coding platform can help to ensure that your risk adjustment department is consistently producing the most complete and accurate risk profiles of your membership.