To restate:

35-year-old man admitted for worsening ascites. We know the patient because he had just left AMA 3 days before.  He had HIV with a low CD4 and did not take any anti-retrovirals.  He also had hep C cirrhosis with encephalopathy and worsening ascites.  He did take spironolactone and lactulose and the lactulose was causing diarrhea.

We go to the ED and find these labs:

Of course we need an ABG on room air

Interpretation:

The patient had respiratory alkalosis, likely secondary to hepato-pulmonary syndrome.

I find this quote on compensation:

Studies have shown an average 5 mmol/l decrease in [HCO3-] per 10mmHg decrease in pCO2 from the reference value of 40mmHg. This maximal response takes 2 to 3 days to reach.

This compensation seems appropriate.  Therefore a chronic respiratory alkalosis.

At morning report we had 3 guesses for normal gap acidosis.  The guesses included:

1. Type IV RTA secondary to spironolactone – the hyperkalemia always comes first, as the hyperkalemia leads to the decrease in NH₃ production
2. Diarrhea – the diarrhea must be a large volume.  Decreased GFR can make this more likely, but he did not have that.  Finally, his diarrhea would not be basic as he was taking lactulose which acidifies the stool
3. Distal RTA secondary to liver disease – I had not heard of that, so I looked up causes of distal RTA

Type I RTA is caused by a variety of conditions, including:

1. Amyloidosis
2. Fabry disease
3. Sickle cell disease
4. Sjogren syndrome
5. Systemic lupus erythematosus
6. Wilson disease
7. Use of certain drugs such as amphotericin B, lithium, and analgesics

To repeat a mantra – you should not assume a normal gap acidosis from the electrolyte panel (you can assume an increased anion gap acidosis).  Similarly when you see an elevated bicarbonate and the clinical situation does not fit metabolic alkalosis, you should think of the possibility of compensation for chronic respiratory acidosis.

I suspected respiratory alkalosis in this case because none of the potential causes of normal gap metabolic acidosis made sense.
 

35-year-old man admitted for worsening ascites. We know the patient because he had just left AMA 3 days before.  He had HIV with a low CD4 and did not take any anti-retrovirals.  He also had hep C cirrhosis with encephalopathy and worsening ascites.  He did take spironolactone and lactulose and the lactulose was causing diarrhea.

We go to the ED and find these labs:

I ask the team why the bicarbonate is low.  I presented this a morning report also and asked students, housestaff and applicants!

I looked at the old labs and this bicarbonate level was present on the previous admission and had remained stable.  What do you think?

50 something patient with combination of restrictive and obstructive disease.  This electrolyte panel caught everyone's attention.

ABG

What likely happened to this patient?  What do you want to do now?

50 something woman admitted for nausea and recurrent vomiting and right flank pain.  She denies hematemesis or melena.  She smokes 1 pack per day and drinks 8-9 beers per day.  She is thin, weighing around 54 kg.

Labs are drawn, and the patient is admitted.

Your job is to identify all the abnormalities in this panel, and suggest the sequence of events most likely to result in these numbers.  What other information do you want?  

This patient was presented at morning report.  I omitted her pulse of 131 and BP of 110/70 (orthostatics not checked)

My analysis at this point:

1. Hyponatremia – need urine osms to distinguish between beer potomania and ADH release from volume contraction.  I suspected that both causes were involved.  This degree of hyponatremia bears attention, but is not dangerous.
2. Increased anion gap acidosis – give the huge (44) anion gap I knew that the patient also had a metabolic alkalosis.  Since I recognized the clinical situation (having seen it many times) I strongly suspected a combination of lactic acidosis and ketoacidosis.  This could represent alcoholic ketoacidosis and/or starvation ketosis.  Regardless, the patient needs volume and glucose (obviously with thiamine).
3. Metabolic alkalosis – virtually all anion gaps above 40 imply metabolic alkalosis.  When you do the delta gap in this patient, you find a delta of approximately 32.  Adding that to 8 (the observed bicarbonate) suggests an underlying bicarbonate of 40.  The hypokalemia supports a metabolic alkalosis from vomiting.
4. The BUN/creatinine ratio supports volume contraction.
5. The glucose is a bit low, but I did not consider that an important problem.

I would have given this patient D5/NS + 20 mEq/L and run at around 200 cc/hr.  

Several hours later they obtained another electrolyte panel and an ABG.  I predicted several more important considerations after seeing these numbers and considering the clinical context.  His serum osms showed no osm gap.  His urine osms were 300.  His lactic acid level was elevated and his urine had positive ketones.

Now further thoughts if you are the hospitalist caring for this patient.  What other tests do you need?

50 something woman admitted for nausea and recurrent vomiting and right flank pain.  She denies hematemesis or melena.  She smokes 1 pack per day and drinks 8-9 beers per day.  She is thin, weighing around 54 kg.

Labs are drawn, and the patient is admitted.

Your job is to identify all the abnormalities in this panel, and suggest the sequence of events most likely to result in these numbers.  What other information do you want?  

I am giving a talk at the ACP meeting this coming year on dangerous electrolyte disorders.  Please vote on which disorders I should emphasize.

My current plans include:

Hyponatremia

Hypophosphatemia

Hypercalcemia

but I am still considering

Hypokalemia

Hyperkalemia

Hypomagnesemia

I need your help – what disorders would interest you most?

Now obviously this is my opinion, and other may differ.

We have an alcohol abusing woman who presented with a Na of 110 mEq/L and depressed mental status.  She awoke as saline increased her sodium quickly to 120 mEq/L.  This occurred because her urine osms of 150.

This patient admitted to a beer diet.  Her history fit beer potomania.  Normally the urine osms in beer potomania are less than 100, but she also presented with volume contraction from vomiting.  I suspect that she had some ADH from the volume contraction.

Given her dilute urine and underlying alcohol abuse, we worried that she had a predisposition for central pontine demyelination or an extra pontine demyelination syndrome.  Therefore the resident made the bold and correct decision to give vasopressin.  The idea behind giving vasopressin in this situation follows from an understanding of the true problem.  This patient would correct her sodium too quickly because she would excrete too much free water.  Therefore, the vasopressin would prevent the free water excretion and slow the sodium rise.

The vasopressin worked and the patient's sodium rose at a slow and desirable pace.

To recap: