December 7, 2009

Myths and Misconceptions in ICD-9 Coding
By Judy Sturgeon, CCS
For The Record
Vol. 21 No. 23 P. 6

Nearly anyone who has ever used e-mail has received some type of digital urban legend. You may have gotten a warning about infected needles hiding in gas pumps or that a super virus will erase your computer if you close your left eye while typing (mine is open just in case). Most people know to check out the veracity of these stories on Snopes.com or other urban legend-debunking Web sites. But where do you turn to uncover the truth behind coding myths and legends?

Some extraordinarily persistent coding myths are passed on from teacher to student to the next student. Unless someone knows of a coding site equivalent to Snopes, it can be quite difficult to sift coding fact from fiction.

If coders have access to software containing all of the American Hospital Association’s Coding Clinics, it is easier to look up official directions and examples that have been published over the years. The Centers for Medicare & Medicaid Services (CMS) Web site for the official ICD-9-CM coding guidelines is updated every year or two as well. But unless you read every direction that gets published and/or have reason to disbelieve the way you were taught, it is unlikely you will even think to reference the resources at hand.

Let’s examine some of the more common myths and misconceptions of ICD-9-CM coding.

• Coding from resident physician documentation: There are strict rules for physicians at teaching hospitals regarding what they can and cannot use as documentation to support billing for their services. Faculty must indicate how much personal involvement they had in the resident physician’s patient evaluation and whether they agree or need to modify the determination. These factors, in turn, determine whether the faculty physician can bill for the services performed.

It is likely that this is the source of the common misconception that the facility coder cannot use a resident’s diagnosis unless the attending physician also documents the same diagnosis. Coding Clinic has provided clarification on this topic and validated that a resident physician is a qualified provider for coding purposes—as are faculty physicians and even physician’s assistants who are practicing within their scope of license. Warning: There should not be contradiction from a faculty physician regarding the same problem or diagnosis. If the attending has updated the information, the coder can likely update the code as well. In cases of unclear documentation, as always, one should query the physician for further clarification.
Also of note is the issue that the facility itself could have bylaws, policies, or procedures that prohibit the practice internally, making it advisable to check those hospital sources to learn whether there exists any contradicting directions. ICD-9-CM, however, does allow the practice with those few caveats.

• Physicians’ documentation of principal diagnosis on the dictated discharge summary: While it is significantly faster to code directly from the dictated discharge summary—especially if the document format includes the term “principal diagnosis”—it still is not a valid method for coding an inpatient hospital chart. While it is critical to consider this documentation source, it is the coder who must correctly assign the codes, sequencing, and the final diagnosis-related group (DRG).

For example, take a discharge summary that reads “Principal diagnosis: pneumonia. Secondary diagnosis: overdose, aspiration, respiratory failure.” Additional documentation elaborates that the patient overdosed on his medication intentionally, aspirated, developed respiratory failure, and was placed on a ventilator. Then he developed a superimposed Pseudomonas pneumonia due to the ventilator.

The information on the discharge summary would never meet criteria for coding and sequencing for DRG assignment. The correct version would be to assign a code for the poisoning as principal diagnosis and then assign codes for the respiratory failure, the aspiration pneumonia, the ventilator-associated pneumonia, a code for the Pseudomonas bacteria itself, and E codes for the poisoning agents and intent. (See the Official Coding Guidelines for ICD-9-CM regarding coding for ventilator-associated pneumonia, Pseudomonas, and E codes and the requirement to sequence the poisoning first in this scenario.)

• Possible/probable/not-ruled-out diagnoses: Inpatient coders should be familiar with another official coding guideline. If, at discharge, a diagnosis is documented by the provider as possible, probable, or other similar terminology, it should be coded as if it exists. At this year’s national AHIMA conference, a critical clarification was made by Coding Clinic’s Nelly Leon-Chisen, RHIA, and Sue Prophet Bowman, RHIA, CCS, who were asked to clarify whether this directive is to be taken literally. Must the “possible” diagnosis actually be documented on the discharge summary? Or does it mean that, at discharge, the possible diagnosis should be documented somewhere in the chart?
The answer will bring dismay to many aggressive DRG coders. To legitimately assign a code for a possible or probable diagnosis to an inpatient chart, the diagnosis must be literally documented at discharge. If no dictated discharge summary is available at coding, the diagnosis must at least be documented in the discharge progress note or on the face sheet, if one is available. Responsibility lies with the DRG coder to confirm that, at discharge, the treating provider still considered the diagnosis to have validity as a probable cause. We all know the alternate choice by now: query the physician for further clarification if needed.

• Symptom with a differential diagnosis: This rule is for principal diagnosis assignment in the Official Coding Guidelines. If a secondary diagnosis has a confirmed symptom and then a list of possible causes, the coder is to assign the symptom alone. If it is a critical factor for the patient’s health or the DRG and the etiology of the symptom has not been clarified at discharge, it will be necessary to break out the physician query process for a confirmed diagnosis.

• Unspecified fifth digits: Do you believe that coders should never assign “unspecified” fifth digits to diagnosis codes—especially on the principal diagnosis—because Medicare won’t pay for unspecified diagnoses? Remember the basic rule of coding and billing: If you code it wrong in order to get it paid, it meets the definition of the “F” word (fraud).
If there is not enough information to assign a specified code, coders must use the unspecified fifth digit. If the payer rejects a claim for lack of detail, it will be necessary to request that the physician be more specific.

• V codes: One urban legend with the greatest longevity is the belief that V codes cannot be assigned to inpatient charts and that they’re only for outpatient encounters. But a review of the chapter-specific Official Coding Guidelines reveals that they may be used on any type of chart in any healthcare setting, and the reason for the visit makes no difference as to whether the V code may be assigned.

This type of code can contain critical information regarding a patient’s need for care. Personal and family history may add justification for tests, medications, and even surgeries. Some V codes must be the principal diagnosis; some are required as additional codes to diagnoses from categories 001 to 999. And while there are other guidelines that apply for any code to be a reportable diagnosis, the fact remains that V codes are most definitely allowed on inpatient claims.

• E codes: If coding staff assign an E code for the cause of the injury every time a patient is processed, they need redirection at the earliest opportunity. The E code is to be assigned at the initial treatment. For transfers, receiving hospitals may code the injury but may not repeat the external cause code. The same is true for subsequent encounters postdischarge.

ICD-9-CM makes an exception for fractures, which may be coded as acute and reported with an E code as long as the fracture is still receiving acute care. (The October update of the Official Coding Guidelines provides details.)

In the future, if you hear one of these coding legends or if you receive coding directions that just don’t seem right, head for the Snopes equivalents. The CMS Web site features links to the Official Coding Guidelines for every year, plus facilities should have either a paper or digital library of Coding Clinics. Professional networking is certainly helpful, and the staff of Coding Clinic in Chicago will often provide a direct response to questions received by their office. Those that cannot be explained simply may be referred to the council for consideration and could even be reflected in a subsequent Coding Clinic issue.

— Judy Sturgeon, CCS, is the clinical coding/reimbursement compliance manager at Harris County Hospital District in Houston and a contributing editor at For The Record. While her initial education was in medical technology, she has been in hospital coding and compliance for 21 years.