Automated HCC enhancement software compliance

One of our physician advisers brought up an interesting perspective about the compliance of automated HCC optimization software that has recently been implemented in his facility.  This program utilizes artificial intelligence (AI) to scan the EMR for previously documented diagnoses that have not been documented for the current year and identifies conditions that have not been specifically documented but are indicated in lab results, medications and procedures. The suspected conditions are flagged and potential higher-weighted HCCs (including RAF scores) are presented to the physician to facilitate decision-making around the condition that maps to an HCC code.  

His major concerns are as follows:

1.       Is this a query?

2.       If it is a query, is it compliant?

3.       Is it leading?

4.       Should the RAF score be shown?

In my research I’ve found nothing that speaks directly to automated HCC optimization compliance, so I’ve attempted to apply other existing guidance. 

1.       Is this a query?

a.       I think it is because the AHIMA compliant query brief states that actions that “actively engage in educating providers in ways that could alter coded data” meets the definition of a query.  Even though the brief provides the examples of case managers and infection control clinicians, the software is educating the physician and the HCC is coded data.

b.       If the software was only presenting the concurrent HCC (without the RAF) based on the existing documentation it would not be a query.  It is the suggestion of additional possible diagnoses to document that result in a higher HCC that makes this a query. 

2.       If it is a query, is it compliant?

a.       I don’t believe it is because the AHIMA outpt CDI brief states, “whether inpatient or outpatient, all queries (no matter the format) should contain clinical indicators to support why the query was initiated”.  For HCC, it must have criteria validation of “MEAT” (monitor, evaluate, assess, treat) to capture.  The software isn’t providing this, only highlighting diagnoses.

3.       Is it leading?

a.       I believe it is leading because it isn’t in a compliant query format and it presents progressively higher level HCCs based on the suggested documentation of the undocumented diagnoses indicated by the record.

4.       Should the RAF score be shown?

a.       I’m not sure about this one.  Although AHIMA states “queries must not contain any information about their current impact on reimbursement”, CMS states the RAF indicates; (1) clinical meaningfulness, (2) medical expenditures/payments, (3) illness level, and (4) coding specificity.   https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Downloads/RTC-Dec2018.pdf  Do we eliminate 3 clinical benefits because of 1 indirect financial element?    

I believe the output of the software would be compliant if it was only showing the physician the current HCC (without the RAF). It would be a useful tool for a CDS who could provide the supporting clinical criteria and compliantly author the query.

Please weigh in on this discussion. 

Thanks,

Steve

Sign In or Register to comment.