Albumin; we won’t get fooled again.

I have already pointed to Unexpected Normal Colloid Osmotic Pressure in Clinical States with Low Serum Albumin from my Twitter feed, and promised a more detailed commentary in this post. There is much nonsense talked about albumin in critical care circles, even though the evidence clearly shows there is essentially no indication for albumin therapy and an understanding of physiology explains why we should not expect any benefit.


The first thing to point out is that albumin is primarily a tissue protein. Intravenously infused albumin will be fully redistributed within 24 hours (within 8 hours in most ICU patients) and is very difficult to shift from the tissues where it will cause oedema. Let’s borrow from one of the most widely cited research papers on albumin ‘therapy’. Even the title misleads; “Albumin replacement in patients with severe sepsis or septic shock”. Sepsis does not cause any significant loss of albumin, so the idea of replacement is fundamentally flawed.  I feel my mentor Ludwig Wittgenstein approving of this very elementary criticism.

The second point is the sort-of-true-but-not-in-critical-illness factoid, stated in the ALBIOS report as “Albumin is the main protein responsible for plasma colloid osmotic pressure”. The Investigators were clearly not familiar with the work of Gilbert Park, who in 1997 reported on systematic measurements of serum albumin concentrations and plasma colloid osmotic pressure in 145 patients treated for a week or more on the ICU at Addenbrooke’s Hospital Cambridge. “Regression analysis of mean colloid osmotic pressure and albumin revealed that albumin only contributed 17% of the colloid osmotic pressure in these patients.” ref; PMID 9797519.

The third point is that the serum albumin concentration is not an indicator of plasma or blood volume; ALBIOS researchers wrongly claimed that “albumin administration during severe sepsis [leads to] larger fluid distribution within the intravascular compartment” only because that is what they erroneously expected, not what they observed. Slightly higher blood pressure (and so slightly less use of inotrope) in the albumin infused patients was more likely due to plasma viscosity increasing the vascular resistance. And they did, of course, get fooled by the Colloid Delusion; “The slight reduction in platelet counts in the albumin group may be a further marker of a larger expansion of the intravascular compartment in this group than in the crystalloid group, with a consequent dilution of the hemoglobin content.”

At last I’m going to get to the point that measurements of serum albumin concentration by bromocresol green (BCG) assay report the low serum albumin concentration to be much lower than it really is. When I was in practice we asked the lab to repeat the albumin assay with bromocresol purple if BCG assay result was albumin lower than 15 g/L. The BCP result was always higher. So why do clinical studies of albumin never report the assay method? We are therefore indebted to Regina Michelis and her colleagues in Israel for their suggestion that the BCG assay does not adequately measure oxidised albumin. And as almost all our ICU patients are experiencing severe oxidative stress, we discover another reason why the serum albumin concentration APPEARS to fall in critical illness. And why the Victorian notion of biophysical osmotic therapy has no role in modern medicine.


Leave a Reply

Your email address will not be published. Required fields are marked *