Q&A: Clearing up confusion about incoorporating consultants’ reports

Q: I’ve been confused recently regarding coding from the pathology or radiology reports for specificity. It seems that in recent years, (I’ve been coding for 25years) a new interpretation of the coding guidelines has come about. I was taught that as long as a physician with direct patient care documents a diagnosis (i.e., ovarian cyst 620.2), that it is appropriate to use the path report to be as specific as possible to code the type of cyst (follicular 620.0, etc.). Or another example the physician documents “uterine fibroid” and pathology report specifies it as “intramural,” so I would code 218.1 rather than just 218.9.

However, it seems now AHA Coding Clinic for ICD-9-CM, Second Quarter, 2013, refers to a question submitted regarding coding from a radiology report to specify the exact site of a fracture is being taken literally and some coders suggest that now we can only pull information from radiology/pathology report if the physician is treating a fracture.

Have I become “set in my ways” and mistaken the coding guidelines all these years?

Of course when the physician only documents “mass” and the pathology report specifies a diagnosis it would not be appropriate to use the pathology report without querying the physician, this is a long standing rule as well, and to my knowledge not changed.

Would you be able to publish your answer and help an “old” coder?

A: I think the core issue is differentiating “clarification” of a diagnosis from a “new” diagnosis. Here is why I make that distinction: consider the diagnosis of “renal insufficiency.”  Many with a clinical background would consider “acute renal failure” clarification of the “type” of renal insufficiency; however, the coding industry has determined that these are two “conflicting” diagnoses so the provider must be queried to clarify if the renal insufficiency can be further clarified as “acute renal failure “before this diagnosis can be assigned by the coder. However, if the provider had documented “renal abnormality” or “renal dysfunction” there would not be a conflict and the “acute renal failure” from a consultant could be coded without further attending clarification. 

Coding is more scrutinized than ever before because so much value is placed on coded data these days. Not only is direct reimbursement tied to it, but we also have indirect reimbursement issues based on CMS’ hospital value based purchasing (HVBP) as well as quality metrics that can affect a hospital’s reputation through publicly reported data that is easily accessible on the internet.

Furthermore, there is so much variation in coding because so much is subjective and subject to the interpretation of an external auditor. According to coding guidelines and UHDDS guidelines, how many times does a code that is supported by a clinical indicator and meets the definition of a secondary or principal diagnosis need to be documented in the medical record to be coded? I ask this question of my CDI Boot Camp students all the time and receive a variety of responses. Some say daily, some say in the discharge summary, some say at least three times, some say it depends on if it is a CC or MCC.

This is an example of the influence of the recovery auditors. Denials are not always based on credible sources (or even established coding guidelines or Coding Clinics). The correct answer is there is not specific guidance, but it is implied that a diagnosis that meets the definition of a principal or secondary diagnosis that is supported by a clinical indicator only has to be documented once.

So we have Coding Clinics that state we can’t code from pathology:

Coding Clinic, Third Quarter, 2008 p. 11

“…This ensures that the documentation and the codes reported are consistent with the attending physician’s interpretation since he/she is responsible for the clinical management of the case.  It is the responsibility of the attending physician to gather and collate all of the findings from the consultants and other providers involved in the care of the patient. . . Although the pathologist provides a written interpretation of a tissue biopsy, this is not equivalent to the attending physician’s medical diagnosis based on the patient’s complete clinical picture.” (Emphasis added.)

If the attending already made the diagnosis and the pathologist just adds clarification is that “coding” from pathology report? It seems reasonable to use another provider’s documentation to further clarify the attending’s diagnosis. However, this has become less tolerated in today’s climate because it isn’t clear as to what is “clarifying” and what is “diagnosing.” Hence examples like the one I mentioned above, with renal insufficiency vs. renal failure is consider a conflict, while coders are generally okay with using consulting provider documentation to clarify the type of anemia if the attending only documents “anemia.”

In my opinion coding guidelines and Coding Clinics are often interpreted literally by auditors even though they have traditionally been extrapolated to other situations by the coding profession.

The Coding Clinic you reference regarding fractures was recently revised for ICD-10-CM in First Quarter, 2013. To paraphrase the question was:

In ICD-9-CM if the record described a fracture of the leg and the radiology report identifies the specific site of the leg, we are allowed to code the more specific site. Will this also be true in ICD-10-CM?

Response: The same advice would apply to more specific coding in ICD-10-CM. If the x-ray report provides additional information regarding the site for a condition that the provider has already diagnosed, it would be appropriate to assign a code to identify the specificity that is documented in the x-ray report.

Due to this type of response, I would agree that we can only use radiographical imaging to provide the site of a diagnosed condition. I don’t think we can extrapolate this to any/all types of imaging like coding the site of a stroke from a CT scan. It would be great if there was a Coding Clinic or coding guideline that would allow specificity for a diagnosed condition to be obtained from any type of imaging since the volume of queries in ICD-10-CM/PCS is going to become overwhelming due to its anatomical specificity. We are more reliant on diagnostics than ever before in healthcare, yet, coding guidelines do not appear to reflect this reality.

Additionally, the expectation that the attending provider always reviews and comments on the findings of consulting providers is also somewhat misguided as there is so much documentation in the medical record from so many consultations that the attending physician could spend their whole day reading progress notes rather than caring for patients. Even the concept of an attending provider is more complex these days as providers are often part of a service and rotate so the attending may, in fact, have the least familiarity with the patient.

In summary, I agree that the “interpretation” of coding guidance has become more literal as auditors often interpret our vague guidance to their advantage.

Hope this helps!