Help keep this newsletter free by visiting our sponsor below:

As summer comes to an end, put yourself first by trying professional therapy done online! BetterHelp can match you to one of over 32,000 licensed therapists so you can do therapy on your schedule. No more getting on waitlists or driving to and from appointments! Try therapy with BetterHelp!

9 people died. But they shouldn’t have. That’s the story their metrics told.

9 is sad. But…9 is…just a number… What she did next was a bold move. But it worked.

On the next slide were pictures of 9 people.

Some were random google images. Some… were people in the audience…

The room went SILENT.

Until the silence was broken by the sniffles of people crying.

Aggressive? Sure. But it got their attention.

Could they sleep at night knowing more people might be dying under their care than expected?

Improving their documentation to improve their mortality score would not just be making a number look better, but assuring them that they ARE providing good, quality care.

This mortality score is known as RAMI.

What is RAMI?

Risk Adjusted Mortality Index.

It’s the “observed to expected mortality ratio” aka “Are more people dying than expected?” This can be broken down by hospital, service line, diagnosis, etc (and even individual provider, though the accuracy breaks down a bit there).

The calculation is “Observed Mortality” divided by “Expected Mortality.”

Less than 1 = Less people are dying than expected.

More than 1 = More people are dying than expected.

It’s the focus of many hospital administrations as we move towards Value Based Care and as hospital quality metrics are increasingly accessible by the public for comparison.

The “observed mortality” (The numerator / top number) is obviously impacted by clinical care and, especially for surgeons, patient selection. Clinicians should obviously always work on improving their care.

But often a RAMI >1 is not due to poor care, but rather poor documentation.

Why?

Because the "expected mortality” (The denominator / bottom number) is affected by documentation.

Why?

Because the “expected mortality” is based on how sick that patient is (and comparison to cohorts, which I won’t get into).

A healthy 35-year-old with pneumonia would not be expected to die.

But an 89-year-old with pneumonia who came from a SNF (following a stroke), has dementia, systolic heart failure, metastatic cancer, CKD and presented with dehydration, hyponatremia, AKI, respiratory failure, encephalopathy and required intubation on arrival…is MUCH more likely to die.

But the “calculator” doesn’t know that unless… it’s documented.

The kicker is that the conditions that impact the “expected mortality” are different for every single principal diagnosis (the overall reason for hospitalization), and it’s impossible to know them all.

So when CDI professionals tell you to “document completely” and query you for seemingly obvious stuff, RAMI is one of the major impacts.

Let’s look back at that previous example.

The Big stuff

My 89-year-old example has some “big diagnoses” that seem obvious…but can be, and often ARE, sabotaged or unclear based on documentation.

First, it’s not always obvious in the documentation WHERE the patient came from. Home? A facility? (Although this should be captured by your intake people)

2nd, I can’t tell you how many times I’ve discovered a reduced LVEF (systolic heart failure) when I reviewed a previously obtained echo, but heart failure was mentioned nowhere in the chart.

3rd, “History of” in coding language means “No longer exists.” Period. Clinicians frequently use “history of” when referring to malignancies which are actually STILL under active surveillance or even on continued therapy, such as tamoxifen for breast cancer. In such situations, that malignancy is still considered active. (I discussed “with” vs “past medical history of” in a previous issue: One important word in the HPI (Part 4 of 4) (beehiiv.com)

4th, chronic diagnoses such as CKD are often overlooked. For these purposes, only what is documented during THAT encounter can be captured AND they cannot be pulled from an auto-generated problem list as it must be clear they’re active (helpful if the documentation suggests they’re being monitored, evaluated, assessed or treated).

Finally, is it just a “simple pneumonia” or… aspiration pneumonia? Or a suspected gram-negative pneumonia? I discussed this in a previous issue: What should you call that pneumonia? (beehiiv.com)

The Little Stuff

The little stuff aint so little...

But clinicians often forget to include these in their documentation as they are seemingly “inconsequential”. These include diagnoses such as 1st degree AV block, PVCs, dehydration, hypokalemia, hyponatremia, thrombocytopenia, pulmonary hypertension, chronic fatigue, age-related physical debility, obesity, malnutrition...to name a few.

And yes, calling it “hypokalemia” and not “low potassium” matters. You must be specific. Coders cannot assume anything (in some cases, it IS just a lab abnormality - such is the case with “high potassium” due to a blood specimen with hemolysis)

As I said, the list of what impacts the expected mortality is different for every principal diagnosis but the “big three” that I see affecting many are:

1. Heart Failure

2. Malnutrition

3. Dehydration / Electrolyte Abnormalities.

I could go on, but for the sake of brevity…I think you get the point.

When I receive a patient, I HUNT in the chart for every diagnoses / chronic condition that I can find. First because I don’t want any surprises while caring for the patient, but second because I’m aware of the impacts on RAMI, etc.

As always, this is not about adding diagnoses that don’t exist but accurately and completely capturing the diagnoses that DO exist…

Because if you were admitted to your hospital under your service line…

would you be one of the 9 pictures?

That’s all for now. I hope that was helpful.

Please feel free to reach out and ask questions as they help inspire future issues!

Cheers,

Robert

When you’re ready, there are two ways I can help you:

1. If you’re a resident, med student, NP, PA or fellow and want to know how to write your notes to get home faster, help you stay organized, impress your attendings, and prepare you for post-graduation with an affordable 1.5 hour video course:

   → Check out The Resident Guide to Clinical Documentation

2. Promote yourself, your business, your conference, etc to 3,974+ subscribers by sponsoring this newsletter.

If you were forwarded this newsletter and would like to subscribe: