Coding-Clinical Disconnect: Type 2 Myocardial Infarction Part II

CCD is when the typical and customary documentation of a condition does not line up with the available codes or the indexing.

Last week, I discussed the Type 2 MI issue. This week, I will address the higher-level problem of the “coding-clinical disconnect,” which I am going to refer to as CCD.

Healthcare providers (HCPs) are tasked with caring for patients. They practice clinical medicine and document to have a record of what transpired during the encounter for clinical communication. The documentation gets translated into ICD-10-CM (International Classification of Diseases, 10^th Revision, Clinical Modification) diagnosis codes by coding professionals according to specific rules (ICD-10-CM Official Guidelines for Coding and Reporting FY 2018). HCPs get precious little training about clinical documentation and know close to nothing about coding.

Coders know coding rules and the classification system, but any clinical knowledge they have comes from book learning or personal experience. They are not clinicians. The clinical documentation integrity professional (CDIP) practices at the intersection of the HCP and the coder. CDIPs who come from nursing have clinical experience, but it is not the exact same knowledge base as the physician or advanced practice provider.

CCD is when the typical and customary documentation of a condition does not line up with the available codes or the indexing. If you have been met with incredulity, irritation, or outright outrage from a physician, you most likely have bumped up against a CCD. The standard clinical documentation might even lead nowhere, like in the case of the common verbiage, “urosepsis.”

The advancement of medicine can also cause CCDs. When early adopters started using HFrEF (Heart Failure, reduced Ejection Fraction) and HFpEF (Heart Failure, preserved Ejection Fraction) instead of systolic and diastolic heart failure, it resulted in a CCD and innumerable queries. The elimination of “severe sepsis,” replaced with only “sepsis,” is wreaking havoc with our institutions. The fact that the ICD-10-CM Coordination and Maintenance Committee only meets bi-annually and makes coding edits on a yearly basis with a prolonged lag does not alleviate the CCDs created by new medical terminology and clinical practice.

When I was preparing ICD-10 modular education, I reviewed over a million diagnoses to see which conditions each subspecialty encountered. Two years later, I informed the section head of vascular surgery that over a single year, not one case of abdominal aortic rupture had been coded. He was irate; “Don’t you look at the codes I put on MY billing sheet?!” he exclaimed. Of course, the answer is no. Inpatient coders do not review professional fee billing claims for diagnosis codes.

This is the quintessential CCD. Coders do not code “I71.3 Abdominal aortic aneurysm, ruptured” unless the clinician uses the term “ruptured,” or “with rupture.” Clinicians often refer to this scenario as “leaking,” “hemorrhaging,” “bleeding,” or “torn,” none of which index to “ruptured.” The problem arises because the HCP is unaware how their documentation was ultimately translated. From that day forward, patients at my institution were noted to have “contained rupture” of their abdominal aortic aneurysms. I had closed that coding-clinical gap.

I also had another vascular surgeon trying to school me that the word “occlusion” implies totality. I don’t disagree, but the coding indexing does. If you want to capture the comorbid condition (CC) of “I70.92, Chronic total occlusion of artery of the extremities,” you must document “chronic occlusion,” “total occlusion,” or “chronic total occlusion.” For vasculopaths undergoing lower extremity bypass surgery, this is often the only CC.

Another common CCD is associated with the word “postoperative.” For clinicians, this is a temporal reference. In coding, it implies a complication unless mitigated, and may trigger a patient safety indicator. This results in numerous queries determining whether the HCP considers the condition unexpected or integral to the typical post-procedural course.

Querying sometimes results in additional exasperation. Occasionally a provider is queried, and his response is, “That is what I said!” It would be nice if the HCP could understand how the coder derives codes and what responsibility they have in the process. They do need to know that the coder may require specific verbiage to land on the correct code. The HCP is under the misunderstanding that an impression list may be interpreted as conveying linkage and causality. Listing “R hip fracture” and “osteoporosis” sequentially may imply osteoporotic hip fracture in the HCP’s mind, but that inference by the coder is not compliant.

Sometimes the problem is imprecise medical terminology. A Type 2 MI is, for all intents and purposes, a “non-ST-segment-elevation” myocardial infarction because there are no EKG changes, but the term, NSTEMI, is designated for a subclass of Type 1 MI. The adjective “toxic” may lead to a poisoning DRG when the provider doesn’t really intend to land there. A urinary tract infection associated with an indwelling Foley may not meet the stringent criteria for the entity known as, “catheter-associated UTI (CAUTI),” even if that seems semantically logical to the HCP.

As we in the CDI community discover coding-clinical disconnects, there are two actions we should take. The first temporizing measure is to try to cajole providers to document accordingly, until the CCD can be eliminated. The second definitive action is to submit it to the governing bodies through American Hospital Association Coding Clinic. If I still worked in a hospital setting, every time I had to query a provider for the word “rupture” I would submit it to Coding Clinic. There are enough unreasonable rules, regulations, and restrictions imposed on clinicians. Making them jump through coding hoops should not be one of them. A “leaking” abdominal aortic aneurysm is not lacerated; it is ruptured. HCPs are welcome to submit questions to CC themselves, but I would not expect many to take advantage of the opportunity.

We should never expect HCPs to change clinical practice to adapt to coding constraints. And since we are not permitted to lead clinicians in queries, we are going to have to resign ourselves to the indignation which results from the occasional CCD.

Keep fighting the good fight—think of it as job security!

Comment on this article