How do CDI professionals review the medical record (and how has tech changed the game)?

By Brian Murphy

 

I’m always amazed at how little the core function of CDI gets discussed.

I’m talking about medical record review.

How it’s actually done, the process. For example:

Where do you begin when reviewing the record? What steps do you take, and in what order?

For inpatients, how long do you wait before beginning to review the record, how soon is too soon? 24 hours? 48?

What records do you prioritize (or do you have prioritization software that handles this)? What’s the most valuable piece of the record, where you often find the most helpful clues or opportunity? How much do you rely on the discharge summary?

Very few people seem to talk about the bread and butter of the job. In my ACDIS days I decided to do something about it. I worked with a half-dozen members of the advisory board to produce the white paper “How to Review a Medical Record,” published in Oct. 2018. See link below.

I just looked at it again after many years and think it’s still pretty good. But I wonder how much technology has changed the process.

The paper offers a nice visual workflow, with a typical example review progressing as follows:

  • ED/EMS notes
  • H&P
  • Consults
  • OR/procedures
  • Diagnostics/medications
  • Progress notes
  • Nursing/ancillary notes
  • Query (if needed)
  • Follow-up and focus

The paper then breaks down each of these into what the CDI might do with a given portion of the record. For example, in diagnostics and medications, it offers the following advice:

“Review your diagnostics together and trend results to save time. These provide a timeline comparison. Both normal and abnormal results may be relevant. Examples include laboratory studies, microbiology reports, radiology reports, EKGs, and echocardiograms. Do these results support, suggest, or rule out a diagnosis? Be sure to query regarding abnormal findings for conditions that are clinically significant (e.g., a clinician may not feel a sodium level of 134 supports a diagnosis of hyponatremia).”

A CDI should look for relevant abnormal trends, which might include:

  • WBC (infection or immunosuppression, cancer)
  • Na, K, Mg (hyper/hypo conditions)
  • Renal enzymes (creatinine, GFR trending/normalization for potential CKD or AKI/ATN)
  • INR, PT, PTT (bleeding, coagulopathy—a mortality risk factor)
  • H/H (anemia, cancers, acute blood loss)
  • Amylase/lipase (pancreatitis)
  • Protein/albumin/prealbumin (note these are NOT presently recommended as a sole indicator in the diagnosis of malnutrition, despite their use by some auditors)
  • LFTs (liver disease)
  • ABG/VBGs (acid-base disturbance and support of respiratory failure; PF ratios)

But now I wonder how much of this advice remains relevant. Are machines are elevating these lab values to the CDI, or just prompting the physician directly as he or she documents, circumventing the need for clinical review? Is chart review giving way to analytics and education?

How much is technology changing the game? Are you still wearing a clinical hat as you perform chart review, and if so, how?

 

Like what you’ve read here? Have questions?

Follow Brian on LinkedIn or send him an email at brian.murphy@norwood.com

 

Reference

Related News & Insights

This is what a good outpatient CDI program looks like

By Jason Jobes  This, right here.  You should be focused on risk score accuracy by simultaneously evaluating how to maximize revenue…

Read More read more

What outpatient CDI transformations are on your 2026 roadmap? Norwood has 45 in its playbook.

By Jason Jobes As 2025 draws to a close, a question: How are you feeling about your…

Read More read more