Intravenous fluids (IVFs) increase right atrial filling (venous return) and are indicated to treat hypovolemia. Crystalloid solutions like saline and Lactated Ringer’s can replace gastrointestinal, urinary, and evaporative fluid losses, and blood products treat hypovolemia from hemorrhage. If poor circulation is not due to hypovolemia, then giving IVFs can be harmful. A good history and physical exam can sometimes distinguish the cause of poor circulation, but uncertainty often remains, leading to empiric IVF boluses to determine if they will help. Unfortunately, feedback from empiric IVF boluses may be misleading, as immediate improvements in markers of cardiac output and tissue perfusion may not necessarily correlate with patient-centered outcomes like all-cause mortality. Thus, PoCUS findings are unlikely to differentiate whether IVFs will be ultimately beneficial. However, if the diagnosis, i.e., suspected hypovolemia, warrants fluid administration, then PoCUS can help identify contraindications, termed fluid intolerance.
Image 13.1 Decision tree summarizing how to incorporate fluid intolerance when determining when or how to give IV fluids to a patient. PoCUS helps determine the diagnosis and whether the patient is fluid intolerant.
Fluid intolerance occurs in patients with an inappropriately elevated central venous pressure (CVP) or a specific vulnerability to tissue edema. If present, avoid administering IVFs or do so judiciously, paying extra attention to the impact of maintenance IVFs, enteral intake, and injection medications, which often make up a majority of volume and sodium intake after a patient has been in the hospital for a few days.1,2 Identifying fluid intolerance requires a way to 1) identify high CVP, 2) distinguish high CVP due to excess RA wall tension from high CVP due to high intrathoracic/pericardial pressure or a cardiac condition that necessitates higher atrial pressure like pericardial tamponade or hypertrophic cardiomyopathy, and 3) identify unique risks for consequential edema. PoCUS can help clinicians do this quickly and accurately.
The right ventricle has a thin wall and relatively large unstretched volume, which allows it to fill without generating much, if any, wall tension. As a result, RV end-diastolic pressure and CVP normally reflect the overlaying intrathoracic and pericardial pressures.3 Thus, high CVP (>12 cm H2O) alone is not specific for volume intolerance. So, when high CVP is identified, consider if the thoracic pressure is high, i.e., from hyperinflation or positive pressure ventilation, then examine the heart to look for pericardial disease and apparent chamber collapse. These findings suggest high CVP may be appropriate, and IVFs are not contraindicated. However, increasing venous return may be harmful if high CVP is due to excess RA wall tension. PoCUS of the heart reveals apparent chamber distension in this setting due to systolic left heart failure, isolated RV overload, or apparent valve insufficiency.
The physical exam can detect elevated CVP with good specificity by assessing visible jugular venous distention. However, the sensitivity of this examination is poor and much worse than evaluating the heart and IVC with PoCUS.4 Even if the IVC and cardiac views are suboptimal, it is possible to use PoCUS to improve the detection of jugular venous distention. This involves using a linear transducer and careful technique to avoid inadvertently compressing the vein.5
The IVC is usually a vertically oriented cylindrical, running just deep to the liver, and normally measures 1.3 to 2.1 cm in diameter. To avoid confusing the IVC with the aorta, note that the aorta is separated from the liver by musculature and does not flow into the heart. PoCUS of the IVC involves primarily a longitudinal (long-axis) subcostal view but should also include a quick transverse (short-axis) view to confirm its identification and appearance. If you cannot obtain an adequate subcostal view, consider the right lateral hepatic window via the midaxillary line with the patient supine. This coronal (flank) view can more easily underestimate IVC collapsibility and should always include short and long-axis views. Measure the maximum IVC diameter at its widest longitudinal plane at end-expiration (before inspiratory collapse). Brennen et al. determined that an IVC diameter of >2cm predicts a CVP >10 mmHg (13.6cm H2O) with 73% sensitivity and 85% specificity.6 If a patient has chronic fluid overload, their IVC may be permanently stretched out to >2 cm in diameter despite having a normal CVP. In addition, the diaphragm can narrow the IVC as it passes into the thoracic cavity. If significant narrowing is present or chronic overload is suspected, consider other parameters like IVC collapsibility and JVP more heavily than IVC size. IVC dilation in this setting may be less specific.
Image 13.2 The longitudinal IVC view (A) shows the IVC directly draining into the right atrium. Do not confuse the IVC with the abdominal aorta, which is separated from the liver by musculature and lymph nodes, and is not seen as directly connected to the heart (B). The transverse IVC view (C) helps confirm the findings of the longitudinal view. Consider obtaining a lateral IVC view (D) if the subcostal window is inadequate. An IVC diameter of 2cm or larger predicts elevated CVP (E). If there is significant narrowing at the level of the diaphragm (F), then avoid relying on IVC diameter to determine CVP. The IVC view may also detect pericardial (+) or pleural (*) effusions (G).
When a patient briefly inhales through their nose to generate negative inspiratory pressure, the IVC usually collapses >50% in diameter. In contrast, the IVC expands in response to higher intrathoracic pressure, such as during mechanical ventilation (positive pressure lung inflation). A larger IVC diameter with reduced inspiratory collapse, despite sufficient respiratory effort, signifies increased CVP.7 Abdominal hypertension from obesity, severe ascites, organomegaly, or poor abdominal wall compliance can shrink the IVC size despite having increased CVP. If abdominal hypertension is suspected, a small or normal IVC that collapses should be considered indeterminate, and decisions should be based on other data. However, a large IVC in this setting is specific for elevated CVP, especially if inspiratory collapse is reduced.
Quantifying the IVC’s inspiratory collapse is tedious and requires averaging multiple measurements. It is more practical to use long and short-axis B-mode views (avoid M-mode) to visually estimate IVC collapsibility after asking the patient to inhale through their nose quickly. Estimates can be grouped into IVC collapse <30%, 30-70%, or >70%. These ranges are comparable to quantifying the collapsibility using measurements.8 Remember to also look at the transverse (short-axis) plane to ensure the change in diameter doesn’t just represent a shift to a different imaging plane. Doppler assessment of hepatic, portal, and renal vein flow can provide additional data regarding the degree of venous congestion,9 but this is beyond the scope of this text and the capabilities of many hand-held ultrasound devices.
Image 13.3 IVC collapsibility can be qualitatively estimated as >70% (A), 30-70% (B), and <30% (C).
To assess whether elevated CVP is due to excess RA wall tension, obtain PoCUS views of the heart as well. Finding plethoric chambers with LV systolic dysfunction, isolated RV overload, or apparent valve insufficiency suggests that high filling pressure (CVP) is inappropriate and not simply a reflection of increased intrathoracic or pericardial pressure. Determining whether RA wall tension is appropriate is more difficult in heart failure with preserved EF, where some atrial wall stretch may be necessary to maintain sufficient forward flow. Gather additional data in ambiguous cases, including a careful assessment of the clinical response to positional changes that affect venous returns, recent fluid administration, and other changes in fluid balance.
Image 13.4 A dilated, poorly collapsing IVC (A) requires additional cardiac views to determine if high CVP is associated with excess RA wall tension i.e., plethoric chambers due in part to systolic heart failure (B) or isolated RV overload (C). However, high CVP may be appropriate if associated with impaired filling due to severe LVH (D), pericardial tamponade (E), or positive pressure ventilation.
Infused IVFs often fail to induce a sustained rise in blood pressure and volume. A primary mechanism for this is that atrial filling promotes the release of atrial natriuretic peptide (ANP), which opposes the renin-angiotensin-aldosterone (RAAS) system to prevent volume-induced hypertension10 and dangerous inflammatory responses.11,12 Experimental evidence suggests that inflammatory cytokines like IL-6 also stimulate ANP release,13,14 and ANP can substantially increase vascular permeability when inflammation is already significant.15 Therefore, when assessing for volume intolerance, consider focal or systemic injury or inflammation that can increase the probability and consequence of extravascular fluid accumulation.
Hypoalbuminemia and pre-existing peripheral edema, ascites, pleural effusions, or pulmonary edema should be considered additional data suggesting fluid intolerance. In addition, patients with injury and inflammation in critical locations may be particularly vulnerable to increasing edema due to IVF administration. These include patients with acute respiratory distress syndrome (ARDS), traumatic brain injury, acute pancreatitis, and trauma patients requiring a damage control laparotomy. Fluid restrictive protocols in these situations are associated with better outcomes.16–20 In patients with severe malaria due to Plasmodium falciparum, where fatal cases often involve cerebral edema, the available evidence also supports a more restrictive fluid strategy.21,22
Sepsis is “life-threatening organ dysfunction caused by a dysregulated host response to infection”23 Microbes, often bacteria spreading in the blood, trigger a chain reaction where macrophages release inflammatory cytokines, such as IL-6 and TNF-alpha, leading to neurohormonal changes that 1) move plasma from the vasculature into the lymphatic system via the spleen,24,25 and 2) disrupt the endothelium and its glycocalyx (surface layer) causing increased vascular permeability and dysregulated vasodilation.26–28 To maintain circulation and tissue perfusion in the setting of decreased preload and afterload, the body ramps up the adrenergic and renin-angiotensin-aldosterone (RAAS) systems, releases vasopressin, and increases cortisol. Patients with sepsis may present with or develop fevers, chills, tachycardia, tachypnea, hypotension, altered mental status, kidney injury, respiratory failure, liver injury, stress cardiomyopathy, leukocytosis, leukopenia, thrombocytopenia, and dysregulated coagulation. Multi-organ failure and death can occur rapidly when blood flow is poorly distributed, or these compensatory mechanisms become inadequate. Prompt diagnosis and treatment improve mortality.29 PoCUS can help rule in/out alternative or additional causes of organ failure. It can also sometimes identify the source of infection, such as soft tissue fluid collections, lung consolidation, ascites or pleural fluid, endocarditis, bowel distention, and urinary tract obstruction.
Treating sepsis involves three simultaneous components, including 1) promptly finding and treating the source(s), 2) immediately infusing effective IV antibiotics, ideally after obtaining blood cultures, and 3) individualized hemodynamic support, such as IVF administration and vasopressor infusion. Sepsis guidelines previously recommended that patients with sepsis immediately receive IVF boluses to improve markers of cardiac output and tissue perfusion. However, there are at least two problems with this. First, the diagnostic criteria for sepsis are optimized to be sensitive (88%) and not specific to avoid missing septic patients who need prompt treatment.23,30,31 As a result, many patients initially considered septic have instead another diagnosis, such as decompensated heart failure, where IVFs are contraindicated.32–34 Second, if present, any benefit from upfront IVF boluses in sepsis is not universal. Multiple large randomized controlled trials in Africa demonstrated increased mortality when patients with septic shock receive upfront IVF boluses despite improving markers of cardiac output and tissue perfusion.35–40 Due to these concerns, sepsis guidelines are evolving to emphasize individualized fluid management. In 2021, the Surviving Sepsis Campaign Guidelines changed the upfront bolus recommendation to only a suggestion, citing a low level of evidence.41
In 2011, the New England Journal of Medicine published the FEAST trial, which compared giving a large fluid bolus or just a maintenance infusion to 3141 children in Sub-Saharan Africa with shock and suspected sepsis. Mortality was significantly higher in the fluid bolus group, with a relative risk of 1.44 (95% CI 1.09 - 1.90, P=0.01), despite more improvement in blood pressure in the treatment group 1 hour after randomization. Thus, the immediate feedback (improvement in blood pressure) can be misleading. The same finding was present across subgroups, including those with or without objective signs of malnutrition, severe anemia (Hgb <5g/dL), or malaria.35,42 In 2017, the Journal of the American Medical Association published a trial enrolling 212 adults with septic shock in Zambia’s central hospital that compared usual care to an immediate 2L IVF bolus and then an additional 2L infusion (4L max) over 4 hours but with protocols to stop if worsening hypoxia, tachypnea, or jugular venous distention were detected. The intervention group received more fluid by hour 6 (3.5L vs. 2L) but less discrepancy by hour 72 (5L vs. 4L). The 28-day mortality was 48% in the intervention group and 33% in the usual care group, with RR 1.48 (95% CI 1.14-1.91). Like the FEAST trial, multiple analyses led to similar results, and the increased mortality was consistent across various subgroups.36 These trials suggest that more aggressive upfront IVFs increase mortality in adults and children with suspected septic shock. A post-hoc analysis of the FEAST trial by Levin et al. revealed that excess mortality in the bolus group was associated with worsening hyperchloremia, metabolic acidosis, and anemia.43 Another analysis showed that the leading cause of excess deaths, which peaked 2-11 hours post-intervention, was circulatory as opposed to respiratory or neurologic.44 In a different effort to elucidate the mechanism behind these findings, Byrne et al. attempted to replicate the FEAST trial in 16 anesthetized sheep by inducing septic shock with a progressive 4-hour infusion of endotoxin. They then compared giving 40mL/kg of IVFs over 1 hour plus vasopressors to giving vasopressors alone. The sheep who received IVFs had greater improvement in cardiac output but also a greater collapse in systemic vascular resistance and required progressively more norepinephrine to maintain adequate arterial blood pressure. Interestingly, they found that ANP significantly increased in response to endotoxin infusion in both groups but rose even higher after fluid resuscitation.45 Additional animal models suggest that ANP activity is deleterious during sepsis,46,47 and ANP levels positively correlate with sepsis mortality.48 If IVF-induced atrial wall stretch increases ANP further during sepsis and ANP promotes systemic vasodilation and vascular permeability, then this could explain why IVFs appear to improve cardiac output and perfusion initially but eventually worsen shock hours later.
Echocardiography studies reveal that about one-third of patients with sepsis who get echocardiograms have RV dysfunction, which is independently associated with mortality.49 RV dysfunction can develop whenever pulmonary hypertension occurs and RV perfusion is limited. These conditions occur in tandem during septic shock due to factors like hypotension, hypoxia, acidosis, and positive pressure ventilation. RV dysfunction promotes atrial-wall stretch and can directly impede LV filling when thoracic or pericardial space is limited. Refractory septic shock may often be a consequence of this ventricular interdependence. Thus, in addition to avoiding excess RV dilation with IVFs, incorporate other RV-protective measures for anyone with septic shock. These include the early use of vasopressors to maintain a sufficient mean arterial pressure and efforts to minimize hyperinflation, high-pressure ventilation, high tidal volumes, hypoxia, and acidosis. The chapter on isolated RV overload addresses this in more detail.
IVFs may also impair circulation by worsening anemia and hypoalbuminemia via hemodilution. In addition, chloride-rich fluids like normal saline and 5% albumin solutions induce hyperchloremia.50 Raising plasma chloride worsens acidosis and suppresses the renin-angiotensin-aldosterone system (RAAS) by causing each kidney to overestimate its GFR and inhibit appropriate renin secretion.51–53 Inadequate RAAS system activation may be particularly harmful in sepsis, where angiotensin II can help maintain systemic blood pressure. Therefore, it is helpful to use chloride-sparing balanced solutions like Lactated Ringer’s instead of normal saline to prevent hyperchloremia-related hypotension54,55 and acute kidney injury.56–58
Image 13.5 Illustration of physiologic changes in treated sepsis with end-organ damage despite compensatory changes (solid green). IV fluid resuscitation sufficient to increase CVP (orange) can improve cardiac output and decrease serum lactate concentration. However, in some patients, these benefits may be outweighed by the harms of hemodilution and increased ANP release. Other interventions remain essential (dashed green) to reverse this process.
The above self-reinforcing interactions explain how IVFs could provoke delayed-onset cardiovascular collapse despite immediate improvements in cardiac output or lactate clearance. These nuanced understandings argue against the routine administration of IVF boluses to everyone with sepsis. The subset of patients that ultimately benefit from upfront IVFs in sepsis was likely under-represented in the cited African studies. This may include older, heavier, more sedentary patients with comorbidities, such as hypertension and diabetes mellitus, that benefit from ACE inhibitors or angiotensin II receptor blockers at baseline. These patients often have underlying diastolic dysfunction and a predisposition to develop an LV outflow tract obstruction in response to drops in both LV afterload and preload. An ICU in France found that 22% of patients in early septic shock had LV hypertrophy, creating a functional LVOT obstruction. They found that fluid infusion decreased the obstruction in most patients with LVOT obstruction and improved cardiac output.59 Thus, it may be reasonable to give IVF boluses to septic patients who do not have an inappropriately elevated CVP and are either 1) dehydrated due to poor intake, gastrointestinal losses, diuretic use, or hyperglycemia, 2) have a pre-existing preload-dependent disorder i.e. diastolic dysfunction or LVOT obstruction, or 3) have comorbidities like hypertension, atherosclerosis cardiovascular disease, and diabetes mellitus that particularly benefit from an ACE inhibitor or angiotensin II receptor blocker, which suggests that they have high baseline RAAS activity and endothelial dysfunction. In contrast, septic patients who are younger, thinner, and without another reason to be dehydrated are likely to benefit from less upfront IVFs and early administration of vasopressors, even if they lack apparent volume intolerance, as they are less likely to decompensate from low preload & afterload and more likely decompensate because they are unable to activate or sustain their compensatory RAAS and adrenergic systems sufficiently.
When considering whether to give IVFs, use PoCUS as part of a more extensive evaluation to identify fluid intolerance, defined as having an inappropriately elevated central venous pressure (CVP) or a specific vulnerability to tissue edema. Large upfront IVF boluses often improve blood pressure and serum lactate in patients with sepsis. However, multiple large randomized controlled trials in Africa suggest that upfront IVF boluses increase mortality even when they immediately improve markers of perfusion. No similar studies have been performed elsewhere, including where patients are predominantly older and more obese. So, it is unknown whether these findings are generalizable to other settings. One can speculate that those patients who do ultimately benefit from upfront IVF boluses in sepsis will have another reason for hypovolemia, a preload-dependent cardiac condition, or high baseline RAAS activation that renders the IVF-induced hypotensive effects of ANP less potent. PoCUS may identify isolated RV overload (iRVO), which may drive a hemodynamic collapse in many patients with septic shock. Therefore, make RV protective measures a core part of sepsis treatment. This includes using chloride-sparing solutions instead of normal saline to minimize hyperchloremia. For a comprehensive guide on managing sepsis, refer to the latest international treatment guidelines or another comprehensive, up-to-date resource.41,60