Opinion: Stop the insanity related to the 2017 coding Guidelines

by Trey La Charité, MD, FACP, SFHM, CCDS

The ICD-10-CM Official Guidelines for Coding and Reporting for the 2017 fiscal year (effective 10/1/2016) added a new directive. In Section I, subsection A, item 19, states that:

“The assignment of a diagnosis code is based on the provider’s diagnostic statement that the condition exists. The provider’s statement that the patient has a particular condition is sufficient. Code assignment is not based on clinical criteria used by the provider to establish the diagnosis.” 

The chilling implication is that our coders have been ordered to code diagnoses documented in the medical record whether they exist or not. In my opinion, these Guideline authors are obviously uninformed and unconcerned about the auditing pressures faced by our hospitals. Similar to a political incumbent who has forgotten his constituents, the coding governing bodies have lost touch with healthcare reality.

No organization can financially afford to follow this Guideline. If your facility does not clinically validate the diagnostic codes reported, the recovery audit vultures that circle your facility will descend like a plague of locusts. Sadly, this new Guideline might result in a deep pool-of-opportunity where auditors can easily hook non-existent diagnostic codes. In my opinion, it is inappropriate and unacceptable to allow variance in the standards between the coders and the auditors. This Guideline establishes the supposition that coders cannot clinically validate but allows auditors to do so. 

Another factor to consider is how those denials are going to be counted by your organization. For example, if a doctor repeatedly reports a condition in the medical record that does not exist, and the coder dutifully codes that condition, but the auditor downgrades the MS-DRG upon discovering that the condition did not exist. Who is at fault? Organizations could deem this example to be a “coding error” and not a “provider documentation error.” Classifying this as a coder problem when the coder’s hands were tied is patently inaccurate and unfair.

Recently, we also received advice published in the July 2016 Journal of AHIMA that we should query to confirm or disprove the existence of that repeatedly documented condition before we consider that chart to be final coded (Denton, Debra Beisel; Endicott, Melanie; Ericson, Cheryl E; Love, Tammy R.; McDonald, Lori; Willis, Daphne. "Clinical Validation: The Next Level of CDI").

Once again, this represents conflicting directives. First, how can you expect a coder to clinically validate the existence of a condition through a query (i.e., review for and recognize clinical criteria for a given diagnosis) if the irreproachable coding guidelines now say “Code assignment is not based on clinical criteria used by the provider to establish the diagnosis.” Second, if the queried provider reaffirms the existence of the diagnosis that the coder knows is not present, the coder is still forced to code the diagnosis as if it existed. This is an untenable position for a coding professional.

This impossible conundrum leaves no clear solution. One course of action is to disregard this new directive. Your loyalty and allegiance lies with your hospital. In today’s environment, you must train your coders to clinically validate certain provider documented diagnoses.

When I say “certain” diagnoses, I refer to those with definitive, easy-to-operationalize objective criteria to serve as concrete landmarks. For simplicity, consider acute renal failure. If your organization follows the KDIGO clinical guidelines for making this diagnosis, your providers should know that creatinine rises of 0.3 mg/dl or more above the patient’s baseline qualify.

Therefore, if one of your providers has documented acute renal failure with a creatinine rise of only 0.2 mg/dl, the criteria have not been met. Your coders can easily make this determination by quickly reviewing the available creatinine levels for the past three months that are available in the patient’s medical record. If the patient’s creatinine did not rise by 0.3 mg/dl, the coder should not code acute renal failure. No need to query. If the diagnosis is not there, don’t code it. The recovery auditors will find nothing. Your organizations and coders no longer face fraud accusations.

As warned, you readers may be becoming nervous. However, it is simply not reasonable to insist that coders code diagnoses they can easily tell do not exist. In my opinion, urging coders to code things they know not to be present results in forced falsification of the medical record.

In the legal world, this situation is characterized as “conspiracy to commit fraud” if coders are forced to code things they know are untrue in an attempt to increase a hospital’s reimbursement. We should not place our coders nor our hospitals in this position. Coding is hard. To make coding harder by adding the risk of a fraud allegation is unacceptable. Coders are smart and hard-working; they are quite capable of handling this expansion of their duties.

Not all diagnoses are amenable to this clear-cut approach. For example, the diagnosis of acute encephalopathy is too complex, clinically variable, and nebulous for this approach. There are simply no numerical or other objective criteria available. Obviously, common sense guidelines must be employed when deciding what diagnoses are easily verified by your coding staff.

The Final Solution?
On the one hand, as one of the four Cooperating Parties, CMS creates the coding guidelines and insists that the coders follow them to the letter. On the other, CMS empowers the recovery auditors by stating they may use “clinical review judgment” to validate diagnoses and issue denials. The implication is that CMS is shackling the coder to code whatever the provider says while simultaneously encouraging the auditors to challenge the diagnostic validity of all submitted codes. Again, reasonableness, fairness, and equity are absent.

The first page of the ICD-10-CM Official Coding Guideline for Coding and Reporting for fiscal year 2017 states that

“A joint effort between the healthcare provider and the coder is essential to achieve complete and accurate documentation, code assignment, and reporting of diagnoses and procedures.” 

No one would argue or find fault with this statement. The paragraph goes on to say that

“These guidelines have been developed to assist both the healthcare provider and the coder in identifying those diagnoses and procedures that are to be reported.” 

However, the healthcare providers that generate the documentation had no input into the development of the ICD-10-CM system. Again, like the incumbent politician analogy, the constituent has been omitted from the legislative process. Saying providers should be involved in the coding process yet omitting them from the development of the process is unacceptable. Compliance without participation and representation is unreasonable and unrealistic. If there is a true and sincere desire for more provider engagement in the coding and reporting of diagnoses, the providers should become the fifth Cooperating Party.

Two clear options exist: Coders should code whatever the providers say in the record and not allow the auditors to challenge the clinical validity of what the providers say or providers must be included in the development of the coding process and establish clinical criteria that all parties (providers, coders, and auditors) can universally abide. There must be consistency in thought and action among all involved parties. When consistency is achieved, positive change and progress can occur.

Editor’s note: La Charité is a hospitalist with the University of Tennessee Hospitalists at the University of Tennessee Medical Center at Knoxville. He is board certified in internal medicine and has been a practicing hospitalist since 2002. He is also a Clinical Assistant Professor with the Department of Internal Medicine and serves as the Medical Director for UTMC’s Clinical Documentation Integrity Program, Coding, and RAC response, as well as a past ACDIS Advisory Board. His comments and opinions do not reflect those of UTMC or ACDIS. Any advice given is general and should be reviewed with management to ensure compliance. Contact him at Clachari@UTMCK.EDU.