Editor’s Note: This post was sent back and forth between the Author, our Staff Reviewer, and the Editorial Team multiple times. Notably, the approach that is outlined is not comprehensive as it fails to evaluate several forms of shock. However, these topics are covered extensively online whereas a classic clinical approach to the patient who is not in shock but may or may not need fluid has been neglected. This approach, outlined by a Nephrology Fellow, relies on the bedside and laboratory assessment rather than many more recently proposed methodologies (e.g. ultrasound). Does it still have a place in the high-tech and probe-heavy environment of the modern ED? I’ll let you decide. I felt it was well worth a read. – Brent Thoma, CanadiEM Editor
Patient’s are often referred to as “wet” or “dry”. What do these terms actually mean? How do we tell if a patient will benefit from fluid administration or diuresis? Through a few clinical cases, this post will explore a practical approach to evaluating a patient’s volume status.
Warning: This post references likelihood ratios. For a crash course on their use, please see our previous post.
Body Fluid Distribution
There are two main body water compartments that make up total body water (TBW): the intracellular (ICF) and extracellular (ECF) fluid compartments. ECF is then subdivided into interstitial fluid and blood plasma. A rough breakdown of volumes in each compartment for a 70kg male is outlined in this chart:
|% if Body Weight||Volume (Litres)|
While these compartments are great as textbook definitions, how can they be assessed in a real patient at the patient’s bedside? The gold standard for the measurement of body fluid requires the injection of radio-labeled compounds, but this is not a practical approach.
The Approach to Volume Assessment
The first step in assessing volume status is to recognize that, clinically, it does not matter whether a patient has too much or to little water compared with a healthy person of equivalent weight. What is relevant is whether taking away or giving water will help optimize a given patient’s physiology and improve organ perfusion. Below is a simplified algorithm that can be used as a starting point for determining whether a patient will benefit from removing or providing fluids.
The first question we must ask ourselves is whether the patient has evidence of septic shock. This is because septic patients may be volume overloaded but still benefit from fluids. If not, we need to ask if the patient could have obstructive or cardiogenic shock. Fluid management is tricky in these patients and is not covered in this post or algorithm [Editor: But is covered extensively elsewhere]. Once septic, obstructive, and cardiogenic shock have been ruled-out, we can now consider whether a patient has hard signs of either volume overload or volume depletion. This type of patient will benefit from treatment to guide them back towards euvolemia. Following the algorithm are four clinical cases that help further explain the rational behind the flow chart.
Case #1: Profuse diarrhea with a normal exam
A 70 year old woman with a 2 week history of profuse diarrhea presents feeling weak and lightheaded. Her exam demonstrates a jugular venous pressure (JVP) of 2 cm, normal blood pressure, moist axilla, no lower extremity edema, and a normal tongue and oral mucosa.
In this scenario, the concern is that this woman’s total body water (as well as plasma volume) is low due to profuse diarrhea and causing her to feel weak. Her physical exam seems to argue against this diagnosis. However, the common physical exam findings for volume depletion are extremely poor at gauging the probability of volume depletion, with average positive and negative LRs no greater than 3 and 0.3 respectively.
But what about the JVP? Hers is normal, which means she can’t be “dry”, right? In fact, this is a very poorly studied finding for predicting volume depletion. Only one study (disclaimer: my own) has analyzed this physical exam finding for volume assessment, and it was specific to peritoneal dialysis outpatients.1 In this study no patients were total body water volume deplete (according to a gold standard bio-impedence study) but many had JVPs of 0. The central venous pressure (CVP) is effectively a direct and more accurate measure of jugular venous pressure. Unfortunately, its usefulness for assessment of volume status is equally poor. A systematic review found absolutely no correlation between CVP and measured blood volume.2 Thus, using “low” JVP as equivalent to volume depletion is called into question.
Now that JVP and the physical exam are out, what are you left with? The answer, obviously, is the kidneys (Disclaimer: being a nephrology fellow, I think this is the answer to all questions). If her body truly is volume deplete, the patient’s kidneys will start holding onto salt and water, causing a rise in plasma urea. While the strength of this finding has not been studied in a clinical setting, a study of healthy volunteers deprived of water for a few days found that they all had an increase in their serum urea.3
As lab values return on our patient, you see that her creatinine is normal. Her urea, however, is 20 mmol/L compared to 3 mmol/L on an admission two years ago. Following the algorithm, you decide to trial fluids and her clinical picture improves.
Case #2: Hypotension with an elevated JVP
A 50-year-old gentleman presents with confusion and a blood pressure of 70/40 mmHg. He left AMA from the hospital a few days ago while being treated for endocarditis. He has no signs of fluid overload on CXR and his JVP is 5 cm. What should you do?
This patient’s fluid status is unclear, but either way we should consider a fluid bolus. In septic shock, the theory is that increasing plasma volume to the heart should increase cardiac output via the Frank Starling mechanism. Don’t be fooled by the JVP: there has been much research showing that changes in CVP and a low CVP are not predictive of fluid responsiveness in sepsis.2 If we were concerned that the fluid would push him into pulmonary edema, we could try a straight leg raise test. If his blood pressure increases with the test, we can be reasonably certain that our fluid bolus will increase his cardiac output.
You may be wondering why a patient may be hypotensive even if they are actually “volume replete.” Sepsis can cause both leaky capillaries (leading to decreased intravascular volume and cardiac output) and vasodilation (resulting in decreased peripheral vascular resistance). Microvascular tone becomes deranged in sepsis with some vessels dilating much more than others, shunting blood away from some tissue areas and causing ischemia. Therefore, once intravascular fluid lost to capillary leak is replaced, any residual hypotension is a consequence of a heart pushing against less resistance (decreased afterload).
Case #3: Pumonary edema with a low JVP
A 45 year old male with a history of a myocardial infarction presents with shortness of breath and chest pain. A chest x-ray confirms pulmonary edema and he has new elevation in his troponins. His blood pressure is 173/95, his JVP is 0, and he has no evidence of end organ ischemia. Should you give him lasix?
Again, in this scenario, we see the limitation of JVP. In an acute MI, the pulmonary vasculature has not had time to adapt to the acute rise in left ventricular pressure and fluid builds in the lungs. It is too early in the disease process for this increased pressure to transmit into the right ventricle and elevate the JVP. Indeed, as shown in this meta-analysis,4 JVP only has a likelihood ratio of approximately 0.7 for ruling out left ventricular failure. This patient will definitely benefit from a dose of lasix to help his kidneys remove the fluid from his lungs.
Moving off the algorithm, imagine this gentleman had a BP of 90/40, cold extremities, and a rising lactate. This patient is now in cardiogenic shock and, even though he may have absolute volume overload, he may benefit from further volume expansion. For example, in right ventricular infarction, the heart is preload dependant and cardiac output can be improved with fluid boluses.5 There is no hard and fast rule for fluid management in cardiogenic shock though and you may need to consider invasive cardiac monitoring to guide fluid management (of course the underlying condition should be treated too!). This is why we are unable to include the fluid management of cardiogenic or obstructive shock in this basic algorithm.
Case #4: Chronic kidney disease and edema
A 65 year old diabetic female with longstanding chronic kidney disease presents to the ER with increasing fatigue, anorexia, vomiting (ie. uremic symptoms), and an increased creatinine. An infectious work-up is negative, and you are reasonably certain she is not in septic shock. She has 1+ pitting edema to her bilateral lower extremities, but you wonder whether she is not truly fluid overloaded and is actually “intravascularly deplete”, leading to decreased renal perfusion. You debate whether you should give her fluids or not.
You will often hear the term “intravascularly deplete,” and it will often be said in the setting of renal failure. However, the only setting where true intravascular water depletion can occur with a concomitant increase in ECF is capillary leak syndrome (ie. leakage of intravascular fluid from capillaries due to systemic inflammation from severe infection, surgery etc.). Heart failure, cirrhosis and nephrotic syndrome (most likely a problem of salt retention, not decreased oncotic pressure) are all causes of fluid retention but will not cause intravascular depletion. The woman in this case has progressive renal decline in the absence of infection. This indicates that her blood vessels are quite full with fluid, and she is probably at the maximum of her Frank-Sterling curve. Therefore, fluids will not help her kidneys function better and could actually make her clinical scenario worse by increasing preload on her heart and potentially even decreasing her GFR by increasing renal venous pressures. She will likely need to move towards initiating dialysis for her fluid overload.
The assessment of a patient’s fluid status is complicated and there is no single rule that can be applied to every patient. The general approach outlined in this algorithm is a tool to help determine the fundamental question that needs to be answered in these patients: Will the patient benefit from increasing or decreasing their body water volume?
Reviewing with the Staff
Clinicians often struggle with volume assessment and management so I genuinely appreciate your efforts to create a reasonably simple algorithm to follow.
Generally speaking, patients with obvious derangements of vital signs are somewhat more straightforward (here, using Point of Care Ultrasound to determine aetiology of shock state is well studied and shown to dramatically improve diagnostic accuracy).
The more challenging patients, as alluded to by this algorithm, are those will near normal vital signs and exam findings of minimal power. Here, you guide the student through the pitfalls of JVP and other common volume assessment measures and help them arrive to a much better approach to fluid administration – your last case is a useful example. As you correctly state, giving fluids to someone at the peak of their F-S curve won’t help, and in fact can worsen a clinical picture.