By Brian Murphy
The new ACDIS/AHIMA Guidelines for Achieving a Compliant Query Practice are out. Why care about this new guidance?
The Guidelines are not law or regulation, but they are the standard that the two organizations recommend everyone follow.
And that means literally everyone involved in the query process—CDI and coding of course, but also case managers, quality improvement professionals, auditors, physician advisors, and technology vendors serving this space too.
Today artificial intelligence-driven products scan the health record and suggest that a patient might have for example a specific type of congestive heart failure, or sepsis, or pneumonia. These machines do this by “reading” documentation, test results, labs, and other clinical indicators, and either a) prioritize those records for the CDI to review manually, or b) send a direct clarification/nudge to the physician to document that diagnosis in the chart, without a human in the middle.
The new brief says, these AI nudges must follow the same rules as us humans.
We don’t believe in hampering technology and view it as a critical tool in driving human progress and health. Physicians are burned out, and machines that get them to the finish line faster seem to hold promise. BUT, these tools are not there yet, and some of the incentives are misaligned.
So, sorry T-1000, Sarah Connor is still in charge, for now.
Some other interesting items from the guidelines.
The brief clarifies the compliant use of multiple-choice queries, stating that there is no mandatory or minimum number of choices necessary to constitute a compliant multiple-choice query. Don’t be loading up your queries with clinically irrelevant choices just to meet some arbitrary number.
The guidelines also state that other answer options may be used including unknown, unable to determine, not clinically significant, integral to, unable to rule out, inherent to, or other similar wording. But these are not required (older versions seemed to imply a choice like this was always needed).
Per Laurie Prescott’s note linked below, apparently some organizations were interpreting the answer of “unable to determine” as an uncertain diagnosis. ACDIS/AHIMA strongly disagreed with this interpretation and wrote, “Unable to determine is defined as the provider being clinically unable to determine if a diagnosis or further clarity can be provided in the documentation. This terminology does not equate to an ‘unable to rule out’ option and does not represent an uncertain diagnosis.” The brief recommends avoiding using query questions/statements and answer options that indicate an uncertain diagnosis (which I know some disagree with, as an uncertain diagnosis can be coded in the IP setting).
Please note that the new brief may undergo further revision; the brief was subject to a comment period and the committee who wrote it will be reviewing those comments to determine if changes are needed.
The new guidelines can be found here: https://acdis.org/resources/guidelines-achieving-compliant-query-practice%E2%80%942022-update
Note from ACDIS Interim Director Laurie Prescott: https://acdis.org/articles/note-acdis-interim-director-new-query-practice-brief-open-review
Related News & Insights
“Bermuda triangle” of Medicare Advantage scrutiny signals alarm bells: Get your risk adjustment house in order
By Jason Jobes Is healthcare prepared to sustain the onslaught of changes to Medicare Advantage payments? How…
Cultivating the right mindset, values will take you far in HIM (and life)
By Ken Macklem, RHIT, CCS My career in healthcare started in the medical records department, filing paper….