Coding diagnoses that only appear after a CDI query.
I believe that the nature of CDI work is mainly through the sending of queries and their impact on subsequent documentation. The case most recently to raise up this issue is about a patient that was only here for three days (8/29-8/31) before he passed away. The CDIS sent three queries on day one and two were answered the next day. One (for consistently elevated creatinine) was forgotten and needed to be escalated to eventually get a response. Several providers were involved, the diagnosis of AKI was added to the Discharge Summary only. The coding department said the documentation didn’t meet the definition of secondary diagnosis to be coded, so our CDI medical director reviewed it again and discussed with the provider again how to document it more completely to meet. The provider added more documentation, again to the discharge summary. Still the coding department will not code this diagnosis. We participate in the Vizient Consortium and this diagnosis is a mortality risk variable, the documentation includes treatment, monitoring and progression of the diagnosis:
Acute kidney injury
- Cr in 2007 was .9
- this new AKI is likely secondary to cardiac arrest
- we monitored and treated with IV fluids.
Three questions:
1. Does the documentation here need to say “POA” if the labs show the creatinine elevation is present on admission?
2. When a diagnosis shows up in a discharge summary, a week after the patient is discharged, with clear documentation from CDI of escalation of said pending query, does that make it not meet criteria for coding or is it still “code-able”?
3. When a CDI department has access to the support of medical director(s), should they be involved in the decision making after these types of disagreements between HIM and CDI? And if so, how?
I guess my real question is if a diagnosis is true, clinically, but only shows up in one note, should it still be captured in the coding? I am concerned that the goal for coding is often times to avoid audits and I believe that we will always see audits and denials as it is part of the work we do. However, if we are always mindful of the goal of coding – to be to code what is true and to improve documentation to reflect that when needed, we will always be able to overcome audits and denials.
Comments
1. POA - if/when a pt is admitted with signs and symptoms of a disease subsequently diagnosed, it is coded with “Y” for POA. In your example, the patient presents with Cr elevation that, after study, met criteria for Acute Kidney Injury. The POA is ‘yes’. Another example would be a pt admitted with troponin elevation, EKG changes and Chest pain. Initial Assessmment might be only the the CP - but, after evaluation of data and the patient, a subsequent documented diagnosis of acute MI would be coded with POA as yes. There are examples of such logic in the POA guidelines for the Official Coding Guidllnes. Coding must recognize the time required to study cases in order to chart diagnoses. Many conditions that are Present On Admit may not be documented as such until data is studied and subsequently a diagnosis is rendered.
2. The summary is the ‘gold standard’ for documentation and coding purposes. Speaking as a former Coding Mgr and current CDI professional, I’d definitely code the N17.9 as the record documents the condition and there is clinical support. This begs an often asked question...”How many times must a condition be noted in order to be coded”?
Example: If a pt is admitted in the ED with Acute Respiratory Failure clearly stated and clinically supported and documented, even though it is noted ‘only once’, it is coded. The condition is documented, met criteria, and may be treated in the ED with subsequent improvemt of the condition by the time the patient goes to the floor. The notion a condition must be stated repeatedly is false...what matters is not how many times a condition is noted, rather it is how well it is stated and the evidence and strength of the clinical support that exists for that condition. There is not necessarily a need to repeat a resolved condition repeatedly, and expecting an MD to do so is futile.
3. I believe CDI should have the ability to work closely with and negotiate issues with coding. The adoption of site approved criteria for diagnoses is useful in this regard. We have coding professionals working as CDI, and this is very helpful in that regard.
Paul Evans, RHIA, CCDS
Then she is ignoring official advice cited in Coding Clinic as this publication specifically states we are to code acute respiratory failure if/when documented in the ED. Of course, it stands to reason, again, the condition must be 'clinically supported'.
Suggest CDI management discuss with Coding Management...does HIM have a policy that officially states they will 'ignore' ED documentation?
Paul
VOLUME 29 THIRD QUARTER
NUMBER 3 2012, Page 22
Emergency Department Physician’s Documentation of Respiratory Failure
Question: The patient presented to the Emergency Department (ED) in full cardiac arrest and respiratory failure due to an acute myocardial infarction. He was resuscitated, transtracheally intubated and placed on mechanical ventilation. The patient was admitted to the intensive care unit and after a short period he expired. The ED physician documented acute respiratory failure. However, the attending physician did not document acute respiratory failure in the health record. Is acute respiratory failure a codeable secondary diagnosis based on the ED physician’s documentation of this condition?
Answer: Yes, code 518.81, Acute respiratory failure, should be assigned based on the ED physician’s diagnosis, as long as there is no other conflicting information in the health record. Whenever there is any question as to whether acute respiratory failure is a valid diagnosis, query the provider.
In this situation, I ask the Attending to decide which is the most appropriate classification at the time of presentation to the ED. This is because the Attending should address any ‘dissonance’. Either way, I feel a coder does have the perogative to chose to ignore the condition that is documented - it needs to be addressed in some way, as per the particulars for each case. I have had many instances in which the Attending has ‘confirmed’ the diagnosis if/when the query cites standard clinical indicators. I’d discuss this with team members as this is a pretty common scenario? Best luck.
I have also had instances in which the Attending has stated the patient had hypoxia, but not ARF. Case by Case, as we like to say.
Paul