The challenge of fluid therapy in sepsis: when less is more

Administration of fluid boluses is considered to be one of the cornerstones of sepsis resuscitation. The surviving sepsis guidelines continue to ardently recommend a fluid bolus of 30 ml/kg within 3 h of presentation to hospital in patients who are hypotensive and considered to have sepsis.1 Let us consider the physiological rationale behind fluid administration in patients with sepsis and hypotension. 

The physiological rationale behind fluid resuscitation 

Fluid resuscitation is based on the Starling curve that correlates preload with stroke volume. The early, ascending part of the curve represents fluid responsiveness; if the preload is increased along this part of the curve, the stroke volume increases. The increase in stroke volume plateaus off towards the latter part of the curve, with no further increase in stroke volume with an additional increase in the preload. Based on this physiological response, bolus fluids are administered, with an expectation that the stroke volume would increase and improve blood flow to the vital organs. However, there are two important considerations with this approach. First, echocardiography based studies have demonstrated that up to 50% of normal subjects may also be fluid responsive;2 second, the stroke volume is normal or high in most patients with sepsis and it may seem counter-intuitive to attempt to increase it further by increasing the preload. Indeed, studies aimed at increasing the cardiac output and oxygen delivery to “supranormal” targets have uniformly demonstrated a lack of benefit or possible harm using this strategy.3

How much fluid, for how long? 

Among patients with sepsis, only 5% of a crystalloid solution may remain in the circulation after an hour of infusion.4 This corroborates with the transient improvement in hemodynamic parameters observed after bolus fluid therapy in septic patients. Hence, the question arises, considering the lack of sustained improvement, how often could fluid boluses be repeated? Inevitably, excessive fluid administration leads to the breakdown of the endothelial glycocalyceal barrier and tissue edema, with impaired organ function. Furthermore, fluid resuscitation may also lead to reduced systemic vascular resistance5 and reduce perfusion pressure to vital organs. There is provocative evidence from a human volunteer study suggesting that an increase in mean arterial pressure may be more related to the temperature of the fluid, compared to the volume administered; administration of intravenous fluid at room temperature (22°C) resulted in a higher mean arterial pressure compared to fluid that was warmed to 38°C.6

Fluid administration, increasing venous pressures, and effects on organ perfusion 

The central venous pressure (CVP) has been recommended to guide intravenous fluid therapy in septic patients. Blood flow to the vital organs is determined by the difference between mean arterial pressure, the upstream pressure, and venous pressure, the downstream pressure. An inappropriate rise in CVP may thus compromise organ perfusion, besides reducing venous return. In fact, the venous pressure may be a more decisive factor that determines organ perfusion. A drop in MAP within the autoregulatory range may not compromise organ perfusion. However, capillary blood flow may be solely determined by venous pressures within the autoregulatory range of arterial pressures.7  

Effect of fluid boluses on cardiac function

Sepsis is generally characterized by a preserved or higher than normal cardiac output. Left ventricular systolic function is usually well maintained; however, there is increasing evidence of significant diastolic dysfunction in septic patients.8 Aggressive fluid resuscitation in patients with impaired diastolic dysfunction leads to raised systemic venous and pulmonary artery pressures, with no significant increase in the stroke volume. Furthermore, excessive fluid resuscitation may exacerbate and perpetuate diastolic dysfunction. 

What is the clinical evidence? 

Measurement of the stroke volume by echocardiography demonstrated fluid responsiveness following 500 ml of a crystalloid solution in only 53% of patients with septic shock.9  A study of over 3000 children with severe sepsis revealed convincing evidence of significantly higher 48 h mortality with the use of bolus intravenous fluid compared to maintenance fluids alone as part of the initial resuscitation strategy.10 In a Zambian study among patients with sepsis and hypotension, early resuscitation with bolus intravenous fluid, vasopressors, and blood transfusion was compared with usual care based on clinician judgment. The intervention group received significantly more intravenous fluid in the first 6 h; however, in-hospital mortality was significantly lower among patients who received treatment based on clinician judgment.11 The CLASSIC study randomized patients with septic shock to receive fluid boluses until circulatory parameters continued to improve or a restrictive strategy with the administration of bolus fluid only if signs of severe hypoperfusion were present. In the fluid-restrictive group, the incidence of worsening of acute kidney injury was significantly less common.12

What may be a more optimal strategy? 

Conventional measures, including CVP, inferior vena caval diameter variation, and central venous oxygen saturation have limited value in the assessment of fluid responsiveness. The passive leg raising test (PLR) coupled with stroke volume monitoring may be more efficacious and safer compared to the administration of bolus fluid as a “challenge dose”. When performed optimally, the PLR mobilizes approximately 300 ml of blood from the peripheral to the central compartment. A change in stroke volume by this intervention may be conveniently evaluated by measuring the velocity-time integral by bedside transthoracic echocardiography. Unlike administration of a fluid bolus, any effect on the hemodynamic status is completely reversed on reassuming the horizontal position of the legs. PLR-induced increase in cardiac output has been shown to be highly predictive of fluid responsiveness in a meta-analysis.13 The importance of reassessment of fluid responsiveness needs to be emphasized prior to the administration of repeated fluid boluses. If fluid responsiveness cannot be reliably assessed, a mini-fluid challenge with 200-500 ml of fluid over 30 min may be much safer compared to larger boluses of 30 ml/kg.14

Early use of vasopressors

Alpha-1 agonists including norepinephrine, at low doses, result in venoconstriction earlier than arterial constriction. Venoconstriction mobilizes unstressed blood volume from the skin and the splanchnic circulation, leading to increased preload and cardiac output. Besides, norepinephrine effectively produces arterial constriction, reversing the vasodilation-induced hypotension among patients with septic shock. Intuitively, a restricted fluid strategy combined with early vasopressor therapy, may be more appropriate in septic shock, characterized by intense vasodilation. Such an approach needs future studies to evaluate clinical outcomes, compared to a fluids-first strategy. 

The bottom line: 

  • Empirical, large volume fluid boluses during resuscitation of septic patients may be ineffective and lead to harm.
  • The hemodynamic effect of bolus fluid administration is usually transient; repeated boluses lead to tissue edema and impaired organ function.
  • Excessive fluid administration has been shown to cause increased mortality compared to a fluid-restrictive strategy.
  • The passive leg raising test combined with echocardiographic assessment of the stroke volume may be a safer, more effective method of evaluating fluid-responsiveness compared to repeated fluid challenges.
  • Early use of vasopressors such as noradrenaline combined with restricted fluid resuscitation may be a more optimal strategy among septic patients. 


1.         Jozwiak M, Monnet X, Teboul J-L. Implementing sepsis bundles. Ann Transl Med. 2016;4(17). doi:10.21037/atm.2016.08.60

2.         Alves DR. Is Fluid Responsiveness A Normal State in Human Beings? A Study in Volunteers. 2016;2017(01):5.

3.         Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330(24):1717-1722. doi:10.1056/NEJM199406163302404

4.         Bark BP, Öberg CM, Grände P-O. Plasma volume expansion by 0.9% NaCl during sepsis/systemic inflammatory response syndrome, after hemorrhage, and during a normal state. Shock Augusta Ga. 2013;40(1):59-64. doi:10.1097/SHK.0b013e3182986a62

5.         Pierrakos C, Velissaris D, Scolletta S, Heenen S, De Backer D, Vincent J-L. Can changes in arterial pressure be used to detect changes in cardiac index during fluid challenge in patients with septic shock? Intensive Care Med. 2012;38(3):422-428. doi:10.1007/s00134-011-2457-0

6.         Wall O, Ehrenberg L, Joelsson-Alm E, et al. Haemodynamic effects of cold versus warm fluid bolus in healthy volunteers: A randomised crossover trial. Crit Care Resusc J Australas Acad Crit Care Med. 2018;20(4):277-284.

7.         Vellinga NA, Ince C, Boerma EC. Elevated central venous pressure is associated with impairment of microcirculatory blood flow in sepsis: A hypothesis generating post hoc analysis. BMC Anesthesiol. 2013;13:17. doi:10.1186/1471-2253-13-17

8.         Sanfilippo F, Corredor C, Fletcher N, et al. Diastolic dysfunction and mortality in septic patients: A systematic review and meta-analysis. Intensive Care Med. 2015;41(6):1004-1013. doi:10.1007/s00134-015-3748-7

9.         AzuRea Group, Roger C, Zieleskiewicz L, et al. Time course of fluid responsiveness in sepsis: The fluid challenge revisiting (FCREV) study. Crit Care. 2019;23(1):179. doi:10.1186/s13054-019-2448-z

10.       Maitland K, Kiguli S, Opoka RO, et al. Mortality after Fluid Bolus in African Children with Severe Infection. N Engl J Med. 2011;364(26):2483-2495. doi:10.1056/NEJMoa1101549

11.       Andrews B, Semler MW, Muchemwa L, et al. Effect of an Early Resuscitation Protocol on In-hospital Mortality Among Adults With Sepsis and Hypotension: A Randomized Clinical Trial. JAMA. 2017;318(13):1233-1240. doi:10.1001/jama.2017.10913

12.       Hjortrup PB, Haase N, Bundgaard H, et al. Restricting volumes of resuscitation fluid in adults with septic shock after initial management: The CLASSIC randomised, parallel-group, multicentre feasibility trial. Intensive Care Med. 2016;42(11):1695-1705. doi:10.1007/s00134-016-4500-7

13.       Cavallaro F, Sandroni C, Marano C, et al. Diagnostic accuracy of passive leg raising for prediction of fluid responsiveness in adults: Systematic review and meta-analysis of clinical studies. Intensive Care Med. 2010;36(9):1475-1483. doi:10.1007/s00134-010-1929-y

14.       Nunes TSO, Ladeira RT, Bafi AT, de Azevedo LCP, Machado FR, Freitas FGR. Duration of hemodynamic effects of crystalloids in patients with circulatory shock after initial resuscitation. Ann Intensive Care. 2014;4:25. doi:10.1186/s13613-014-0025-9

9 thoughts on “The challenge of fluid therapy in sepsis: when less is more”

  1. Sir, in your practice what are the clinical or lab end points u take to say that Pt is resuscitated.

    Thank you.

  2. This analysis is appropriate while looking at this common issue in our daily practice. I think ”excessive’ fluid resuscitation in early septic shock has not had favourable outcome, and understandably so, in the final analysis. Dr Jose’s analysis is therefore very appropriate.

  3. Absolutely great over view of what we should be doing in sepsis/septic shock patients. Thanks for this Jose

  4. Great ! Nice practical approach. I really liked PLR in resource limited setting

  5. It was nice and quick to read review about fluid boluses in septic shock. We should conduct large randomized study on restricted fluid strategy combined with early vasopressor therapy compared to classic liberal fluid strategy combined with late vasopressor therapy. Now, many questions remain to be answered; is it possible to start early vasopressor through peripheral line?, how long to wait before switching to a central line?, would this approach (early vasopressor) not expose the patient to higher risks of central line access? If someone thinks of using EJV for such approach, what’s firm evidence?

    1. Hi Mohammed

      Thanks for reading my blog and for the important questions you ask.

      Although it makes intuitive sense to administer vasopressors early in a state of severe vasodilation, as you rightly point out, it needs to be systematically investigated. My own feeling is that most clinicians commence noradrenaline support early if the blood pressure is low, instead of waiting for signs of fluid-responsiveness to disappear. That is definitely the way we practice by the bedside.

      Over the years, we have been extensively using peripheral IV lines when we commence noradrenaline support. If the dose required for a satisfactory blood pressure keeps rising, we would, of course, insert a central venous catheter. By this time, most patients would require a central line anyway.

      Having said that, the safety of vasopressor infusion through peripheral lines has been investigated in observational studies. These studies suggest that infusion of vasopressors through peripheral lines may be safe and effective, corroborating real-world practice (see below).

      1. Cardenas-Garcia J, Schaub KF, Belchikov YG, Narasimhan M, Koenig SJ, Mayo PH. Safety of peripheral intravenous administration of vasoactive medication. J Hosp Med. 2015;10(9):581-585. doi:10.1002/jhm.2394

      2. Datar S, Gutierrez E, Schertz A, Vachharajani V. Safety of Phenylephrine Infusion Through Peripheral Intravenous Catheter in the Neurological Intensive Care Unit. J Intensive Care Med. 2018;33(10):589-592. doi:10.1177/0885066617712214

      Thanks once again,


  6. Hello sir,
    Appreciate your good observation and clinical practice put in words . It is very important to understand that no two individuals are alike and the practice of fluid and vasopressors need should be individualized based on presence of other co-morbidity (ies). choice of Predominant therapy FLUID or VASOPRESSOR again depends on the cause of sepsis. Usually first 24 hours patient in septic shock with low pulse volume and low urine output receives large amount of fluid irrespective of Vasopressor dose. EARLY START OF NORAD is big YES! titrate upwards every 3-15 min with watch on MAP also helps in achieving goals and prevention of progression of organ dysfunction. PLR as you said is good bed side test to know fluid responsivness. Fluid dose adjustment on day 2 -3 (48 hours > ) should be considered to negate large positive balance and dilutional lab values. Probably Fluid therapy will be seen as LIBERAL – CONSERVATIVE- RESTRICTIVE regimens on our charts.

Leave a Reply