Predicting Volume Responsiveness in the Emergency Department

Authors: Paolo N. Grenga, MD (EM Resident Physician, University of Rochester Emergency Medicine Residency), Ryan P. Bodkin, MD (Program Director, University of Rochester Emergency Medicine Residency), and Jason M. Rotoli, MD (Assistant Program Director, University of Rochester Emergency Medicine Residency // Edited by: Alex Koyfman, MD (@EMHighAK) and Brit Long, MD (@long_brit)

The ED physician commonly encounters undifferentiated shock.  As is the case with many shock states, rapid intravenous (IV) fluid administration is often the initial therapy of choice during initial resuscitation.  Yet, it has been established in the literature that only about ~50% of hemodynamically unstable patients will respond favorably to fluid administration [1-4].  What options exist to predict which patients will respond favorably to a fluid challenge?  Recent debate over various indices and ultrasonographic techniques for indirectly assessing volume responsiveness in critically ill patients suggests that there may be fewer options than previously proposed.


You are working in the critical care bay of a tertiary medical center when you receive a transfer from an outside hospital.  The patient is a 67-year-old female with advanced atherosclerosis who presented to her local hospital with multiple days of abdominal pain, nausea, vomiting, and diarrhea.  Computed tomography of the abdomen and pelvis showed non-specific but extensive small and large bowel inflammatory changes including bowel wall edema concerning for colitis versus obstruction.  She arrives to your facility toxic appearing – she is febrile, hypotensive, tachycardic, and tachypneic.  On review of the accompanying medical records you find that she has been under-resuscitated, having received minimal intravenous fluids prior to transport.


Undifferentiated Shock – Workup and Resuscitation Occur in Tandem

One of the challenges in caring for a patient in shock is that the workup must occur simultaneously with the resuscitation.

The approach suggested by Winters et al. can help to narrow the differential diagnosis, which begins with a rapid and deliberate physical examination [5].  Distended neck veins in the hypotensive patient may represent elevated central venous pressure in patients with obstructive shock, as is seen in pulmonary embolism, tension pneumothorax, and pericardial tamponade.  This same clinical presentation can also be seen in patients with pump failure secondary to cardiogenic shock from a variety of etiologies.  Urticaria, angioedema, facial swelling, and respiratory distress may accompany hypotension in anaphylaxis leading to a vasodilatory shock.  In the setting of hypotension, warm extremities may suggest pathologic vasodilation, as is seen in sepsis and anaphylaxis.  Cold extremities, however, are a result of sympathetic compensation for hypotension and may point toward hemorrhagic, hypovolemic, or cardiogenic shock.  Hematemesis or a rectal examination positive for blood (melena/gastrointestinal blood) may suggest a hemorrhagic cause of hypovolemic shock.

Early diagnostic testing is also helpful and should include the following:

  • Blood glucose
  • ECG – dysrhythmia and ischemia
  • Chest radiograph – pneumothorax, hemothorax, pulmonary edema, pneumonia
  • Pregnancy test – ectopic pregnancy


Case (Continued)

The patient’s blood pressure improves modestly with the administration of two liters of IV normal saline.  Shortly after fluid administration you are called to the bedside by her nurse, who is concerned about the patient’s work of breathing.  You arrive at the bedside and find the patient in tripod posture, tachypneic with increased accessory muscle use.  She is hypoxic.  A non-rebreather is applied, but before her oxygen saturations improve she suddenly has an episode of emesis and becomes unresponsive.  She loses pulses and CPR begins.  She receives two rounds of CPR and epinephrine for pulseless electrical activity in the setting of hypoxic respiratory arrest, the patient is intubated, and ROSC is achieved.  Post-intubation chest x-ray shows the endotracheal tube in appropriate position and near-complete opacification of the right lung.  She remains hypotensive.


Fundamentals of Fluid Responsiveness

Only ~50% of hemodynamically unstable patients will respond favorably to a fluid challenge, defined as an increase in stroke volume after fluid loading [1-4].  Further, overly aggressive fluid administration is associated with worse outcomes [1, 6-9].  Responsiveness of the heart to a fluid challenge can be anticipated by where its output lies on the Frank Starling curve.

As right atrial pressure (RAP, a function of ventricular preload) increases, stroke volume and therefore, cardiac output (recall that CO = HR x SV) increases.  Other parameters affect the shape of the curve. For example, the curve will move upward with decreased afterload and downward with decreased contractility and decreased heart rate.

Ventricular preload is determined by venous return, which is comprised of three parts, 1) mean systemic filling pressure (MSFP), or the volume of blood not in capacitance vessels that is able to create a transmural pressure above zero, 2) RAP, and 3) the resistance to venous flow (Rv).  The relationships of these components is shown in the following equation:

Venous return = (MSFP – RAP)/Rv

Venous return can be increased by 1) increased MSFP, 2) decreasing RAP, and 3) decreasing resistance to venous flow (Rv).  The MSFP drives venous return and may be increased by fluid loading or by venoconstriction, whereas options to modulate the latter two parameters are limited in the ED [1].

Cardiac output cannot increase without a corresponding increase in venous return, which can be directly augmented by fluid loading.  This is only the case, however, when the heart is operating on the steep portion of the curve.  For example, consider the decreased cardiac function curve seen below.  Point II indicates that, despite an increase in venous return, there is no corresponding increase in cardiac output.  This is the case in cardiogenic shock, as the heart does not have the ability increase its contractility regardless of IV fluid administration.

Case (Continued)

The patient continues to be hypotensive despite receiving multiple liters of intravenous fluids.  She requires initiation of a norepinephrine drip to keep her MAP > 65 mmHg.  Hours later, however, her MAP begins to fall again despite increasing the dose of the norepinephrine, administration of calcium, and stress dose steroids. 

Should this patient receive further intravenous fluid administration?  How can we determine whether her hemodynamics will improve with a volume challenge?


Predictors of Fluid Responsiveness

Despite their continued use in many ICU settings, static markers of central venous pressure (i.e. ventricular end diastolic volume, pulmonary artery occlusion pressures, etc.) have consistently been shown not to correspond to volume responsiveness [1, 10, 11].  Central filling pressures (classically considered to be cardiac filling pressures) are poor predictors of volume responsiveness for three reasons.  First, CVP is an intramural pressure only, which does not take extramural pressures into account. This can be problematic in patients with elevated intrathoracic pressures, such as in mechanically ventilated patients.  Second, preload is made up by CVP and ventricular end-diastolic radius. The latter is determined by ventricular compliance, which varies widely between patients and disease states.  Lastly, even low CVP does not imply that a patient can be fluid responsive (consider point I on the previous figure) [1].  For these reasons, dynamic markers of CVP have been deemed preferable to static markers and have been described in the literature for the past 20 years.  So, which of these, then, would be of use to the ED physician?

Pulse Pressure Variation/Stroke Volume Variation

Pulse pressure (difference between the systolic and diastolic blood pressures) during mechanical ventilation has been used in the past to estimate stroke volume.  Cyclic variation of the pulse pressure during mechanical ventilation (or pulse pressure variation) can predict response of cardiac output to volume expansion.  For example, insufflation during positive pressure ventilation decreases preload of the right ventricle.  If this decrease in preload is transmitted to the left ventricle, then both ventricles may be thought of as preload dependent.  Pulse pressure variation has the largest body of evidence supporting its use, with a recent meta-analysis showing a pooled sensitivity of 88% and specificity of 89% for predicting volume responsiveness [10, 12].  The usefulness of this technique in patients with ARDS is questionable, however, as these patients are ventilated with smaller tidal volumes that may not be able to generate enough insufflation to appreciably modify the pulse pressure.  This technique is also less accurate in the spontaneously breathing patient and in the setting of dysrhythmia.

Variations of Vena Caval Dimensions

Variation in the diameter of the inferior vena cava (IVC) measured by transthoracic echocardiography has been used to predict fluid responsiveness.  Compared to patients using PPV, evaluation of vena caval dimension in spontaneously breathing patients is less studied and less accurate.  In a 2014 meta-analysis of 8 studies, the pooled sensitivity was 76%, and the pooled specificity was 86% [10, 13].  Unlike pulse pressure variation, it can be used in patients with cardiac dysrhythmias.

End Expiratory Occlusion Test

As mentioned earlier, positive pressure ventilation decreases ventricular preload by way of decreased venous return.  Temporarily stopping mechanical ventilation, end-expiratory and/or end-inspiratory holds, depending on the strategy being employed, can stop this impediment to venous return.  If cardiac output increases during this brief pause in positive pressure ventilation, then it can be expected to be volume responsive.  This test is easy to perform (simply a < 15 second pause in mechanical ventilation) and is valid in patients with ARDS.  However, the change in cardiac output is measured by pulse contour analysis in studies validating its use.  Pulse contour analysis is based on the observation that the area under the arterial pressure waveform is proportional to stroke volume.  Systems that perform this calculation (i.e. FloTrac, PiCCO) are not always immediately available in the ED, and depending on the system, requires central venous and/or arterial cannulation [14].  Furthermore, only recently has echocardiography been studied as a means for measuring the change, and implementing the appropriate technique may require advanced training in echocardiography [10, 15].

Fluid Challenge

Perhaps the most direct way to predict fluid responsiveness is to administer a bolus of IV fluids and then assess for improved cardiac output. This has the benefit of being easy to perform; however, it is a therapeutic endeavor just as much as it is diagnostic and may lead to deleterious effects if given when contraindicated.  There is evidence that using a small fluid bolus (~100 mL) can produce changes in left ventricular outflow velocities (detected by transthoracic cardiac ultrasound), but this requires exceedingly sensitive cardiac output monitoring [10]. 

Passive Leg Raise

The passive leg raise (PLR) is essentially a fluid bolus of about ~300 mL that can be repeated interminably without infusing any additional fluid.  This test has the benefit of being easy to perform and highly accurate in predicting volume responsiveness.  Two meta-analyses of these studies have been recently published showing a pooled sensitivity of 85% and specificity of 91% for predicting volume responsiveness [10, 16-17].  However, invasive, direct measures of cardiac output must be used in order to accurately make this determination.  When substituting direct measurement of cardiac output for pulse pressure, specificity is unchanged but sensitivity is decreased.

Case Resolution

The patient’s MAP improved with PLR and fluid challenge, showing volume responsiveness.  She received additional crystalloid with improvement in her hemodynamics.  An emergent surgical consult was made, and after further stabilization, the patient was brought to the OR for exploratory laparotomy, where she underwent partial small bowel resection for obstruction.


The Verdict? No Substitute for Clinical Judgment

So how does one predict which patients will be responsive to administration of IV fluids?  Like the start of many investigations in medicine, the answer begins with a thorough history and physical exam and is dependent on clinical judgment as well as other additional pieces of information.  Indeed, there are instances where the expected response to fluid administration is obvious, such as the initial phase of septic shock (prior to IV fluid administration) or hemorrhagic shock (as may be seen in trauma) [10, 18].  Yet, emergency physicians must be cognizant of the adverse effects of fluid overload, as has been observed in septic patients [19].

No single method for predicting volume responsiveness is without limitation.  Further, none of the techniques have 100% sensitivity or specificity for detecting which patients will respond favorably to fluid administration.  Emergency physicians increasingly use bedside ultrasound in clinical decision making. However, recent literature review of the ability of IVC diameter to predict fluid responsiveness demonstrates unreliability and unpredictability in certain patients.  Ultrasonographic protocols like the RUSH exam may be more useful in sorting out undifferentiated shock and which patients could benefit from fluid administration [5].  In the meantime, the PLR, a low risk and simple maneuver, appears to have stronger evidence supporting its use and may help emergency physicians at the bedside decide whether to continue to administer intravenous fluids. Ultimately, the evaluation of the patient in undifferentiated shock should not be based on a single technique.  Instead, it should be a multi-faceted approach that includes as much information as necessary to positively influence the patient’s care.


References / Further Reading:

  1. Cherpanath, T. G. V., B. F. Geerts, W. K. Lagrand, M. J. Schultz, and A. B. J. Groeneveld. “Basic concepts of fluid responsiveness.” Netherlands Heart Journal21.12 (2013): 530-36. Web.
  2. Marik PE, Cavalazzi R, Vasu T, et al. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med. 2009;37:2642–7.
  3. Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest. 2002;121:2000–8.
  4. Marik, Paul. “ISepsis – Vena Caval Ultrasonography – Just Don’t Do It!” EMCrit. N.p., 08 July 2017. Web. 28 July 2017.
  5. Winters, Michael E., Deblieux, Peter, Marcolini, Evie G., Bond, Michael C., Woolridge, Dale P. Emergency Department Resuscitation of the Critically Ill. Dallas: American College of Emergency Physicians, 2011.
  6. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–77.
  7. Holte K, Kehlet H. Fluid therapy and surgical outcomes in elective surgery: a need for reassessment in fast-track surgery. J Am Coll Surg. 2006;202:971–89.
  8. Wiedemann HP, Wheeler AP, Bernard GR, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006;354:2564–75.
  9. Boyd JH, Forbes J, Nakada TA, et al. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality. Crit CareMed. 2011;39:259
  10. Monnet, Xavier, Paul E. Marik, and Jean-Louis Teboul. “Prediction of fluid responsiveness: an update.” Annals of Intensive Care 6.1 (2016): n. pag. Web.
  11. Kory, Pierre. “COUNTERPOINT: Should Acute Fluid Resuscitation Be Guided Primarily by Inferior Vena Cava Ultrasound for Patients in Shock? No.” Chest151.3 (2017): 533-36. Web.
  12. Yang X, Du B. Does pulse pressure variation predict fluid responsiveness in critically ill patients? A systematic review and meta-analysis. Crit Care. 2014;18:650.
  13. Zhang Z, Xu X, Ye S, Xu L. Ultrasonographic measurement of the respiratory variation in the inferior vena cava diameter is predictive of fluid responsiveness in critically ill patients: systematic review and meta-analysis. Ultrasound Med Biol. 2014;40:845–53.
  14. Mehta, Yatin. “Newer methods of cardiac output monitoring.” World Journal of Cardiology 6.9 (2014): 1022. Web.
  15. Jozwiak M, Teboul JL, Richard C, Monnet X. Predicting fluid responsiveness with echocardiography by combining end-expiratory and inspiratory occlusions (abstract). Ann Intensive Care. 2016 (in press).
  16. Cherpanath TG, Hirsch A, Geerts BF, Lagrand WK, Leeflang MM, Schultz MJ, et al. Predicting fluid responsiveness by passive leg raising: a systematic review and meta-analysis of 23 clinical trials. Crit Care Med. 2016;44:981–91.
  17. Monnet X, Marik P, Teboul JL. Passive leg raising for predicting fluid responsiveness: a systematic review and meta-analysis. Intensive Care Med. 2016.
  18. Grenga, Paolo. “ED Management of the MVC Patient: Pearls and Pitfalls.” – Emergency Medicine Education. N.p., 14 June 2017. Web. 28 July 2017.
  19. Genga, Kelly, and James A. Russell. “Early Liberal Fluids for Sepsis Patients Are Harmful.” Critical Care Medicine 44.12 (2016): 2258-262. Web.

Leave a Reply

Your email address will not be published. Required fields are marked *