Let me start with a warning. Hardly a day goes by that we don’t hear a story about the wonders of artificial intelligence (AI). And maybe it’s because I am an old guy and remember the days of actually having to turn a dial to make a phone call that I have some skepticism, but last week showed what can happen when you rely too heavily on AI for answers.
On an online case management discussion group, someone asked if a stay of three days or longer at an inpatient rehabilitation facility (IRF) allowed the patient to access the Part A benefit to go to a skilled nursing facility (SNF) – and a participant quickly answered “no.” Now, I knew that was wrong, so I nicely pointed out that I disagreed, and asked for the source of that information.
The user posted a screenshot that showed the three sources the AI tool used to generate its answer. And when I went to each of those documents (two of which were Medicare pages about SNF care), not one included any mention of IRFs. The third one was a blog post from an ED physician that not only did not mention IRFs but also contained markedly inaccurate information about patient cost sharing, throwing the reliability of anything in that article into question.
So, what happened here? It appears the AI system determined that an IRF was not mentioned in any of its sources, so it just made up an answer and said “no, an IRF stay does not qualify the patient for a covered SNF stay.” And that is wrong. The Centers for Medicare & Medicaid Services (CMS) allows any inpatient hospital stay of three or more days to qualify the patient for the Part A SNF benefit, including IRF care, long-term acute-care hospitals, inpatient psychiatric hospitals, and of course, short-term acute-care hospitals, assuming all other conditions are met.
This is a perfect example of AI hallucinating to sound authoritative. So please, be careful out there. The adage “trust but verify” is doubly important if you want to trust a computer’s answer to a question.
Next, I want to talk a bit more about the Aetna Medicare Advantage policy I discussed last week, whereby they will approve all inpatient admissions, but only pay the inpatient rate if MCG inpatient criteria are met. First, I am gratified to have seen the anger that has been spreading across social media, and let’s hope CMS takes action. I will also hope that the rumors of other payers adopting this “legal but evil” policy are not true.
And while the policy is just abhorrent, there are some interesting nuances. First, for teaching hospitals, every inpatient admission brings with it a payment direct from Medicare for medical education. So, more inpatient admissions does mean more medical education money. Now, that won’t come close to the loss of inpatient revenue when those admissions are paid at the observation rate, but it’s something.
Next, with the increase in inpatient admissions comes the risk of more readmissions and a higher readmission penalty for the hospital. And we know Aetna doesn’t pay hospitals for readmissions, so it’s a double whammy. The first inpatient admission gets paid as an observation stay, then the second admission is not paid at all. That’s simply unacceptable.
But at the same time, that increase in readmissions also counts against Aetna; Medicare Advantage (MA) plans are scored on their readmission rates, and an increase in readmissions could adversely affect their Star rating and their bonus from CMS. So maybe we can all look at that as a positive to this policy.
And finally, there is the fact that this is a total misuse of MCG. Will Aetna only run MCG once, or if the patient remains hospitalized or worsens clinically, will they check again to see if inpatient criteria are met and approve the inpatient payment rate? And what about the MCG criteria that allows inpatient admission if the patient requires continued hospital care after a trial of observation? Will that be honored?
It is also interesting to note that as Aetna was getting all the attention, Cigna released a policy that will allow them to downgrade professional services claims with high level evaluation and management (E&M) codes. Their policy states, “Cigna may adjust the E/M CPT® code 99204- 99205, 99214-99215, 99244-99245 to a single level lower when the encounter criteria on the claim does not support the higher level E/M CPT® code reported.”
Really? Since when does a claim for a physician visit indicate the degree of medical decision-making or time spent on the visit? Those two elements are how a physician chooses their E&M code, and the claim simply includes the diagnosis codes, with no indication of the medical decision-making or records reviewed or tests ordered, etc. Just as Aetna’s policy has no basis as an ethical business practice, Cigna now joins Aetna, and UnitedHealthcare (UHC), which for years has been using their proprietary Optum ED Claim Analyzer to automatically downgrade facility ED visit claims, in putting corporate financial gain above all else.
EDITOR’S NOTE:
The opinions expressed in this article are solely those of the author and do not necessarily represent the views or opinions of MedLearn Media. We provide a platform for diverse perspectives, but the content and opinions expressed herein are the author’s own. MedLearn Media does not endorse or guarantee the accuracy of the information presented. Readers are encouraged to critically evaluate the content and conduct their own research. Any actions taken based on this article are at the reader’s own discretion.