Sharing insights on assumptive coding
When I was a physician advisor, I used to offer a diabetic Charcot joint as an example of why we must be explicit with linkage.
Years ago, if a provider listed diabetes mellitus and a Charcot joint (neuropathic osteoarthropathy) as concomitant diagnoses without documenting clear linkage, the encoder would lead a coder to a syphilitic Charcot joint. I used to joke with my audiences that this would be quite a revelation to patients and their family members, in this day and age of medical record transparency.
Then a very astute neurologist friend of mine pointed out that I had taught her that the documentation must support the diagnoses. How could we arrive at a syphilitic Charcot joint as a diagnosis if the provider never mentioned syphilis?!
There are many conditions for which providers assume linkage would seem obvious. A query regarding whether I “think there is a relationship between the patient’s tobacco use and COPD,” or whether “the fact that the patient is anticoagulated is affecting their epistaxis” would make me roll my eyes in exasperation. My answer would be, “duh!”
Guideline I. A. 15, allowing presumptive linkage for code titles sporting “with” or “in,” was most welcome. Allow me some liberty in paraphrasing:
The word “with” or “in” (“in” first appeared in the 2018 Guidelines) should be interpreted to mean “associated with” or “due to” when it appears in a code title, the Alphabetic Index, or an instructional note in the Tabular List. The 2019 Guidelines added “either under a main term or subterm” after the Alphabetic Index phrase.
The classification presumes a causal relationship between the two conditions linked by these terms. The conditions are coded as related even in the absence of explicit linkage, unless the documentation clearly states the conditions are unrelated or when there is a superseding guideline demanding explicit linkage (like in sepsis-related organ dysfunction).
Diabetes mellitus is the poster child for this concept. There are many resultant conditions from the endocrinopathy that are overwhelmingly likely to be causally related. It would be the exception to the rule to see retinopathy in a diabetic that is unrelated to the diabetes. It likewise would be unusual to see peripheral neuropathy.
The causal relationship is established by the presence of “with” in the code title. If the condition were not related to the diabetes, the provider would have to make that explicit; otherwise, the relationship is assumed.
Inflammatory bowel diseases have assumptive linkage with rectal bleeding, intestinal obstruction, fistula, and abscess. The code title reads “with.” A perirectal fistula in Crohn’s disease is a result of Crohn’s, unless otherwise specified.
Here are some examples of “in.” If a patient has postpartum hemorrhoids, since the title of the code is “hemorrhoids in the puerperium,” you assume automatic linkage if hemorrhoids are diagnosed in the appropriate timeframe. Pericarditis in systemic lupus erythematosus is considered to be from lupus and coded to M32.12, unless specified as being due to something else.
If a patient has a diagnosis of arsenic poisoning and the manifestation of neuropathy, the indexing leads us to G62.2, Polyneuropathy due to other toxic agents, because the “in” is found in the Alphabetic indexing. The “due to” verbiage is listed in parentheses, making it non-essential.
I can’t take credit for having discovered this next one, which I think you will find very useful. A very astute friend and colleague of mine, Kathy Murchland, pointed this out to me and I am passing it along to you.
Prior to the adjustment in the guidelines, if a provider noted a positive culture in a patient with pneumonia, we would have to query for explicit linkage. We couldn’t presume that the organism was necessarily causal.
However, the Alphabetic Index lists “in (due to)” after the “hypostatic” entry, and then catalogs almost every organism you can imagine. It is our contention that the “with/in” convention renders querying unnecessary. If the provider documents pneumonia and a bacterial (fungal/viral) agent (e.g., pneumonia, cultures positive for E. coli.), these conditions can be inferred as being related; that is, the specific pneumonia would be the correct code. This may land the principal diagnosis into complex pneumonia as opposed to simple pneumonia. This is not gaming the system – it is appropriately representing the patient encounter.
I am sure there are other conditions that I haven’t stumbled across yet. I would love to hear other obscure examples of “with” and “in” linkage. If you have some to share, please email us at email@example.com.
Listen to Dr. Remer every Tuesday on Talk Ten Tuesday, 10 a.m. ET.