Feb 142017

Jon-Emile S. Kenny [@heart_lung]

Read part 1 here

Fluids and the Glycocaylx

Critically-ill patients all likely have endothelial dysfunction to some degree.  This perturbation in microvascular physiology may be underpinned by abnormal glycocalyx structure and function.  Sepsis, trauma, surgery and ischemic insults are all known to disrupt the glycocalyx which will increase vascular fluid capacitance.  Indeed, at 90 minutes, 20% of crystalloid volume remains within the vascular space, while in sepsis only 5% does.  Further, in sepsis, precapillary arteriolar dilation increases capillary filtration pressure which favours interstitial edema.  Thus, the use of alpha agonists such as norepinephrine may protect the microvascular beds from excessive filtration pressure.

With respect to fluid type, there is great debate within the literature.  Currently, colloids such as albumin and hydroxyethyl starches [HES] are to be withheld in most circumstances.  Albumin use has waxed and waned in the face of various observational and randomized controlled studies [e.g. SAFE & ALBIOS].  In totality, there is probably little clinical benefit to albumin use in most disease states; it should be avoided in the traumatic brain injury population and used in sepsis and septic shock only if ‘excessive’ amounts of crystalloid are required [which shouldn’t be often] as suggested by the Surviving Sepsis Guidelines update.  Starches have received bad press in the last few years including the large CHEST and 6S trials, both of which detected an increased risk of acute kidney injury [AKI].  While there may be some very select instances where starches may be useful, it is hard to recommend their use at all.

Lastly, the chloride debate continues to plague anyone providing volume resuscitation.  While there is certainly observational data and biological plausibility for untoward renal effects of hyperchloremia, not all fluid physiologists believe that balanced solutions [e.g. lactated ringers, Hartmann’s] are superior to ‘normal’ saline.  Additionally, the recent SPLIT trial didn’t seem to answer the question of whether or not large volumes of chloride rich crystalloid affords harm; these answers will – hopefully – soon be established [see this update on the SMART investigation].

Excessive Fluid Volume and Organ Dysfunction

There are many associations between excessive fluid accumulation and adverse outcome in the ICU.  While these observations relate fluid administration and poor ICU prognosis, they do not prove causality.  It is still possible that high fluid requirements are a marker of bad disease.  Nevertheless, there is biological conceivability in that organ edema drives dysfunction.  The liver, lungs, brain, gut, kidneys, skin and even heart are known to function abnormally in the setting of volume overload.  This is particularly true for encapsulated organs such as the kidneys where the underlying pathophysiology may lie in ‘nephrosarca,’ increased renal interstitial pressure, collapsed glomeruli and diminished renal blood flow.  Notably, removing the renal capsule has been shown to improve renal blood following an ischemic insult. Increased renal venous pressure, diminished portal vein flow and intra-abdominal hypertension may all play separate roles in the AKI seen following excessive volume administration.  Importantly, in such setting, the BUN-creatinine ratio may rise and urine sodium fall.  Consequently, ‘pre-renal’ biochemical markers do not necessarily imply volume depletion and definitely should not mandate blind administration of additional intravenous fluids.

In addition to biological plausibility there is some observational data which shows that limiting fluids in certain ICU populations can be beneficial.  For instance, in the FACTT trial, patients randomized to conservative fluid management had less ICU days, more ventilator-free days and there was a non-statistically significant 4% absolute risk reduction in the use of renal replacement therapy in those who were kept on the ‘drier side.’  Then, a post-hoc analysis of the FACTT trial demonstrated an association between excessive fluid and AKI.

Further, in patients with septic shock and lung injury, both initial resuscitation and delayed conservative fluid management were independently associated with improved mortality – lending credence to the ‘4 phases’ of volume resuscitation described above.

De-Escalation Therapy

Loop diuretics remain the mainstay for fluid removal in the ICU and on the general medical floor.  They cause a notable natriuresis, which removes volume from the body.  It is suggested that urine output of 3-4 mL/kg per hour be obtained as this draws out fluid at a rate which prevents capillary under-filling.  Whether this be via bolus dosing of loop diuretics, or as an infusion is controversial with data supporting either approach.  The oft-quoted DOSE trial suggests that bolus dosing, rather than infusion, may achieve greater diuresis – at least in acute heart failure patients.  Importantly, in AKI, furosemide may have diminished delivery to the distal loop of Henle and there may be down-regulation of the Na-K-2Cl transporter rendering this loop diuretic less effective; this may require a change in loop diuretic and/or addition of a thiazide.  For a must-read perspective on diuresis without dehydration, please see this excellent post by Josh Farkas.

The question of early renal replacement in patients with AKI has been posed and investigated.  Payen and colleagues found that renal replacement initiated in septic patients with AKI within 5 days produced more favourable clinical outcomes as compared to later initiation.  However, the data regarding early versus delayed renal replacement remains cloudy without clear direction.  For some more detail within this realm, please turn to this brief overview of the recent AKIKI trial.


The patient was removed from the bioreactance functional hemodynamic monitor; this removed the temptation to check for fluid responsiveness.  She was also initiated on a furosemide infusion with a goal to keep her hourly urine output 100 mL more negative than her hourly total input.  Over a period of 5 days her oxygenation improved; as well both her work of breathing and BUN-creatinine fell.  She left the ICU for the general medical floor while embarking on an aggressive physical therapy regimen.



Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


A Primer on the Perils of Intravenous Fluids – Part 2