Query Practice Brief Update: Clinical Indicators and Clinical Validation

Creating industry guidance is a challenging task. It is often difficult to remove ambiguity that could lead to unintended interpretations.

I think the contributors to the updated Guidelines for Achieving a Compliant Query Practice Brief did a great job drafting the document, but I also think there are opportunities for improvement regarding the concept of clinical indicators and clinical validation.

Clinical indicators are necessary to support why a query is needed, as well as to determine appropriate choices when using a multiple-choice query format. Additionally, query compliance “requires that clinical indicators are specific to the patient and episode of care, support a more complete or accurate diagnosis or procedure, and are free from subjective interpretation by the query professional (p. 13).”

In other words, clinical indicators create the foundation of compliant querying, necessitating the need for a clear definition used consistently throughout the Practice Brief.

Clinical indicators are defined within the draft as “documentation elements that support a diagnosis as reportable or establish the presence of a condition. These may include provider assessments, diagnostic findings, treatments, medications, clinical trends, and relevant documentation from ancillary healthcare professionals (p. 5).”

I am confused by the use of the words “documentation elements” within this context.

I do not consider diagnostic results, which are objective data, as a documentation element. In my opinion, clinical indicators include objective data such as diagnostic results that can be interpreted as findings by the physician, but documentation will not change the result.

As I have discussed in prior articles, how much and what clinical evidence a provider requires to make a diagnosis is variable. The draft brief supports this concept by reading “the number and type of clinical indicators required may vary based on the clinical scenario.”

It might be helpful for the authors to explicitly state that even medically accepted clinical criteria are not definitive, and clinical indicators can vary by patient populations.

Newsweek recently published an article titled Medical Misdiagnosis Costs Billions – Women Often Pay More, wherein they reported that women are “50% more likely to be misdiagnosed following a heart attack and 33% more likely to be misdiagnosed during a stroke.” The authors attribute these findings to the fact that “a significant portion of modern medicine is based on the study of men.”

Overgeneralizing the significance of clinical indicators developed from research on limited populations can lead to patient harm. Clinical documentation integrity (CDI) and coding should not be used as tools to limit provider judgment, which could influence treatment.

The draft authors briefly address this concept under the Clinical Validation Queries section, where they assert that “because each patient is unique, the purpose of organizational definitions is not to limit how a diagnosis is defined, but rather to promote consistency in the determination of a minimal threshold that has been met to support the presence of the diagnosis.”

I think it would be beneficial if the authors clearly explained that clinical indicators are unique to each patient, and may not meet guidelines established by payers. This transitions nicely with the existing statement, “providers make the final determination as to what indicators define a diagnosis,” but would add “within each encounter for each individual patient.”  It is important to emphasize that a provider is not universally defining a diagnosis when assessing a patient.

They are reviewing the clinical scenario specific to the patient and developing a treatment plan to support their findings, which may or may not alight with payer clinical indicators. This is the practice of medicine. Payers have the advantage of hindsight. Treating providers do not.

The Code Assignment and Clinical Criteria Section of ICD-10-CM Official Guidelines for Coding and Reporting notes that “code assignment is not based on clinical criteria used by the provider to establish the diagnosis.” This means that determining the clinical significance of a diagnosis, which allows it to be reportable, is not based upon the associated clinical criteria.

The reality of healthcare is that some providers are quick to make a diagnosis, while others may wait for more clinical evidence for a higher level of certainty. But there has never been a regulatory or binding requirement that a diagnosis must be proved beyond any doubt for it to be reportable on a claim.

Quite the opposite, in fact, when coding guidelines instruct coders to assign a diagnosis code “based on the provider’s diagnostic statement that the condition exists.”

Therefore, it is not appropriate to eliminate a diagnosis from a claim if it includes reasonable clinical criteria, even if it does not match payer guidelines, and especially when it is associated with women or people of color. Adherence to coding guidelines is required under the Health Information Portability and Accountability Act (HIPAA), but payer clinical criteria guidelines may not be legally binding.

If it were to become a legal issue, industry data has demonstrated that hospitals often prevail when appealing denials at the administrative law judge (ALJ) level.

As a cooperating party for ICD-10-CM, I think the American Health Information Management Association (AHIMA) should suggest revising coding guidelines to better clarify the relationship between clinical criteria, clinical indicators, and clinical significance. Perhaps part of the problem is that we are using so many different terms for the same concept: deciding if a condition exists.

However, that is the wrong approach. It is not about determining if a condition exists; it should be about deciding if the physician has a reasonable expectation to treat the condition. If given a choice, most people would want to be diagnosed as quickly as possible, for better outcomes and less suffering.

You would think it would benefit the payers as well, because it would avoid future associated healthcare needs.

Waiting until arbitrary clinical indicators defined by insurance plans are met could delay care and result in patient harm. It is antithetical to quality healthcare. In regard to the Medicare population, it also violates the spirit of the Inpatient Prospective Payment System (IPPS), which is predicated upon hospital resource expenditure associated with particular patient populations.

If the hospital is monitoring and treating a patient as if a documented condition exists, then the clinical significance threshold is met, and hospitals should be paid accordingly.

Unfortunately, healthcare is being driven by conflicting paradigms. The hospital paradigm is intervening as quickly as possible to prevent adverse outcomes (e.g., code sepsis, code stroke, etc.) This stands in stark contrast to the payer paradigm of fiscal responsibility (yes, I am being generous here), where clinical certainty is the priority, but that clinical certainty is based upon clinical indicators representing a minority of those who receive healthcare services (e.g., white men of European decent.)

The draft addresses many of the gaps associated with earlier versions, but since it is industry guidance, not a regulatory requirement, it is unlikely to change payer behavior. Instead, it is likely that as a profession, we need to advocate for changes in coding guidelines, since adherence to them is required under HIPAA.

Although it may appear as I have digressed, this brings me to my next point: how clinical validation is defined within the draft brief. It is currently defined as “the process of ensuring that a documented diagnosis is supported by clinical indicators within the health record.

When a diagnosis is documented but lacks sufficient supporting evidence, clarification may be necessary to confirm its clinical validity (p. 5).”

This definition is followed up with a section that centers on clinical validation queries. In my opinion, the following statement is a little confusing: “The purpose of these (clinical validation) queries is not to simply confirm or negate the presence of a documented diagnosis, but to also request documentation of additional clinical data or clinical decision-making by the provider to support the diagnosis.” I think this guidance could be misinterpreted by payers and weaponized.

Traditional Medicare allows medical necessity to be inferred from provider documentation that links the service to the patient’s diagnosis, treatment plan, and medical need without additional clarification. Medicare defers to provider judgment when determining if services are reasonable and necessary, as defined by the Social Security Act.

Requiring additional documentation can place an undue burden on the provider, and changes clinical validation from determining if the condition exists to determining if there is sufficient documentation to support the diagnosis. These are not clinically equivalent concepts, and brings me back to my first point involving the use of “documentation elements” when describing clinical indicators. Documented clinical reasoning is evidence of the clinical significance of diagnostic criteria: two separate but equally important concepts that are necessary to support a claim as medically necessary and accurately coded.

Rather than asking if the condition exists, CDI and coding professionals should determine if it “was it reasonable for the provider to diagnose and treat the condition as if it exists.”

Reframing our goal will increase the importance of clinical reasoning as a defense when payer-specific clinical criteria are unmet.