Yes, it's a little pocket of insanity but does it have something to teach us?

A few months ago there was a screen shot of a clinic note floating around the internet. It was, in the words of another nephrologist the most passive-aggressive nephrology consult note I have read in a long time:

Patient's sodium dropped further to 120 in the evening. He has had a precipitous drop that I suspect is due to over-diuresis, which does not seem to be a diagnosis within the lexicon of heart failure cardiologists. It is possible that he could have developed SIADH, through a drug side effect. In any case, we have reached the usual place where attempts to fix the heart have blithely interfered with renal physiology, and I am not willing to let his serum sodium decline into the 110s. If we give NS, and he has SIADH, we will worsen his serum sodium. We could use 3% NS, but he is not having mental status changes, yet, and this is bad form for a patient in heart failure. If he is volume depleted, and we use conivaptan, he could develop hypotension which would be difficult to fix. So I seem to have been finagled into ordering tolvaptan, which will hopefully prevent any further decrease tonight. Tolvaptan fixes a number and has not been shown to improve clinical outcomes with chronic use.

Clearly there is a large dose of crazy in the assessment and plan, but it highlights a number of real issues in hyponatremia. Let's dissect the note a bit and tease out the best parts.

The first clam that over-diuresis does not seem to be a diagnosis within the lexicon of heart failure cardiologists seems to be true. A brief survey of google finds a paucity of relevant hits for the phrase and most of those are from nephrologists or family practitioners. Given the frequency that I see patients suffering from this I was a bit shocked at these results.

The next sentence seems a bit preposterous, It is possible that she could have developed SIADH, through a drug side effect. Presuming that a patient with heart failure induced hyponatremia now has a second denovo disease seems a bit of a stretch, but we don't have access to the clinical data and so it is hard to determine if this is true. However the definition of SIADH requires that patients be euvolemic and judging from as much of the story as we know it seems like this patient is clinically hypervolemic. This rules out a clinical diagnosis of ADH, because the release of ADH in heart failure is due to physiological trigger for ADH, a decrease in perfusion. The disease of SIADH is specifically reserved for patients in which there is no physiologic stimuli for ADH release. The presence of heart failure and volume overload, definitionally rule out SIADH.

The next sentence is interesting: In any case, we have reached the usual place where attempts to fix the heart have blithely interfered with renal physiology, and I am not willing to let his serum sodium decline into the 110s. Diuretics increase water and sodium loss, but the cation content of the urine is almost always significantly lower than the plasma cation content, urinary sodium with loop diuretics is typically around 70 mmol/L. So use of loop diuretics cause loss of relatively more water than sodium and result in hypernatremia, except in heart failure. To understand why, one needs to understand electrolyte free water clearance (and an example of using it in the treatment in hypernaatremia is here). The higher the free water clearance, the less prone patients are to hyponatremia. Here are the calculations for electrolyte free water clearance for a patient with hyponatremia due to CHF before and after the addition of loop diuretics:

Before diuretics

In CHF, the patient is actually doing a pretty good job clearing free water. More than half of the urine output is electrolyte free water, the character of the urine is appropriate for correcting the hyponatremia. The problem is not the character of the urine but the amount. The patient just doesn't make enough urine to generate adequate electrolyte free water to account for the water the patient is drinking. Water restriction will be effective for these patients.

So, if the problem is an inadequate amount of urine the logical next step would be to increase the volume of urine with a diuretic:

With diuretics

Unfortunately, though the diuretic increases the volume of urine, it also changes the character of the urine. In this case, it dramatically increases the urine sodium content. This makes the urine almost completely ineffective at removing electrolyte free water and the net result is that the electrolyte free water clearances actually falls with the addition of the diuretic. This is the trap our poor nephrologist is raging against.

The next sentence of the rant: If we give NS, and he has SIADH, we will worsen his serum sodium. We could use 3% NS, but he is not having mental status changes, yet, and this is bad form for a patient in heart failure. I have a problem with this sentence and do not think it is well thought out. The nephrologists clearly believes this patient is over diuresed, we see this situations all the time and they respond briskly to additional fluids. Don't complain that the patient has been overdosed with diuretics and then refuse to provide the antidote for this overdose. I am skeptical of his theory that the patient has SIADH and it is a bit unreasonable to even give a trial of 0.9% saline. In regards to 3% saline, our good doctor might want to take a look at some of the data on the use of 3% saline in heart failure: SMAC-HF (PDF) or Seminars in Nephrology summary. While it is not standard of care the data are certainly intriguing and when the traditional approach is not helping the patient, as apparently is occurring in this patient, it may be worth a look.

The next sentence is regarding conivaptan: If he is volume depleted, and we use conivaptan, he could develop hypotension which would be difficult to fix. Conivaptan is a non-selective vasopressin antagonist, as opposed to tolvaptan which is a selective V2 receptor antagonist. Blocking V1 could cause hypotension as has been reported in multiple case reports. See this open access review.

So I seem to have been finagled into ordering tolvaptan, which will hopefully prevent any further decrease tonight. Tolvaptan fixes a number and has not been shown to improve clinical outcomes with chronic use. This complaint that tolvaptan fixes a number seems a bit obtuse for a nephrologists who was presumably consulted to fix a number. It also implies that tolvaptan is unique in that it has only been shown to fix a number. Well unfortunately, all the therapies of hyponatremia, outside of acute symptomatic hyponatremia have only been shown to fix a number. However given the profound morbidity associated with low sodium, it seems judicious to correct hyponatremia until data proves it is unhelpful. Additionally, to claim in one sentence that you refuse to allow the sodium to fall below 120 and in the next sentence to rail against a therapy that has only has been shown to fix a number seems to belie a profound lack of self awareness.

But keep fighting the good fight against the cardiologists, somebody needs to keep their egos in check. #PracticallySurgeons