Determine when and how to query physicians
March 31, 2014
CDI Blog - Volume 7, Issue 3
Coders and clinicians seem to speak different languages. CDI specialists often serve as the translators between clinicians and coders, so it's important that all three groups work together.
Both coders and CDI specialists can query physicians when documentation:
- Is conflicting, imprecise, incomplete, illegible, ambiguous, or inconsistent
- Describes or is associated with clinical indicators without a definitive relationship to an underlying diagnosis
- Includes clinical indicators, diagnostic evaluation, and/or treatment not related to a specific condition or procedure
- Provides a diagnosis without underlying clinical validation
- Is unclear for present-on-admission (POA) indicator assignment
Query guidance requires continual review and updates to reflect best practices. "Everyone is getting more sophisticated in their querying," says Cheryl Ericson, RN, MS, CCDS, CDIP, AHIMA-approved ICD-10-CM/PCS trainer, CDI education director for HCPro, a division of BLR, in Danvers, Mass. "Patients are getting sicker. There are a lot of complexities that we need to capture."
In February 2013, AHIMA and the Association for Clinical Documentation Improvement Specialists (ACDIS) jointly released Guidelines for Achieving a Compliant Query Practice, which provides information about how and when to query.
Coders can't start coding a record until all of the documentation is in and any of the above problems are resolved. That's where concurrent clinical queries from CDI specialists can help.
"You need to have someone alongside the clinical staff," says Darice M. Grzybowski, MA, RHIA, FAHIMA, president of HIMentors, LLC, in La Grange, Ill. "Queries should go down if CDI specialists are doing their job."
Physician's perspective
Physicians are ultimately responsible for their own documentation, but what is obvious to a coder may not be obvious to a physician, says Jonathan Elion, MD, president and CEO of ChartWise Medical Systems in Wakefield, R.I.
For example, physicians tend not to differentiate between levels of malnutrition. "It's either malnutrition or it isn't," Elion says. But to coders, malnutrition looks much more complex. ICD-9-CM includes the following codes for malnutrition:
- 260, kwashiorkor
- 261, nutritional marasmus
- 262, other severe protein-calorie malnutrition
- 263.x, other and unspecified protein-calorie malnutrition
- 264.x, vitamin A deficiency
- 265.x, thiamine and niacin deficiency states
- 266.x, deficiency of B-complex components
- 267, ascorbic acid deficiency
- 268.x, vitamin D deficiency
- 269.x, other nutritional deficiencies
Some forms of malnutrition are CCs, while others are classified as MCCs; the difference between CC and MCC affects MS-DRG selection.
Should coders and/or CDI specialists query for more specific information about the malnutrition? If the physician documents moderate-severe malnutrition, coders or CDI specialists should query to determine which of these apply. If the answer is "severe," assign code 261 (nutritional marasmus). If the answer is "moderate," assign code 263.0 (malnutrition of moderate degree). See Coding Clinic, Third Quarter 2012, p. 10.
When to query
Coders and CDI specialists should only query for a limited number of reasons, and reimbursement isn't one of them, Elion says.
The purpose of the query process is to ensure appropriate documentation appears in the health record. "Oftentimes, CDI departments are implemented for financial incentives, but the bottom line for most CDI departments is accuracy of documentation in the health record to support the coded data and the DRG assignments," Ericson says.
Coders should be careful not to query for something so often that physicians begin to document it even when it's not present, says Kathy DeVault, RHIA, CCS, CCS-P, senior director of HIM practice excellence for AHIMA in Chicago.
For example, if a physician suddenly starts documenting that every patient with pneumonia has bacterial pneumonia, he or she may be doing so to avoid a query. That doesn't benefit anyone because it introduces incorrect information into the medical record. A CDI specialist may not question that information, and the coder might end up coding bacterial pneumonia for a patient who may not actually have it.
When a practitioner documents a diagnosis that does not appear to be supported by the clinical indicators in the health record, coders or CDI specialists should query.
The reason for a query is in some ways more important than how the query is formed, says William E. Haik, MD, FCCP, CDIP, director of DRG Review, Inc., in Fort Walton Beach, Fla. "Above all, we should remain intellectually honest. Don't lead the doctor to document a condition that you know is not clinically significant, such as an isolated radiographic finding of atelectasis when there is no specific therapy or evaluation, just to get it documented and coded, and oftentimes reported inappropriately."
Clinically insignificant conditions
Coders can report conditions that aren't clinically significant but that satisfy criteria for reporting additional diagnoses as listed in the ICD-9-CM Official Guidelines for Coding and Reporting. The guidelines state:
For reporting purposes, the definition of 'other diagnoses' is interpreted as additional conditions that affect patient care in terms of requiring and of the following:
- Clinical evaluation
- Therapeutic treatment
- Diagnostic procedures
- Extended length of hospital stay
- Increased nursing care or monitoring
Conditions that are clinically significant to a physician almost always meet one or more of the criteria for reporting additional diagnoses, says Joel Moorhead, MD, PhD, CPC, adjunct assistant professor at the Rollins School of Public Health at Emory University in Atlanta. However, conditions that a physician designates as "not clinically significant" can also meet these criteria, Moorhead says. "For example, a clinically insignificant pleural effusion is a reportable secondary diagnosis if additional diagnostic testing, such as an x-ray with a lateral decubitus view, is required to clarify the diagnosis. The pleural effusion is reportable because the condition was specifically evaluated." See Coding Clinic, Third Quarter 1991, p. 20.
"Now we're starting to move into the arena of not only do we query when something is vague, missing and incomplete, illegible, and so forth, but when we don't see it meeting the definition of a secondary diagnosis," Ericson says. For example, the physician may not have documented treatment, evaluation, length of stay, impact on the length of stay, or increase in nursing care. The record may contain no clinical indicator to support that particular diagnosis.
When a practitioner documents a diagnosis that does not appear to be supported by clinical indicator(s), coders or CDI specialists should query to address the conflict, Ericson says.
Coders also need to remember that a clinical condition does not need to meet Coding Clinic criteria. Coding Clinic is not an authoritative source for establishing clinical indicators, Haik says.
Query formats
Queries can be either concurrent, meaning at the time the patient is being treated, or retrospective, meaning after the fact, Elion says. Retrospective queries are very different because coders or CDI specialists are asking physicians to reopen a record and try to remember a particular patient or a particular thought process. Whenever possible, concurrent queries are better, says Elion.
Queries can be in multiple-choice format or yes/no format. Multiple-choice query formats should include clinically significant and reasonable options supported by clinical indicators in the health record.
Coders and CDI specialists should not use a yes/no query when clinical indicators of a condition are present but the physician has not documented the condition/diagnosis in the record. In short, don't introduce a new diagnosis with a yes/no query format. However, yes/no queries, per the AHIMA/ACDIS guidelines, are appropriate in the following circumstances:
- Substantiating or further specifying a diagnosis that is already present in the health record (e.g., findings in pathology, radiology, and other diagnostic reports) with interpretation by a physician
- Establishing a cause-and-effect relationship between documented conditions such as manifestation/etiology, complications, and conditions/diagnostic findings (e.g., hypertension and congestive heart failure, diabetes mellitus, and chronic kidney disease)
- Resolving conflicting documentation from multiple practitioners
The response to a written query might not occur on the query form itself, Ericson says.
If documentation occurs "directly into the health record and there is lack of supporting clinical information, it is recommended the practitioner provide the clinical rationale for the diagnosis (e.g., patient transfused four days ago due to acute blood loss anemia) unless the query is maintained as a permanent part of the health record," according to the AHIMA/ACDIS practice brief.
The goal is to achieve a transparent relationship between the query and additional documentation, Ericson says. "If the practitioner documents his or her response only on the query form, then the query form should become part of the permanent health record."
With the increase in CDI programs, CDI specialists are performing more concurrent reviews, and as a result frequently query physicians verbally for information.
Verbal queries should contain the same clinical indicators as written queries and be retained in the same manner and format to ensure compliance and consistency with policies and processes, Ericson says. "There is nothing against verbal querying; there just has to be some kind of record or memorialization of what transpired during that conversation. We want to make sure that we're not leading the physician or introducing new information during that verbal interaction."
The documentation of the verbal query may be condensed to reflect the stated information, but it should always identify the clinical indicators. "You're going to see that over and over again, the importance of having the clinical indicators included with the actual query," says Ericson.
It's also a good rule of thumb to make sure that the clinical indicators in a query are specific to the patient addressed in the query. All queries should not read exactly the same. Coders and CDI specialists should include some unique identifiers to link a query to a particular patient or episode of care.
"You want to have those clinical indicators to support the query documented when you are doing a verbal query," Ericson says. If the clinical indicators can't be documented during the conversation, the best practice would be to document them immediately afterward.
Leading queries
A leading query is one that is not supported by the clinical elements in the health record that makes one think a more complete or specified diagnosis or procedure is present.
"The bottom line, a query is not a fishing expedition," says Haik. Many times CDI specialists and coders know the answer to the query when they ask it, so they need to have a firm clinical rationale for the query to justify it not being leading in nature. "I think oftentimes why you query is more important than how you query."
Coders and CDI specialists only need one clinical indicate present in order to support a query, but more than one may be present.
Editor’s note: This article was originally published in the March issue of Briefings on Coding Compliance Strategies.
Found in Categories:
Physician Queries