Want to be a better CDI or coding professional? You must read the chart

By Brian Murphy
You’ve got to read. Whether it’s a blog post, a book, or in the case of the professions I serve, the patient’s chart.
Reading results in comprehension and eventually, understanding. It takes time, but it’s time well-spent.
I believe very strongly in this, which is why in my ACDIS days I resisted national productivity requirements for CDI chart review. It’s not about how many charts you review, it’s about the impact you make.
Yes, you can skim to meet productivity requirements … but what is the output?
Today we face more barriers than ever to deep reading. Digital distractions, notifications, slack. EHRs are built for billing, not to tell the patient’s story in a comprehensible way. In fact this post was prompted by me reading (pun intended) an article from MedPage Today, “How Complete Is Your Patient’s Electronic Health Record, Really?” Link below.
Here is a telling comment left on the article: “In my organization I’m the old guy and known for writing inclusive notes … unfortunately more and more the younger staff cannot be troubled to read the note.”
If a provider assumes the responsibility for documenting in the chart, other providers—and the CDI and coder—must assume the responsibility of reading it and acting upon the information.
Good CDI specialists and coders read the chart and assemble the final diagnoses or query opportunity from the entirety of the evidence, like a puzzle. The best love doing this.
In fact I wonder if reading comprehension isn’t the most important skill.
You might wonder (fairly), ‘what is the point, if an AI can do it and serve you up what you need to know?’ Even the Medscape author admits, “A discharge summary could probably be better generated by a smart AI that read the charts and assembled a comprehensive story about what just happened.”
… but that still means we have to read the discharge summary and act on it. And if point of care AI removes a lot of the basic queries (HF specificity, laterality, type, etc.) then deep reading becomes only more important to unravel the truly complex cases.
Machines aren’t ready to take over yet. They still need to be prompted, guided. They still hallucinate. They don’t ever say, “I don’t know.”
Which means we still need to read.
Besides, I don’t want to surrender to the machines.
So I’ll keep reading. I hope you do, too.
Reference
MedPage Today, “How Complete Is Your Patient’s Electronic Health Record, Really?” https://www.medpagetoday.com/opinion/patientcenteredmedicalhome/114169
Related News & Insights
Could AI increase clinician burnout? The answer is yes… and the solution is human beings
By Brian Murphy Could artificial intelligence (AI) increase physician burdens—and burnout? This headline from Healthcare Dive (see…
New sepsis aftercare code Z51.A: How would you assign to this clinical scenario?
By Joanne Wilson, RN, ADN, CCDS, Senior Director, Solutions We have a new ICD-10-CM code to bill…