Q&A: Determining when coders should query for definitive diagnoses

CDI Blog - Volume 11, Issue 9


Have a CDI question?
Send it to larchibald@acdis.org!

Q: We are struggling with coders querying certain diagnosis because they feel there is conflicting documentation in the record, and then we are “losing” that diagnosis. For example, the physician documents acute respiratory failure a couple of times and dyspnea in the discharge summary and the coders query confirming whether respiratory failure was ruled in or out. Same thing with respiratory failure and hypoxia. Same with bacteremia and sepsis (that’s a little stickier) and delirium and encephalopathy. Can a symptom be conflicting with a diagnosis? I’d love to have ACDIS’ opinion on this.

A: When coders see a diagnosis downgraded to a symptom in the discharge summary they assume the attending decided “after careful study” that the patient had symptoms present without diagnoses. This underscores the paramount importance of doing extensive physician education regarding how their conditions can, and are, being reported.

I believe you will have to approach this from the physician side. The coders are not likely to change their indexing protocols based on what the coders feel the attending’s diagnosis in the final diagnostic statement actually meant. They are going to stick with the literal indexing process.

If the physician wants respiratory distress/hypoxia/dyspnea reported as a diagnosis—all of which indicate the patient had mild symptoms and minimal visible distress but would generally not require hospitalization as the diagnosis would be self-limiting—then I would say you are in a good place. Have you asked the physician if they wanted a diagnosis with such a meaning reported for patients who are admitted, treated, and documentation elsewhere as being in failure? This would be the same with bacteremia. Ask the physician if they wanted a diagnosis of a positive blood culture without clinical significance, without the need for treatment and generally self-limiting to be reported as the diagnosis (because that is what you get with bacteremia).  

Let’s review some of the guidance.

“A basic rule of coding is that further research is done if the code title suggested by the index does not identify the condition correctly,” according to AHA Coding Clinic for ICD-10-CM/PCS, Fourth Quarter, 2017.  This could mean that questionable, symptom-based records should not be finalized until a query is done to clarify the discrepancy and that coders could be in violation of this Coding Clinic if they report symptoms when they appear incorrect. (Good luck with that.)

Consider Coding Clinic, Third Quarter, 2016, p. 26, which states “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. The plan of care is based on the attending physician’s evaluation, interpretation, and collation of all the findings (i.e., pathology, radiology, and laboratory results).” 

While probably not legally enforceable with the chief medical officer, the above guidance does at least reinforce the long-held policy that it is the attending’s responsibility to fully address all conditions which were managed during the stay, even when the attending physician did not directly interact or address certain conditions. 

Both CMS in their official guidance from MedLearn Matters as well as the AHA in further Coding Clinics have repeatedly stated that if there is conflicting or contradictory information in the medical record, a coder should query the attending physician to clarify the correct diagnosis. While terms such as dyspnea and failure may not seem contradictory to a clinician, they are contradictory for ICD-10 code assignment. This means a coder should (if following these guidelines) hold the record and place a query or have the CDI do a secondary review to determine in a query is needed prior to diagnosis assignment.

AHIMA’s “Standards of Ethical Coding” also states that coding professionals are expected to consult physicians for clarification and additional documentation prior to code assignment when there is conflicting or ambiguous data in the health record.

The reasons for submitting a query to a physician have been addressed through the history of guidance published by both AHIMA and ACDIS with the latest, Guidelines for Achieving a Compliant Query Practice (released in 2013 and revised for ICD-10 in 2016) calling for queries to be generated when:

  • The documentation is not clear enough to support the rationale for tests that were performed
  • An extended length of stay due to conditions is documented by other patient caregivers
  • A response is needed for a condition identified by another clinician
  • Added specificity is needed for the codes being applied

Previous AHIMA guidance called for query generation if:

  • Clinical indicators are present without a definitive diagnosis
  • Clinical evidence of higher degree of specificity or severity present exists
  • A cause-and-effect relationship between two conditions or organism is present without explicit diagnosis or relationship documentation
  • Documentation lacks an underlying cause for patient symptoms
  • Documentation includes treatments without diagnosis
  • Present on admission status is unclear

This scenario clearly meets not only the justification of clarifying conflicting diagnoses, but also meets the need to clarify based on clinical evidence for a higher degree of specificity or severity. Per the Official Guidelines for Coding and Reporting, diagnoses must always be assigned to the highest degree of specificity possible and diagnoses are always preferred over signs and symptoms.   

Respiratory failure certainly is more specific than dyspnea. Assuming the diagnosis of respiratory failure has clinical indicators and treatment to support it, it would obviously be the more appropriate code and with the highest level of specificity. This situation also relates to the justification of an underlying cause when admitted with symptoms. That justification for a query is clearly present when a physician is documenting a symptom as a diagnosis and there is reasonable support for that diagnosis within the record. 

Failure to address the discrepancy represents a partial violation of the CMS Quality Improvement Organization (QIO) case review policy which states that the goal of the ICD-10 reporting is to ensure the record as coded and reported by the hospital on its claim matches both the attending physician’s description and the information contained in the patient’s medical record. Where we reported the indexing from ICD-10 matching the physician’s description (dyspnea) but completely failed in the second half of that direction which was to capture the information contained in the record (a patient who may have full justification for reporting a more specific respiratory illness). Dyspnea, does not accurately report the information contained in a medical record of a patient who is extremely hypoxic, meets respiratory failure criteria, needed to be in the hospital to avoid potential negative outcomes and received treatment and/or monitoring which could not be safely done at home.

Guideline 19 states that coding is not based on indicators but simply the providers statement. Be that as it may, it does not supersede the preponderance of guidance suggesting that a query should be done prior to that code assignment. It simply means that if the physician refuses to change or modify his diagnosis and or refuses to answer a query, the coder cannot independently edit the reported codes. It does not mean no query should be placed prior to getting to that point. 

Editor’s note: Allen Frady, RN, BSN, CCDS, CCS, CRC, CDI education specialist for BLR Healthcare in Middleton, Massachusetts, answered this question. Contact him at AFrady@hcpro.com. For information regarding CDI Boot Camps visit http://hcmarketplace.com/clinical-doc-improvement-boot-camp-1.

Found in Categories: 
Ask ACDIS, Clinical & Coding