A Primer on the Perils of Intravenous Fluids – Part 1
Jon-Emile S. Kenny [@heart_lung]
“To every (wo)man is given the key to the gates of heaven. The same key opens the gates of hell. And so it is with science.”
-Richard Feynman
A rich, frosty wind etherizes my face; this cool gust rips through the medieval, labyrinthine passageways of Old Stockholm like frayed edges of an exhausted Swedish flag. I am lashed with cold, but I press on towards a warm Fika with my classmates; I am deeply happy. As I pass by vociferous water fowl perched on shifting ice shards of the Lilla Värtan, I turn my thoughts to fluid resuscitation – especially in light of the recent Surviving Sepsis Guidelines and this observational study by Marik and colleagues.
A Case
A 68 year old woman has a rising BUN:Cr ratio with diminished urine output following an ax-fem bypass surgery complicated by urosepsis. She has known diastolic dysfunction and has received many litres of lactated ringers. Her physical examination reveals bilateral crackles and notable lower extremity and abdominal wall edema. A portable chest x-ray demonstrates indistinct hila, vascular congestion and new bilateral pleural effusions; since admission she has gained 10 kg. Using bioreactance functional hemodynamic monitoring and repeat passive leg raise assessments she is found, repeatedly, to be ‘fluid responsive.’ Accordingly, the overnight house officer continues to challenge the patient with litre after litre of crystalloid, yet the patient continues with worsening urine output, rising BUN and worsening dyspnea.
Introduction
The provision of intravenous fluids is no trivial intervention. Indeed, one eminent nephrologist has called for medical students to receive, not a ‘white coat ceremony’ at the outset of their education, but instead a ‘normal saline ceremony.’ This pomp is an occasion whereby the fledgling physician imbibes a reverence for sodium chloride. To hold a bag of resuscitation fluid should rouse veneration no different than grasping an ampule of morphine or vial of piperacillin-tazobactam. Thus, a zealous awareness of the balance between therapy and toxicity is required for all things we inject into another beings’ veins.
Yet, the administration of intravenous fluid is so ubiquitous and commonplace that we take the aforementioned for granted; it certainly isn’t – emotionally – on par with mixing mannitol or readying dantrolene. We ‘bolus’ and ‘challenge’ patients continuously with crystalloid so much so that these terms lose meaning; what, exactly, is indicated when one says ‘bolus’ or ‘challenge?’
Simply, a bolus connotes therapy whilst a challenge implies diagnosis. When a patient requires intravenous fluids to restore embarrassed hemodynamics, one should give at least 500 mL of fluid over no more than 15 minutes – a fluid ‘bolus.’ By contrast, when a clinician wants to provide fluids, but is uncertain as to whether a patient will augment his or her cardiac output in response to increased preload, a fluid ‘challenge’ is undertaken. This occurs when the clinician administers 500–1,000 ml of crystalloids [300– 500 ml of colloids] over no more than 30 min to assess the impact of increased preload on cardiac output and, ostensibly, tissue perfusion.
Yet, summative data reveals that when we challenge patients in the intensive care unit [ICU], roughly 50% of them are volume non-responsive. In other words, roughly half of the patients in the ICU receive an intervention which provides little benefit and leads, simply to volume overload. Within the literature, volume overload, is defined as an increase in body weight by 10% from admission.
Given the increased concern over excessive fluid administration in the ICU, a recent conceptual model was proposed to make explicit the use of intravenous fluids during the course of critical illness. Here, there are 4 stages which include: rescue, optimization, stabilization and de-escalation. The first two stages occur over minutes to hours and involve firstly rescuing the patient and secondly optimizing organ and tissue perfusion. The final two phases evolve over days to weeks and require close monitoring of fluid balance. Ideally, during stabilization, there is a net even-to-negative fluid balance. A rational approach to fluid provision – that is, the use of dynamic predictors of fluid responsiveness – are likely best utilized during the optimization and stabilization phases.
Dynamic Approaches to Fluid Responsiveness
While it is beyond the scope of this brief overview to dissect the nuances of predicting fluid responsiveness, some salient points deserve mention. The last 15 years has witnessed an explosion of methods by which a clinician can attempt to determine if increasing preload will increase cardiac output; that is, determine fluid responsiveness. While initial indices such as pulse pressure variation, stroke volume variation and inspiratory inferior vena cava [IVC] dilation showed promise, they were derived in extremely controlled clinical scenarios which included the absence of inspiratory effort and large tidal volumes. The complexity of mechanical heart-lung interaction degrades these indices greatly when there are other perturbations of cardiopulmonary physiology such as: spontaneous breathing, tachypnea, cardiac dysrhythmia, small tidal volumes, high right ventricular afterload, altered chest wall compliance, etc. These difficulties have, further, translated to impaired diagnostic accuracy of inspiratory IVC collapse as well.
Given the above, the current, most accurate process by which a clinician can estimate fluid responsiveness requires a transient increase in venous return, followed by an immediate [within 1-2 minutes] assessment of left ventricular output [or a surrogate thereof]. An increase in venous return may be triggered by a passive leg raise, mini fluid challenge or end-expiratory occlusion test; an instantaneous surrogate of left ventricular output can include Doppler assessment of flow through a major artery.
Importantly, fluid ‘responsiveness’ does not equate with fluid [or volume] ‘status.’ As in the introductory case, a patient may be massively volume overloaded yet remain fluid responsive. Functional hemodynamic monitoring can only inform the clinician if the venous return curve intersects the ascending portion of the starling curve; it does not, however, tell the clinician if giving fluids is the correct intervention – the latter quandary is solely a clinical assessment.