Are Leading Queries Prohibited by Law or Lore?

AHIMA released its CDI Practice Brief Monday.

At Yom Kippur services, I found myself thinking about the question Dr. Ronald Hirsch posed to me the day before. My rabbi was talking in her sermon about the difference between halacha and minhag.

Halacha is law; it is the prescriptions and proscriptions set in the Bible and other rabbinic legislation. Thou shalt not murder; Honor thy father and mother; rules for celebrating post-Biblical holidays. Minhag is tradition or custom, and, although the practices may have become so commonplace as to be near universal, they are not derived from law, per se. Wearing a head covering (i.e., a yarmulke or kippah) in the synagogue and avoiding rice, lentils, and legumes at Passover in the Ashkenazi tradition are examples of minhagim (the plural of minhag).

The question that was posed by a listener of Monitor Mondays was, “Does the law prohibit leading questions in queries?” and they asked for a source. There was no time to answer on air, and Ron (Dr. Ronald Hirsch, MD) kicked the football to me. Behind the scenes, healthcare attorneys David Glaser and Knicole Emanuel both concurred that there is no law, rule, or regulation prohibiting leading queries, and I agree. However, there is more to this.

One of the hardest things I had to do as a physician advisor was wrap my head around the term, “compliance.” I finally settled on it indicating going along with the rules, which to me implied the government’s rules, regulations, and laws. However, in addition to external requirements, compliance can reflect following your organization’s internal rules, policies and procedures, and acting in accordance with ethical practices. Laws, rules, and regulations are halacha; industry standards and organization’s internal rules, policies and procedures are minhag.

What is a leading query? It is a request for clarification that is not supported by the clinical elements in the health record and/or directs a provider to a specific diagnosis or procedure. If it is not supported by the clinical indicators, then it will not be a valid condition, even if documented in the record.

Attempting to guide the provider into desired documentation for reimbursement or quality reasons is also noncompliant. Only the clinician caring for the patient is qualified to make diagnoses; clnical documentation integrity specialists (CDISs) must not overreach.

An example of a leading query would be, “This patient is bedbound and needs assistance in all activities of daily living. Wouldn’t you characterize him as having functional quadriplegia?” The CDIS is putting words in the physician’s mouth. In this case, it probably is true, but that is not considered relevant to whether the query is leading or not. However, there are situations where a CDIS may be asking a provider about a condition which may not be present. “This patient meets SIRS criteria with a urinary tract infection. Don’t they have sepsis?” The provider may go against their better judgment and answer in the affirmative, thinking that is the outcome the CDIS wanted. They have been led to a diagnosis.

Even if the condition being pointed out or queried about is legitimately present, it is not kosher for the CDIS or coder to ask the provider to please document it. They must offer a set of choices, all of which could match the clinical indicators, and allow the provider to make their own diagnosis.

The Association of Clinical Documentation Integrity Specialists (ACDIS) and the American Health Information Management Association (AHIMA) just released the 2022 updated version of Guidelines for Achieving a Compliant Query Practice, which Melissa Potts reviewed for us in her segment. The guidelines are considered industry standard (minhag) and establish best practice for anyone querying. The new guidelines state the Industry Practice Brief “should serve as an essential resource for coding and clinical documentation integrity, and other professionals in all healthcare settings, who participate in query (documentation clarification) processes and/or functions.” The Practice Brief also states that the guidance is to be used by payers, auditors, and compliance agencies.

This is interesting: under whose authority do ACDIS and AHIMA set the standard for case managers or quality folks? I don’t have an answer for that, but I can tell you that I always caution physician advisors from posing queries. I instruct them to do whatever education their medical staff needs to be able to respond intelligently to compliant queries from the CDISs or coders. Give them what they need to be able to make an informed, considered, thoughtful decision. Then leave the formal compliant querying to the CDISs and coders.

Similarly, I teach CDISs and coders to be sure that verbal queries are not leading, and to record them accurately in a non-leading fashion. Admittedly, this can frustrate providers who sometimes implore CDISs, “Just tell me what you want me to write!” My recommended response is something along the lines of, “I can’t tell you what to write. Only you, as the clinician, can determine the diagnoses. But those clinical indicators may have a corresponding diagnosis (here, education ensues in a non-leading fashion). Do you think this patient has a diagnosis that explains those clinical indicators?”

This is why yes-no constructed queries have a limited utility. They can result in a leading query. For example, “The patient is short of breath, required IV furosemide, has an ejection fraction of 25%. Does she have acute systolic heart failure?” is leading. There are other choices which could fit the clinical indicators: acute on chronic systolic heart failure, potentially the patient could have combined systolic and diastolic heart failure. If the provider just said, “Yes,” to the query, they might not be telling the whole story. However, yes-no queries may be appropriate to verify information like pathology findings or present on admission determinations. The Practice Brief offers lots of appropriate scenarios.

There is no law, rule, or regulation which explicitly prohibits CDISs from generating leading queries. The closest reference I found was in a the Centers for Medicare & Medicaid Services (CMS) transmittal from 2014 (https://www.cms.gov/Regulations-and-Guidance/Guidance/Transmittals/Downloads/R18QIO.pdf) which said, regarding DRG validation:

The purpose of DRG validation is to ensure that diagnostic and procedural information and the discharge status of the patient, 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. Refer the case for a physician review if medical judgment is needed when changing the narrative diagnosis that the codes were based upon. Your reviewer must use his or her professional judgment and discretion in considering the information contained on a hospital’s physician query form along with the rest of the medical record. If the physician query form is leading in nature or if it introduces new information, the nonphysician reviewer must refer the case to the physician reviewer.

(A little further down:)

Base your DRG validation upon accepted principles of coding practice. (However, it also instructs to “Use only these guidelines [that is, ICD Coding Guidelines] and updates when performing DRG validation.”)

This does not say that if the query is leading, consider it null and void. It says a physician reviewer must be recruited to render an opinion. It also is contradictory because accepted principles of coding practice include ethical and compliant coding practices which may not be based solely upon the ICD Guidelines for Coding and Reporting.

Here are my recommendations for the CDIS:

• Only give choices which match the clinical indicators
• Give all choices which match the clinical indicators (don’t cherry pick)
• Don’t coerce (or give the appearance that you are trying to force) the provider into making a specific diagnosis
• Never give quality or reimbursement implications in the query
• Always give an option for the provider to give an alternate explanation for the clinical indicators, like “other __________.”
• Only use “unable to determine” when that is appropriate (“What was the status of the COVID-19 infection?” Did they have COVID-19, did they have a sequela of a previous infection, or did they not have COVID-19? What would “unable to determine” MEAN?! How would a coder code THAT?)

My strongest recommendation is read for yourself the newly released version of Guidelines for Achieving a Compliant Query Practice. Even if it is minhag, it is our industry’s best practice standard, and we must follow it. It will save us a lot of grief on the back end trying to justify diagnoses obtained by non-compliant queries.

Program note: Listen to Dr. Erica Remer when she cohosts Talk Ten Tuesdays at 10 Eastern with Chuck Buck.