Q&A: Clinical validation query type

Q: We struggle with the use of the clinical validation query type. Would others be able to share their best practices for this type of query? For example: 

• You are clinically validating a diagnosis of respiratory failure when the criteria do not support that diagnosis. Do you use a respiratory failure query type or the clinical validation query type?

• Do you classify the response always as an “agree” regardless of the actual response?
• Under what circumstances do you go straight to the clinical validation query template? 

Response #1: We use the clinical validation template because our options include:

• "...is ruled in (please provide evidence used to support this diagnosis"
• "...has been ruled out"

We've struggled with agree versus disagree responses as well and have therefore provided education to the staff.

Ultimately, this is how I lead my team:

• If you send a clinical validity query (because it is not a validated diagnosis) and the provider rules it in still, it's a “disagree”: think of as "I personally don't agree that the diagnosis is valid."
• If you send a clinical validity query and they rule out the diagnosis (because they realize it is not valid), then it's “agree.”
• If you send a query for a noted opportunity (acidosis, electrolyte imbalance, etc.) that is clinically valid and is a reportable diagnosis, and they respond with "no clinical significance" or something similar, then it's “disagree”: you gave the provider the opportunity to pick up the diagnosis, and they didn't.
• If you send a query for a noted opportunity (acidosis, electrolyte imbalance, etc.) that is clinically valid and is a reportable diagnosis, and they rule it in the diagnosis, it's “agree.”

Response #2: We clinically validate when something is documented but lacks clinical support. We classify the response as “agree” unless they respond with "unable to determine" or something similar, then it would be a “disagree.”

Our template choices include:

• After study, please clarify the clinical significance of _____. 
• Confirmed and valid as demonstrated by the following clinical indicators (please provide additional support/rationale):_____.
• Ruled out
• Other explanation for clinical findings (please specify):_____.

Response #3: We only have specific clinical validation query types for sepsis and respiratory failure, otherwise we use typical query types for clinical validation or for adding a diagnosis.

For example, we would use a malnutrition query type either for adding the diagnosis or clinically validating an unsupported malnutrition diagnosis.

We classify the response as agreed when they rule the condition in or out. Additionally, we use clinical validation templates when documented diagnoses are not supported by guidelines or clinical criteria.

Response #4: We have several clinical validation query templates that are specific to diagnoses, such as sepsis, respiratory failure, acute kidney injury, encephalopathy, etc. We use them when a provider documents a diagnosis that does not meet our organization's established criteria. Our query response options are: diagnosis ruled out after study; diagnosis confirmed (add clinical indicators to support the diagnosis below); other.

As far as responses, we developed specific response types for validation queries in our electronic health record and they include: validation query (diagnosis confirmed without additional indicators); recipient agreed (diagnosis ruled out); alternate diagnosis.

Response #5: We have two clinical validation templates (the rest are condition specific): unsupported and inconsistent.

The unsupported is used to rule out or provide additional information to support the diagnosis. If the provider rules out an unsupported, it's considered an “agree.” If they provide supporting evidence, it's also an “agree.” If the "supporting" evidence is "read my note," then it's a “disagree.”

The inconsistent is used to confirm if a diagnosis that's been dropped from the record has been ruled out, resolved prior to admission, or confirmed and treated during admission. If the response is anything other than "disagree" or "unable to determine," it's considered an “agree.”

Response #6: We have taken the approach that if there are no clinical indicators to support a documented diagnosis then it is not clinically present. Providing both the option of ruled in or ruled out does not seem to align well with the principle that clinical indicators need to support the query.

Therefore, our queries are to "rule out" the diagnosis. If the diagnosis is ruled out, then the query is agreed. In the rare instance that the provider responds by adding meaningful clinical indicators to support the diagnosis, we would also note that as agreed.

Response #7: At my organization we use clinical validation type queries when a diagnosis is documented but it does not appear this diagnosis is supported with clinical indicators and/or treatment from our guidebook. These queries are not used when a diagnosis is clinically supported but not carried through the medical record. 

Here are some examples with choices:

• Acute hypoxic respiratory failure is documented in the electronic health record. Please provide clinical evidence or criteria to support this diagnosis. 
• Acute hypoxic respiratory failure ruled out due to lack of clinical evidence or criteria. Acute hypoxic respiratory failure ruled in, please provide documented clinical evidence or criteria to support this diagnosis. 

If they provide any clinical evidence in response to the query (even if it is not the evidence in our guidebook) to support the diagnosis in question or rule the diagnosis out due to lack of clinical evidence, we mark the query as agreed. If they just ruled in the diagnosis without additional clinical support, we mark it as disagreed.

Response #8: We use clinical validation queries specifically in cases where the documented condition lacks clinical evidence to support the coding and reporting of diagnosis. We do not have a template; we simply draft a query presenting all the clinical indicators that the patient lacks (e.g., in a sepsis clinical validation, we make sure to include the normal vital signs, the lack of antibiotic treatment, etc.). The question is neutral (e.g., "sepsis is documented in the medical record, your help is needed to determine the clinical validity of the condition such as...").

The option of "ruled in" always includes a note in parenthesis that reads "if choosing this option, please document the clinical indicators used to support the condition.”

We do not use "agree" nor "disagree" in answer options for these types of queries. I can't take credit for the doctor confirming something already documented, and if he removes the condition, it is typically a downgrade. All clinical validation query responses will either be "other explanation" or "no response.”

I do not include the negative financial impact in our metrics, but I do keep tabs on providers who do not address these types of questions.

Response #9: At our facility, we have a clinical validation query, which asks the following questions:

Please provide additional clinical indicators supportive of the documented diagnosis of ***.

[ ] Clinical indicators include: (Please specify) ____________________

[ ] After further study this diagnosis was no longer valid. 

[ ] Diagnosis is valid. Unable to provide additional clinical indicators regarding the diagnosis.

[ ] Other diagnosis (Please specify) ___________________

The response is not always agreed. If clinical indicators are added to support the diagnosis or is correctly ruled out, it is agreed. If needed, poor responses may be sent to physician advisors or chief medical officers for education. 

Response #10: At our facility, we currently utilize four distinct query templates for clinical validation:

• Clinical validation – Acute kidney injury
• Clinical validation – Sepsis
• Clinical validation – Respiratory Failure
• Clinical validation -- General

When a provider documents one of the specific diagnoses listed above and the clinical criteria are not present, staff are directed to use the corresponding diagnosis specific template.

For diagnoses not covered by these templates, if documentation review reveals that the diagnosis does not meet clinical criteria, we use the general clinical validation query template. This approach supports tracking and denial management, as we can reference query types in reporting. From a denials perspective, having a validation query type documented in the EHR is beneficial for our denials team.

Currently, all clinical validation query responses are classified as “agree.” However, we’ve recently implemented a partial response workflow. If a provider responds to a validation query by simply confirming the diagnosis without providing clinical indicators, we follow up with a second query specifically requesting those indicators.

We’ve provided education to our providers on this new process and the purpose of clinical validation queries. So far, we’ve seen success in obtaining appropriate and complete responses.

Editor’s note: This question was previously answered by members of the ACDIS CDI Leadership Council, and originally appeared in the CDI Leadership Insider, the exclusive eNewsletter for members of the ACDIS CDI Leadership Council. For the purposes of this article, all Council member answers have been deidentified.