Applied physiology of intravenous fluid replacement

This chapter is relevant to Section I2(i) of the 2023 CICM Primary Syllabus, which expects the exam candidates to develop or at least fake that they "understand the pharmacology of colloids and crystalloids". This learning outcome feels somewhat belated in the middle of a critical care training program. That prescribing IV fluid is usually expected from doctors long before this understanding develops is a serious issue, as an average intern may prescribe a hundred litres of saline over the course of a night shift, perhaps without having the slightest idea of what it contains or what it does. 

The question "what happens when you infuse me with " is much easier to answer when one has a well-defined volume and composition for x, and much harder to answer in generic terms. All the more reason to prepare an answer, as that is exactly the kind of devilry the CICM examiners are known for. Therefore:

  • Pre-infusion conditions
    • Total body water (60% of body mass) is divided into intracellular (40%) and extracellular (20%)
    • Of the extracellular fluid (20% body mass), one quarter is intravascular (8.3% of total body water; 3.5 litres in the 70kg adult)
    • The immediate effects of a fluid bolus are felt through its effects on the composition and volume of the intravascular compartment
  • The rationale for fluid resuscitation is to increase the circulating volume, thereby increasing cardiac preload, thus increasing the stroke volume and cardiac output
  • Infusion kinetics
    • Infused fluids have a variable half-life in the circulating volume
    • This ranges from 10 minutes to 24 hrs
    • The transcapillary escape rate is determined by Starling forces and depends on the colloid and electrolyte content of the fluid, the balcne of hydrostatic pressures, and the permeability of the capillaries
  • Cardiovascular response to the change in blood pressure
    • Atrial stretch receptors sense the increase in venous pressure, and briefly increase the heart rate (Bainbridge reflex).
    • Baroreceptors sense the increase in arterial pressure generated by the increase in stroke volume, and decrease the heart rate and the systemic vascular resistance, thereby renormalising the blood pressure
    • Cardiac output remains stable under conditions of normovolemia or hypervolemia, or trends towards normal in hypovolemia
  • Humoral responses
    • to the increase in renal perfusion
      • Renin, angiotensin and aldosterone secretion is depressed
      • As the result, the reabsorption of sodium and water in the distal nephron is decreased
    • to the change in atrial stretch
      • Natriuretic peptide release increases, as atrial stretch stimulates their release, which in turn stimulates natriuresis and diuresis
    • to the change in osmolality
      • The change in osmolality is sensed by the hypothalamus
      • Vasopressin secretion from the posterior pituitary is then altered to produce the desired change in the urine concentrating function of the distal nephron
  • Response to the constituents of the infused fluid
    • Variable: eg. hyperchloraemia and acidosis with saline, or alkalosis with balanced crystalloid
  • Effect of the rate of administration
    • Rapid boluses expand the intravascular volume faster, and are eliminated sooner
    • Slow infusions maintain the intravascular volume expansion by matching the rate of elimination. 

Now, as for these humoral and sympathetic reflexes responses, already so many pages have been written here that the reader is spared their duplication, and instead offered the following internal links:

Additionally, discussions of physiological responses to specific fluids are also available locally:

What would be left to discuss of the physiology of intravenous fluids, if you removed everything listed there from the topic?  Probably the acute kinetics of a generic "bolus". Consider: what happens immediately when you pump the manual giving set is going to be rather similar irrespective of what kind of fluid product you are pumping, i.e. there is going to be a change which is mainly related to the volume of the infusate rather than its physicochemical characteristics. Then, later you can think about the hyperchloraemia or oncotic games the albumin is playing, but these will be largely immaterial in the first five minutes or so. The rest of this chapter will therefore focus on the immediate effects of a fluid bolus, and how that differs from a slow infusion.

The best overview for the details your really want to know is probably Hanh & Warner (2010), as their paper is an attempt to model fluid pharmacokinetics. 

There is also a paper in Nature from 2018, authored by some of the most preeminent members of the CICM pantheon, which would no doubt be useful to the casual reader (but probably not useful for answering any specific exam question per se).

A bolus by any other name

That the word "fluid bolus" has no fixed scientific meaning in medicine will come as no surprise to the returning reader. The word itself (βῶλος) literally means "clod" or "lump" and the ancient Greeks certainly would raise an eyebrow at this abuse, as surely there must have been better words (for example "clysis", which means "flood").

Not unexpectedly, any given room full of intensivists will disagree on what defines a fluid bolus, down to the content rate and volume. To characterise this cacophony of opinions,  Glassford et al (2016) tried to survey the critical care world, and came up with some common ground. Most intensivists seem to agree that a bolus should be given over 30 minutes at most, and that its volume should be at least 250 ml.

Immediate effects of a rapidly infused volume

This volume is (usually) something you administer intravenously, which means your fluid is entering the venous circulation; often the right atrium, if you are in the ICU and using a central line. Considering the hydrostatic and oncotic forces in the lung are usually set up in a way that prevents water from exiting the circulation, one can reasonably expect that volume to remain in the intravascular compartment, at least until it reaches the peripheral capillaries (which could take 30-60 seconds). So, for that duration, the whole volume is doing some positive haemodynamic work.

The rate of administration will obviously vary, depending on how you're giving the bolus, and through what. In a standard scenario, a 18g peripheral cannula will usually be rated to deliver a flow rate of around 100-120 ml/min when running freely under gravity, but this can be greatly enhanced with a manually pumped giving set or a rapid infuser pressure bag, as well as higher caliber access, to the point where probably something like 100-150ml could be given within the span of a single circulation time.

What happens as the result depends on whether you are empty or full, and depends entirely on the opposing influences of the Bainbridge reflex and the arterial baroreceptor reflex. Without revisiting the content of other chapters, these influences can be summarised as follows:

  • The Bainbridge reflex is an atrial stretch reflex, and increases the heart rate in response to a fluid bolus
  • The baroreceptor reflex is an arterial stretch reflex, and decreases the heart rate in response to an improved blood pressure
  • When the patient is hypovolemic, the baroreceptor reflex overrides the atrial stretch reflex, and the heart rate decreases in response to the restoration of volume
  • When the patient is euvolemic, the Bainbridge reflex increases the heart rate in response to an increase in volume. 

Is this supported by the evidence? Hard to say.  Ukor et al (2017) infused some healthy volunteers with a litre of saline over 30 minutes and found that the heart rate actually dropped (slightly, by an average of 2 beats per minute), which may have been too slow to trigger much atrial stretch. Muir et al (1975) gave 2000ml over twenty minutes to four of his co-authors (physicians at the Royal Infirmary of Edinburgh) which is more representative of a vigorous bolus, and found that some of the volunteers experienced an increase in heart rate (by a maximum of 10), whereas others had a trivial drop in heart rate.  In this cohort, the cardiac output increased (from 5 to 7L/min), which was largely the result of an increase in stroke volume (from 60-70ml up to 100ml). The haemodynamic data from this experiment was cleaned and sliced for general consumption:

Effects of a fluid bolus on haemodynamics from Muir et al, 1975

The readers' attention is drawn to the abrupt return to status quo which follows the fluid challenge. Some of this was probably due to the effects of compensatory mechanisms, as Muir's healthy volunteers were euvolemic. The rest was due to the distribution of the fluid out of the circulation, which is the topic of the next section.

Distribution of infused volume

The distribution of water through the body is a fairly fixed and predictable thing, governed by such incommutable factors as the osmolality of the compartments and the permeability of various barrier membranes. The fate of any intravenously infused water is to distribute among these compartments according to the influence of the abovementioned factors, and the extent of the distribution is affected mostly by the dissolved electrolytes which are given along with the fluid. Thus, it is usually held that, where an isotonic amount of sodium is co-administered, the water will largely distribute to the extracellular space (as that is where the sodium stays), whereas a hypotonic fluid would also distribute into the intracellular water. 

This distribution is not instantaneous, but the rate of distribution is relatively rapid. When Drobin & Hahn (1999) infused some healthy volunteers with 25ml/kg of Ringer's acetate over 30 minutes, only 50% of the infused volume remained in the intravascular space at the end of the 30 minute period. Similarly, Svensen et al (2009) gave 15ml/kg (about 1L) over 10 minutes, and found that about 70% of the infused fluid was still in the intravascular volume at the end of the infusion, and the full distribution took another 30 minutes to complete, with the final intravascular volume expansion being something like 20-25% of the infused volume (exactly as you would predict from the sodium contentof their Ringers'). The graphical representation of a model based on their data represented this matter so well that this author was powerless against the urge to steal it:

distribution of a rapid fluid bolus from Svensen et al (2009)

As one can see from the end of this graph, the intravascular volume expansion from this infusion (about 200-250ml at the end of one hour) is not sustained, and decreases to 100ml at the end of the next hour. The infused fluid is removed at roughly the same rate from both the interstitial and intravascular compartments, because of the effect of mechanisms that defend the body fluid volume, and this is the underlying mechanism for the difference between the volume expanding effects of a bolus as compared to a continuous infusion.

Difference between a fluid bolus and a slow infusion

Slow infusions actually cause a larger intravascular volume expansion.Ukor et al (2017) compared a 30 minute infusion to a 120 minute infusion of the same volume, and determined that the slower infusion was in fact more effective at increasing cardiac output. Hahn et al (1997) demonstrated that blood volume expansion at the end of a 80 minutes of infusion was greater than with a shorter infusion. 

The main reason for this is that infusions continuously replenish the intravascular space that is being depleted by elimination - which can be so rapid that a one-volume kinetic model is enough to describe it. At the end of the experiment by Drobin & Hahn (1999) mentioned above, after two hours, the healthy volunteers had already produced enough urine to eliminate 50% of the infused volume. This is obviously not always the case - the volume-depleted patient would surely be more reluctant to lose the volume again, and they have all of the aforementioned volume-defending mechanisms well activated at the time of the fluid bolus, leading to better fluid retention. Drobin & Hahn were also able to demonstrate this in volunteers who received a fluid bolus after having 900mls of their blood withdrawn - these patients preferred to retain the infused fluid instead of turning it into urine. To borrow a more professional turn of phrase from Drobin & Hahn, the retention of the bolus "increases in the presence of hypovolemia, which can be attributed chiefly to a reduction of the elimination rate constant". That constant, from the sample of relatively healthy female volunteers in their 1997 study, was about 95ml/hr. 

References

Hahn, Robert G., and David S. Warner. "Volume kinetics for infusion fluids." The Journal of the American Society of Anesthesiologists 113.2 (2010): 470-481.

Finfer, Simon, John Myburgh, and Rinaldo Bellomo. "Intravenous fluid therapy in critically ill adults." Nature Reviews Nephrology 14.9 (2018): 541-557.

Glassford, Neil J., et al. "Defining the characteristics and expectations of fluid bolus therapy: a worldwide perspective." Journal of critical care 35 (2016): 126-132.

Drobin, Dan, and Robert G. Hahn. "Volume kinetics of Ringer's solution in hypovolemic volunteers." The Journal of the American Society of Anesthesiologists 90.1 (1999): 81-91.

Muir, A. L., et al. "Cardiorespiratory effects of rapid saline infusion in normal man." Journal of applied physiology 38.5 (1975): 786-775.

Kumar, Anand, et al. "Effect of large volume infusion on left ventricular volumes, performance and contractility parameters in normal volunteers." Intensive care medicine 30.7 (2004): 1361-1369.

Svensen, Christer H., et al. "Arteriovenous differences in plasma dilution and the distribution kinetics of lactated Ringer’s solution." Anesthesia and analgesia 108.1 (2009): 128.

Hahn, R. G. "Volume effect of Ringer's solution in the blood during general anaesthesia." European journal of anaesthesiology 15.4 (1998): 427-432.

Ukor, Ida F., et al. "The haemodynamic effects of bolus versus slower infusion of intravenous crystalloid in healthy volunteers." Journal of critical care 41 (2017): 254-259.

Hahn, R. G., D. Drobin, and L. Ståhle. "Volume kinetics of Ringer's solution in female volunteers." British journal of anaesthesia 78.2 (1997): 144-148.

Kumar, Anand, et al. "Pulmonary artery occlusion pressure and central venous pressure fail to predict ventricular filling volume, cardiac performance, or the response to volume infusion in normal subjects." Critical care medicine 32.3 (2004): 691-699.