Note from the Instructor: Real CDI is about solving communication misunderstandings, not just queries

By Allen Frady, RN-BSN, CCDS, CCS, CRC

When AHA Coding Clinic for ICD-10-CM, Second Quarter 2017, came out last year I found myself repeating a conversation I’d had with physicians many times before. Such is the life of a CDI professional—if it were one-and-done none of us would have job security.

On page 9 of that Coding Clinic, the AHA tells us that encephalopathy is not intrinsic to a stroke. “But how can that be?” A physician asked me during a Clinical Documentation Improvement Boot Camp at the time. “The stroke caused the encephalopathy!” He considered the encephalopathy to be intrinsic to the CVA because without the stroke, encephalopathy would have never occurred. This is in sharp contrast to how I used the term intrinsic, which was to say “a routine finding.” At that moment I realized the communication barrier stemmed simply from how each of us used the word “intrinsic.”

In the reporting world, a diagnosis is intrinsic if it can be expected that most or all patients with condition Y would also have condition X. If that is the case, then condition X is intrinsic to Y. In the coding world, we seek to report only as many codes as is needed to describe a condition. If X condition is always assumed with Y condition, then you report Y only and there is no need to report X as X would then be intrinsic to Y. In the case of encephalopathy with a CVA, this would imply that nearly every stroke patient can be expected to present with encephalopathy.

In other words, if someone said to you, “my friend had a stroke,” a response of “how are you doing with her encephalopathy” would actually be a reasonable question. However, this is not the case. There are many CVA victims who do not present with encephalopathy but rather weakness, facial drooping, or vision changes. Encephalopathy is not just assumed to be present for all stroke victims so it is certainly not “intrinsic” to a stroke.

To fully report the patient’s presentation, both codes are required. You cannot convey the patient status accurately with only the CVA code. I had been saying this for years. Fortunately, Coding Clinic finally caught up.

My real epiphany here was that in many cases, CDI professionals focus on the indicators but not the definitions and communication barriers. We often assume we just have a problematic physician and keep repeating the same non-productive behavior, sending the same query for the same documentation over and over. All the while, the real problem is a fundamental communication failure setting up a never ending road block.

Just this week, my co-workers and I faced similar confusion regarding the new code for a Type III myocardial infarction (MI). After what seemed like dozens of emails (each of us probably thinking the other was crazy), I realized that I was referencing the 2012 Universal Definition of MI which states that a Type III MI is simply one in which a patient expires without obtained biomarkers but with signs (patient complaints or electrocardiogram changes) of ischemia. The third definition states that such a fatal scenario should be assumed to be cardiac in nature and would be classified as a Type III MI.

My co-workers used a more “real world” definition they’d seen many cardiologists apply. They were looking at a Type III as only describing a situation when there was no previous cardiac history.

Now you can almost imagine the conversation. Me, insisting there is no reason why a patient who had an MI in the last 28 days can’t come in as a subsequent encounter and expire before biomarkers are drawn, and them saying no, that can’t be right because they had an MI in the past 28 days. Neither party in the conversation could figure out what the malfunction was with the other. We could have gone on for days and never figured it out. Fortunately, I work with a highly insightful and dedicated group of instructors and we considered this internal confusion a challenge and set to work to resolve it.

Many working in CDI, however, are not so fortunate.

This sort of thing happens all the time between CDI and coding staff, CDI and CDI colleagues, and especially between CDI specialists and providers. Perhaps worst of all, this sort of break down no doubt also occurs between Recovery Auditors/commercial auditors and facilities.

CDI professionals face an abundance of such clinical and coding communication challenges—it’s why we have jobs. For example, should you differentiate between AIDS related illnesses (illnesses which occur in people with healthy immune systems but which are known to be especially associated with compromise; e.g., simple pneumonia) versus AIDS defining illnesses (illnesses which are known to only occur in compromised patients; e.g., PCP pneumonia) when determining if you should follow the coding guidelines and jump to AIDS versus HIV status? I say the differentiation is important, but many people just go by the World Health Organization list from the 1980s, catching large swaths of patients with pneumonia, thrush, and cold sores as having AIDS.

How about sepsis? Is systemic inflammatory response syndrome (SIRS) the same thing as sepsis? What about immunocompromised people who are septic but do not present with SIRS? If you just go by SIRS, you may miss a potentially fatal diagnosis. If you’re a recovery auditor, you just denied a legitimate sepsis case. The reverse is also true for people expecting to see a positive sequential organ failure assessment (SOFA) score on every patient who is septic. To that I say: No. Every patient with sepsis will not always score in one of those very limited six body site categories.

These examples underscore one very important point: CDI professionals cannot succeed if focused only on clinical criteria without sitting down with clinical staff to make sure we use the same definitions and language required to determine reportable conditions. It’s at least equally important to our role as record reviewers and query writers. If we are really serious about documentation improvement educating physicians about how the data is reported and working with them closely to make sure we use the same language as they do in the first place is imperative.

Editor’s note: Frady is a CDI education specialist for BLR Healthcare in Middleton, Massachusetts. For more information regarding the October 6 webinar on the IPPS changes, click here. Contact him at AFrady@hcpro.com. For information regarding CDI Boot Camps visit http://hcmarketplace.com/clinical-doc-improvement- boot-camp-1. This article originally appeared on the ACDIS Blog.