AAPACN is dedicated to supporting post-acute care nurses provide quality care.

Coding Diagnoses in MDS Section I: The Art of the Physician Query

The coding instructions for MDS section I (Active Diagnoses) in the Long-Term Care Facility Resident Assessment Instrument 3.0 User’s Manual require that every diagnosis coded in this section be documented by a physician or a physician extender (i.e., a nurse practitioner, physician assistant, or clinical nurse specialist when allowed under state law), points out Carol Maher, RN-BC, RAC-MTA, RAC-MT, RAC-CTA, RAC-CT, CPC, director of education for Hansen, Hunter & Co. PC in Vancouver, WA.

“This RAI User’s Manual requirement applies whether the diagnosis is captured by an ICD-10-CM diagnosis code or checkbox, and ICD-10 codes entered on the MDS also must meet the coding guidelines in the ICD-10-CM Official Guidelines for Coding and Reporting,” explains Maher.

In many nursing homes, there is a gap between the documentation in the medical record and the diagnoses that are coded on the MDS—and that gap is the physician-documented diagnosis, says Maher. “For example, many residents are at risk for malnutrition. However, it can’t be assumed from other clinicians’ documentation. The diagnosis has to be written by the physician. The way to fill in that gap is the physician query.”

A physician query is a communication tool, long used by coders and clinical documentation improvement (CDI) specialists in other healthcare settings, to request that the physician clarify patient diagnoses or procedures, explains Maher. “A query can be a powerful method of clarifying documentation in the medical record and achieving accurate ICD-10 code assignments. In nursing homes, it also can be used to confirm diagnoses—or identify additional diagnoses—for checkbox items in section I.”

It’s important that nurse assessment coordinators (NACs) and other MDS assessors not assume a diagnosis from something the physician has written if it is not clear, adds Maher. “You need to query the physician because it is not up to the coder to try to figure out what that diagnosis could have been or should have been. Failing to query is a significant risk for providers.”

Why a strong query process matters

Ensuring that section I is accurate and up-to-date has always been important. Active diseases and infections have “a direct relationship to the resident’s current functional status, cognitive status, mood or behavior status, medical treatments, nursing monitoring, or risk of death,” says the Centers for Medicare & Medicaid Services (CMS) in the Intent statement (page I-1 in chapter 3) for section I in the RAI User’s Manual. That direct relationship means that section I affects care planning, quality of life, and multiple quality measures (QMs).

However, accuracy has become even more critical under the Medicare Part A Patient-Driven Payment Model (PDPM) for the following reasons:

  • ICD-10 codes captured in item I0020B (ICD Code/Resident’s Primary Medical Condition) determine the resident’s default primary diagnosis clinical category for the physical therapy (PT), occupational therapy (OT), and speech-language pathology (SLP) components. The specificity of the diagnosis can make the difference between a code assignment that maps correctly into PDPM vs. one that is designated return to provider (i.e., rejected).
  • Certain ICD-10 codes entered in I8000 (Additional Active Diagnoses) count as comorbidities for the SLP component or the non-therapy ancillary (NTA) component.
  • Certain diagnoses captured by a section I checkbox can affect the NTA component, the SLP component, and even the nursing component. For example, I5200 (Multiple Sclerosis (MS)) can impact PDPM case-mix classification in both the NTA and nursing components:
  • NTA comorbidity (2 points); and
  • Nursing Special Care Low qualifier (with nursing function score <=11).

Note: To learn more, review the PDPM Calculation Worksheet for SNFs in chapter 6, “Medicare SNF PPS,” of the RAI User’s Manual in combination with the current PDPM ICD-10 Mappings tool.

Establish a query process

“NACs should work with medical records staff or a medical records consultant, as well as the medical director, to develop an effective physician query process that can be used to add a new diagnosis or to clarify an existing diagnosis for greater specificity,” says Maher. “You want to define a query procedure from start to finish, including who will make queries and how forms will be transmitted (e.g., fax or another HIPAA-compliant method).”

Create facility-specific query guidelines

“As part of setting up a query process, providers should establish facility-specific guidelines for when it is appropriate to query,” suggests Maher. “You want to feel comfortable knowing when a query is the right step, and you also don’t want to overwhelm physicians with so many queries that they ignore them. Guidelines help you establish that balance where you can get the diagnostic information that you need and keep the process efficient and effective.”

Common situations in which queries are necessary include the following, according to Maher:

  • A culture lab report indicates an infectious organism, and there is no documentation substantiating the significance of the organism in the disease process.
  • The resident has a diagnosis of urosepsis, which is nonspecific and does not support coding urinary tract infection (UTI) or sepsis.
  • The physician has documented a diagnosis of anemia without etiology.
  • The resident has a dietitian-documented body mass index (BMI) greater than 40 without a diagnosis of overweight, obesity, or morbid obesity.
  • The resident has a dietary assessment of low body weight and poor intake without a diagnosis of malnutrition or at risk for malnutrition.
  • A negative outcome occurred, but there is no clear indication in the medical record whether the outcome was a complication of care.

Develop nonleading query templates

Query templates can help providers ensure that physician documentation requests are consistent, says Maher. “Having routine query forms also allows physicians to get used to filling them out.”

The one hard-and-fast rule of query forms is to avoid leading the physician to a diagnosis, says Maher. “You shouldn’t signal to the physician, ‘Here is the diagnosis that we want,’ by asking, “Do you agree that the resident has this?,’ ‘Are you treating the resident for that?,’ or ‘Can your diagnosis of X be further specified as this?’ You want your query forms to give the physician the ultimate authority to make a diagnosis. Leading query forms can contribute to inaccurate diagnoses. They can also give medical reviewers, state surveyors, and other auditors the perception that you are telling the physician what to do.”

That can have a negative impact on the resident, the individual nursing home, and even the entire sector. For example, in a May 2021 report, the Office of Inspector General found that “nearly one-third of residents who were reported in the MDS as having schizophrenia—a diagnosis that excludes them from CMS’s measure of antipsychotic drug use—did not have any Medicare service claims for that diagnosis.”

A September 2021 New York Times investigation followed up with accusations of phony diagnoses that spurred the overutilization of antipsychotic medications, and that media spotlight led to a new OIG audit, slated to be released by this fall, to assess trends over time related to “(1) the use of psychotropic drugs for elderly nursing home residents; (2) citations and civil monetary penalties assessed to nursing homes regarding psychotropic drugs; and (3) the presence of diagnoses that exclude nursing home residents from CMS’s measure of the use of antipsychotic drugs.”

One way to combat any potential tendency to lead the physician is to use a multiple-option format, says Maher, “For example, if the dietitian’s nutrition assessment indicates that the resident is at risk for malnutrition, you could concisely present these clinical assessment findings (e.g., height, weight, albumin levels, etc.) at the top of the query form. Then, you could ask, ‘Is there an appropriate diagnosis that you would like to add to the medical record based on these findings?’”

This is where the options could come into play, says Maher. “Continuing the malnutrition example, you may want to offer yes or no checkbox options for malnutrition and at risk for malnutrition, as well as giving the physician the option to document another diagnosis or explain why a diagnosis cannot be clinically determined. In addition, some query forms ask the physician to enter the diagnosis in a progress note.”

Maher pulls together those tips to offer the following example of a nonleading query form:

Query Example  

This resident has a documented diagnosis of urosepsis with a temperature of 102 degrees Fahrenheit, confusion, and low blood pressure. Per the ICD-10-CM coding guidelines, urosepsis is a nonspecific term. It is not to be considered synonymous with sepsis. It has no default code in the Alphabetic index. Based on your clinical judgment, would you add any of the following diagnoses to the resident’s clinical record?  
A. Urinary Tract infection    Yes_____  No_______
B. Sepsis    Yes_______  No _______
C. Other_________________
D. Clinically undetermined _________________  

Signature________________    Date _____________

“It should be relatively easy for the medical director to introduce attending physicians to the query forms and to stress the importance of responding in a timely manner to keep diagnoses up-to-date for MDS and payment purposes,” adds Maher.

“When healthcare switched from ICD-9 to ICD-10 in 2016, CMS educated physicians on the need to be more specific when making and documenting diagnoses—and that they would be asked to provide clarifications more frequently,” she notes. “So, they are now used to supplying additional documentation in response to queries from hospitals and other providers. Nursing homes just need to catch up.”

Monitor to push improvements

Once the query process is up and running, whoever is overseeing that process needs to do some monitoring to assess opportunities for improvement, suggests Maher. “You should look for patterns of queries. In other words, does your team make repeated queries on the same topic, such as anemia or pneumonia? If you find such patterns, you should plan to educate both coders and physicians on the query process and documentation issues.”

It’s also important to identify top “offenders,” says Maher. “If certain physicians are being queried frequently, investigate the reasons for the queries to determine what additional education may be needed.”

Last but not least, the responses to queries need to be checked, says Maher. “This review can reveal discerning use of queries by the coding staff (i.e., whether they are querying unnecessarily) and/or poor physician documentation practices. It can also reveal whether necessary reports (e.g., the discharge summary, operative reports, etc.) are included in the medical record before coding.”

Note: For additional insights, see the American Health Information Management Association’s (AHIMA) Guidelines for Achieving a Compliant Query Practice (2019 Update) and Physician Query Examples.

For permission to use or reproduce this article in full or in part, please complete a permissions form