Assessment of fluid balance


Estimation of the fluid balance in the ICU is generally held to be important because it offers some idea of whether or not the patient is "fluid overloaded", and by how much. The clinical significance of fluid status changes are not debated in the college papers; the past paper SAQs focus on the methodology ("how to measure", rather than "why would you bother") - the clinical utility of fluid balance data being an unquestioned truth. However, as it turns out the mindless worship of water content has little foundation in the published evidence.

With appropriate attention to this topic, the college have invited the candidates to tabulate the the advantages and limitations of the various ways of estimating fluid balance in Question 16 from the second paper of 2009 and Question 16 from the second paper of 2006. The typical methods are by fluid balance record, by daily weights, and by clinical examiantion. More exotic methods range from PA catheter wedge pressures though PiCCO and CVP and to weird stuff like bioimpedance and tritium dilution. The tabulated response offered as the college "model answer"  is difficult to improve upon, as it  strikes a balance between detail and brevity. As such, the time-poor exam candidate may be well advised to limit their revision to the college answer. 

If, however, brevity were sacrificed to detail, an answer to Question 16 would resemble the following table.

Methods used to Estimate  Fluid Balance
in the Critically Ill Patient




Clinical estimates

  • Cheap
  • Easily performed at the bedside

Fluid balance chart

  • Cheap
  • Easily performed at the bedside
  • Accuracy depends on accuracy or recording
  • Usually, cumulative balance records are inaccurate and tend to disagree with body weight measurements
  • This technique fails to estimate losses into incontinence pads, spilled secretions, sweat, evaporative losses from wounds, and losses via the lungs; in short "insensate" losses are forgotten.

Daily weights

  • Easily performed in the presence of specialised bed equipment
  • Accuracy depends on accuracy of recording
  • Requires expensive bed equipment
  • Requires attention to detail - one must ensure the same amount of bedding and on-bed equipment is with the patient each time, otherwise fluctuations in weight may occur. Usually, there is about 3.5kg of nonremovable hardware in the bed together with the patient.
  • Correlates poorly with bedside charts
  • Lack of evidence for cost-effectiveness


  • Easily performed at the bedside
  • Constant monitoring is possible



  • Interpreter-dependent
  • Serial assessments across a series of clinicians may yield variation purely due to technique
  • Yields information regarding chamber filling volumes rather than total body fluid volume - and then you infer the fluid balance from this.
  • Not universally accepted as a method to assess intravascular volume
  • Only accurate when compared to the (known to be useless) clinical examination by an expert.


  • Invasive
  • Labour-intensive (thermodilution measurement)
  • Association with cardiac function makes it difficult to use lung water to estimate whole-body fluid balance


  • Experimental technique, yet to be validated
  • Does not agree with thermodilution measurements
  • Most of the available methods measure transthoracic bioimpedance, which relies on the absence of pleural effusion, and is usually useless in cardiac surgery or thoracic trauma

Tritium indicator dilution

A critical evaluation of fluid balance measurements

The overall merit of fluid balance calculation (nevermind their accuracy) is questionable. After all, what precisely are we measuring, and does it matter? Consider. The normal healthy person, whose physiology is supposedly well understood, is still subject to fluid balance changes which may be measured in the litres, and which may not be accounted for by the usual cylinder diagrams.

Normal fluid balance breakdown for the healthy human

There he is, the smug Healhy Person. Arrogantly sucking down an irresponsibly unmeasured volume of water while exercising with his iPod earphones. Probably listening to some sort of abominable Nu-Metal. He clearly has no idea of how much moisture he is losing per breath, nor of how much he is drinking, or how long he has been exercising for. Moreover, the surface of his body will produce evaporative losses which are even harder to estimate, because they are contingent on body surface covering, surface temperature, sun exposure, ambient air humidity and temperature, convection by way of wind, and what if it starts raining? In short, fluid balance estimates are difficult.

Worse yet, they may be pointless. Consider a situation where one has precise control of all the abovementioned variables, and is able to measure every droplet of intake and output very accurately. In such an implausible scenario, at the end of a day one ends up generating a number (eg. "negative balance 500ml") which represents the fluid balance difference between this morning and this afternoon.

What significance does this balance have in the healthy person?

One may posit that it has none.

Consider: the smug jogger has changed their body fluid volume by perhaps 1% of the total. At this level, neither the tonicity of the body fluid nor the pressure of the circulating compartment are affected, which means that the normal autoregulation mechanisms have not had to exert any homeostatic effect. In short, the man's own pitutary doesn't care about this fluid shift. So why should we?

Yes, there may be situations where fluid balance has clinical significance. It would be silly to completely ignore the value of input and output measurement, because ultimately a certain fluid input must be maintained for normal health, and in the intensive care unit the doctor has complete control over this variable. In a reductio ad abdsurdum scenario such as a man dehydrating to death in the desert the importance of input and output volumes becomes readily apparent. However, the question is not whether inputs and outputs are important, but whether there is a clinical significance in the fastidious attention to their measurement.

Consider now the critically ill patient.

Fluid balance in the complex critically ill patient

Let's say you are measuring some of those inputs and some of those outputs.

Let's say youre measurements are accurate; all of the fluid boluses are signed and accounted for. Still there will be some fluid losses and gains - potentially, vast ones - which remain impossibl to measure. Within the context of the situation, with severe trauma, organ system failure, rampant infection, missing limbs and potentially weeks of ICU stay, the relevance of the daily fluid balance becomes questionable, even if it is accurate. It is hard to discern its effect on survival, that signal being lost in the deafening noise of critical illness.

So, if a fluid balance assessment question ever comes up again, it may come in the form of a miny-essay. "Critically evaluate the use of the cumulative fluid balance in estimating fluid gains and losses in critically ill patients", they might ask. In preparation for such a question, one should have a plan.

Rationale for fluid balance measurements

  • Critically ill patients are often subjected to large volumes of fluid resuscitation or fluid loss (eg. by haemorrhage, diuresis or in burns).
  • The trend in fluid balance can guide therapy (in terms of further fluid resuscitation or diuresis)
  • Fluid balance records can help determine the dose of ultrafiltration.
  • Positive fluid balance has been associated with worse outcomes in certain disease states, eg. sepsis (Vincent et al, 2006)
  • For some conditions (eg. ARDS) a neutral or negative fluid balance has been associated with better outcomes (Wiedemann et al, 2006)
  • Neglect of fluid balance calculations can lead to dehydration or fluid overload
  • Often, it is impractical to perform daily weight measurements (though these are acknowledged as the gold standard).

Advantages of fluid balance calculations

  • Quick and easy
  • Can be outsourced to nursing staff
  • Can be easily automated by computerised information systems
  • Requires no special bed equipment or scales
  • Requires only fluid input and output data, which is already collected in routine observation charts

Disadvantages of fluid balance calculations

  • Subject to human error (or computer error)
  • Only as accurate as the records on which they are based
  • Non-recorded losses (eg. evaporative, or losses into dressings or pads) will not be reflected in the calculated fluid balance
  • To increase accuracy (eg. pre and post use weight of pads or absorbent dressings) increases nursing workload
  • The practice of calculating fluid balance is wildly inaccurate. For example, Perren et al (2011) found that in their prospective case series the fluid balance recording were only complete in 38% of cases, and that calculated fluid balances agree very poorly with daily weight measurement, differing by up to 1kg. Moreover, about one third of them was arithmetically incorrect (i.e. somebody performed the calculation in a sloppy manner), with errors ranging from -3606 ml to +2020 ml. The authors were forced to conclude that "this reduced attention for the FB might theoretically hide a general sense of futility regarding the conventional practice" of fluid balance calculation. 

The theoretical benefits of attention to fluid balance

  • Frequently, ICU stay is associated with vigorous fluid resuscitation.
  • The resulting positive fluid balance has been associated with poor outcomes
  • It is thought that tissue oedema contributes to organ dysfunction: "Tissue oedema impairs oxygen and metabolite diffusion, distorts tissue architecture, impedes capillary blood flow and lymphatic drainage, and disturbs cell-cell interactions" (Malbrain et al, 2014)
  • Other mechanisms of harm from positive fluid balance include raised intra-abdominal pressure, worse perfusion of (swollen) encapsulated organs, oedema of anastomotic sites, and damage to the endothelial glycocalyx.
  • On the basis of studies which demonstrated a harmful impact of positive fluid balance on outcomes, "de-resuscitation" has been suggested as a valid practice- to aim for a neutral or negative fluid balance by days 3-7 of critical illness.
  • However, it must be mentioned that the bulk of these studies have been performed using calculated fluid balances, with all the inaccuracy thereof.
  • Moreover, the authors of the "de-resuscitation" strategy (Malbrain et al, 2014) could not make any strong recommendation regarding how one should aim for this negative fluid balance, as there is no evidence to support the routine use of albumin or diuretics, versus letting the patient mobilise the fluid by normal homeostatic mechanisms.

The unclear clinical significance of changes in total body water

  • Under ideal circumstances, the fluid balance calculations and body weight measurements will give a (more or less accurate) representation of the fluctuations of total body water. However, the significance of this variable is questionable.
  • Total body water is dispersed into a variety of compartments, and not all compartments have an equal significance for patient care.
  • For instance, in shock states the intravascular compartment volume is of greatest interest, but this occupies only a small fraction of the total body water volume. In ARDS, total body water only has importance insofar as it reflects the volume of extravascular lung water, which contributes to the poor lung compliance and gas exchange problems. In cardiogenic shock or post-operative cardiac surgery scenarios, the volume of interest is the end-diastolic volume which represents preload; total body water is even less related to this variable.
  • In states of critical ilness, the physiological processes which govern the movement of water between compartments are grossly deranged:
    • The oncotic pressure gradient is gone (your albumin is low)
    • The musculoskeletal pump is gone (venous and lymphatic return is impaired by the prolonged immobility of a sedated intubated patient)
    • Capillaries are leaky due to inflammation (encouraging water movement out of the circulatory volume into the interstitium)
    • Serosal surfaces are leaky due to inflammation, allowing the collection of fluid in "potential spaces" such as the pleural cavity
  • In short, it is unclear how the manipulation of total body water affects clinically relevant fluid compartments. And if it does, then it crudely affects all compartments,  not only the ones of clinical interest.
  • One might argue that fluid-related therapies (eg. fluid resuscitation, dialysis dosing, diuretics) should be directed on the basis of other measurements.


Schneider, Antoine G., et al. "Estimation of fluid status changes in critically ill patients: Fluid balance chart or electronic bed weight?." Journal of critical care27.6 (2012): 745-e7.

Schoeller DA, van Santen E, Peterson DW, Dietz W, Jaspan J, Klein PD: Total body water measurement in humans with 18O and 2H labeled water. Am J Clin Nutr 1980, 33(12):2686-2693

Charra, Bernard. "Fluid balance, dry weight, and blood pressure in dialysis."Hemodialysis International 11.1 (2007): 21-31.

Stephan, F., et al. "Clinical evaluation of circulating blood volume in critically ill patients—contribution of a clinical scoring system†." British journal of anaesthesia 86.6 (2001): 754-762.

Chung, Hsaio-Min, et al. "Clinical assessment of extracellular fluid volume in hyponatremia." The American journal of medicine 83.5 (1987): 905-908.

Schneider, Antoine Guillaume, et al. "Electronic bed weighing vs daily fluid balance changes after cardiac surgery." Journal of critical care 28.6 (2013): 1113-e1.

Perren, A., et al. "Fluid balance in critically ill patients. Should we really rely on it?." Minerva anestesiologica (2011).

Wilson, John N., et al. "Central venous pressure in optimal blood volume maintenance." Archives of Surgery 85.4 (1962): 563-578.

Piccoli, Antonio, et al. "Relationship between central venous pressure and bioimpedance vector analysis in critically ill patients." Critical care medicine28.1 (2000): 132-137.

Marik, Paul E., Michael Baram, and Bobbak Vahid. "Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares." CHEST Journal 134.1 (2008): 172-178.

Mitchell, John P., et al. "Improved outcome based on fluid management in critically III patients requiring pulmonary artery catheterization." American Review of Respiratory Disease 145.5 (1992): 990-998.

Bethlehem, Carina, et al. "The impact of a pulmonary-artery-catheter-based protocol on fluid and catecholamine administration in early sepsis." Critical care research and practice 2012 (2012).

Schwann, Nanette M., et al. "Lack of effectiveness of the pulmonary artery catheter in cardiac surgery." Anesthesia & Analgesia 113.5 (2011): 994-1002.

Wheeler, A. P., et al. "Pulmonary-artery versus central venous catheter to guide treatment of acute lung injury." N Engl J Med 354.21 (2006): 2213-2224.

Nguyen, Viviane TQ, et al. "Handheld echocardiography offers rapid assessment of clinical volume status." American heart journal 156.3 (2008): 537-542.

Schuller, D., et al. "Fluid balance during pulmonary edema. Is fluid gain a marker or a cause of poor outcome?." CHEST Journal 100.4 (1991): 1068-1075.

Marik, Paul E. "Hemodynamic parameters to guide fluid therapy." Transfusion Alternatives in Transfusion Medicine 11.3 (2010): 102-112.

Monnet, Xavier, et al. "Assessing pulmonary permeability by transpulmonary thermodilution allows differentiation of hydrostatic pulmonary edema from ALI/ARDS." Intensive care medicine 33.3 (2007): 448-453.

Mattar, J. A. "Application of total body bioimpedance to the critically ill patient. Brazilian Group for Bioimpedance Study." New horizons (Baltimore, Md.) 4.4 (1996): 493-503.

Foley, Kieran, et al. "Use of single-frequency bioimpedance at 50 kHz to estimate total body water in patients with multiple organ failure and fluid overload." Critical care medicine 27.8 (1999): 1472-1477.

Barry, Ben N., et al. "Lack of agreement between bioimpedance and continuous thermodilution measurement of cardiac output in intensive care unit patients."Critical Care 1.2 (1997): 71.

House, Andrew A., et al. "Volume assessment in mechanically ventilated critical care patients using bioimpedance vectorial analysis, brain natriuretic peptide, and central venous pressure." International journal of nephrology 2011 (2010).

Vincent, Jean-Louis, et al. "Clinical review: Update on hemodynamic monitoring-a consensus of 16." Crit Care 15.4 (2011): 229.

Wiedemann HP, Wheeler AP, Bernard GR et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006; 354: 2564–2575.

Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, Moreno R, Carlet J, Le Gall JR, Payen D; Sepsis Occurrence in Acutely Ill Patients Investigators. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med. 2006 Feb;34(2):344-53.

Malbrain, Manu LNG, et al. "Fluid overload, de-resuscitation, and outcomes in critically ill or injured patients: a systematic review with suggestions for clinical practice." Anaesthesiology intensive therapy 46.5 (2014): 361-380.