Q: Our cardiologists like to document “biventricular heart failure.” Is a query needed to clarify systolic and/or diastolic?
A: A query should be issued for the chronicity and type of heart failure when the physician states only “biventricular heart failure.” Biventricular represents that both ventricles have mechanical problems, but it doesn’t specify which phase of the cardiac cycle (systole or diastole) is encountering a pumping/filling issue.
If you have access to an encoder and type in “biventricular heart failure” you will receive prompts to add in participating factors, and asked to choose what type of heart failure (systolic, diastolic, combined, etc). If you pick “unspecified,” you end up with code “428.0 Congestive heart failure, unspecified.”
Editor’s Note: Vicki Sullivan Davis is CDI manager at Cone Health System at Alamance Regional in Burlington, North Carolina, and past-speaker at ACDIS National Conference. Contact her at firstname.lastname@example.org.
Q: If a is patient admitted with malnutrition and the physician documented the patient to be malnourished from mild to severe, would the CDI team use DRG 641, Severe Malnutrition as a working DRG or should we query the physician to clarify the severity or type of malnutrition.
A: DRG 641, Severe Malnutrition would require use of ICD-9-CM code 261, Nutritional Marasmus, which is a high-risk diagnosis vulnerable to denial. The same is true for ICD-9-CM code 260, Kwashiorkor. These conditions describe a very specific type of severe malnutrition typically found in third-world countries and doesn’t typically exist in the U.S.
Even if it wasn’t a vulnerable diagnosis the difference between mild and severe malnutrition constitutes the difference between a CC and an MCC designation. So first look to see what clinical indicators and treatment support were documented in the record. In my opinion, the treatment is often what separates mild nutrition from severe malnutrition. If the clinical indicators and treatment support severe protein calorie malnutrition (ICD-9-CM code 262), I would query the provider to clarify their documentation as to the type.
The following are some additional articles that might help shed a little more light on these conditions and their relative controversy over the years:
- Q&A: Review clinical criteria for malnutrition
- Tip: Ensure malnutrition documentation and coding meet current clinical standards
- News: Hospital settles allegations of False Claims Act violations related to malnutrition and leading queries
- New malnutrition criteria could help ensure consistent coding
- Dietitian involvement resolves malnutrition query quandary
- Body Mass Index and malnutrition: Interrelated comorbidities
- Take a closer look at clinical indicators of malnutrition
Q: What should I do if I see a non-compliant query in the chart? Should I remove it, let my co-worker know, or just leave it in the chart?
A: Addressing non-compliant queries can be tricky. The best course of action would be to share your concerns with your supervisor who can then either confirm your perception of the query being non-complaint or could let you know why he or she feels the query is acceptable. Ask your manager or supervisor to go over any internal query policies to help you better understand your facility’s compliance parameters.
Most facilities have standard query policies and procedures which reflect national standards (such as the 2013 AHIMA/ACDIS “Guidelines for Achieving a Compliant Query Practice” brief). They also have processes in place to help co-workers handle questionable query processes.
If there are no policies and procedures in place (or if you and your coworker are only the two CDI staff querying physicians at your facility) you may want to review the latest query practice information together and approach whatever management team is in place to develop such policies yourselves.
If the query is truly non-compliant, I would definitely want the supervisor to address it rather than you doing so on your own. It may be that the individual needs additional training or it may become a potential performance issue. In which case your manager or supervisor needs to know about the situation and may even need to have a documented conversation with the CDI team member who left the query.
You wouldn’t want to remove the query. The physician may have already reviewed it and responded in his or her progress note. If auditors or internal staff later question where that diagnosis came from, no query trail would exist and you may not be privy to those subsequent questions. If the supervisor or program manager determines the query was indeed non-compliant he or she may need to also circle back to discuss the situation with the physician and/or coding team.
Q: I was told that a multiple choice query should have at least four options. Keeping in mind that there may be only one reasonable option in a multiple choice query, what would be a good fourth option for a query about hyperkalemia if the other options are:
A: There are many myths concerning compliant query practices so before automatically accepting a dictum of query parameters go back to the official sources to ensure compliance. By this I mean first reference the most recent guidance from the Association of Health Information Management Association (AHIMA) and the Association of Clinical Documentation Improvement Specialists (ACDIS). AHIMA is one of the four cooperating parties (along with CMS, American Hospital Association, and the National Center for Health Statistics) so its recommendations have additional credence should auditors or other investigators question your CDI program practices.
According to the 2013 “Guidelines for Achieving a Compliant Query Practice:”
“Multiple-choice query formats should include clinically significant and reasonable options as supported by the clinical indicators within the medical record, recognizing that there may be only one reasonable option. Multiple-choice query formats should also include additional options such as ‘clinically undetermined’ and ‘other’ that would allow the provider to add free text. Additional options such as ‘not clinically significant’ and ‘integral to’ may be included on the query form if appropriate.”
If you still feel a fourth choice is needed perhaps the choice of “not clinically significant” could be offered. But this would depend on the circumstances of the particular patient encounter.
The 2013 practice brief also provides an option for yes/no queries. However, the brief does recommend that even in yes/no queries that additional options be included, similar to those recommended for multiple-choice queries.
“The ‘yes/no’ query format should be constructed to include the additional options associated with multiple-choice queries (i.e., ‘other,’ ‘clinically undetermined,’ and ‘not clinically significant’ and ‘integral to’). Yes/no queries may not be used in circumstances where only clinical indicators of a condition are present and the condition/diagnosis has yet to be documented in the health record. Also, new diagnoses cannot be derived from a yes/no query.”
Again, refer to the practice brief for additional circumstances where yes/no queries may be warranted and read up on previous practice brief recommendations for a better understanding of how queries should be formatted.
Q: I understand that most CDI departments develop a standardized list of clinical indicators/criteria to support query efforts. Is this something we need to develop or is it available in the encoder process? If we need to develop this, how do we go about that?
A: While the AHA’s Coding Clinic for ICD-9-CM (ICD-10-CM/PCS) often lists clinical indicators for specific diagnoses, the publication should not be used as a stand-in for the provider’s own clinical judgment, as reiterated in Coding Clinic, First Quarter, 2014, p. 11.
CDI programs should work with the physician team to develop a standardized list of clinical indicators for the team to use in query creation, CDI and physician training, and record review. Such mutually developed criteria is particularly helpful for highly vulnerable or often miss-documented conditions such as levels of malnutrition severity, acute and chronic respiratory failure, acute kidney injury, encephalopathy, etc.
Research nationally established guidelines for these common, core conditions (e.g., ASPEN criteria for malnutrition, or RIFLE or NKIDO criteria for renal failure), then work with the specialty most closely related to that diagnosis (e.g., pulmonologists for acute respiratory failure). This criteria could then be consistently used by CDI and coding staff to initiate a query to support the diagnosis.
Q: At what stage should an established program most likely experience a reimbursement plateau? One may naturally expect the physicians to improve as CDI programs hammer them with education. After we’ve gathered all the low-hanging fruits and go for the mangos? We ran the top principal diagnoses and also top diagnoses for our system. We are a home-grown program, about three years old now. Our team made $6 million last year. I know there are many query opportunities and ideas for program expansion but how do we find the right areas for our facility?
A: I am unaware of any industry standards that identifies a timeline of expectations for a plateau of reimbursement/physician documentation improvement. If you have a relatively stable medical staff with few changes, the program should mature and demonstrate physician documentation improvement more quickly than an organization that experiences high turnover of physicians, such as a university or teaching hospital. A teaching hospital may never plateau as the influx of residents and the constant rotation among specialties means educating physicians and capturing the “low hanging fruit” never ends. Each organization will have their own rate of turnover and educational needs for medical staff.
Such programs would have a lower query rate but maintain increased levels of CC/MCC and severity of illness/risk of mortality (SOI/ROM) capture. In other words, the physicians have retained and applied the education, they require less questioning but their documentation supports higher reimbursement levels.
The second variable in this equation is that as a CDI program matures the staff will find “different trees of low-hanging fruit” to pick. At the beginning, you learn to recognize potential documentation opportunities amongst the apples and oranges and begin to see a decrease in queries related to these but you learn there are opportunities in the lemon trees and the mangoes.
Lastly, organizations are constantly adding new services, new procedures and with each change CDI programs may identify entirely new opportunities. In other words, I have not seen in my experience a leveling off but more of an evolving focus as a program matures.
Self-education and participation in CDI networking are important in advancing your own career, your own knowledge, and your CDI program efforts. If you do not currently have tracking systems in place for individual CDI specialists’ and individual physicians’ query behaviors, you may want to. This might identify specific learning needs for individuals that could be targeted. For example, what diagnoses are the CDI specialists querying for? Does one CDI staff member miss sepsis opportunities or are there opportunities the entire team needs to learn about? If there a specific physician that needs intervention on a particular diagnoses?
Most successful CDI programs work closely with their coding teams. Expand on this collaboration by having the coding staff bring forward any trends or difficulties they’re seeing in daily practice. Ask to review any retroactive queries for trends and trouble spots. Identify any documentation improvement opportunities the CDI staff may have missed. The idea being, you and your CDI team won’t miss that opportunity next time.
If you do not have access to your organization’s Program for Evaluating Payment Patterns Electronic Report (PEPPER), seek out access. PEPPER is produced by CMS and it compares your organization to like organizations within your region. It identifies where you maybe an outlier for specific diagnoses and CC/MCC capture. You may be able to identify improvement opportunities in areas where you are a low outlier compared to your peers.
I often found my new fruit by reviewing the code set. Just opening up the code book and seeing what specificity was needed in code assignment for specific diagnoses often demonstrated for me areas of needed improvement. I would suggest doing that with theICD-10-CM code book. You can start asking questions related to ICD- 10 now so that the learning curve will not be so steep come October of next year.
Lastly, an exercise I suggest for new CDIs and one that might require repeating as the definitions do change is to take the list of CC/MCCs and highlight those diagnoses that are often seen within your population. You may find there is a “fruit basket” just sitting there that you never considered. If you identify codes that you have not thought to ask for look them up in the code books and learn what terms are needed to support their documentation. I promise you this effort will bring to light at least one diagnosis common to your population that you might not be capturing on a regular basis.
Wikipedia tells me there are over two thousand different fruits in the world so you have many to harvest!