How Many Clicks Does It Take to Document a Patient’s Morbidities?
[post_content]
Disclaimer: This article has been automatically aggregated from
At the end of my day, it’s time to finally close all my notes, wrap things up after all those patient visits. I settle back down in my desk chair, and start to get things to the point where I can close (and bill) those notes. I’ve opened the encounters, gone through the chart, and confirmed their medications, allergies, past medical history, problem list, surgical history, social history, family history.
I remembered the dot phrase that inserts the ambient listening text generated from our interaction, then spend quite some time editing it to make it make sense and, in some ways, resemble the visit.
I enter the patient’s vital signs, confirm all the orders and referrals and imaging and labs we talked about are there and signed.
Then my assessment and plan.
Then finally comes the time to sign and bill the visit, decide the complexity of care I provided that day, the ICD-10 codes, and then I’m ready to sign.
But wait, there’s more!
The system prompts me that it’s time to go to the Best Practice Alerts section of the electronic health record (EHR) and verify the Hierarchical Condition Category (HCC) codes for the patient’s medical problems.
As most of you know, this is a system designed to tell the insurance companies, most often Medicare, how sick our patients are, how complex their care is, what risk modifiers need to be added to their claims to fairly reimburse us for the care we provide, during this visit and throughout the year.
Historically, at the time of billing, we only put in the codes for the things we are actively managing, the problems we are evaluating, new symptoms, new diagnoses, and any chronic conditions the patient has that were taken care of that day.
For some time now, we also needed to remind Medicare how sick these patients were, at least once a year. So, at the time of closing the note, we’re often presented with a list of codes, along with options about whether to add them to the diagnosis list for today, defer, or resolve the problem.
Often these are arcane, very complicated conditions, such that it’s not so easy to do without a deep dive into the chart.
Some subspecialist once entered this code for a patient, and now it persists, and it’s up to us to add it on to the visit (if still relevant) to make sure the insurance companies know about it. And often, these codes represent something someone else is managing, and we don’t include much information in our notes, don’t order many related tests, and often don’t prescribe any medications for these conditions.
And yet we, the primary care doctors, are the ones left responsible for telling the government how sick these patients truly are.
You’d think the government and the insurance companies would know by now.
Haven’t they seen those thrice weekly bills for hemodialysis sessions, the admission for a defibrillator placement, the ICU stay for respiratory failure leading to a stay in pulmonary rehab, that bone marrow transplant and chemotherapy?
Often, we address these problems with a brief summary in our notes. But many of these incredibly sick patients are also seeing their specialists, so isn’t it enough that the insurance companies know they’re getting care for these conditions somewhere else?
Under certain insurance plans, such as our accountable care organization for Medicare-attributed patients and others that use quality risk-based adjustment payments, we get a small portion of any shared savings as a bonus at the end of the year. But a portion of this is dependent on how many times we add these codes for conditions we’re not really managing.
Don’t we have enough to do?
Yes, these conditions are usually on the patient’s problem list, and sometimes we say things like, “follow-up with their nephrologist, their cardiologist, their pulmonologist, their oncologist.” But we’ve got so much else going on that this feels like lip service to add these codes on when we’re not really managing these issues today.
When these programs first rolled out, I remember someone telling us we needed to tell the government every year about every problem the patient had, including things such as amputations.
I remember one snide comment from the audience: “Did Medicare think that, from one year to the next, the amputation of their leg was no longer relevant?”
There’s got to be a better way to calculate a risk score, to figure out how sick these incredibly sick patients are, to more fairly reimburse us for care without making us go through a lot of silly clicking in the chart.
Here’s where I hope artificial intelligence (AI) can begin to help us.
Perhaps we teach the systems looking through our patient’s charts to search out medications, labs, imaging, test results, specialist visits, and so much else buried in claims data, to get a more accurate picture of how sick a patient really is.
Sure, it’s not always a lot of work. Sometimes it’s pretty easy to confirm that yes, they had breast cancer in the past, yes, they have major depression, yes, they’re still on hemodialysis.
But I bet someone much smarter than me can write some code with AI to churn through all this data and get a better picture of how complex these human beings are, and how hard it is to take care of them.
For now, we’ll keep doing this, clicking a bunch more boxes, to make sure our institution globally gets reimbursed appropriately.
In the future, with all the information stored in the EHR and pharmacy records and claims data and radiology servers and on and on, a more efficient truth about how sick our patients really are will come out. Without us having to do all these clicks.
for informational purposes only. We do not claim ownership, accuracy, or liability for the content provided. All rights belong to the original publisher.