Q&A: Querying for clinical validation of a diagnosis

Q: I attended the ACDIS conference in Nashville. (It was excellent!) I had a question about whether or not queries can be used to question the documentation of a condition or procedure where the clinical picture in the record does not appear to support a given diagnosis. I thought the new ACDIS/AHIMA query practice brief, “Guidelines for Achieving a Compliant Query Practice,” states a query can be used in this manner. Can you clarify for me?

A: I was lucky enough to be a contributor to the two most recent AHIMA practice briefs related to CDI efforts and physician queries, so I can offer my perspective. You are correct, the recent practice brief addresses concerns of clinical validity of diagnoses and suggests possible procedures for query escalation. (You can read about the process and see a couple of sample policies for this in the July edition of the CDI Journal.)

Best practice is to obtain a second opinion from a clinician, who could be the CDI department physician advisor or the department head familiar with the particular condition before querying. Some hospitals develop organizational definitions including applicable clinical indicator(s) for particular high risk diagnoses as agreed upon by the medical staff to help the CDI specialist and/or coder to know when these types of cases should be escalated.

Because these types of queries could potential jeopardize the working relationship between CDI and coding staff with the medical staff (who are you to question how I care for my patient), the 2013 ACDIS/AHIMA practice brief alternatively suggests that such these types of queries could come from the CDI manager, coding manager, physician advisor or someone of similar stature.

In situations where the provider’s documented diagnosis does not appear to be supported by clinical findings, a healthcare entity’s policies can provide guidance on a process for addressing the issue without querying the attending physician.

CDI and coding staff have done such a good job educating providers that sometimes physicians may be a bit overzealous in their documentation and write a diagnosis that may not meet the definition of a secondary diagnosis and/or use templates with the same statements (or macros) for all patients.

The Official Guidelines for Coding and Reporting define an additional diagnosis as one which:

“affects patient care in terms of requiring: clinical evaluation; or therapeutic treatment; or diagnostic procedures; or extended length of hospital stays; or increased nursing care and/or monitoring.”

Consequently, before a CDI specialist or coder queries for clarification of a diagnosis, they need to ensure it meets the definition of an additional or “secondary” diagnosis.  Support for some diagnoses may be hidden in the nursing notes. That is often the case with neurological conditions. Since increased nursing care (according to the definition above) supports the reporting of a condition as an additional/secondary diagnosis so look to these for evidence to support your query efforts.

Remember that the medical record needs to contain clinical evidence of a condition to warrant a query. The most recent query practice brief, Guidelines for Achieving a Compliant Query Practice, purposely uses the term “clinical indicator(s).” The (s) was an intentional addition because there may be only one indicator to support a diagnosis. This document also clarifies whose indicator(s) can be used to support a diagnosis.

If the provider documented, “monitoring for acute blood loss anemia,” are they documenting a diagnosis or making a statement?  Some would not consider this statement a diagnosis even though the provider used the term “acute blood loss anemia” because the statement can be perceived as the provider demonstrating that the patient has the potential for acute blood loss anemia.

Contrast this with the statement, “monitoring acute blood loss anemia.”  Does the absence of the word “for” make this a diagnosis rather than a possibility?  It may seem like I’m splitting hairs, but CDI and coding efforts are very much about the nuances. We want the documentation to be clear, so that anyone who reviews the record comes to the same conclusion, the same conclusion the physician had.  What was the provider’s intent with these two different statements?

If you and your colleagues disagree as to whether either statement represents a diagnosis or not, then it may be best to query the provider to determine if “acute blood loss anemia,” as documented in the progress notes (always specify where the documentation occurred in the medical record) was:

• Validated
• Ruled out
• Unable to determine
• Clinically insignificant
• Other (with a line for additional documentation)

The prevalence of electronic problem lists, a component of meaningful use, can cause confusion as to which conditions relate to this particular episode of care and which have been previously resolved.

Once we agree that the provider has documented an additional diagnosis, then we must consider if the health record contains clinical indicator(s) that support the diagnosis.  In other words, would other providers presented with the same information come to the same conclusion/diagnosis?  Is this a reasonable conclusion?  This is not the same as asking if the CDI specialist and/or coder would use the same clinical indicator(s) as we are not diagnosticians.

If a diagnosis does not appear to be supported by clinical indicator(s) then the CDI specialist/coder should query to substantiate the validity of the diagnosis. The provider can use any clinical indicators he/she deems appropriate to make a diagnosis. That’s why the Guidelines for Achieving a Compliant Query Practice states clinical indicators listed by Coding Clinic for ICD-9-CM do not define a diagnosis; rather these indicators are to assist a coder/CDI specialist in understanding a diagnosis.

In other words, let’s say your organization decides to query for malnutrition based on the new ASPEN criteria, which does not include an albumin or prealbuin level as an indicator, but the provider documents malnutrition without being prompted by a query based on an abnormal albumin value. It is not for the coder/CDI specialist to question this diagnosis as the provider does have a rationale for the diagnosis even if it isn’t the criteria the coder/CDI specialist would prefer to see.  Although this diagnosis may be challenged by an auditor, it is defensible as other healthcare professionals would likely come to the same conclusion.

There is a difference between the criteria we, as CDI specialists/coders, must use to support when issuing a query compared to those the provider uses to make a diagnosis.  I know this seems like a subtle distinction, but I think this is what was meant by not questioning the provider’s medical judgment. We are not “armchair quarterbacks” validating the diagnoses determined by the provider; rather we are ensuring there is a basis for making such a determination within the health record, not evaluating the quality of those clinical indicator(s).

Yes, there is a lot of pressure to only report those diagnoses that can withstand audit; however, we must abide by coding guidelines and industry standards as defined by the practice briefs so we balance when a query is genuinely warranted and when it is not. We don’t want to brow beat providers into adding documentation we think should be in the record or discount valid diagnoses because we disagree with the criteria used to define the condition. To resolve this conflict (and to help facilitate physician support for CDI program efforts) many facilities have asked for physician assistance in developing query guidelines and identifying appropriate clinical indicators for frequently queried diagnoses such as malnutrition, acute renal failure, and the like.

The 2008 AHIMA Managing an Effective Query Process says a query should not be used:

“for every discrepancy or unaddressed issue in physician documentation. Insignificant or irrelevant findings may not warrant a query regarding the assignment of an additional diagnosis code, for example. Entities must balance the value of collecting marginal data against the administrative burden of obtaining the additional documentation…

“Providers often make clinical diagnoses that may not appear to be consistent with test results. For example, the provider may make a clinical determination that the patient has pneumonia when the results of the chest x-ray may be negative. Queries should not be used to question a provider’s clinical judgment, but rather to clarify documentation when it fails to meet any of the five criteria listed above—legibility, completeness, clarity, consistency, or precision.”

Editor’s Note: Cheryl Ericson, MS, RN, CCDS, CDIP, answered this question. At the time of this article's original release, she was the CDI Education Director for HCPro Inc.