Response to 1 unit of packed red blood cells

These are the physiological effects of infusing one unit of packed red blood cells into a patient.

What am I giving?

In short, no idea. This is a discussion of 1 unit of packed red cells, which is a bag with about 250-350ml of fluid in it. So its not exactly a carefully calibrated volume. Plus, the hematocrit is anywhere from 0.50 to 0.75. Not only that, but the duration of storage before transfusion will influence all sorts of trivial (electrolyte concentration, in-bag pH) and non-trivial (oxygen-carrying capacity, erythrocyte lifespan) consequences of transfusion.

So, as for the validity of this discussion, there are going to be problems. I will endeavour to support my assertions with some sort of evidence.

The infusion of 1 unit of packed cells

Change to the initial conditions

The volume of the intravascular compartment increases by 250ml. The typical electrolyte and non-electrolyte contents of a unit of packed cells has been described elsewhere. For the remainder of this thought experiment, the unit of packed cells will have a hematocrit of 0.60, therefore containing 150ml of cells and 100ml of fluid.

That 100ml of water sloshing around the red cells in the bag has a total osmolality of about 340mOsm/L. So, by giving 100ml one gives 34mmol of dissolved electrolytes.

Thus, each unit of packed cells contains the following dose of electrolytes:

Na+ 15mmol
K+ 2mmol
Cl- 15mmol
HCO3- 1mmol
Lactate 0.9mmol

These figures make it difficult to believe that electrolyte derangement might arise as a consequence of transfusion. So what if your patient gets 2 mmol of potassium? True. It is a small amount. But, this is merely the floating dissolved potassium in 100ml of water. The cells themselves also have a large amount inside them (at an average intracellular K+ concentration of 100mmol/L, the 150ml of “dry” erythrocytes will contain 15mmol of potassium).  So maybe they don’t release all of it. Maybe they only let go of 10% of their potassium. That would still contribute an additional 1.5mmol per unit of packed cells.

In any case, the contribution from 1 unit of packed red blood cells is merely 34 mmol of electrolytes.

Response to 1 lunit of packed cells

The sodium and chloride stay in the extracellular fluid, potassium sloshes around according to the intracellular concentration, and the bicarbonate and lactate engage in a complex metabolic dance the details of which are outside the scope of this discussion (and they are safely ignored, given that the dose is insignificant).

The water distributes itself according to where the osmotic forces drag it. The extracellular compartment is now slightly richer in solutes, and so some of the intracellular water is dragged out (a miniscule amount).

How long does this distribution take? Not long, according to a study of non-haemorrhaging patients. In fact the majority of the equilibration of water occurs within the first 15 minutes. The hematocrit they measured at 15 minutes was much the same as the hematocrit measured at 24 hours post transfusion.

Osmoreceptor response

Because it has received a load of essentially isoosmolar fluid, the plasma and the extracellular fluids both do not experience much of a change in osmolality. (it does rise, but only by about 0.095mOsm/L).

Baroreceptor response

The intravascular compartment volume increases by 176ml – from 5000ml to 5176ml. The increase in intravascular volume is 3.5% - outside the volume receptor sensitivity threshold.

I don't care about the electrolytes; I gave erythrocytes, so what happens to them?

The initial conditions in this thought experiment give us a hematocrit of 0.30 (1500ml divided by 5000ml).

This is somewhat lower than the normal hematocrit of an adult, but hey- we are working with a system of imaginary cylinders. But this sort of an admission hematocrit is not completely out of the question for an ICU patient who has already had some sort of fluids on the way to their crisis.

The 1500mls of intravascular cell volume gets an extra 150mls of cells.

The 3500mls of intravascular water expands by only 26ml.

Thus: the new hematocrit should be 1650ml divided by 5176ml, or 0.318

The hematocrit should increase by 6%.

Does experimental evidence confirm this calculation?

In fact, it does not. Only about one half to one third of this predicted increase is observed in trauma patients receiving blood transfusion. A study of patients recovering from pelvic fractures suggests that the real change in hematocrit from a single unit of packed red cells is more like 1.9% (varying from 0.4% to 3.1%).

I also don't care about the hematocrit. Its the hemoglobin concentration that matters.

As The ICU book by Merino advises us, hematocrit is a lousy transfusion trigger. Hemoglobin concentration increase in patients receiving transfusion has been studied, and the findings suggests that for every unit of packed red cells transfused, the Hb rises by 10g/L.

In summary;

The fluid balance of the body and the electrolyte concentration is not significantly affected.

The hematocrit rises by somewhere around 2% to 3%

The Hb concentration increases by 10g/L per unit.

Why the hypocalcemia?

Why indeed. The citrate used to prevent clotting in packed cells is present in a tiny concentration, about 1800mg per bag ( which is about 1 mmol, given that citrate has a molar mass of 189g/mol). At maximum activity, this will only chelate 1 mmol of calcium. Given that the extracellular fluid has about 30.8 mmol of calcium in it, one can expect the calcium to drop from 2.2mmol/L to 2.12 mmol/L, which is not a lot.

TACO: transfusion-associated circulatory overload

Let us consider what the capacitance vessels do in response to the loss of blood.

They constrict, of course. If they didnt, the loss of a large blood volume would result in a proportional loss of preload. The decrease in cardiac output would be catastrophic. So, the venous capacitance vessels constrict to maintain sufficient right ventricular filling pressure to produce an adequate cardiac output.

transfusion-associated circulatory overload

If you then go and dump a massive amount of blood products into this constricted venous network, you will cause a dramatic and sudden increase in central venous pressure. The increase in preload will be out of proportion to the increase in volume. The result is pulmonary oedema.


Thomas C. Clifford and William Beautyman. Changes in the Erythrocyte Potassium in Patients with Cardiac Failure Treated with Digitalis. Clinical Chemistry August 1958 vol. 4 no. 4 311-315

Elizalde JI, Clemente J, Marín JL, Panés J, Aragón B, Mas A, Piqué JM, Terés J. Early changes in hemoglobin and hematocrit levels after packed red cell transfusion in patients with acute anemia. Transfusion. 1997 Jun;37(6):573-6.

Elzik ME, Dirschl DR, Dahners LE. Correlation of transfusion volume to change in hematocrit. Am J Hematol. 2006 Feb;81(2):145-6.

Andrew R. Wiesen, MD; Duane R. Hospenthal, MD, PhD; John C. Byrd, MD; Kevin L. Glass, MD; Robin S. Howard, MA; and Louis F. Diehl, MD Equilibration of Hemoglobin Concentration after Transfusion in Medical Inpatients Not Actively Bleeding Ann Intern Med. 1994;121:278-280.

Robert I. Weed, Claude F. Reed and George Ber IS HEMOGLOBIN AN ESSENTIAL STRUCTURAL COMPONENT OF HUMAN ERYTHROCYTE MEMBRANES? J Clin Invest. 1963;42(4):581–588. doi:10.1172/JCI104747. 

Van Beekvelt MC, Colier WN, Wevers RA, Van Engelen BG. Performance of near-infrared spectroscopy in measuring local O(2) consumption and blood flow in skeletal muscle. J Appl Physiol. 2001 Feb;90(2):511-9.

Australian Red Cross Blood Service, Blood Component Information(2012)

Popovsky MA. Transfusion reactions. 2nd ed. Bethesda: American Association of Blood Banks; 2001.