A Tale as Old as Time: Heart Failure & Fluid Overload Part 2

In the last instalment, Dr Felicity Sinclair-Ford went on a quest to find all the coding standards and advice relevant to heart failure and fluid overload. This time, she applies a clinical perspective to these rules.

Now that we’ve found the relevant standards, let’s dig into the implications, contradictions, and clinical considerations of this plethora of advice.

Coding Matters Volume 7 Number 3 (Retired)

While this rule has been retired, I do think it’s useful to consider the evolution of the coding rules in this area.

By advising coders that coding fluid overload or pulmonary oedema is not necessary in heart failure patients, the rule seemed to be suggesting that these sequelae are part and parcel of the clinical picture of heart failure.

However, this isn’t always the case. And it’s important that the codes assigned describe the patient journey.

Let’s consider two patients admitted with neck of femur fractures with a past medical history of heart failure.

Patient A is reviewed prior to surgery by the anaesthetic registrar. The registrar requests an echocardiogram to assess the patient’s heart failure. The echocardiogram findings are similar to one two years prior, and the patient has a successful surgery and anaesthetic.

Patient B experiences an exacerbation of their heart failure following surgery. They become profoundly fluid overloaded and develop significant pulmonary oedema requiring a short duration of non-invasive ventilation.

You can see that if both these patients are only assigned the code for heart failure, they look quite similar on paper, but were vastly different in reality.

Another possibility is that this coding rule was considering fluid overload a “symptom” of the heart failure. However, describing fluid overload as a “symptom” or “sign” does not adequately reflects its complex relationship with heart failure.

In seeming agreement with this perspective, fluid overload is not in the “symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified” chapter, but rather in the endocrine chapter.

It would therefore seem that fluid overload is not just a “sign” of heart failure and should be considered as a distinct clinical entity when devising coding rules.

So, we can all agree that the retirement of this rule was definitely a step in the right direction.

VICC Coding Query 3089 and VICC Coding Query 3118 September

The advice for both of these queries is pretty much the same, so we can consider them together.

These rules state that if both fluid overload and heart failure “independently” meet the criteria for coding, both can be assigned.

But what does “independently” mean? Can a distinction such as this apply to such a complex pathophysiological relationship? We need to consider some possibilities for how this might work in practice.

1. Does the treatment for fluid overload DIRECTLY treat heart failure? Or does the treatment of fluid overload SECONDARILY treat heart failure?

Usually, diuretics are the mainstay of treatment for fluid overload. So, the question we need to consider is whether the diuretics themselves independently improve cardiac function? Or are we improving cardiac output by treating the fluid overload?

Trying to research this topic, I ended up bouncing around from PubMed to physiology websites, therapeutic guidelines to the national library of medicine.

The only definitive statement I can make that this is a very complex topic. Helpful, I know.

For example, when considering if improving fluid overload improves cardiac function it seems that the reduction in afterload would indeed achieve this. However, we are also reducing preload on the heart, which might reduce stroke volume in right sided heart failure. Furthermore, aggressive diuresis, as well as intravascular fluid depletion, can further reduce kidney function, which can negatively affect cardiac function.

You can see the conundrum.

The interaction between cardiac and renal pathophysiology, and the advantageous and deleterious effects of diuretics seem unable to be completely teased apart.

To me this suggests that fluid overload and heart failure are so intertwined that considering them as separate entities, and insisting that they each “independently” meet coding criteria, is reductive.

2. Are we treating fluid overload or are we treating exacerbation of heart failure?

In a completely well patient with normal physiology who has become fluid overloaded (for example, iatrogenically) we would not need to treat. The patient’s cardiovascular and renal systems will restore homeostasis. Therefore, in a heart failure patient with fluid overload we are treating the exacerbation of heart failure, rather than just the overload.

And I think most doctors would consider this clinical picture an exacerbation of heart failure rather than an isolated “fluid overload”.

3. Are the treatments we give to treat the fluid overloaded heart failure patient treating BOTH fluid overload and heart failure?

And what do the coding standards say about ONE treatment treating TWO conditions. Application of ACS0002 would seem to say that both codes could be assigned based on the following documentation.


  • Exacerbation of heart failure treated with frusemide
  • Fluid overload treated with frusemide

Even though the treatment is actually the same, each condition has been linked with a commencement of treatment; therefore, we can code them both.

It seems startling that the coding standards do not have specific direction regarding whether one treatment can meet ACS0002 for two conditions. (Is it okay to assign a code for cellulitis and UTI if both are treated with cephalexin? We think so?)

Coding Rule Q2730 and Western Australian Coding Rule 1017/07

This rule advises to code fluid overload as per “problems and underlying conditions”.

So if a patient with no previous medical history presents in fluid overload, and investigations during the admission establish a diagnosis of heart failure, then heart failure is the principal diagnosis, and if we follow the seizures/brain tumour example in ACS0001, then the “problem” is not coded if the underlying diagnosis is established during the episode of care.

However, (there’s always a however), in this example “seizures” is an R code, a “symptom, sign or abnormal finding”).

And we’ve established that fluid overload in heart failure is much more than a “sign”, it’s a clinical problem in its own right! So not getting the fluid overload code in this instance doesn’t reflect the clinical picture.

If we apply ACS0001 to patients with known heart failure, doesn’t this mean that if a patient with a known history of heart failure presents in fluid overload, we would code the fluid overload as the principal diagnosis, and the heart failure, as the known underlying condition, is an additional diagnosis?

But we don’t really do that? Do we? And how does this stack up clinically?

In AR-DRG version 10, a principal diagnosis of fluid overload puts you into the DRG L65… Kidney and Urinary Tract Signs and Symptoms. Yikes.

Personally, I think most cardiologists would be startled to hear that their “exacerbation of CCF” patients were being assigned to this DRG.

Considering this, I think the WA coding rule’s recommendation to respect the clinician’s designated principal diagnosis is a good one. And most doctors will consider the principal diagnosis in these cases to be “exacerbation of heart failure”.

What about for additional diagnoses? Well, according to the “problems and underlying conditions” section of ACS0002, both codes can be assigned as long as the link is made between the problem and the underlying condition. Consider the following documentation.

Fluid overload secondary to heart failure --> frusemide, 1.5L fluid restriction, daily weights.

If we apply “problems and underling conditions” to the letter, both codes are assigned.

Circling Back

“A recurring issue for coding heart failure, is whether or not to code fluid retention as an additional diagnosis. Would be interested in your and others' opinion on this.”

How would I answer this now (after my down-the-rabbit-hole approach to a simple question)?

In terms of the conditions meeting the coding criteria “independently” as per the VICC coding advice, these conditions are so closely linked, and the management so overlapped, that, in my clinical opinion, to treat one is to treat the other, and both codes should be assigned.

In terms of “problems and underlying conditions”, if the problem we’re treating is fluid overload, and the underlying cause is heart failure, both codes can be assigned.

However, this all hinges on one gargantuan assumption: that clinical documentation is adequate.

Usually, if the codes assigned are not reflecting the clinical truth, it’s not a coding issue, and it’s certainly not an issue of tiny nuances in coding rules. It’s almost always an issue with the clinical documentation.

So the moral of the story?

CDI is a journey, not an event, it’s here to stay, and we all need to get on board.


Addendum: A Suggestion from Left Field

When considering, researching, and obsessing over this topic (I’m a GREAT dinner party guest), I began to consider if the issue isn’t with the coding rules, but with the codes themselves.

I was thinking how most doctors would consider the principal diagnosis to be “exacerbation of CCF”. And I began to consider that for other conditions, we have codes that distinguish between those “with exacerbation” and those “without exacerbation”.

As I discovered in my podcast with American physician advisor Erica Remer, the ICD-10-CM has over twenty codes regarding heart failure. And while I do think that’s a little bit much (sorry America), it certainly got me wondering as to whether our measly three codes (I50.0, I50.1, and I50.9) can reflect our ever-changing understanding of this complex disease.

Will this be improved in the ICD-11?

The ICD-11 alphanumeric system is a marked change from the ICD-10; however, it now allows specificity to be added to the code regarding the New York Heart Classification, the ejection fraction, the cause, and the chronicity.

And while this might be a good guide, surely we have our own “Australian Modification” for a reason. Wouldn’t it be great if new heart failure codes reflected how Australian clinicians actually speak, and we could put this whole debate around coding heart failure and fluid overload to rest.

Imagine if we had the codes “Acute exacerbation of heart failure with fluid overload without mention of pulmonary oedema” or “Heart failure without exacerbation”.

Ah well. A girl (CDI Medical Projects Consultant) can dream.


We invite you to share your ideas, experiences, and achievements in CDI by submitting content to the CDIA Community!  Contact community@cdia.com.au to learn more.

Want to read and view more?

This content is available to CDIA Community members only.

Log In Register Now