ICD-10 queries: Where real time is likely to be lost

By Carl Natale, Editor, for ICD10Watch

Are you worried that ICD-10’s granularity will take a hefty toll on productivity?

Then you’re not alone. One of the major issues with ICD-10 implementation is the concern that the increased specificity will create more work for medical practices and damage productivity. To document that increased specificity, physicians are going to have to spend more time with documentation.

Of course, physicians could refuse to take that time documenting too many details. Then they could devote that time to treating patients. And answering queries from medical coders.

That’s where the real time is going to be lost. Queries aren’t liked by anyone now. ICD-10 implementation isn’t going to improve popularity. Which makes it worthwhile to start improving how queries are written.

When to query

The Journal of AHIMA suggests writing a query when clinical documentation:

• “Is conflicting, imprecise, incomplete, illegible, ambiguous or inconsistent”
• Describes clinical indicators that don’t clearly support the underlying diagnosis
• Includes clinical indicators, evaluation, and/or treatment that does not seem related to any medical condition or procedure
• Does not support or validate a diagnosis
• Does not support the present on admission indicator

If the medical coder needs to write a query, then there are three principals to help coax — not coerce — the information from physicians.

Be written in clear, concise and precise language

Some medical coders prefer to write their own queries so they can keep the queries concise. Others prefer standardized queries based on templates. The second option could lead to more consistent physician responses.

Note that if a medical practice or hospital uses query templates, those need to be converted to ICD-10-CM/PCS coding language:

Use ICD-10 coding manuals and other industry references.
Teams should consist of:

• Clinical documentation specialists
• Medical coders
• Physicians
• Plan to break it into manageable chunks each week.

Which ever format, the queries need to be individualized and addressed to a specific physician. The medical coder needs to provide name and contact info with each query.

Contain evidence specific to the case

Richard D. Pinson, MD, FACP, CCS, principal of HCQ Consulting and coauthor of the CDI Pocket Guide, tells For the Record that queries need to have three things:

1. The condition or diagnosis that the medical record already cites.
2. Any data in the record or supporting documentation that pertain to the question being asked.
3. The actual question.

The goal is to give the physicians enough information so they don’t have to look up the medical records themselves.

Be non-leading

Don’t ask if the patient has a certain condition. Ask if the details in the documentation support a more specific or different diagnosis than what is initially documented.

And multiple choice questions would not be considered leading questions as long as the options are medically reasonable.

How to cut down on queries

Now is the time to start improving clinical documentation so there is less need for queries after Oct. 1, 2015. And the queries used now can be part of that effort.

Introduce physicians to ICD-10 terminology by using those terms in queries now. It will improve clinical documentation and help find the right diagnosis codes.

If time is spent now improving queries and clinical documentation for ICD-9 coding, reimbursements will improve. Healthcare providers will be able to bill for all the conditions they are entitled to.

And if the ICD-9 diagnosis is incorrect, it’s not going to correct itself in ICD-10 coding. The diagnoses codes need to be correct now for a better chance of getting it right after Oct. 1.
In addition to using ICD-10 language now in queries, ask for details now that support ICD-10 levels of specificity. Those two practices will train physicians to use the kind of documentation that will support ICD-10 coding.

All these steps will make physician queries less likely after ICD-10 implementation. If not, at least medical coders will have practice writing good queries that extract the needed information to support ICD-10 codes.