Ventilator-associated events: are we losing the plot?

Hospital-acquired infections are generally considered preventable and used as a quality assessment tool in health care by regulatory bodies. Ventilator-associated pneumonia (VAP) is one of the quality indicators employed by accreditation bodies, including the National Accreditation Board for Hospitals and Healthcare Providers (NABH) in India. It is not unusual for hospital administrators and ICU staff, in particular, to dedicate considerable time and effort to ensure compliance to the satisfaction of the assessors involved with accreditation. Increasing emphasis is placed on the new definitions and surveillance methodology in the pursuit of evaluating the incidence of VAP.  

The new approach to surveillance

Let us consider the changes that have evolved over the past few years in the surveillance of VAP in critically ill patients. VAP Surveillance and public reporting commenced in the US several years ago, at the behest of the  Center for Disease Control and Prevention (CDC) and the National Healthcare Safety Network (NHSN). An inexplicable decline in the incidence of VAP was observed with the commencement of surveillance. The most likely reason for the steep decline may have arisen from biased interpretation and reporting by hospitals faced with unwelcome consequences related to a high incidence. The highly subjective criteria employed for the diagnosis of VAP may have led to bias and underreporting. Against this background, the CDC, along with the Critical Care Societies Collaborative, attempted to develop a more objective and credible definition of VAP.1 A three-tiered algorithmic approach was introduced under the broad category of ventilator-associated events (VAE). The first step was defined as ventilator-associated conditions (VAC), followed by infection-related, ventilator-associated conditions (IVAC), and finally, possible or probable VAP(Fig. 1). Radiological confirmation of consolidation was not considered as an essential requirement for the diagnosis of VAP considering its lack of specificity. 

Fig 1. The new ventilator-associated events (VAE) algorithm. VAC: ventilator-associated conditions; IVAC: infected ventilator-associated conditions; VAP: ventilator-associated pneumonia

Problems with the new approach

The step-wise approach requires stable or decreasing FiOand PEEP levels for two calendar days. A sustained FiOincrease by more than 0.2 or an increase of PEEP by more than 3 cm of H2O for at least two consecutive days must follow the period of stability before a diagnosis of IVAC or VAP can be considered. However, a pre-defined period of stable oxygenation may not occur in patients with worsening lung disease at the outset, such as pneumonia or acute respiratory distress syndrome (ARDS). Furthermore, VAP can clearly occur without a sustained period of diminished oxygenation requiring an increase of FiOor PEEP.2 Although the new definitions were intended to add objectivity to VAP surveillance, they have not been validated by clinical evidence. 

What does the evidence suggest? 

A Dutch cohort study revealed that the VAE algorithm could identify only 32% of patients with VAP, diagnosed by prospective surveillance.3 In another retrospective study that evaluated 165 episodes of VAP diagnosed according to the conventional definition, only 12.1% were identified as probable VAP and 1.2% as possible VAP based on the new diagnostic algorithm.4 Clearly, the new definitions are poorly sensitive to the diagnosis of VAP; besides, worsening of oxygenation is often due to non-infective conditions such as cardiogenic pulmonary edema, ARDS, and atelectasis. Furthermore, some episodes of VAP may be missed out by the exclusion of radiological findings from the VAE algorithm. Fan et al. performed a meta-analysis of 61,489 patients from 18 studies to evaluate consistency between surveillance using conventional VAP criteria and the VAE algorithm. The VAE algorithm underestimated the true incidence of VAP as diagnosed by conventional criteria, with a pooled sensitivity and positive predictive value of less than 50%.5

The new algorithm as a quality indicator

The creators of the VAE algorithm argue that it is meant solely for surveillance and not for clinical management of patients. However, it appears to be an imperfect screening tool and lacks sensitivity for the diagnosis of VAP based on the clinical evidence available. The new algorithm is clearly ineffective in the early identification of mechanical ventilation-related complications by focussing solely on deterioration of oxygenation. 

In India, if you seek NABH accreditation, toeing the CDC line with the VAE algorithm appears to be a key ICU quality indicator. Hospitals take great pains to furnish VAE data to the satisfaction of NABH assessors who are often far removed from the real world of bedside clinical practice. Besides, there is often an irrational expectation to seek a “root cause” for everything under the sun and corrective action that will “eliminate” the problem, with little realization that many of the common complications of clinical medicine may not be entirely preventable. 

The bottom line

  • The new algorithm-based surveillance defines VAE as sustained deterioration in oxygenation after a period of stability; the new definitions were meant to add more objectivity to VAP surveillance. 
  • Lack of sensitivity to diagnose VAP is one of the main drawbacks of the new algorithm; many patients who develop VAP do not experience stable oxygenation followed by sustained deterioration. 
  • Absence of radiological correlation may also lead to underestimation of the true incidence of VAP.
  • The validity of the new algorithm has not been established in clinical studies.
  • Regulatory bodies need to be mindful of the relative lack of evidence to support the VAE algorithm-based surveillance methodology as a quality assessment tool. 


1.         Magill SS, Klompas M, Balk R, et al. Executive summary: Developing a new, national approach to surveillance for ventilator-associated events. Ann Am Thorac Soc. 2013;10(6):S220-223. doi:10.1513/AnnalsATS.201309-314OT

2.         Lilly CM, Landry KE, Sood RN, et al. Prevalence and test characteristics of national health safety network ventilator-associated events. Crit Care Med. 2014;42(9):2019-2028. doi:10.1097/CCM.0000000000000396

3.         Klein Klouwenberg PMC, van Mourik MSM, Ong DSY, et al. Electronic implementation of a novel surveillance paradigm for ventilator-associated events. Feasibility and validation. Am J Respir Crit Care Med. 2014;189(8):947-955. doi:10.1164/rccm.201307-1376OC

4.         Chang H-C, Chen C-M, Kung S-C, Wang C-M, Liu W-L, Lai C-C. Differences between novel and conventional surveillance paradigms of ventilator-associated pneumonia. Am J Infect Control. 2015;43(2):133-136. doi:10.1016/j.ajic.2014.10.029

5.         Fan Y, Gao F, Wu Y, Zhang J, Zhu M, Xiong L. Does ventilator-associated event surveillance detect ventilator-associated pneumonia in intensive care units? A systematic review and meta-analysis. Crit Care. 2016;20(1):338. doi:10.1186/s13054-016-1506-z


Oxygen : the elixir of life or the kiss of death?

The administration of supplemental oxygen is ubiquitous in medical practice, especially among critically ill patients. Hyperoxia is fairly common during oxygen therapy, and generally considered to be less deleterious than the potential harm that may arise from hypoxia. However, there has been an increased understanding of the detrimental effects of hyperoxia in recent times. How does the clinician balance the lifesaving effect of oxygen therapy while minimizing possible harm? 

Hypoxemia and compensatory mechanisms

A PaO2 level of less than 60 mm Hg is generally considered to be the lower limit of acceptable oxygenation. Several adaptive mechanisms are triggered in the presence of hypoxia. Downregulation of mitochondrial uncoupling has been demonstrated in human volunteers; hypoxia may enable more efficient ATP generation and mitochondrial protection.1 The release of hypoxia-inducible factors result in the activation of glycolytic enzymes leading to a preponderance of anerobic metabolism. Mitochondrial hibernation may occur, resulting in a reduction in the oxygen demand.2 Other compensatory mechanisms that enable acclimatization include hypoxic pulmonary vasoconstriction, increased cardiac output, polycythemia, and increased production of 2-3 diphosphoglycerate with a shift of the oxygen-dissociation curve to the right, enabling offloading of oxygen to the tissues.3

Potential harmful effects of hyperoxia

Hyperoxia results in the generation of reactive oxygen species (ROS). Excessive production of ROS resulting from high levels of oxygen may lead to cellular death by necrosis or apoptosis. Furthermore, ROS are the chief mediators of reperfusion injury. Hyperoxia has been well known to lead to pulmonary damage similar to acute respiratory distress syndrome, inhibition of mucociliary transport, and atelectasis. Supranormal arterial partial pressures of oxygen also lead to a fall in the cardiac output due to generalized vasoconstriction and increased afterload. Besides, coronary vasoconstriction may occur, and predispose to myocardial ischemia.4

Evidence of harm or lack of benefit from hyperoxia

In a randomized controlled study of normoxic patients with acute ST-elevation myocardial infarction (STEMI), supplemental oxygen at 8L/min was compared to no oxygen therapy. Recurrent myocardial infarction was significantly more common in the supplemental oxygen group; besides, there was an increase in the incidence of cardiac arrhythmias with oxygen supplementation. Infarct size assessed by cardiac magnetic resonance imaging (MRI) was significantly higher with oxygen therapy.5 In a similar study of normoxic patients with STEMI, supplemental oxygen was administered until the completion of percutaneous coronary intervention and compared to controls who received no supplemental oxygen. No difference was observed in the myocardial salvage index assessed by cardiac MRI or in the infarct size.6 These studies suggest that supplemental oxygen therapy in normoxic patients with acute myocardial infarction offers no benefit and may be potentially harmful. 

Does supplemental oxygen improve outcomes in acute stroke patients who are not hypoxic? Patients with acute stroke were randomized to receive continuous supplemental oxygen for 72 h, nocturnal oxygen only for three consecutive nights, or no oxygen in the Stroke Oxygen Study.7 No significant difference was observed in death or disability at 3 months, assessed using the modified Rankin score. 

The association between supranormal oxygen levels and in-hospital mortality was examined in a multicenter cohort study among survivors of cardiac arrest. A PaO2 level of > 300 mm Hg carried a significantly higher in-hospital mortality compared to 60–300 mm Hg or < 60 mm Hg.8 This study suggests that supranormal oxygen levels may worsen reperfusion injury following cardiac arrest. 

Three contemporaneous studies have compared a conservative to a more liberal oxygenation target in mechanically ventilated patients. The Oxygen ICU study randomized mechanically ventilated patients targeting a PaOlevel between 70–100 mm or oxygen saturation (SpO2) of 94–98% in the conservative group.9 In the liberal group, PaO2 levels of up to 150 mm Hg was allowed, with an oxygen saturation between 97–100%. ICU mortality was significantly lower in the conservative group (25% vs. 44%, p = 0.01). Panwar et al. studied patients who were likely to require mechanical ventilation for more than 24 h.10 In the conservative group, the target SpO2 was 88–92% compared to more than 96% in the liberal group. No significant differences were observed in new onset organ dysfunction, intensive care or 90-day mortality between groups.

An observational cohort study involving three Dutch ICUs assessed the association between predefined metrics of hypoxia and clinical outcomes.11 Mild hyperoxia was defined as a PaOranging between 120 to 200 mm of Hg and severe hyperoxia as a PaOmore than 200 mm Hg. Logistic regression analysis was performed with each metric of hyperoxia using specific thresholds. A significant association was observed between severe hyperoxia and mortality rates; ventilator-free days were also less in patients with hyperoxia. A linear relationship was seen between the duration of hyperoxia and hospital mortality. 

Permissive hypoxia 

A U-shaped relationship has been demonstrated between PaO2 and hospital mortality; the lowest mortality being between 110-150 mm Hg. Mortality was higher at a PaOof <67 mm Hg and >225 mm Hg.12 There has been increasing focus on “permissive” hypoxia, in patients who might suffer adverse outcomes from interventions aimed to maintain oxygenation in the physiological range. It is also important to consider lung injury resulting from high FiOlevels besides the harmful effects of systemic hypoxemia. Permissive hypoxia must ensure a favorable balance between oxygen consumption and delivery. The parameters that may help assess adequacy of oxygen delivery include lactate levels, central venous oxygen saturation (>70%), central venous to arterial PaCOdifference (<6 mm Hg), and the ratio between the central venous and arterial venous PaCO2difference and the arterial to central venous PaOdifference (<1.23).13 Clinical evidence to support permissive hypoxia is lacking; the threshold levels below which harm may ensue is unknown and likely to vary depending on age, the underlying clinical condition, and the chronicity of hypoxia. In practice, targeting lower values of PaO2 (55-80 mmHg) may be safer compared to the adverse outcomes that may ensue from efforts at maintaining normoxia. 

The bottom line

  • High levels of inspired oxygen and PaOhave been well established to cause harm. Aiming for physiological levels of oxygenation with injurious ventilatory strategies are likely to be counterproductive. 
  • Several damage-limiting compensatory mechanisms are triggered by hypoxia; however, hyperoxia may lead to unmitigated harmful effects. 
  • Supplemental oxygen therapy is not beneficial and may lead to adverse outcomes in normoxic patients with acute myocardial infarction and acute stroke. 
  • The is a growing body of evidence that supports the use of conservative targets for oxygenation in critically ill patients. 
  • Permissive hypoxia may be appropriate in selected patients; however, the safe thresholds for permissible oxygen levels are unknown. 


1.        Martin DS, Grocott MPW. Oxygen Therapy in Critical Illness: Precise Control of Arterial Oxygenation and Permissive Hypoxemia*. Crit Care Med. 2013;41(2):423-432. doi:10.1097/CCM.0b013e31826a44f6

2.         Schumacker PT. Hypoxia-inducible factor-1 (HIF-1). Crit Care Med. 2005;33(12 Suppl):S423-425.

3.         MacIntyre NR. Tissue Hypoxia: Implications for the Respiratory Clinician. Respir Care. 2014;59(10):1590-1596. doi:10.4187/respcare.03357

4.         Farquhar H, Weatherall M, Wijesinghe M, et al. Systematic review of studies of the effect of hyperoxia on coronary blood flow. Am Heart J. 2009;158(3):371-377. doi:10.1016/j.ahj.2009.05.037

5.         Stub D, Smith K, Bernard S, et al. Air Versus Oxygen in ST-Segment–Elevation Myocardial Infarction. :8.

6.         Khoshnood A, Carlsson M, Akbarzadeh M, et al. Effect of oxygen therapy on myocardial salvage in ST elevation myocardial infarction: the randomized SOCCER trial. Eur J Emerg Med Off J Eur Soc Emerg Med. 2018;25(2):78-84. doi:10.1097/MEJ.0000000000000431

7.         Roffe C, Nevatte T, Sim J, et al. Effect of Routine Low-Dose Oxygen Supplementation on Death and Disability in Adults With Acute Stroke: The Stroke Oxygen Study Randomized Clinical Trial. JAMA. 2017;318(12):1125. doi:10.1001/jama.2017.11463

8.         Kilgannon J. Hope, Jones Alan E., Parrillo Joseph E., et al. Relationship Between Supranormal Oxygen Tension and Outcome After Resuscitation From Cardiac Arrest.Circulation. 2011;123(23):2717-2722. doi:10.1161/CIRCULATIONAHA.110.001016

9.         Girardis M, Busani S, Damiani E, et al. Effect of Conservative vs Conventional Oxygen Therapy on Mortality Among Patients in an Intensive Care Unit: The Oxygen-ICU Randomized Clinical Trial. JAMA. 2016;316(15):1583. doi:10.1001/jama.2016.11993

10.       Panwar R, Hardie M, Bellomo R, et al. Conservative versus Liberal Oxygenation Targets for Mechanically Ventilated Patients. A Pilot Multicenter Randomized Controlled Trial. Am J Respir Crit Care Med. 2016;193(1):43-51. doi:10.1164/rccm.201505-1019OC

11.       Helmerhorst HJF, Arts DL, Schultz MJ, et al. Metrics of Arterial Hyperoxia and Associated Outcomes in Critical Care*: Crit Care Med. 2017;45(2):187-195. doi:10.1097/CCM.0000000000002084

12.       de Jonge E, Peelen L, Keijzers PJ, et al. Association between administered oxygen, arterial partial oxygen pressure and mortality in mechanically ventilated intensive care unit patients. Crit Care. 2008;12(6):R156. doi:10.1186/cc7150

13.       He H, Liu D, Long Y, Wang X. High central venous-to-arterial CO2 difference/arterial-central venous O2 difference ratio is associated with poor lactate clearance in septic patients after resuscitation. J Crit Care. 2016;31(1):76-81. doi:10.1016/j.jcrc.2015.10.017

Decompressive Craniectomy for Severe Traumatic Brain Injury: Does It Make Life Worth Living?

Severe traumatic brain injury (TBI) due to focal or diffuse lesions leads to raised intracranial pressure (ICP). The normal ICP is less than 15 mm Hg; if the ICP remains persistently high, cerebral perfusion is compromised. Unrelieved intracranial hypertension culminates in irreversible neurological deterioration leading to fatal brain herniations. Raised ICP may be controlled by surgical evacuation of focal lesions when feasible and medical therapy including head end elevation, sedation, ventilation to normocarbia, osmotherapy, induced hypothermia, and metabolic suppression using barbiturate infusion. In refractory intracranial hypertension, a decompressive craniectomy (DC) is being increasingly performed, involving removal of a large segment of the skull and opening of the underlying dura mater.

It makes physiological sense to perform DC; if excessive pressure within the skull squashes the brain and leads to fatal herniation, opening the skull to relieve the pressure seems intuitive. Historically, the pressure-relieving effect of DC has been well known. In 1901, Kocher proposed DC for patients with raised ICP following TBI, followed by sporadic use of this technique over the years. Following its established efficacy in acute ischemic stroke,there has been an upsurge of interest in DC for severe TBI with intractable intracranial hypertension in recent times.

How does it work?  

According to the Munro-Kellie hypothesis, the intracranial compartment in inelastic. Increase in volume of any of the intracranial contents (brain, blood, or cerebrospinal fluid), is compensated by a shift of the remaining contents out of the skull. However, when compensatory mechanisms fail, the ICP begins to rise. Opening the rigid skull allows room for the intracranial contents to expand, thus relieving pressure. Maintenance of ICP between 20–25 mm Hg improves outcomes following TBI.2,3

Technique of DC

DC is performed to reduce ICP and prevent brain herniation. Craniectomy must be extensive allowing enough room for the brain to expand. A unilateral frontotemporoparietal craniectomy is commonly carried out for a unilateral focal lesion. The incision is approximately 2 cm lateral to the midline to prevent injury to the superior sagittal sinus. The anteroposterior diameter of the bone flap must be more than 15 cm and descend down to the base of the temporal fossa to enable adequate decompression (Figure 1). Bifrontal craniectomy is performed for diffuse lesions with generalized cerebral edema. Through a bicoronal skin flap, a frontotemporal bone flap is removed including the bone over the superior sagittal sinus (Figure 2). The dura is opened after removal of the bone flap followed by augmentative duroplasty.


Does DC improve clinical outcomes?

Although relief of life-threatening intracranial hypertension may seem rational and intuitive, does it really improve clinical outcomes in TBI? Two randomized controlled studies have addressed this all-important question. The DECRA study recruited 155 patients with severe, diffuse TBI between 2002–2010.Patients with an ICP of more than 20 mm Hg for more than 15 min during a 1-h period, either continuously or intermittently, despite optimal first-tier medical management were enrolled. DC was performed within the first 72 h of injury in the study group, while standard medical care was continued in the control group. The surgical technique involved a bifrontotemporoparietal craniectomy with bilateral dural opening. Predictably, patients who underwent DC had significantly lower ICP compared to those who received standard medical management alone. The composite primary outcome comprised of death, vegetative state, or severe disability corresponding to a score of 1–4 on the extended Glasgow outcome scale (GOS-E). At 6-month follow-up, the composite unfavorableoutcome was significantly higher among patients who underwent DC compared to those who received standard medical therapy alone (70% vs. 51%). The study had several limitations including a higher number of patients with non-reactive pupils in the DC arm. The ICP threshold for DC was considered too low and out of tune with real-world care. Furthermore, bifrontal craniectomy without sectioning of the falx may not have been the most optimal DC technique. Notwithstanding these limitations, DECRA suggested that relief of ICP with DC may not translate to improved clinical outcomes.

The RESCUE-ICP study enrolled 398 patients with severe TBI between 2004–2014 who had an ICP of more than 25 mm Hg for 1–12 h in spite of stage I and II treatments.Patients were randomized to undergo DC or continued medical treatment, including barbiturate infusion. In contrast to the DECRA study, patients who had undergone evacuation of an intracranial hematoma were included if a craniectomy had not been performed previously. A large frontotemporoparietal DC was carried out for unilateral lesions and bifrontal DC was performed for diffuse lesions, based on the judgment of the operating surgeon. The primary outcome was assessed using the GOS-E at 6 months. Patients were followed up at for 24 months after randomization. In the DECRA study, a favorable outcome was defined as a score of 5 or higher on the GOS-E; however, in the RESCUE-ICP study, a score of 4 or higher was defined as a favorable outcome. (It is important to note this difference; a score of 4 on the GOS-E indicates that the patient is able to be alone at home, without assistance, for a period of up to 8 h). At 6 m follow-up, mortality was significantly lower with DC; however, there was a higher incidence of vegetative state, lower and upper severe disability compared to medical therapy. On sensitivity analysis, at 6 months, a favorable GOS-E (4 or higher) was not significantly different between patients who underwent DC and those who did not.

The RESCUE-ICP study suggests that DC improves survival at the expense of severely disabled life. However, the culture, beliefs, and values of the community may be crucial in determining what may be an acceptable level of functional ability. For instance, a GOS-E of 4 indicates the ability to be alone at home, but unable to travel or shop independently. Many patients and families may consider this an acceptable outcome, especially among the Indian population. Another important consideration is the validity of these studies in settings where ICP monitoring is not routine, with decision-making based primarily on clinical examination and CT imaging.

The bottom line

  • DC as a relatively early intervention does not improve outcomes; it improves survival as a later intervention when first and second stage medical therapies fail.
  • Improved survival with DC may result in a life with significant disability; however, the degree of acceptable disability varies among populations.
  • A considered, patient and family-centered decision is important in deciding the most optimal therapy.
  • A subset of patients may benefit from DC; it may be challenging to identify or study this subgroup in randomized controlled trials.



  1. Vahedi K, Hofmeijer J, Juettler E, et al. Early decompressive surgery in malignant infarction of the middle cerebral artery: a pooled analysis of three randomised controlled trials. Lancet Neurol. 2007;6(3):215-222. doi:10.1016/S1474-4422(07)70036-4
  2. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons, et al. Guidelines for the management of severe traumatic brain injury. VI. Indications for intracranial pressure monitoring. J Neurotrauma. 2007;24 Suppl 1:S37-44. doi:10.1089/neu.2007.9990
  3. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons, et al. Guidelines for the management of severe traumatic brain injury. VIII. Intracranial pressure thresholds. J Neurotrauma. 2007;24 Suppl 1:S55-58. doi:10.1089/neu.2007.9988
  4. Cooper DJ, Rosenfeld JV, Murray L, et al. Decompressive Craniectomy in Diffuse Traumatic Brain Injury. N Engl J Med. 2011;364(16):1493-1502. doi:10.1056/NEJMoa1102077
  5. Hutchinson PJ, Kolias AG, Timofeev IS, et al. Trial of Decompressive Craniectomy for Traumatic Intracranial Hypertension. N Engl J Med. 2016;375(12):1119-1130. doi:10.1056/NEJMoa1605215


Damage control resuscitation: redefining trauma management



Many conventionally held dogmas in trauma resuscitation have been disproven in the past decade. Generous use of crystalloids during the early resuscitation phase of trauma, recommended by the Advanced Trauma Life Support Course, based on earlier studies, has been largely shown to be harmful. Coagulopathy was considered unlikely with up to transfusion of six units of packed red cells. Supplementation with plasma and platelets was thought to be unnecessary unless there was clinical or laboratory evidence of coagulopathy. However, we know today that trauma-induced endogenous coagulopathy occurs at the onset of shock; besides, conventional tests of clotting function are poorly sensitive in coagulopathy associated with trauma. There has been a major shift in the resuscitation strategy of major trauma over the years. One of the most important tenets that have emerged in the management of severe trauma has been damage control resuscitation.

Let us consider the key elements of damage control resuscitation in major trauma.

Recognition of the actively bleeding patient

Damage control resuscitation applies specifically to the patient who is actively bleeding. Hypovolemia from ongoing blood loss is compensated by vasoconstriction, with the maintenance of normal heart rates and blood pressures. However, a fall in cardiac output and compromised organ perfusion occurs in this compensated state. In this state of cryptic shock, hemodynamic parameters may appear normal; however, it is associated with increased mortality (1). The response to a fluid bolus may help identify patients who are actively bleeding. Patients who respond to a fluid bolus and continue to remain stable may not have ongoing bleeding; however, non-responders and transient responders are likely to be actively bleeding. In practice, a considered decision on the likelihood of ongoing bleed must be made based on the nature of injuries and the response to the initial administration of a fluid bolus. If ongoing bleeding is suspected, packed red cells are preferred over crystalloids as the initial bolus of fluid.

Cut down crystalloid use

Administration of large volumes of crystalloids leads to major adverse outcomes in severe trauma. Dilution of clotting factors commonly occurs, leading to the perpetuation of coagulopathy associated with major trauma. Hypothermia is another adverse effect of the administration of large volumes of crystalloids. The excessive use of normal saline may lead to metabolic acidosis. Abdominal compartment syndrome with multiorgan failure may be associated with overzealous crystalloid resuscitation. Delayed fluid resuscitation was studied in patients with penetrating injury to the torso, who presented with a pre-hospital systolic blood pressure of 90 mm Hg or less. Patients were randomized to receive pre-hospital standard resuscitation or delayed resuscitation after reaching the operating room. Hospital mortality was significantly less with delayed resuscitation; the duration of hospital stay was also shorter in the delayed resuscitation group (2).

Permissive hypotension

The key to damage control resuscitation is the prioritization of hemorrhage control over perfusion. The blood pressure is maintained at a low level to prevent hydrostatic disruption of the clot. The lowest blood pressure required to maintain coronary perfusion is targeted during damage control resuscitation. Considering that coronary flow occurs mainly during diastole, a diastolic pressure of 25–30 mm Hg may be the minimum target; this may be extrapolated to a systolic pressure target of 60–70 mm Hg (3). For practical purposes, the presence of a radial pulse may be a rough indicator of an adequate level of perfusion (4). Measurement of lactate levels and the base deficit also offers important information on the circulatory state. A drop in the end-tidal carbon dioxide (ETCO2) levels is closely associated with decreasing cardiac output. Continuous monitoring of ETCOmay be useful in monitoring the hemodynamic status in patients who are intubated and ventilated.

It is unclear whether a strategy of permissive hypotension is appropriate in hemorrhaging patients with traumatic brain injury (TBI). There is strong evidence to suggest that even transient hypotension may lead to worse outcomes in TBI (5). No definite recommendations are available in this situation; however, it may be appropriate to follow a hypotensive strategy if the predominant clinical problem is ongoing major bleeding. However, if TBI appears to be the dominant pathophysiology, adequate cerebral perfusion pressures must be maintained (3).

Hemostatic resuscitation

Coagulopathy occurs in trauma by two different mechanisms. An early, endogenous coagulopathy occurs due to activation of protein C, resulting in the cleaving of factors Va and VIIIa with reduced thrombin generation. Degradation of fibrinogen and excessive fibrinolytic activity also contributes to the early, endogenous coagulopathy. During the later phase, dilutional coagulopathy secondary to resuscitation further exacerbates impairment of clotting function. Dilutional coagulopathy may be prevented using a balanced transfusion regimen, by administering an equal ratio of red cells, plasma, and platelets. Protocols using a balanced transfusion strategy have been shown to reduce mortality and mortality. The PROPPR trial compared the efficacy of a 1:1:1 ratio of plasma, platelets, and RBCs with a 1:1:2 ratio; although there was no difference in the 24 h and 28 d mortality, there were a lower number of deaths due to exsanguination with a 1:1:1 strategy (6). Fibrinogen levels may drop in spite of a 1:1:1 resuscitation strategy; it is important to maintain levels >2g/L with cryoprecipitate or fibrinogen concentrate.

Tranexamic acid has been shown to reduce blood loss and mortality in bleeding trauma patients. The CRASH II trial revealed a mortality reduction of 15% in patients with exsanguinating hemorrhage. Tranexamic acid must be administered within 3 h of injury; the initial dose is 1 gm followed by a repeat dose of 1 gm infused over 8 h (7).

It is important to note that conventional parameters of clotting function, including APTT and INR are poorly sensitive for the coagulopathy associated with major trauma; besides, they can only be tested intermittently. Point of care testing of coagulation function with thromboelastography (TEG) is being increasingly used to monitor trauma-induced coagulopathy. It is possible to assess the entire coagulation cascade including platelet function with TEG.

The bottom line

  • The traditional strategy of liberal crystalloid resuscitation aimed at hemodynamic optimization leads to adverse outcomes in trauma patients with active bleeding.
  • In patients who are actively bleeding, resuscitation should ideally commence with blood products.
  • Damage control resuscitation prioritizes hemorrhage control over optimization of hemodynamic parameters; permissive hypotension is an important element of this strategy.
  • Hemostatic resuscitation involves the transfusion of plasma and platelets in a 1:1:1 ratio. Conventional tests of coagulation are poorly sensitive in major trauma.
  • Future research is required to evaluate hemodynamic targets during damage control resuscitation and the management of actively bleeding patients with traumatic brain injury.



  1. Meregalli A, Oliveira RP, Friedman G. Occult hypoperfusion is associated with increased mortality in hemodynamically stable, high-risk, surgical patients. Crit Care Lond Engl. 2004 Apr;8(2):R60-65.
  2. Bickell WH, Wall MJ, Pepe PE, Martin RR, Ginger VF, Allen MK, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med. 1994 Oct 27;331(17):1105–9.
  3. Harris T, Davenport R, Mak M, Brohi K. The evolving science of trauma resuscitation. Emerg Med Clin North Am. 2018 Feb;36(1):85–106.
  4. Giannoudi M, Harwood P. Damage control resuscitation: lessons learned. Eur J Trauma Emerg Surg. 2016 Jun;42(3):273–82.
  5. Spaite DW, Hu C, Bobrow BJ, Chikani V, Sherrill D, Barnhart B, et al. Mortality and prehospital blood pressure in patients with major traumatic brain injury: Implications for the hypotension threshold. JAMA Surg. 2017 Apr 1;152(4):360–8.
  6. Holcomb JB, Tilley BC, Baraniuk S, Fox EE, Wade CE, Podbielski JM, et al. Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 Ratio and mortality in patients with severe trauma: The PROPPR randomized clinical trial. JAMA. 2015 Feb 3;313(5):471–82.
  7. Roberts I, Shakur H, Coats T, Hunt B, Balogun E, Barnetson L, et al. The CRASH-2 trial: a randomised controlled trial and economic evaluation of the effects of tranexamic acid on death, vascular occlusive events and transfusion requirement in bleeding trauma patients. Health Technol Assess Winch Engl. 2013 Mar;17(10):1–79.


The endothelial glycocalyx, the modified Starling principle, and rational fluid therapy

History has witnessed intense debates on the behavior of intravenously administered fluids in critically ill patients. The basic concepts of capillary permeability have changed in recent times. There has been an increasing understanding of the crucial role played by the glycocalyx that lines the endothelium on the behavior of intravenously administered fluids.

The Starling principle

According to the Starling principle, movement of fluid across the capillary barrier is driven by the difference in the hydrostatic and colloid osmotic pressures between the capillaries and the interstitial space. The higher hydrostatic pressure within the capillaries tends to drive fluid out to the interstitium; the higher colloid osmotic pressure of the plasma tends to retain fluid within the capillaries. Thus, according to the Starling principle, the filtered volume per unit area is as follows:

Jv/A = Lp [(Pc–Pis) – σ (πc–πis)] (Fig. 1)

fig 1 mod


Fig 1. The Starling principle that governs the net movement of fluid from the capillaries to the interstitium. The difference between the capillary and interstitial hydrostatic pressures (Pc–Pis) tends to drive fluid out; the difference in colloid osmotic pressures (πc–πis) tends to retain fluid within the capillary lumen. 

Jv/A: filtered volume per unit area; Lp: hydraulic conductance; Pc: capillary hydrostatic pressure; Pis: interstitial hydrostatic pressure; σ: osmotic reflection coefficient; πc: capillary colloid osmotic pressure; πis: interstitial colloid osmotic pressure

The endothelial glycocalyx layer

Recent changes in the understanding of vascular permeability have led to changes in the concepts that guide transendothelial fluid movement. The key to the revised concept is the presence of the glycocalyx layer that lines the lumen of the vascular endothelium. The existence of this layer was unknown at the time that Starling proposed his hypothesis.

The glycocalyx is composed of proteoglycans including syndecans and glypicans. Attached to the central core of the proteoglycans are glycosaminoglycan (GAG) side chains, including heparan, chondroitin, dermatan, and keratin. The endothelial glycocalyx plays a crucial role in regulating vascular permeability. Furthermore, it regulates coagulation, fibrinolysis, and modulates shear stress on the vasculature. The glycocalyx suffers damage in several types of critical illnesses including sepsis, trauma, and following major surgery. Besides, ischemia-reperfusion injury, hypervolemia, hyperglycemia, and hypernatremia are known to cause shedding of the glycocalyx layer. Damage to the glycocalyx results in increased vascular permeability, platelet aggregation, adhesion of leukocytes, and leads to a prothrombotic state, typically seen is sepsis.

Revisions to the Starling principle

Based on the pivotal function of the glycocalyx, several important modifications have been made to the Starling principle.

Capillary filtration is less than predicted by the Starling principle

It has been consistently observed that filtration out of the capillaries to the interstitial space is much less than predicted by the Starling principle. This phenomenon is due to the presence of the sub-glycocalyx layer, a protein-free layer between the glycocalyx and the endothelium. The absence of protein essentially means that the sub-glycocalyx layer has a negligible colloid osmotic pressure. Thus, the Starling equation for calculation of the filtered fluid needs to be revised; specifically, the force that opposes filtration is the difference between the capillary (πc) and sub-glycocalyx (πg) colloid osmotic pressures. Clearly, (πc – πg) is much higher than (πc – πis), thereby, reducing the filtered volume per unit area, Jv/A.

Jv/A = Lp [(Pc– Pis) – σ (πc – πg)] (Fig. 2)

Fig 2 mod


Fig 2. The modified Starling principle. The difference between the capillary and interstitial hydrostatic pressures (Pc–Pis) tends to drive fluid out; however, the force that tends to retain fluid within the capillaris is the difference in colloid osmotic pressure between the capillary and the subglycocalyx layer (πc – πg)

Jv/A: filtered volume per unit area; Lp: hydraulic conductance; Pc: capillary hydrostatic pressure; Pis: interstitial hydrostatic pressure; σ: osmotic reflection coefficient; πc: capillary colloid osmotic pressure; πg: colloid osmotic pressure of the glycocalyx 

In clinical practice, at normal or low levels of Pc, the filtered volume is similar during intravenous administration of both crystalloids and colloids. This explains why in recent studies, resuscitation volumes required to achieve hemodynamic endpoints were similar with both crystalloids and colloids. In two trials of critically ill patients, 100 ml of normal saline was comparable to 62–76 ml of human albumin (1)and 63–69 ml of pentastarch, a synthetic colloid (2).

At higher levels of Pc, more filtration occurs with crystalloids compared to colloids, resulting in more efficient filling of the intravascular compartment with colloids.

The interstitial COP is not a major determinant of filtered volume

The interstitial COP is higher than originally assumed and plays very little role in determining the filtration rate. The main driver of filtration is the transendothelial hydrostatic pressure difference and the difference in COP between the plasma and the sub-glycocalyx layer.

Under most circumstances, there is no reabsorption of fluid from the interstitium to the capillaries

According to the Starling principle, fluid is filtered from the arterial end of capillaries and reabsorbed from the venous end. However, under most circumstances, there is no reabsorption of fluid from the interstitium to the capillaries. On the contrary, the filtered fluid is absorbed by the lymphatic system. Reabsorption of fluid back into the capillaries does not occur because the driving force for filtration, Pc, is higher than the forces that oppose filtration [Pis + (πc – πis)] throughout the capillary system.

Pc > Pis + (πc – πis)

Transient fluid reabsorption of about 500 ml may occur from the interstitium to the capillaries, for about 15–30 min after a sudden decrease in Pc. This may occur due to rapid loss of intravascular volume, for instance, from hemorrhage. However, equilibrium is reached after a transient phase of reabsorption, followed by continued filtration.

In clinical practice, this implies that the administration of colloids does not result in reabsorption of fluid from the interstitium back into the intravascular compartment; resolution of tissue edema does not happen with colloid use. On the contrary, if the glycocalyx is damaged, colloids may leak out from the capillaries and worsen edema.

Measures that may protect the glycocalyx

Several studies have shown improved clinical outcomes when a higher ratio of fresh frozen plasma to packed red cells is administered during major transfusion (3). A more liberal use of fresh frozen plasma may thus help preserve the glycocalyx following major transfusion. Adverse clinical outcomes associated with overzealous fluid resuscitation have emerged from many recent trials (4,5). Hypervolemia due to aggressive resuscitation may lead to damage to the glycocalyx and increased vascular permeability. There has been an intense debate on the optimal degree of glycemic control among critically ill patients (6,7). Avoidance of hyperglycemia may be one of the important strategies that may preserve the glycocalyx. Hydrocortisone may preserve the glycocalyx by attenuating possible damage by inflammatory mediators, especially TNF-alpha (8). It is possible that the purported clinical benefits observed with corticosteroid use in sepsis may be related to a favorable effect on the glycocalyx. Volatile anesthetic agents, especially sevoflurane may conserve the glycocalyx in post-ischemic coronary artery beds and improve coronary flow (9).

The bottom line

  • The presumed superiority of colloid resuscitation based on the Starling principle is a flawed concept.
  • In critically ill patients, the glycocalyx layer suffers damage, leading to leakage of both colloids and crystalloids.
  • Transendothelial filtration of fluid occurs to a similar extent with both crystalloids and colloids at normal or low capillary pressures.
  • Filtration occurs throughout the capillary bed; reabsorption does not occur except transiently following acute hypovolemia.
  • Reabsorption of fluid from the interstitium to the intravascular compartment and resolution of edema does not occur with the infusion of colloids.
  • Maintenance of normovolemia and normoglycemia may protect the glycocalyx; hydrocortisone may attenuate damage to the glycocalyx in septic patients. The glycocalyx-preserving effect of fresh frozen plasma during major transfusion needs further investigation.



  1. Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R, et al. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med. 2004 May 27;350(22):2247–56.
  2. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008 Jan 10;358(2):125–39.
  3. Gonzalez EA, Moore FA, Holcomb JB, Miller CC, Kozar RA, Todd SR, et al. Fresh frozen plasma should be given earlier to patients requiring massive transfusion. J Trauma. 2007 Jan;62(1):112–9.
  4. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, et al. Mortality after Fluid Bolus in African Children with Severe Infection. N Engl J Med. 2011 Jun 30;364(26):2483–95.
  5. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006 Jun 15;354(24):2564–75.
  6. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al. Intensive Insulin Therapy in Critically Ill Patients. N Engl J Med. 2001 Nov 8;345(19):1359–67.
  7. Intensive versus Conventional Glucose Control in Critically Ill Patients. N Engl J Med. 2009 Mar 26;360(13):1283–97.
  8. Chappell D, Hofmann-Kiefer K, Jacob M, Rehm M, Briegel J, Welsch U, et al. TNF-alpha induced shedding of the endothelial glycocalyx is prevented by hydrocortisone and antithrombin. Basic Res Cardiol. 2009 Jan;104(1):78–89.
  9. Annecke T, Chappell D, Chen C, Jacob M, Welsch U, Sommerhoff CP, et al. Sevoflurane preserves the endothelial glycocalyx against ischaemia-reperfusion injury. Br J Anaesth. 2010 Apr;104(4):414–21.





Fluid resuscitation in septic patients: Is it a case of “less is more”?


In patients with septic shock, one of the key initial interventions is fluid resuscitation. The Surviving Sepsis Guidelines recommend an initial volume of resuscitation of 30 ml/kg, followed by additional boluses guided by volume responsiveness (1). In fact, most patients with septic shock receive around 5 liters of fluid in the first few hours of resuscitation (2).

Septic shock is characterized by widespread vasodilatation of vascular beds leading to hypotension. Besides, intravascular fluid leaks into the interstitial space due to capillary endothelial dysfunction. Administration of intravenous fluid enables filling up of the intravascular compartment and compensates for the loss of fluid into the interstitial space. Fluid replenishment may also improve perfusion pressures to the vital organs.

However, do the putative benefits of targeted fluid resuscitation lead to improved clinical outcomes in septic patients?

The evidence behind fluid resuscitation

In a landmark study, Rivers at al. randomized patients presenting to the emergency department with severe sepsis or septic shock to a protocolized, early, goal-directed therapy (EGDT) or to clinician-guided resuscitation targets (3). In the study arm, the resuscitative interventions included fluid boluses of 500 ml of crystalloid every 30 min targeting a central venous pressure of 8–12 mm Hg and vasopressors to achieve a mean arterial pressure of 65 mm Hg. The central venous oxygen saturation (ScvO2) was monitored continuously using a special catheter; packed red cells were transfused if the ScvOdropped below 70%. If the ScvOremained low in spite of all the above measures, a dobutamine infusion was commenced aiming to improve the cardiac output. Patients who received goal-directed therapy had significantly larger volumes of fluid resuscitation in the first 6 h (5.0 L vs 3.5 L). In-hospital mortality, the primary endpoint, was significantly lower in patients who received goal-directed therapy (30.5 vs. 46.5%, p = 0.009).

Several subsequent studies seem to ratify the findings of this single-center study from an emergency department in the US (4). However, it remained unclear whether the improvement in clinical outcomes was related to the overall impact of a “bundled” approach or due to specific elements of the bundle. Particularly, were the favorable effects due to earlier recognition of sepsis and appropriate antibiotic administration? Indeed, a recent meta-analysis suggests that the favorable effect on mortality may be more likely due to earlier administration of appropriate antibiotics compared to the effects of fluid resuscitation or attainment of hemodynamic targets (5).

More than a decade after the original EGDT trial, three randomized controlled studies were published that compared EGDT with usual care. The ProCESS (6), ARISE (7), and ProMISe (8)studies did not find any difference in any clinical outcomes, including mortality, with the use of EGDT. The volume of fluids administered during the first 6 h of resuscitation was similar across all the groups in these studies. However, it is important to note that the volume of fluid received by all the groups in the first 6 h was less than in either arm of the original EGDT study.

 Can fluid resuscitation cause harm?

The Fluid Expansion as Supportive Therapy (FEAST) study was conducted among children in sub-Saharan Africa,  who presented with a severe febrile illness and signs of impaired perfusion. In this provocative three-armed study, fluid boluses of 20 to 40 ml of 5% albumin or normal saline solution were compared with a control group who received no bolus fluid (9). The 48 h mortality was significantly higher in children who received bolus fluids compared to those who did not (10.5% vs. 7.3%, p = 0.003). Interestingly, although shock resolution occurred more often in the bolus groups, improved blood pressures did not translate to better survival.

The mean cumulative positive fluid balance has ranged from 5–11 liters in the first week of illness in previous studies among septic patients (10). Several observational studies have reported increased mortality with a higher cumulative fluid balance. The Vasopressin in Septic Shock Trial (VAAST) revealed a doubling of mortality among patients with the highest cumulative fluid balance (10). The Sepsis Occurrence in Acutely Ill Patients (SOAP) study also revealed similar findings, with a 10% excess mortality for every liter of positive fluid balance at 72 hours (11).  A fluid-conservative strategy was evaluated during the first week of illness in patients with acute respiratory distress syndrome (12). The mean cumulative fluid balance was –136 ± 491 ml vs. 6992 ± 502 ml (p < 0.001) in the fluid-conservative and fluid-liberal groups. A fluid-conservative strategy resulted in significantly less duration of mechanical ventilation and ICU stay.

A fluid-restrictive approach

An alternative strategy to liberal fluid resuscitation would be to limit fluid administration and use vasopressors early to maintain perfusion pressures and preserve organ function. Excessive fluid administration may raise venous pressures, leading to a reduction in the perfusion pressure to vital organs. Vasopressors decrease the venous capacitance, thereby, effectively increasing the venous return and cardiac output without overloading vital organs with excessive fluid and tissue edema. A restrictive approach may be justified as the response to intravenous fluid boluses is often transient; furthermore, fluid accumulation occurs over time leading to organ dysfunction involving the lungs, kidneys, and the heart. Besides, the effect of a fluid bolus among patients in circulatory shock may be transient; further fluid boluses may not improve the hemodynamic status even among apparent fluid-responders (13).

A forthcoming multi-center randomized controlled trial (CLOVERS) aims to compare a conventional resuscitation strategy with a fluid-restrictive approach during the first 24 h in patients with septic shock. The liberal strategy will use conventional fluid therapy similar to the control arms of the ProCESS, ARISE, and ProMISe studies. The fluid-restrictive approach will involve early vasopressor administration with lower volumes of initial resuscitation fluid.

The bottom line

  • The landmark EGDT study and subsequent accumulated evidence has led to widespread adoption of targeted fluid resuscitation in septic patients.
  • Excessive fluid administration and a cumulative positive balance may lead to fluid overload, edema, and dysfunction of vital organs.
  • Recent evidence supports the possibility of clinical harm from overzealous fluid administration.
  • An alternative approach may be to restrict the volume of resuscitation fluid and use vasopressors early to maintain perfusion pressure to vital organs.
  • The optimal volume of initial fluid administration and hemodynamic targets remain unclear and require further studies.


  1. Levy MM, Evans LE, Rhodes A. The Surviving Sepsis Campaign Bundle: 2018 update. Intensive Care Med. 2018 Jun;44(6):925–8.
  2. Nguyen HB, Jaehne AK, Jayaprakash N, Semler MW, Hegab S, Yataco AC, et al. Early goal-directed therapy in severe sepsis and septic shock: insights and comparisons to ProCESS, ProMISe, and ARISE. Crit Care Lond Engl. 2016 Jul 1;20(1):160.
  3. Emanuel R, Bryant N, Suzanne H, Julie R, Alexandria M, Bernhard K, et al. Early Goal-Directed Therapy in the Treatment of Severe Sepsis and Septic Shock. N Engl J Med. 2001;10.
  4. Trzeciak S, Dellinger RP, Abate NL, Cowan RM, Stauss M, Kilgannon JH, et al. Translating research to clinical practice: a 1-year experience with implementing early goal-directed therapy for septic shock in the emergency department. Chest. 2006 Feb;129(2):225–32.
  5. Kalil AC, Johnson DW, Lisco SJ, Sun J. Early Goal-Directed Therapy for Sepsis: A Novel Solution for Discordant Survival Outcomes in Clinical Trials. Crit Care Med. 2017 Apr;45(4):607–14.
  6. ProCESS Investigators, Yealy DM, Kellum JA, Huang DT, Barnato AE, Weissfeld LA, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med. 2014 May 1;370(18):1683–93.
  7. ARISE Investigators, ANZICS Clinical Trials Group, Peake SL, Delaney A, Bailey M, Bellomo R, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med. 2014 Oct 16;371(16):1496–506.
  8. Mouncey PR, Osborn TM, Power GS, Harrison DA, Sadique MZ, Grieve RD, et al. Trial of early, goal-directed resuscitation for septic shock. N Engl J Med. 2015 Apr 2;372(14):1301–11.
  9. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, et al. Mortality after Fluid Bolus in African Children with Severe Infection. N Engl J Med. 2011 Jun 30;364(26):2483–95.
  10. Boyd JH, Forbes J, Nakada T, Walley KR, Russell JA. Fluid resuscitation in septic shock: A positive fluid balance and elevated central venous pressure are associated with increased mortality*. Crit Care Med. 2011 Feb 1;39(2):259–65.
  11. Vincent J-L, Sakr Y, Sprung C, Ranieri V, Reinhart K, Gerlach H, et al. Sepsis in European intensive care units: Results of the SOAP study*. Crit Care Med. 2006 Feb 1;34(2):344–53.
  12. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006 Jun 15;354(24):2564–75.
  13. 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 [Internet]. 2014 Dec [cited 2019 Apr 6];4(1). Available from:







Renal Replacement Therapy in Acute Kidney Injury: It Is All About Timing!

Intensive care physicians often face the conundrum of deciding when to consider renal replacement therapy (RRT) in acute kidney injury (AKI). RRT may be commenced for the early correction of metabolic complications and prevention of volume overload. However, an early strategy may entail unnecessary therapy for some patients who might recover renal function otherwise. Even more concerning is the perpetuation of damage that may occur due to the hemodynamic instability that often accompanies RRT. The proximal tubules located at the corticomedullary junction are particularly vulnerable to ischemic insult. The partial pressure of oxygen at this region is less than 40 mm Hg, which increases the susceptibility of the tubular cells to RRT-induced damage (1) (Fig 1). Hence it is critically important to decide on the requirement for RRT and to initiate therapy at the optimal time.


Fig 1. The corticomedullary junction (marked in red) has a low partial pressure of oxygen (less than 40 mm Hg). The relatively low partial pressure of oxygen makes the proximal tubular cells vulnerable to ischemic insult that may occur during renal replacement therapy

Many of the older studies that support early RRT have been observational in nature (2); besides, the overall management of critically ill patients has changed considerably over the years. Hence, it is appropriate to reconsider the impact of the timing of RRT, particularly with new information available from the more recent randomized controlled studies.

The ELAIN study involved 231 postoperative patients (47% cardiac surgical) with AKI from a single center in Germany (3). The early group was randomized to receive RRT within 8 h of the diagnosis of AKI, KDIGO stage 2; the late group was randomized to undergo RRT within 12 hours of reaching KDIGO stage 3. Besides, at least one of the following conditions had to be present, including severe sepsis, catecholamine use, refractory fluid overload, and new-onset or worsening of non-renal organ dysfunction. All 112 patients randomized to the early group received RRT, compared to 108/119 patients in the late group. The 90-d mortality, the primary outcome, was significantly lower in the early group. Approximately 75% of all patients had evidence of fluid overload according to the pre-defined criteria. This raises the question of whether RRT may have been delayed among patients who had evidence of fluid overload but were randomized to the late group, leading to worse outcomes in this group of patients. Besides, the fragility index for the primary outcome was only 3 patients (the results would have been non-significant if 3 extra patients had suffered 90-day mortality in the early group). Besides, it is hard to explain a mortality benefit at 90 d, with no difference in the 28 and 60 d mortality.

Patients in KDIGO stage 3 were randomized to receive early or late RRT in the AKIKI study (4). In the early group, RRT was initiated immediately after randomization. In the late group, RRT was performed if any of the following criteria were present: K+> 6.0 mmol/l (> 5.5 mmol/l after corrective measures), BUN > 112 mg/dl, pH < 7.15, and acute pulmonary edema due to fluid overload. In the early group, 305/311 (98.1%) received RRT, while only 157/308 (51%) patients in the late group underwent RRT. There was no significant difference in mortality between groups on day 60, which was the primary outcome of the study. Thus, in the AKIKI study, nearly half of all patients who were randomized to the late group did not require RRT but did not suffer adverse outcomes.

The IDEAL-ICU investigators, in a multicentric French study, randomized patients in early-stage septic shock who were in the failure stage (stage F) of the RIFLE classification (5). The early group received RRT within 12 h of documentation of stage F; the late group underwent RRT after 48 h if renal recovery had not occurred in the meantime. The primary outcome was 90-d mortality. The study was stopped for futility after a second interim analysis. No difference was observed in the 90-d mortality between the early and late groups (58% vs. 54%, p = 0.38). In the late group, 38% of patients did not receive RRT.

The AKIKI and the IDEAL-ICU studies suggest that in the absence of life-threatening complications, a delayed strategy may be appropriate in most patients with AKI. Such a strategy may facilitate avoidance of RRT, with its associated complications. In fact, these studies reinforce the view that initiation of RRT should not be based on the AKI stage; it may be more appropriate to be guided by the presence of complications, including fluid overload and metabolic abnormalities. Furthermore, a strategy of watchful waiting may enhance the possibility of renal recovery. In contrast, RRT-related hemodynamic instability may lead to worsening or perpetuation of renal damage and adversely impact recovery.

It is also important to assess the trajectory of the underlying disease process and the presence of failure of other organs while considering RRT. If the patient is clinically improving with the resolution of other organ failures, perhaps a waiting strategy would be more appropriate. However, if the metabolic demands continue to remain high with a continued requirement for fluid resuscitation, it may be necessary to consider early RRT. In such situations, the metabolic and fluid demands far exceed the capacity of the kidney necessitating prompt supportive therapy. The “frusemide test” was used to identify patients who are likely to require renal replacement therapy in a previous study (6). It involves the administration intravenous frusemide, 1.0 mg/kg to frusemide-naïve patients and 1.5 mg/kg to those already on frusemide. A urine output of less than 200 ml in the ensuing 2 h was highly predictive of patients who required RRT.

The bottom line

  • Previous observational studies of AKI seemed to suggest that an early-RRT strategy may lead to more favorable clinical outcomes.
  • Recent randomized controlled trials reinforce the view that a delayed strategy may be appropriate in the absence of life-threatening complications of AKI.
  • Apart from the complications directly related to therapy, hemodynamic perturbations during RRT may adversely impact recovery of renal function.
  • The decision to perform RRT will also depend on the trajectory of the underlying illness and metabolic demand.
  • Fastidious conformance to physiological or biochemical parameters is unlikely to be helpful with the timing of initiation of RRT.
  • A holistic, individualized strategy based on clinical, physiological, and biochemical parameters is more likely to be beneficial compared to a “one-size-fits-all” approach.



  1. Gaudry S, Quenot J-P, Hertig A, Barbar SD, Hajage D, Ricard J-D, et al. Timing of Renal Replacement Therapy for Severe Acute Kidney Injury in Critically Ill Patients. Am J Respir Crit Care Med [Internet]. 2019 Feb 20 [cited 2019 Mar 21]; Available from:
  2. Bagshaw SM, Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, et al. Timing of renal replacement therapy and clinical outcomes in critically ill patients with severe acute kidney injury. J Crit Care. 2009 Mar;24(1):129–40.
  3. Zarbock A, Kellum JA, Schmidt C, Van Aken H, Wempe C, Pavenstädt H, et al. Effect of Early vs Delayed Initiation of Renal Replacement Therapy on Mortality in Critically Ill Patients With Acute Kidney Injury: The ELAIN Randomized Clinical Trial. JAMA. 2016 May 24;315(20):2190.
  4. Gaudry S, Hajage D, Schortgen F, Martin-Lefevre L, Pons B, Boulet E, et al. Initiation Strategies for Renal-Replacement Therapy in the Intensive Care Unit. N Engl J Med. 2016 Jul 14;375(2):122–33.
  5. Barbar SD, Clere-Jehl R, Bourredjem A, Hernu R, Montini F, Bruyère R, et al. Timing of Renal-Replacement Therapy in Patients with Acute Kidney Injury and Sepsis. N Engl J Med. 2018 Oct 11;379(15):1431–42.
  6. Lumlertgul N, Peerapornratana S, Trakarnvanich T, Pongsittisak W, Surasit K, Chuasuwan A, et al. Early versus standard initiation of renal replacement therapy in furosemide stress test non-responsive acute kidney injury patients (the FST trial). Crit Care Lond Engl. 2018 Apr 19;22(1):101.



Superbugs vs. superdrugs: are we waging a losing battle?


Intensive care units (ICUs) are the breeding grounds for resistant microorganisms. The use of invasive devices that breach physiological defensive barriers predispose to nosocomial infections in ICUs. A state of “immunoparalysis” often accompanies critical illness, including sepsis, trauma, and major surgery. Furthermore, therapy with powerful, broad-spectrum antibiotics, appropriate or otherwise, is common in the ICU, contributing to the preponderance of drug-resistant microorganisms.

The most frequently encountered resistant pathogens in the ICU include:

  • Carbapenem-resistant Enterobacteriaceae
  • Multidrug-resistant Pseudomonas aeruginosa, Acinetobacter spp, and Stenotrophomonas maltophilia
  • Methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE)

In this review, we shall focus on resistant gram-negative pathogens that are commonly encountered in ICUs.

Carbapenemase-producing Enterobacteriaceae

The family of Enterobacteriaceae includes Escherichia coli, Klebsiella pneumoniae, and Enterobacter species, and commonly cause gram-negative bacteremia, intra-abdominal, and urinary tract infections. Infection with resistant Enterobacteriaceae results in poor clinical outcomes, including high mortality, in addition to the increased cost of care. Carbapenem resistance is conferred by the production of the hydrolyzing enzyme carbapenemase, an extended-spectrum beta-lactamase (ESBL). Carbapenemases are carried on genetic material including transposons and plasmids and transmitted extensively to other bacterial genera. Pseudomonas aeruginosa and Acinetobacter baumanni can also acquire carbapenem resistance through mechanisms other than carbapenemase production.

Beta-lactamases are classified based on the amino acid sequence into 4 types (Ambler classification).

  • Class A: Penicillinases, including Klebsiella pneumoniae carbapenemase (KPC). KPC confers resistance to most beta-lactams; besides, KPC can be transmitted from Klebsiella to other Enterobacteriaceae and Pseudomonas aeruginosa.
  • Class B: Metallo-beta-lactamases (MBL), including the New Delhi metallo-beta-lactamase (NDM-1). NDM-1-producing bacteria have been isolated from tap water and sewage effluents from New Delhi (1). Metallolactamases mediate resistance to all beta-lactamase inhibitors
  • Class C: Cephalosporinases
  • Class D: Oxacillinases, which hydrolyze oxacllin and hence referred to as OXA-type; some beta-lactamase inhibitors can neutralize them. Enterobacteriaceae and Acinetobacer baumanni are known to produce OXA-type carbapenemases.

Carbapenems are generally considered as the most effective therapy against resistant pathogens; however, there is an ever-increasing proliferation of carbapenem-resistant pathogens in ICUs all over the world. The mortality rate associated with carbapenem resistance may be 24–70% (2).

Faced with a daunting milieu of highly resistant, potentially lethal pathogens, what are our therapeutic options?  Several new antimicrobials have been added to our therapeutic armamentarium; besides some older antibiotics that have re-emerged as possible treatment options.

A common mechanism of resistance to carbapenem is through the production of Klebsiella pneumoniae carbapenemases (KPC). KPC is produced by several enterobacterial species; besides carbapenems, they confer resistance against penicillins, cephalosporins, and monobactams. Polymyxins, generally colistimethate (polymyxin E) is the preferred treatment for KPC-producing organisms if the isolate is sensitive. Polymyxins are often used in combination, usually with meropenem, if the minimum inhibitory concentration (MIC) for meropenem is less than 8–12 mcg/ml. Combination therapy with meropenem may not be superior to monotherapy with polymyxin if the MIC for meropenem is more than 16 mcg/ml (3). An emerging therapeutic option is the combination of ceftazidime, a third-generation cephalosporin with the novel beta-lactamase inhibitor, avibactam. Avibactam is more potent and active against a wider range of beta-lactamases including Amber class A, C, and D. The ceftazidime-avibactam combination was non-inferior to imipenem-cilastatin in complicated urinary tract infections, including acute pyelonephritis (4). In another study, a combination of ceftazidime-avibactam and metronidazole was shown to be equivalent to meropenem in the treatment of complicated intra-abdominal infections (5). This combination is an emerging therapeutic option for infections caused by carbapenemase-producing Enterobacteriaceae and has been approved by the FDA for the treatment of complicated urinary tract and intra-abdominal infections. The combination of meropenem with vaborbactam, another novel beta-lactamase inhibitor, may also be effective in the treatment of infection with KPC-producing organisms, although it requires further studies to establish clinical efficacy.

A polymyxin-based combination as described previously is generally preferred for metallo-lactamase producers. If the isolate is colistin resistant, the options are limited, as beta-lactamase inhibitors are ineffective against metallo-lactamases. However, ceftazidime-avibactam combined with aztreonam has been successfully used in this situation (6). This combination may have a synergistic effect as avibactam may inhibit beta-lactamases other than metallo-lactamases and enhance the effectiveness of aztreonam.

The new aminoglycoside antibiotic, plazomycin is another new antibiotic that may be useful in the treatment of carbapenemase-producing Enterobacteriaceae. Unaffected by carbapenemase, it is effective against beta-lactamase producing K. pneumoniae, E. coli, and Enterobacter species. However, it may be less effective against NDM-1 producers. Plazomycin was shown to be equally effective compared to levofloxacin in the treatment of complicated urinary tract infections (7). Besides, the nephrotoxic and ototoxic effects may be less compared to other aminoglycosides.


Intravenous fosfomycin has excellent penetration into tissues and body fluids including lungs, soft tissue, bone, urinary bladder, and the central nervous system. It has a broad spectrum of activity against gram-positive and gram-negative bacteria. The efficacy against carbapenemase-producing Enterobacteriaceae is variable. It is mainly used in urinary tract infections as it is excreted in the urine and works optimally in an acidic environment. There are anecdotal reports of its efficacy against NDM-1 producing Enterobacteriaceae (8). Fosfomycin is moderately effective against pseudomonas; however, Acinetobacter and anaerobic organisms are poorly susceptible. It is advisable to use fosfomycin as part of combination therapy as evidence for its efficacy as monotherapy is currently unclear.


Another re-emerging antibiotic, minocycline has evinced renewed interest in the treatment of multi-drug resistant infections. A tetracycline class of antibiotic, minocycline has been showed to be effective against infections caused by Acinetobacter baumanni. It has been successfully used in combination with colistin, meropenem, and aminoglycosides in carbapenem-resistant bloodstream infections caused by Klebsiella pneumoniae. It may result in higher serum levels compared to tigecycline; besides, it has relatively few side effects and carries the option of scaling down to oral therapy. The FDA has approved its use in urinary tract infections.

Pseudomonas aeruginosa and Acinetobacter baumanni resistant to carbapenems

Carbapenem-resistant A. baumannii and P. aeruginosa are usually resistant to all beta-lactams and fluoroquinolones. Polymyxins, usually colistimethate, are commonly used solely or in combination to treat multi-drug resistant infections due to these organisms. Besides, for ventilator-associated pneumonia and tracheobronchitis, aerosolized colistin may be administered. Aerosolized colistin has minimal systemic absorption and hence may be safely used in renal dysfunction. However, it is doubtful if aerosolized therapy alone is efficacious in lung infections.

Colistin resistance in A. baumanni and P. aeruginosa is bad news. Carbapenem-resistant A. baumanni may be susceptible to tigecycline; another option is eravacycline, the novel synthetic fluorocycline, which is 2–4 times more potent than tigecycline. Eravacycline was comparable to imipenem in the treatment of complicated intra-abdominal infections (9). Although sulbactam is another therapeutic option for multi-drug resistant A. baumanni, most isolates have reduced susceptibility. Ceftazidime-avibactam may not be effective against most isolates of carbapenem-resistant P. aeruginosa and A. baumanni. Aztreonam may be a potential therapeutic option against carbapenem-resistant P. aeruginosa; however, corroboratory clinical evidence regarding efficacy is lacking.

The bottom line

  • Carbapenemase-producing gram-negative pathogens pose an increasing threat worldwide and therapeutic options are limited. These organisms are resistant to most other classes of antibiotics. Among the most worrisome are KPC-producing pathogens.
  • The ceftazidime-avibactam combination is preferred for KPC-producing organisms.
  • Possible therapeutic options are re-emergent, older drugs, including colistin, fosfomycin, and minocycline; novel agents that may have an increasing role to play include ceftazidime-avibactam, plazomycin, and eravacycline.
  • The optimal dosing regimen is unclear with many of the newer drugs.
  • Metallo-lactamases, including NDM-1 have proliferated across ICUs in India.
  • Beta-lactamase inhibitors are largely ineffective against metallo-lactamase producing pathogens; colistin remains the preferred therapeutic option.
  • Although combination therapy is often preferred in resistant infections, there is no corroboratory clinical evidence.
  • Clinical studies are urgently required to adequately assess the efficacy of newer agents in the real world.



  1. Walsh TR, Weeks J, Livermore DM, Toleman MA. Dissemination of NDM-1 positive bacteria in the New Delhi environment and its implications for human health: an environmental point prevalence study. Lancet Infect Dis. 2011 May;11(5):355–62.
  2. Thaden JT, Pogue JM, Kaye KS. Role of newer and re-emerging older agents in the treatment of infections caused by carbapenem-resistant Enterobacteriaceae. Virulence. 2017 May 19;8(4):403–16.
  3. Tumbarello M, Viale P, Viscoli C, Trecarichi EM, Tumietto F, Marchese A, et al. Predictors of mortality in bloodstream infections caused by Klebsiella pneumoniae carbapenemase-producing K. pneumoniae: importance of combination therapy. Clin Infect Dis Off Publ Infect Dis Soc Am. 2012 Oct;55(7):943–50.
  4. Vazquez JA, González Patzán LD, Stricklin D, Duttaroy DD, Kreidly Z, Lipka J, et al. Efficacy and safety of ceftazidime-avibactam versus imipenem-cilastatin in the treatment of complicated urinary tract infections, including including acute pyelonephritis, in hospitalized adults: results of a prospective, investigator-blinded, randomized study. Curr Med Res Opin. 2012 Dec;28(12):1921-31
  5. Mazuski JE, Gasink LB, Armstrong J, Broadhurst H, Stone GG, Rank D, et al. Efficacy and Safety of Ceftazidime-Avibactam Plus Metronidazole Versus Meropenem in the Treatment of Complicated Intra-abdominal Infection: Results From a Randomized, Controlled, Double-Blind, Phase 3 Program. Clin Infect Dis Off Publ Infect Dis Soc Am. 2016 01;62(11):1380–9.
  6. Shaw E, Rombauts A, Tubau F, Padullés A, Càmara J, Lozano T, et al. Clinical outcomes after combination treatment with ceftazidime/avibactam and aztreonam for NDM-1/OXA-48/CTX-M-15-producing Klebsiella pneumoniae infection. J Antimicrob Chemother. 2018 Apr 1;73(4):1104–6.
  7. Connolly LE, Riddle V, Cebrik D, Armstrong ES, Miller LG. A Multicenter, Randomized, Double-Blind, Phase 2 Study of the Efficacy and Safety of Plazomicin Compared with Levofloxacin in the Treatment of Complicated Urinary Tract Infection and Acute Pyelonephritis. Antimicrob Agents Chemother. 2018;62(4).
  8. Denis C, Poirel L, Carricajo A, Grattard F, Fascia P, Verhoeven P, et al. Nosocomial transmission of NDM-1-producing Escherichia coli within a non-endemic area in France. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis. 2012 May;18(5):E128-130.
  9. Solomkin J, Evans D, Slepavicius A, Lee P, Marsh A, Tsai L, et al. Assessing the Efficacy and Safety of Eravacycline vs Ertapenem in Complicated Intra-abdominal Infections in the Investigating Gram-Negative Infections Treated With Eravacycline (IGNITE 1) Trial: A Randomized Clinical Trial. JAMA Surg. 2017 01;152(3):224–32.












Tidal volume and plateau pressure vs. driving pressure targeted ventilator management in ARDS


Mechanical ventilation in acute respiratory distress syndrome (ARDS) aims to maintain gas exchange and support respiratory muscles during the critical phase of illness. It is important to prevent possible harm from ventilation-induced lung injury (VILI) during this period. Limiting tidal volumes to 6 ml/kg of predicted body weight and plateau pressures to 30 cm of H2O are considered to be crucial in preventing VILI (1). However, ARDS is characterized by non-homogenous involvement of the lung; the extent of relatively normal lung available for ventilation is highly variable and has been conceptualized as the “baby lung”(2). Hence, a ventilation strategy targeting tidal volume based on the predicted body weight may result in hyperinflation of relatively normal lungs and lead to VILI (3).

Cyclic strain leads to mechanical stress and triggers VILI. The difference between the plateau pressure (Pplat) and PEEP, known as the driving pressure (ΔP) may be a surrogate for the degree of cyclic strain.

Driving pressure (ΔP) = Pplat – PEEP

Lung compliance (CRS) = ΔV/ ΔP

Hence, ΔP = ΔV/ CRS

As evident from this equation, to maintain a constant driving pressure, the tidal volume (ΔV) must decrease as the compliance decreases. Hence, the tidal volume will need to be adjusted based on compliance to control the driving pressure. Targeting tidal volume based on predicted body weight, on the contrary, may lead to excessive driving pressures and cause VILI.

Amato et al. re-analyzed data involving 3562 patients with ARDS from nine randomized controlled trials (4). A statistical methodology called multilevel mediation analysis was used to evaluate the impact of the driving pressure as an independent predictor of survival. Four parameters, including tidal volume, Pplat, PEEP, and driving pressure were tested for possible association with survival. Three important findings were observed. 1. For a constant level of PEEP, mortality increased as the driving pressure increased (Fig. 1); 2. A higher Pplat did not increase mortality if the driving pressure was maintained constant (Fig. 2); 3. Even with similar Pplat, the mortality decreased as the driving pressure decreased (Fig. 3) (for the same Pplat, the driving pressure becomes lower with an increase of PEEP). In simple terms, even if the Pplat rose, the mortality was higher only if driving pressure increased. Another interesting finding was that the tidal volume predicted survival only if it was adjusted for compliance. The tidal volume did not predict survival when adjusted for the predicted body weight.



Fig. 1 For a constant level of PEEP (red  bracket) the mortality increased with an increase of driving pressure (green bracket)



Fig. 2 A higher Pplat did not increase mortality if the driving pressure (green bracket) remained constant 



Fig 3. At constant Pplat, the mortality decreased with lower driving pressures (green bracket) (Figures adapted from Amato MBP, Meade MO, Slutsky AS, Brochard L, Costa ELV, Schoenfeld DA, et al. Driving Pressure and Survival in the Acute Respiratory Distress Syndrome. N Engl J Med. 2015 Feb 19;372(8):747–55)

Aoyoma et al. performed a meta-analysis including secondary analyses of five randomized controlled and two observational studies to assess the mortality risk associated with high driving pressures (5). They found a significant association of higher driving pressures with increased mortality among patients with ARDS (RR, 1.44; 95% CI, 1.11–1.88). From a sensitivity analysis of three studies with similar driving pressure thresholds, the authors suggested a target driving pressure of 13–15 cm of H2O.

Lung stress (calculated from esophageal and airway pressures), driving pressure, and lung and chest wall elastance were studied in 150 patients with ARDS at PEEP levels of 5 and 15 cm H2O (6). At both PEEP levels, increasing driving pressures were significantly associated with lung stress. The optimal cutoff value of driving pressure was 15 cm of H2O to prevent significant lung stress.

The driving pressure may also be useful as a tool to titrate PEEP for optimizing lung recruitment. Lower driving pressures for a constant tidal volume with increasing levels of PEEP suggests improved compliance and potential for recruitment (7).

VILI may be conceptualized to occur from excessive mechanical power applied to the lung (8). The mechanical power may be directly related to the driving pressure; high tidal volumes probably contribute to the damage indirectly through an increase in driving pressures. The mechanical power applied to the lung may also be a function of the respiratory rate. The lack of benefit with high-frequency oscillation ventilation, despite minimal tidal excursions, may be related to the rate-related increase in mechanical power.

There are several caveats to the application of the concept of driving pressure to clinical practice. First, the driving pressure may be significant only within the range of airway pressures commonly employed in clinical practice (Pplat less than 40 and PEEP more than 5 cm of H2O). Second, measurement of Pplat and driving pressure in spontaneously breathing patients is unreliable; hence driving pressure-guided ventilation is applicable only in patients who do not have spontaneous breathing efforts. Third, most of the available data are based on secondary analysis from previous randomized controlled studies. Controlled studies comparing driving pressure-guided with tidal volume and Pplat-guided ventilatory management are required to validate the applicability of driving pressure in clinical practice; besides, optimal cutoff levels for driving pressure need to be defined.

The bottom line

  • A ventilatory strategy targeting tidal volume and plateau pressure may not prevent VILI.
  • Conventionally accepted limits for “low” tidal volume may lead to hyperinflation and injury to relatively normal lung regions in ARDS.
  • The driving pressure (Pplat – PEEP) may be a more significant determinant of the mechanical power applied to the lungs.
  • Secondary analysis of previous randomized controlled studies of ARDS suggests that a higher Pplat does not increase mortality at constant driving pressures.
  • Similarly, a lower driving pressure may reduce mortality for the same level of Pplat.
  • Based on the limited evidence available, a target driving pressure of 13-15 cm of H2O may be appropriate to prevent VILI.
  • Controlled studies are required to validate the superiority of driving pressure targeted ventilator management with a tidal volume and Pplat-guided approach.



  1. Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome. N Engl J Med. 2000 May 4;342(18):1301–8.
  2. Gattinoni L, Pesenti A. The concept of “baby lung.” Intensive Care Med. 2005 Jun;31(6):776–84.
  3. Terragni PP, Rosboch G, Tealdi A, Corno E, Menaldo E, Davini O, et al. Tidal hyperinflation during low tidal volume ventilation in acute respiratory distress syndrome. Am J Respir Crit Care Med. 2007 Jan 15;175(2):160–6.
  4. Amato MBP, Meade MO, Slutsky AS, Brochard L, Costa ELV, Schoenfeld DA, et al. Driving Pressure and Survival in the Acute Respiratory Distress Syndrome. N Engl J Med. 2015 Feb 19;372(8):747–55.
  5. Aoyama H, Pettenuzzo T, Aoyama K, Pinto R, Englesakis M, Fan E. Association of Driving Pressure With Mortality Among Ventilated Patients With Acute Respiratory Distress Syndrome: A Systematic Review and Meta-Analysis*. Crit Care Med. 2018 Feb;46(2):300–6.
  6. Chiumello D, Carlesso E, Brioni M, Cressoni M. Airway driving pressure and lung stress in ARDS patients. Crit Care [Internet]. 2016 Aug 22 [cited 2019 Mar 4];20. Available from:
  7. Aoyama H, Yamada Y, Fan E. The future of driving pressure: a primary goal for mechanical ventilation? J Intensive Care [Internet]. 2018 Dec [cited 2019 Mar 4];6(1). Available from:
  8. Gattinoni L, Quintel M. How ARDS should be treated. Crit Care [Internet]. 2016 Dec [cited 2019 Mar 4];20(1). Available from:


The rocket science behind PEEP titration in ARDS


The acute respiratory distress syndrome (ARDS) was first described half a century ago by Ashbaugh et al. (1). They considered several therapeutic options to combat refractory hypoxemia and proposed that appropriate titration of positive end-expiratory pressure (PEEP) may be the sole effective intervention. Ever since the publication of this seminal paper, the pursuit of an adequate level of PEEP has never ceased to intrigue intensivists. Let us consider the physiology behind the application of PEEP and how it may be titrated to an ideal level.

Titration of PEEP to improve oxygenation

PEEP may be targeted to improve oxygenation; many intensivists aim for this easily discernible endpoint by the bedside. PEEP may also be used as part of an overall strategy aimed at lung protection. Conventionally, the inspiratory limb of the pressure-volume curve has been the focus of attention in an attempt to find the optimal pressure that will enable an open lung approach. However, lung recruitment is not confined to a single point such as the lower inflection point of the pressure-volume curve; it occurs throughout the inspiratory limb. The extent of lung recruitment increases with increasing inspiratory pressure. Thus, recruitment is essentially an inspiratory phenomenon. In contrast, PEEP is an expiratory phenomenon that prevents de-recruitment during the expiratory phase. Besides, it is important to note that for any level of airway pressure, the lung volume is higher during expiration compared to inspiration (Fig. 1). Hence, it may be more appropriate to titrate PEEP based on the deflation limb of the pressure-volume curve (2). Aiming for improvement in oxygenation as the sole target may also have several pitfalls. ARDS is a non-homogenous disease process; attempts to recruit non-ventilated regions of the lung invariably leads to overdistension of areas of the lung that are better ventilated. Overdistension may lead to a rise in the pulmonary artery and right ventricular pressures, leading to a fall in cardiac output. Any fall in cardiac output and impaired oxygen delivery may offset the advantage gained by improvement in arterial oxygen levels.


Fig 1. The pressure-volume curve of the lung. For a given level of pressure, the lung volume is higher during expiration compared to inspiration. Green arrow: Lower inflection point; red arrow: upper inflection point

Titration based on compliance

This method involves the upward titration of PEEP until the static compliance ceases to rise. Pintado et al. increased PEEP levels in a step-wise manner based on static compliance (3). The PEEP level that resulted in the highest static compliance was chosen in the intervention group. In the control group, PEEP was applied based on the FiO2, as recommended in the ARDS-net study. Patients who received best compliance-based PEEP levels had significantly more multiorgan dysfunction-free days. There were non-significant improvements in the PaO2/FiO2ratios in the initial 14 days of treatment and the 28-day mortality.

CT and ultrasound-guided lung recruitment

Would CT imaging help to assess recruitability and appropriate titration of PEEP? Cressoni et al. performed CT scans at PEEP levels of 5 and 45 cm of H2O and attempted to calculate the optimal level of PEEP (4). This theory assumes that PEEP constitutes the pressure required to elevate the chest wall and non-dependent lung that leads to compression and gravity-induced collapse of the dependent lung regions. However, no correlation could be found between CT-guided PEEP values and the extent of lung recruitment. Besides, repeated transfer of sick patients for CT makes this an unwelcome option. Repeated ultrasound imaging may also guide the clinician in choosing the appropriate level of PEEP. A progressive reduction in the extent of lung collapse with increasing PEEP is relatively easy to visualize by ultrasonography (Fig. 2). A significant limitation of this strategy may be the relative inability to view deeper regions of the lung by ultrasonography.


Fig 2. Ultrasonography-based titration of PEEP. Progressive recruitment of collapsed lung is seen from panel A to C

PEEP titration based on transpulmonary pressure

Maintenance of positive airway pressure alone may be insufficient to prevent alveolar collapse, especially at end-expiration. A higher pleural pressure compared to applied PEEP may lead to lung collapse. A non-compliant chest wall and increased intra-abdominal pressure may lead to variable pleural pressures in mechanically ventilated patients.

The distending pressure of the lung (which maintains the lung open) is the transpulmonary pressure:

Transpulmonary pressure = airway pressure – pleural pressure

The esophageal pressure is considered to be a reasonable surrogate of the pleural pressure. The transpulmonary pressure needs to be positive throughout the respiratory cycle to prevent derecruitment. Beitler et al. titrated PEEP levels to maintain the end-expiratory transpulmonary pressure equal to or higher than the esophageal pressure (between 0–6 cm H2O higher than the esophageal pressure) and compared it with PEEP settings based on a PEEP–FiOtable (5). The composite primary outcome of death and ventilation-free survival through day 28 did not differ significantly between groups. Pre-specified secondary outcomes, including 28-day mortality, ventilation-free days, and the need for rescue therapy were also not significantly different. Although titration of PEEP based on end-expiratory pleural pressures is based on strong physiological grounds, it requires further investigation for efficacy in severely hypoxic patients and during prone ventilation.

High vs. low PEEP

Three large randomized controlled studies have compared empirical high vs. low PEEP levels in patients with ARDS (6–8). The ARDS-net study used PEEP/FiO2tables for PEEP adjustment, while the ExPress study adopted a compliance-based PEEP strategy. None of these studies revealed a significant mortality benefit with the application of a higher PEEP level. Briel et al. performed a meta-analysis of the three studies (9). They observed that the treatment effect was variable and depended on the severity of ARDS, with a suggestion that a higher PEEP level may be beneficial in patients with more severe illness, with a PaO2/FiO2ratio < 200.

The bottom line

  • The holy grail of optimal PEEP lies in the grey area between maintenance of effective alveolar recruitment and the potential for harmful overdistension.
  • Strategies based solely on improving arterial oxygenation are unlikely to be beneficial; a more comprehensive lung-protective approach may be more appropriate.
  • Several methods have been studied to choose an optimal level of PEEP; however, no significant improvement in clinical outcomes have been demonstrated with any specific titration strategy.
  • Arbitrarily chosen higher levels of PEEP may not benefit in heterogeneous groups of ARDS patients; however, a higher PEEP may improve outcomes in patients with severe ARDS.
  • A ballpark PEEP level may be 5–10 cm of H2O in mild to moderate ARDS; a higher level of 15–20 cm of H2O may be more appropriate in severe ARDS.
  • Time-consuming and cumbersome techniques in the pursuit of titration of PEEP levels may not lead to a discernible benefit.



  1. Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory distress in adults. Lancet Lond Engl. 1967 Aug 12;2(7511):319–23.
  2. Gattinoni L, Carlesso E, Cressoni M. Selecting the ‘right’ positive end-expiratory pressure level: Curr Opin Crit Care. 2015 Feb;21(1):50–7.
  3. Pintado M-C, de Pablo R, Trascasa M, Milicua J-M, Rogero S, Daguerre M, et al. Individualized PEEP Setting in Subjects With ARDS: A Randomized Controlled Pilot Study. Respir Care. 2013 Sep 1;58(9):1416–23.
  4. Cressoni M, Chiumello D, Carlesso E, Chiurazzi C, Amini M, Brioni M, et al. Compressive forces and computed tomography-derived positive end-expiratory pressure in acute respiratory distress syndrome. Anesthesiology. 2014 Sep;121(3):572–81.
  5. Beitler JR, Sarge T, Banner-Goodspeed VM, Gong MN, Cook D, Novack V, et al. Effect of Titrating Positive End-Expiratory Pressure (PEEP) With an Esophageal Pressure–Guided Strategy vs an Empirical High PEEP-F io2Strategy on Death and Days Free From Mechanical Ventilation Among Patients With Acute Respiratory Distress Syndrome: A Randomized Clinical Trial. JAMA [Internet]. 2019 Feb 18 [cited 2019 Feb 24]; Available from:
  6. Brower RG, Lanken PN, MacIntyre N, Matthay MA, Morris A, Ancukiewicz M, et al. Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med. 2004 Jul 22;351(4):327–36.
  7. Writing Group for the Alveolar Recruitment for Acute Respiratory Distress Syndrome Trial (ART) Investigators, Cavalcanti AB, Suzumura ÉA, Laranjeira LN, Paisani D de M, Damiani LP, et al. Effect of Lung Recruitment and Titrated Positive End-Expiratory Pressure (PEEP) vs Low PEEP on Mortality in Patients With Acute Respiratory Distress Syndrome: A Randomized Clinical Trial. JAMA. 2017 10;318(14):1335–45.
  8. Mercat A, Richard J-CM, Vielle B, Jaber S, Osman D, Diehl J-L, et al. Positive end-expiratory pressure setting in adults with acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. JAMA. 2008 Feb 13;299(6):646–55.
  9. Briel M, Meade M, Mercat A, Brower RG, Talmor D, Walter SD, et al. Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta-analysis. JAMA. 2010 Mar 3;303(9):865–73.


High flow nasal oxygen therapy: A breath of fresh air?

Supplemental oxygen is conventionally delivered through nasal prongs or various types of masks. Although these devices increase the inspired oxygen concentration, they have significant limitations. The inability to generate adequate flows in patients who are dyspneic is a major drawback. Respiratory failure is characterized by high peak inspiratory flow rates ranging from 30–120 L/min (1). However, most conventional devices are limited to a maximum flow of 15 L/min, resulting in a significant mismatch between the peak inspiratory and delivered flows. The flow mismatch results in variable entrainment of atmospheric air and failure to deliver a constant FiO2.

Bubble humidifiers are commonly used to humidify the inspired gas. However, the humidification imparted by bubble humidifiers is negligible (2). Dry inspired gas, delivered at room temperature, results in reduced water content of the airway mucus. Lack of humidification and low temperature of supplemental oxygen impairs ciliary activity and inhibits mucus clearance (3). To make matters worse, many patients in respiratory failure are unable to cough forcefully, leading to retention of secretions. There has been increasing interest in the use of high flow nasal cannula (HFNC) to deliver optimally warmed and humified oxygen at high flows following an initial report by Dewan et al (4). Several commercially available devices are currently in use that can deliver fully conditioned gas at flows of up to 60 L/min, at a fixed FiO2. (Figure 1). These systems use a built-in flow generator and an air-oxygen blender that generates inspired gas flows at a pre-set FiO2.  A heated passover type of humidifier is utilized to ensure adequate humification. The pre-warmed, humidified gas mixture flows through a heated wire embedded, single limb corrugated tubing. The heated tubing prevents condensation of water and any significant drop in temperature of the inspired gas as it reaches the patient end. A soft, wide-bored nasal cannula is used to deliver the inspired gas.


Figure 1. A typical high-flow device with nasal cannula

How does HFNC deliver a constant FiO2?

The flow rates delivered by HFNC exceed the peak inspiratory flow rate of patients who are dyspneic. This minimizes atmospheric air entrainment and enables a constant, predictable FiO2.  Most devices allow a maximal gas flow of 60L/min, at a fixed FiO2 ranging between 0.21 to nearly 1.0.   

Does HFNC enable CO2 washout?          

High gas flows result in flushing out of CO2 from the nasopharynx with less rebreathing of dead space gas and improved alveolar ventilation (5). A lower partial pressure of carbon dioxide in the inspired gas mixture results in an increase in the FiO2 (6). As the alveolar ventilation improves, a lower respiratory rate and a greater sense of comfort ensue (7).

Does HFNC offer positive end-expiratory pressure?

HFNC generates positive pressure in the airways, directly proportional to the flow rate. In postoperative patients, the airway pressure increased with increasing inspiratory flow. Mean airway pressures with the mouth closed ranged from 1.52 ± 0.7, 2.21 ± 0.8, and 3.1 ± 1.2 cm H2O at flow rates of 40, 50, and 60 L/min respectively in postoperative patients (8). The low level of PEEP generated by HFNC may facilitate alveolar recruitment and an increase in the end-expiratory lung volume.

HFNC and airway resistance

The nasopharynx is a narrow part of the supraglottic airway that offers resistance to airflow. The application of positive pressure may splint and widen the airways during inspiration and reduce airway resistance. Furthermore, inspiratory gas flows that closely match the peak inspiratory flow may contribute to decreased inspiratory resistance. Both these mechanisms may contribute to a decrease in the resistive work of breathing (9).

The beneficial effects of humidification

Warming and humidification of the inspired gas involve an energy expenditure of 156 calories/min for a tidal volume of 500 ml at a respiratory rate of 12/min. Delivery of pre-conditioned gas conserves the energy thus expended and may reduce the metabolic cost of breathing. Furthermore, warming and humidification results in an improved sense of comfort and reduces resistance to breathing. Besides, optimal humidification adds to the aqueous content of the mucosal layer, thus improving ciliary activity and secretion clearance (10).

The aforementioned physiological effects of HFNC are clearly advantageous. However, do they translate to improved clinical outcomes in the real world?

Does HFNC help in patients with acute hypoxemic respiratory failure?  

Early studies with the use of HFNC in patients with acute hypoxemic respiratory failure have shown improvement in physiological endpoints including oxygenation and respiratory rate, besides increasing the level of comfort (5,11). The FLORALI study evaluated HFNC therapy in patients with acute hypoxic respiratory failure (12). In this three-armed randomized controlled study, HFNC was compared with non-invasive ventilation (NIV) and oxygen administered through a non-rebreather mask. There was no significant reduction in the requirement for endotracheal intubation and invasive ventilation with HFNC use. The overall intubation rate was lower than that assumed for power calculation; this may have led to a type II error. In patients who required intubation and mechanical ventilation, HFNC use resulted in more ventilator-free days and improved 90-day survival after adjusting for simplified acute physiology score II (SAPS II) scores. On post hoc analysis, a significantly lower rate of endotracheal intubation was noted in with PaO2/FiOratios of less than 200. This finding suggests that severely hypoxic patients may benefit with HFNC; however, this hypothesis needs validation. The 90-d mortality was lower in the HFNC group; however, being a secondary endpoint, the study was not powered to demonstrate a difference.

In a randomized controlled study (RCT) of postoperative patients who underwent cardiac surgery under cardiopulmonary bypass, HFNC was compared with standard oxygen therapy through a face mask or nasal prongs. There was no significant difference in the SpO2/FiOratios between groups on day 3 after enrolment. However, NHFC resulted in fewer episodes of desaturation. The number of patients requiring escalation of respiratory support was significantly lower in the HFNC group (13).

HFNC was used in patients with severe ARDS in a single center, 1-y observational study. The intubation rate was 40% in this study; failure of HFNC was associated with higher SAPS II scores, the presence of extra-pulmonary organ failures, a lower PaO2/FiOratio, and a higher respiratory rate (14).

Concerns have been raised regarding persistence with HFNC for too long with delayed intubation leading to adverse outcomes. In a propensity-matched, retrospective study, patients who required endotracheal intubation after failure of HFNC were analyzed. Delayed intubation after 48 hours of HFNC use was associated with significantly higher mortality compared to intubation within 48 hours (15).

Is HFNC beneficial as postextubation therapy?

Earlier studies showed improvement in physiological endpoints including heart rate, respiratory rate, and dyspnoea scores with postextubation use of HFNC (16). HFNC was compared with ventimask in a randomized controlled trial of patients with a PaO2/FiOratio of less than 300 prior to extubation (17). There were fewer episodes of desaturation with HFNC use. The need for NIV support and endotracheal intubation was also significantly less in patients who received HFNC. In patients who were at high risk of reintubation following cardiothoracic surgery, HFNC was compared to NIV delivered as bilevel positive airway pressure following extubation. There was no difference in treatment failure (reintubation, switch to the other treatment modality, or premature discontinuation of the study intervention) or intensive care mortality(18). HFNC was compared with standard oxygen therapy using nasal prongs or facemask after extubation in patients who underwent major abdominal surgery (19). The incidence of hypoxemia 1 h after extubation and after discontinuation of treatment was not different between groups. Pulmonary complications at 1 week after surgery were also similar. In a large, multicentre randomized controlled study, HFNC was compared with standard oxygen therapy following extubation among patients who were at low risk for reintubation (20). Reintubation within 72 h of extubation was significantly lower with HFNC (4.9% vs 12.2%, p = 0.004). Postextubation respiratory failure was also less common with HFNC.

Other uses

HFNC may be the preferred mode of support in terminally ill patients in whom endotracheal intubation and invasive ventilation may be inappropriate. In addition to an increased sense of comfort, HFNC perseveres the ability to communicate and allows oral intake. HFNC may be the ideal modality of continuous oxygen administration during attempts at intubation with better maintenance of oxygen saturation. It may also be useful to prevent desaturation in awake, spontaneously breathing patients who undergo bronchoscopy. 

The bottom line

  • HFNC delivers warmed and humified gases. As the delivered flow is higher than peak inspiratory flow rates, a constant FiOis achieved
  • The high flow rate enables maintenance of positive airway pressure in the nasopharynx with augmentation of the end-expiratory lung volume
  • The positive airway pressure distends the airways and reduces airway resistance
  • Improved physiological parameters have been consistently demonstrated with HFNC
  • HFNC may reduce the requirement for intubation and invasive mechanical ventilation in patients with low PaO2/FiOratios
  • Improvement in oxygenation, reduced requirement for NIV use and reintubation are possible advantages with postextubation HFNC use
  • As with any other form of respiratory support, persistence with HFNC in non-responders and delayed intubation may lead to adverse outcomes



  1. L’Her E, Deye N, Lellouche F, Taille S, Demoule A, Fraticelli A, et al. Physiologic effects of noninvasive ventilation during acute lung injury. Am J Respir Crit Care Med. 2005 Nov 1;172(9):1112–8.
  2. Chanques G, Constantin J-M, Sauter M, Jung B, Sebbane M, Verzilli D, et al. Discomfort associated with underhumidified high-flow oxygen therapy in critically ill patients. Intensive Care Med. 2009 Jun;35(6):996–1003.
  3. Kilgour E, Rankin N, Ryan S, Pack R. Mucociliary function deteriorates in the clinical range of inspired air temperature and humidity. Intensive Care Med. 2004 Jul;30(7):1491–4.
  4. Dewan NA, Bell CW. Effect of low flow and high flow oxygen delivery on exercise tolerance and sensation of dyspnea. A study comparing the transtracheal catheter and nasal prongs. Chest. 1994 Apr;105(4):1061–5.
  5. Sztrymf B, Messika J, Bertrand F, Hurel D, Leon R, Dreyfuss D, et al. Beneficial effects of humidified high flow nasal oxygen in critical care patients: a prospective pilot study. Intensive Care Med. 2011 Nov;37(11):1780–6.
  6. Spence CJT, Buchmann NA, Jermy MC. Unsteady flow in the nasal cavity with high flow therapy measured by stereoscopic PIV. Exp Fluids. 2012;52(3):569–579.
  7. Schmidt M, Banzett RB, Raux M, Morélot-Panzini C, Dangers L, Similowski T, et al. Unrecognized suffering in the ICU: addressing dyspnea in mechanically ventilated patients. Intensive Care Med. 2014 Jan;40(1):1–10.
  8. Ritchie JE, Williams AB, Gerard C, Hockey H. Evaluation of a humidified nasal high-flow oxygen system, using oxygraphy, capnography and measurement of upper airway pressures. Anaesth Intensive Care. 2011 Nov;39(6):1103–10.
  9. Ricard J-D. High flow nasal oxygen in acute respiratory failure. Minerva Anestesiol. 2012 Jul;78(7):836–41.
  10. Sim M a. B, Dean P, Kinsella J, Black R, Carter R, Hughes M. Performance of oxygen delivery devices when the breathing pattern of respiratory failure is simulated. Anaesthesia. 2008 Sep;63(9):938–40.
  11. Roca O, Riera J, Torres F, Masclans JR. High-flow oxygen therapy in acute respiratory failure. Respir Care. 2010 Apr;55(4):408–13.
  12. Frat J-P, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, et al. High-Flow Oxygen through Nasal Cannula in Acute Hypoxemic Respiratory Failure. N Engl J Med. 2015 Jun 4;372(23):2185–96.
  13. Parke R, McGuinness S, Dixon R, Jull A. Open-label, phase II study of routine high-flow nasal oxygen therapy in cardiac surgical patients. Br J Anaesth. 2013 Aug 6;aet262.
  14. Messika J, Ben Ahmed K, Gaudry S, Miguel-Montanes R, Rafat C, Sztrymf B, et al. Use of High-Flow Nasal Cannula Oxygen Therapy in Subjects With ARDS: A 1-Year Observational Study. Respir Care. 2015 Feb;60(2):162–9.
  15. Kang BJ, Koh Y, Lim C-M, Huh JW, Baek S, Han M, et al. Failure of high-flow nasal cannula therapy may delay intubation and increase mortality. Intensive Care Med. 2015 Apr;41(4):623–32.
  16. Rittayamai N, Tscheikuna J, Rujiwit P. High-flow nasal cannula versus conventional oxygen therapy after endotracheal extubation: a randomized crossover physiologic study. Respir Care. 2014 Apr;59(4):485–90.
  17. Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, et al. Nasal High-Flow versus Venturi Mask Oxygen Therapy after Extubation. Effects on Oxygenation, Comfort, and Clinical Outcome. Am J Respir Crit Care Med. 2014 Aug;190(3):282–8.
  18. Stéphan F, Barrucand B, Petit P, Rézaiguia-Delclaux S, Médard A, Delannoy B, et al. High-Flow Nasal Oxygen vs Noninvasive Positive Airway Pressure in Hypoxemic Patients After Cardiothoracic Surgery: A Randomized Clinical Trial. JAMA. 2015 Jun 16;313(23):2331–9.
  19. Futier E, Paugam-Burtz C, Godet T, Khoy-Ear L, Rozencwajg S, Delay J-M, et al. Effect of early postextubation high-flow nasal cannula vs conventional oxygen therapy on hypoxaemia in patients after major abdominal surgery: a French multicentre randomised controlled trial (OPERA). Intensive Care Med. 2016 Dec;42(12):1888–98.
  20. Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, et al. Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapy on Reintubation in Low-Risk Patients: A Randomized Clinical Trial. JAMA. 2016 Apr 5;315(13):1354.



Early vasopressors or a fluid-liberal resuscitation strategy​ in sepsis-related hypotension?

Fluid resuscitation is the cornerstone of the established management of sepsis-related hypotension. Guidelines recommend an initial bolus of 30 ml/kg of crystalloids with administration of repeated boluses if the hemodynamic parameters continue to improve (1). The physiological rationale behind this approach is the extensive vasodilatation and capillary leak that characterize sepsis. Adequate intravascular volume expansion is considered to be the appropriate initial management, followed by the administration of vasopressors if severe hypotension persists. Rivers et al., in their landmark study, administered intravenous fluid boluses targeting central venous pressure; this resulted in a much larger volume of fluid being administered to the intervention group compared to controls during the initial hours of resuscitation (2). Subsequently, three randomized controlled studies further bolstered the “fluids first” approach; both control and intervention arms of these studies received similar volumes of fluid in the first few hours, suggesting that this approach has firmly entrenched into clinical practice over the years (3–5). However, a “fluid liberal” initial approach may have adverse consequences, with overfilling of the intravascular and interstitial compartments due to capillary leakage. Excessive fluid in the interstitial compartment may lead to tissue edema and precipitate organ failure (6). Furthermore, a central venous pressure targeted resuscitation strategy may result in increased venous pressures, thereby decreasing perfusion pressures in vital organs such as the kidneys (7). Is it time to re-evaluate the traditional fluid liberal resuscitation strategy considering the putative harm that may emanate from such an approach? Would it be appropriate to limit fluid resuscitation and administer vasopressors early in patients with sepsis? Such an approach may accelerate the process of resuscitation and result in a lower volume of fluids being administered.

In a study of African children with severe febrile illness and signs of impaired perfusion, an initial fluid bolus of normal saline or 5% albumin was compared to no bolus fluid. Intravenous maintenance fluid was administered to all children. Mortality at 48 h and 4 weeks was significantly lower in children who received no bolus fluid (8). Although this study was performed among children in resource-limited settings and may not be directly extrapolatable to adult practice, it underlines potential harm from excessive fluid resuscitation in the initial stage of sepsis.

Bai et al. conducted a retrospective cohort study on the timing of noradrenaline administration in patients with septic shock (9). They hypothesized that later administration of noradrenaline may lead to more prolonged hypotension resulting in impaired organ perfusion compared to early administration. Mortality at 28 days was significantly lower in patients who received noradrenaline within 2 h of the onset of septic shock compared with those who received it later. The mortality increased by 5.3% for every 1 h of delay in noradrenaline administration within the first 6 h of onset of septic shock. On multivariate logistic regression analysis, the time to noradrenaline administration was one of the independent predictors of mortality.

A multicentre randomized controlled study was conducted to assess the feasibility of a fluid-restrictive strategy after initial resuscitation. In the fluid-restricted group, after initial resuscitation, additional boluses were administered only if signs of severe hypoperfusion were noted. In the control group, continued boluses were administered if there was an improvement in hemodynamic parameters assessed using static or dynamic indices. Resuscitation volumes on day 5 and during ICU stay were the co-primary outcomes. The resuscitation volumes at both time points were significantly less with a restrictive strategy. The study was not powered to evaluate patient-centered outcomes; however, it suggested that a restricted fluid strategy was feasible among patients with septic shock.

In a recent randomized controlled study, noradrenaline was administrated in a dose of 0.05 mcg/kg/min within 1 hr of the diagnosis of septic shock (mean arterial pressure of less than 65 mm Hg and diagnosis of sepsis based on the Surviving Sepsis Guidelines) (10). The infusion was continued for a period of 24 h; the control group received a placebo infusion. Fluid resuscitation, antibiotic therapy, and organ support were based on physician discretion. If the mean arterial pressure remained below 65 mm Hg, open-label vasopressors were administered. The primary outcome was “shock control” at 6 h after the diagnosis of septic shock. Shock control was determined by a sustained mean arterial pressure of more than 65 mm Hg and evidence of adequate tissue perfusion, defined as a urine output of more than 0.5 ml/kg/h for 2 consecutive hours or a decrease in the serum lactate by more than 10% of the baseline levels. In this study, shock control after 6 h of initial resuscitation was significantly higher with early noradrenaline compared to placebo. The median time to noradrenaline administration was significantly less in the early noradrenaline group (93 vs. 192min; P<0.001). There was no difference in the 28-day and hospital mortality between groups. This study demonstrated that the early use of noradrenaline leads to more rapid attainment of resuscitation goals.

The Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis (CLOVERS) study is a  multi-center randomized controlled trial that is currently recruiting patients (11). It aims to compare the impact of a restrictive fluid strategy with a liberal fluid strategy on the 90 d mortality. In the restricted fluids group, patients are randomized to receive noradrenaline after an initial fluid bolus of 1L; further boluses are administered as a rescue measure. In the fluid liberal group, repeated boluses of fluid are administered as the initial intervention; rescue noradrenaline is commenced only after either 5 L of fluid has been administered, signs of acute volume overload develop, or other predefined rescue criteria are met. The results of this study are eagerly awaited.


  • The conventional, guideline-based approach of the “fluids first” strategy may lead to excessive capillary leakage, interstitial edema, and impaired organ perfusion.
  • Early use of vasopressors may enable more rapid attainment of resuscitation goals and improved organ perfusion.
  • The optimal volume of fluid resuscitation prior to the commencement of vasopressors may be variable.
  • Clinical studies suggest that a fluid restricted resuscitation strategy with early use of vasopressors may improve outcomes.
  • Further prospective controlled studies are warranted to assess the impact of early vasopressor therapy in sepsis-related hypotension.



  1. Rhodes A, Evans LE, Alhazzani W, Levy MM, Antonelli M, Ferrer R, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Intensive Care Med. 2017 Mar;43(3):304–77.
  2. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001 Nov 8;345(19):1368–77.
  3. ProCESS Investigators, Yealy DM, Kellum JA, Huang DT, Barnato AE, Weissfeld LA, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med. 2014 May 1;370(18):1683–93.
  4. ARISE Investigators, ANZICS Clinical Trials Group, Peake SL, Delaney A, Bailey M, Bellomo R, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med. 2014 Oct 16;371(16):1496–506.
  5. Mouncey PR, Osborn TM, Power GS, Harrison DA, Sadique MZ, Grieve RD, et al. Trial of early, goal-directed resuscitation for septic shock. N Engl J Med. 2015 Apr 2;372(14):1301–11.
  6. Marik PE. Iatrogenic salt water drowning and the hazards of a high central venous pressure. Ann Intensive Care. 2014;4:21.
  7. Legrand M, Dupuis C, Simon C, Gayat E, Mateo J, Lukaszewicz A-C, et al. Association between systemic hemodynamics and septic acute kidney injury in critically ill patients: a retrospective observational study. Crit Care Lond Engl. 2013 Nov 29;17(6):R278.
  8. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. 2011 Jun 30;364(26):2483–95.
  9. Bai X, Yu W, Ji W, Lin Z, Tan S, Duan K, et al. Early versus delayed administration of norepinephrine in patients with septic shock. Crit Care Lond Engl. 2014 Oct 3;18(5):532.
  10. Permpikul C, Tongyoo S, Viarasilpa T, Trainarongsakul T, Chakorn T, Udompanturak S. Early Use of Norepinephrine in Septic Shock Resuscitation (CENSER) : A Randomized Trial. Am J Respir Crit Care Med [Internet]. 2019 Feb [cited 2019 Feb 7]; Available from:
  11. Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis – Full Text View – [Internet]. [cited 2019 Feb 8]. Available from:



When the most important muscle in the body fails…


The diaphragm is the principal muscle of respiration and is innervated by the phrenic nerves through the C3–C5 nerve roots. There is a high prevalence of diaphragmatic weakness among critically ill patients. No correlation seems to exist between weakness of the limbs and diaphragmatic weakness; in fact, diaphragmatic dysfunction may be twice as common as paresis of the limbs due to polyneuropathy or myopathy related to critical illness (1). Diaphragmatic dysfunction may be a stronger predictor of mortality than the severity of multiorgan failure among critically ill patients (2). Furthermore, weakness of the diaphragm is much harder to recognize compared to limb muscle weakness. Hence, it is important to have a high index of suspicion in patients who may be at risk of developing diaphragmatic dysfunction.

Clinical indicators

The possibility of diaphragmatic dysfunction must be considered in patients who seem to have recovered sufficiently from the underlying illness but fail repeated attempts to wean. Failure of adequate spontaneous ventilation in spite of reasonable lung mechanics and good gas exchange heightens the suspicion of diaphragmatic dysfunction. Paradoxical breathing, when the abdomen seems to get sucked in during inspiration and moves outwards during expiration is an important clinical indicator. The chest radiograph may show elevation of the diaphragm on one or both sides, depending on the extent of involvement.

The incidence of diaphragmatic dysfunction

The most sensitive test of diaphragmatic dysfunction is through magnetic stimulation of the phrenic nerves in the neck and measurement of the twitch pressure generated at the proximal end of the endotracheal tube. Using trans-diaphragmatic twitch pressure measurement, 34/43 (79%) patients developed diaphragmatic dysfunction during their stay in the ICU (3). Obviously, mild dysfunction may often be missed by clinical examination alone.

Ventilation (VIDD) and sepsis-induced diaphragmatic dysfunction (SIDD)

Both mechanical ventilation and sepsis can independently, and in combination, lead to signification diaphragmatic dysfunction. The degree of dysfunction may be closely related to the duration of mechanical ventilation. Histopathological examination of the diaphragm was carried out in brain dead organ donors who were on mechanical ventilation for 18–79 h and compared with patients who underwent 2–3 h of ventilation during thoracotomy for lung cancer. Inactivity of the diaphragm combined with mechanical ventilation resulted in marked atrophy of myofibers compared to controls (4). Activation of oxidative stress appears to be the trigger for VIDD. The generation of reactive oxygen species appears to initiate oxidative stress, resulting in DNA fragmentation and proteolysis leading to impairment of diaphragmatic contractility.

Ultrasonography- a simple bedside tool to identify diaphragmatic dysfunction

The identification of diaphragmatic dysfunction by the bedside may be reliably carried out by ultrasonography. Two parameters are used to assess diaphragmatic function: (1) diaphragmatic thickening during a maximal inspiratory effort (thickening fraction) and (2) amplitude of diaphragmatic excursion during tidal breathing.

Diaphragmatic excursion

A low frequency (2–5 MHz) curvilinear probe is placed in the subcostal region between the midclavicular and anterior axillary lines using the liver or the spleen as the acoustic window. Once the diaphragm is visualized, the M-mode cursor is positioned perpendicular to the posterior aspect of the diaphragm and the excursion is measured (Fig 1). A distance of excursion of less than 1 cm during quiet breathing or paradoxical movement (cranial movement on inspiration) is indicative of diaphragmatic weakness (5).

fig 1

Fig 1. Measurement of diaphragmatic excursion on M-mode view using a 5 MHz curvilinear transducer. The M-mode cursor is positioned perpendicular to the posterior aspect of the diaphragm. The amplitude of excursion measured in this image is 3.15 cm (double-headed arrow)

Thickening fraction

To assess thickening during inspiration, a high frequency (8–12 mHz) linear probe is used. The probe is placed between the 7–9thintercostal spaces in the midaxillary line. The diaphragm is seen as a non-echogenic zone sandwiched between the diaphragmatic pleura and the peritoneum. Thickness is measured at end-inspiration and end-expiration (Fig 2). The thickening fraction is calculated using the formula: Thickness at end-inspiration (maximal thickness) – thickness at end-expiration/thickness at end-inspiration. A thickening fraction of less than 20% is suggestive of diaphragmatic dysfunction (6).

Fig 2

Fig 2. Ultrasonographic imaging of the diaphragm using a 12 MHz linear probe. The diaphragm is seen as a hypoechoic zone between the pleura and the peritoneum, seen as hyperechoic lines. The image on the left shows thickness at end-expiration (green line); the image on the right shows thickness at end-inspiration (orange line). Thickening fraction = End-expiratory thickness – End inspiratory thickness/End-expiratory thickness

Prevention and treatment of diaphragmatic dysfunction

Maintenance of spontaneous respiratory effort during mechanical ventilation is the key to prevention of diaphragmatic dysfunction. Clearly, this strategy may be counterproductive and harmful in patients with a pronounced respiratory drive, particularly in severe acute respiratory distress syndrome. Modest use of sedatives and muscle relaxants for a limited duration is appropriate to during the initial period of mechanical ventilation. However, regular interruption of muscle relaxants and sedation should be carried out to assess the respiratory drive with a view to allowing spontaneous respiratory efforts as early as possible. Electrolyte abnormalities, including hypocalcemia, hypomagnesemia, and hypophosphatemia are well known to provoke or worsen diaphragmatic weakness; levels must be estimated regularly and maintained within the normal range. Endocrine abnormalities, especially myxedema is known to cause muscle weakness and must be considered in patients with diaphragmatic dysfunction. There is emerging interest in improving diaphragmatic function by phosphodiesterase inhibition using theophylline and with the use of levosimendan, a calcium sensitizer. Phrenic nerve stimulation and antagonization of oxidative stress using antioxidants including vitamin C and E may be therapeutic modalities for the future.


  • Diaphragmatic dysfunction is common among critically ill patients and may often be unrecognized.
  • Diaphragmatic dysfunction leads to adverse clinical outcomes including mortality.
  • Mechanical ventilation and sepsis are independent triggers of diaphragmatic dysfunction and may potentiate each other in combination.
  • It is important to have a high index of suspicion in patients at risk of diaphragmatic dysfunction.
  •  Ultrasonography is a simple and sensitive tool to diagnose diaphragmatic dysfunction by the bedside in critically ill patients.
  • Maintenance of spontaneous respiratory efforts in mechanically ventilated patients with limited use of sedatives and muscle relaxants may prevent diaphragmatic dysfunction.
  • Electrolyte and endocrine abnormalities must be sought for and corrected.



  1. Dres M, Dubé B-P, Mayaux J, Delemazure J, Reuter D, Brochard L, et al. Coexistence and Impact of Limb Muscle and Diaphragm Weakness at Time of Liberation from Mechanical Ventilation in Medical Intensive Care Unit Patients. Am J Respir Crit Care Med. 2017 01;195(1):57–66.
  2. Supinski GS, Westgate P, Callahan LA. Correlation of maximal inspiratory pressure to transdiaphragmatic twitch pressure in intensive care unit patients. Crit Care. 2016 Mar 23;20(1):77.
  3. Demoule A, Molinari N, Jung B, Prodanovic H, Chanques G, Matecki S, et al. Patterns of diaphragm function in critically ill patients receiving prolonged mechanical ventilation: a prospective longitudinal study. Ann Intensive Care. 2016;6:75. Available from:
  4. Levine S, Budak MT, Sonnad S, Shrager JB. Rapid Disuse Atrophy of Diaphragm Fibers in Mechanically Ventilated Humans. N Engl J Med. 2008;9.
  5. Kim WY, Suh HJ, Hong S-B, Koh Y, Lim C-M. Diaphragm dysfunction assessed by ultrasonography: influence on weaning from mechanical ventilation. Crit Care Med. 2011 Dec;39(12):2627–30.
  6. Gottesman E, McCool FD. Ultrasound evaluation of the paralyzed diaphragm. Am J Respir Crit Care Med. 1997 May;155(5):1570–4.


Does pantoprazolization prevent gastrointestinal bleeding in critically ill patients?


The efficacy of acid suppression in the prevention of stress ulcers using antacids was first investigated more than four decades ago among burns patients (1). Many different classes of acid suppressant medication have evolved since then, but their utility in the prevention of stress ulcers in critically ill patients remains unresolved. It has been hypothesized that stress ulcers may be more related to impaired mucosal blood flow and ischemia reperfusion-related injury; hence, in contrast to peptic ulceration, the evolution and propagation of such ulcers may be less related to gastric acidity.

The reported incidence of clinically important gastrointestinal bleeding is variable and probably related to the heterogeneity among patient populations studied, non-uniform definitions, and the lack of distinction between true stress ulceration and other unrelated causes of bleeding. Although not approved by the Food and Drug Administration for this indication, proton pump inhibitors (PPIs) are among the most common medications administered to critically ill (2) and perhaps to all hospitalized patients. In a study from two Australian ICUs, acid suppressant medication was inappropriately continued in 63% of patients after ICU and in 39% of patients after hospital discharge (3).

Most of the current recommendations regarding stress ulcer prophylaxis are based on results from studies and risk factors identified several decades ago. Changes in clinical practice, particularly with the increased emphasis on expeditious resuscitation and early commencement of enteral nutrition may have reduced the risk of stress ulceration and clinically important bleeding in the ICU. Do we need to rethink and modify our approach with the use of acid suppressant medication in the light of more contemporaneous evidence?

Harmful effects of acid suppression

A cohort study of patients on mechanical ventilation for more than 24 hours compared the use of PPI with H2 receptor antagonists (H2RA). After propensity-matched multivariate analysis, PPIs were associated with a significantly higher incidence of pneumonia and C. difficile infection; surprisingly, the incidence of gastrointestinal hemorrhage was also higher with PPI use (4). A recent meta-analysis investigated the risk of C. difficile infection associated with the use of PPIs. Among the 23 observational studies included, 10,307 cases of C. difficile infection were observed, with a significantly higher incidence with the use of PPIs (5). There is concern regarding adverse cardiovascular events associated with the use of PPIs. In a cohort study of 56,406 patients from Denmark, PPI use was associated with a significantly increased risk of cardiovascular death and readmission to hospital with myocardial infarction or stroke (hazard ratio: 1.29; CI, 1.21 to 1.37) (6). Among 244,679 subjects who underwent elective gastroscopy, there was a dose-related increase in the risk of first-time ischemic stroke with PPI use. No such association was observed with the use of H2RAs (7). Concern has also been raised regarding thrombocytopenia in patients with upper gastrointestinal bleeding who are treated with continuous infusions of PPIs (8).

Early enteral nutrition as prophylaxis

Continuous enteral feeds may be more effective at maintaining a higher gastric pH (above 3.5) compared to PPIs and H2RAs. In a randomized controlled trial (RCT), intravenous pantoprazole was compared to placebo in 214 mixed medical-surgical patients on mechanical ventilation who were expected to be commenced on enteral nutrition within 48 hours of ICU admission. No clinically significant gastrointestinal bleeding was observed in either group of patients (9). In a similar RCT of critically ill patients on mechanical ventilation who were expected to receive enteral nutrition within the first 24 hours of ICU admission, there was no overt or significant gastrointestinal bleeding with intravenous pantoprazole administration compared to placebo (10).

The SUP-ICU trial investigated the effects of intravenous pantoprazole in critically ill adult patients at high risk of gastrointestinal bleeding (11). The study was conducted across 33 ICUs in six countries between January 2016 to October 2017. In this randomized controlled trial, 3298 patients were enrolled to receive intravenous pantoprazole or placebo. The primary outcome, which the study was powered for, was the 90-d mortality. There was no significant mortality difference between pantoprazole (31.1%) and placebo (30.4%). The secondary outcome, a composite endpoint of gastrointestinal bleeding, pneumonia, C. difficile infection, and acute myocardial ischemia, was not different between groups. Clinically important bleeding was lower with pantoprazole compared to placebo (2.5 vs. 4%); the significance of this finding was not evaluated, as no adjustments were made for multiple comparisons. The power calculation of this study may have been flawed because it was based on a previous report that revealed no increase of mortality secondary to gastrointestinal bleeding. The composite secondary endpoint has also not been evaluated previously. The source of bleeding was not confirmed endoscopically, to differentiate from causes other than stress ulceration. Besides, there was no information on whether enteral nutrition had been established at baseline. This study confirms that the overall incidence of gastrointestinal bleeding is very low in today’s world (3.3% overall); adequately powered studies to prove a difference in the incidence may be difficult to carry out in practice.

The bottom line

  • With an increased focus on early resuscitation, the incidence of clinically important gastrointestinal bleeding has probably decreased compared to previous reports.
  • Widespread use of acid-suppressant medication has been linked to several adverse effects, including pneumonia, C. difficile infection, and cardiovascular events in critically ill patients.
  • Routine use of stress ulcer prophylaxis may be unnecessary in patients who can be initiated on early enteral nutrition within the first 24–48 hours.
  • Risk factors identified in earlier studies, including mechanical ventilation for more than 48 hours, may not necessitate routine prophylaxis.
  • Future studies should be directed towards identification of patients who may be truly “at risk” for stress ulcer-related bleeding.



  1. McAlhany JC, Colmic L, Czaja AJ, Pruitt BA. Antacid control of complications from acute gastroduodenal disease after burns. J Trauma. 1976 Aug;16(08):645–8.
  2. Barletta JF, Lat I, Micek ST, Cohen H, Olsen KM, Haas CE, et al. Off-label use of gastrointestinal medications in the intensive care unit. J Intensive Care Med. 2015 May;30(4):217–25.
  3. Farley KJ, Barned KL, Crozier TM. Inappropriate continuation of stress ulcer prophylaxis beyond the intensive care setting. Crit Care Resusc J Australas Acad Crit Care Med. 2013 Jun;15(2):147–51.
  4. MacLaren R, Reynolds PM, Allen RR. Histamine-2 receptor antagonists vs proton pump inhibitors on gastrointestinal tract hemorrhage and infectious complications in the intensive care unit. JAMA Intern Med. 2014 Apr;174(4):564–74.
  5. Arriola V, Tischendorf J, Musuuza J, Barker A, Rozelle JW, Safdar N. Assessing the Risk of Hospital-Acquired Clostridium Difficile Infection With Proton Pump Inhibitor Use: A Meta-Analysis. Infect Control Hosp Epidemiol. 2016;37(12):1408–17.
  6. Charlot M, Ahlehoff O, Norgaard ML, Jørgensen CH, Sørensen R, Abildstrøm SZ, et al. Proton-pump inhibitors are associated with increased cardiovascular risk independent of clopidogrel use: a nationwide cohort study. Ann Intern Med. 2010 Sep 21;153(6):378–86.
  7. Sehested Thomas S, Fosbøl Emil L, Hansen Peter W, Charlot Mette G, Torp-Pedersen Christian, Gislason Gunnar H. Abstract 18462: Proton Pump Inhibitor Use Increases the Associated Risk of First-Time Ischemic Stroke. A Nationwide Cohort Study. Circulation. 2016 Nov 11;134(suppl_1):A18462–A18462.
  8. Binnetoğlu E, Akbal E, Şen H, Güneş F, Erbağ G, Aşık M, et al. Pantoprazole-induced thrombocytopenia in patients with upper gastrointestinal bleeding. Platelets. 2015;26(1):10–2.
  9. Selvanderan SP, Summers MJ, Finnis ME, Plummer MP, Abdelhamid YA, Anderson MB, et al. Pantoprazole or Placebo for Stress Ulcer Prophylaxis (pop-up): Randomized Double-blind Exploratory Study*. Crit Care Med. 2016 Oct 1;44(10):1842–50.
  10. El-Kersh K, Jalil B, McClave SA, Cavallazzi R, Guardiola J, Guilkey K, et al. Enteral nutrition as stress ulcer prophylaxis in critically ill patients: A randomized controlled exploratory study. J Crit Care. 2018 Feb;43:108–13.
  11. Krag M, Marker S, Perner A, Wetterslev J, Wise MP, Schefold JC, et al. Pantoprazole in Patients at Risk for Gastrointestinal Bleeding in the ICU. N Engl J Med. 2018 Dec 6;379(23):2199–208.




Does prone ventilation make them less prone to​ ECMO in severe ARDS?

Extracorporeal membrane oxygenation (ECMO) is widely used in patients with severe acute respiratory distress syndrome (ARDS) and refractory hypoxemia. With the technological refinement of pumps and circuitry, along with increasing clinical expertise, many centers across the world seem to have adopted ECMO as an early treatment strategy. However, is this necessarily the most optimal management of patients with severe ARDS?

Li et al. analyzed 17 studies that utilized veno-venous ECMO between 1995–2017.(1) A total of 672 patients were included in these studies. The vast majority of patients (69%) did not undergo a trial of prone ventilation prior to the initiation of ECMO. In 2013, the PROSEVA study was published, which revealed improved survival with prone ventilation in severe ARDS.(2) However, bewilderingly, the utilization of prone ventilation before initiation of ECMO was even lower in studies published after PROSEVA.

The EOLIA study compared early ECMO with conventional care and rescue ECMO.(3) At the time of randomization, only 56% of patients had received prone ventilation in the “early” ECMO group. Would it be possible that the results may have been different had prone ventilation been resorted to more often among these patients? The study was stopped early for futility; besides, it aimed to demonstrate a fairly unrealistic 20% absolute mortality reduction. Crossover to rescue ECMO was largely based on clinician judgment, which may have led to bias. Although the results were eagerly awaited, the EOLIA study probably raised more questions than it could answer; importantly, would a more liberal proning strategy be advisable prior to initiation of ECMO?

An observational study from 11 intensive care units in Korea compared patients who received prone ventilation prior to ECMO with those who did not.(4) The 30-day mortality was similar among both groups of patients. Successful weaning from ECMO and from mechanical ventilation was more frequent in patients who were prone ventilated prior to ECMO. This study suggests that a trial of prone ventilation may be appropriate prior to consideration of ECMO; such a strategy may even have a protective effect.

Why are some clinicians reluctant to prone ventilate their patients with ARDS? Unfamiliarity and unjustified concerns regarding device dislodgement may be one of the factors. However, no study of prone ventilation has reported harm that could be directly attributed to this technique. Insertion of cannulae and overall management of ECMO would appear to be far more complex than adopting the prone position. Perhaps there is a general tendency to believe that a more complex treatment modality may be more efficacious. Perhaps the ready availability of equipment, skills, and possible financial incentive from an expensive treatment strategy are inevitable inducements.

After several randomized controlled trials (2,5–7) and two meta-analyses,(8,9) there is fairly robust evidence that prone ventilation saves lives in ARDS. It is needless to emphasize that ECMO will play an increasing role in the future in the management of the sickest of patients with ARDS. However, perhaps we should consider prone ventilation prior to contemplating ECMO therapy in every patient with severe ARDS.




  1. Li X, Scales DC, Kavanagh BP. Unproven and Expensive before Proven and Cheap: Extracorporeal Membrane Oxygenation versus Prone Position in Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med. 2018 Apr 15;197(8):991–3.
  2. Guérin C, Reignier J, Richard J-C, Beuret P, Gacouin A, Boulain T, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013 Jun 6;368(23):2159–68.
  3. Combes A, Hajage D, Capellier G, Demoule A, Lavoué S, Guervilly C, et al. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome. N Engl J Med. 2018 May 24;378(21):1965–75.
  4. Kim W-Y, Kang BJ, Chung CR, Park SH, Oh JY, Park SY, et al. Prone positioning before extracorporeal membrane oxygenation for severe acute respiratory distress syndrome: A retrospective multicenter study. Med Intensiva [Internet]. 2018 Jul [cited 2019 Jan 21]; Available from:
  5. Guerin C, Gaillard S, Lemasson S, Ayzac L, Girard R, Beuret P, et al. Effects of systematic prone positioning in hypoxemic acute respiratory failure: a randomized controlled trial. JAMA. 2004 Nov 17;292(19):2379–87.
  6. Mancebo J, Fernández R, Blanch L, Rialp G, Gordo F, Ferrer M, et al. A multicenter trial of prolonged prone ventilation in severe acute respiratory distress syndrome. Am J Respir Crit Care Med. 2006 Jun 1;173(11):1233–9.
  7. Taccone P, Pesenti A, Latini R, Polli F, Vagginelli F, Mietto C, et al. Prone positioning in patients with moderate and severe acute respiratory distress syndrome: a randomized controlled trial. JAMA. 2009 Nov 11;302(18):1977–84.
  8. Sud S, Friedrich JO, Taccone P, Polli F, Adhikari NKJ, Latini R, et al. Prone ventilation reduces mortality in patients with acute respiratory failure and severe hypoxemia: systematic review and meta-analysis. Intensive Care Med. 2010 Apr;36(4):585–99.
  9. Gattinoni L, Carlesso E, Taccone P, Polli F, Guérin C, Mancebo J. Prone positioning improves survival in severe ARDS: a pathophysiologic review and individual patient meta-analysis. Minerva Anestesiol. 2010 Jun;76(6):448–54.




Adrenaline in cardiopulmonary resuscitation: time to rethink?


The potential efficacy of adrenaline in cardiac arrest was first highlighted by Criley and Dolley in 1901.(1) In a study of anesthetic agent or asphyxiation induced-cardiac arrest in dogs, the infusion of a therapeutic dose of adrenaline resulted in improved aortic blood pressures and enabled resuscitation. Subsequent animal studies by Redding et al. further emphasized the administration of adrenaline as a critical intervention for resuscitation from cardiac arrest.(2) The guidelines of the American Heart Association recommend intravenous administration of 1 mg adrenaline for every 3–5 minutes of cardiopulmonary resuscitation (CPR).(3)

Currently, there is strong evidence that early defibrillation and rapid initiation of effective chest compressions are important predictors of favorable clinical outcomes after cardiac arrest. However, there is considerable debate on whether the administration of boluses of adrenaline during CPR improves outcomes. In cardiac arrest, adrenaline is considered to cause alpha adrenergic receptor-mediated vasoconstriction with an increase the aortic diastolic pressure. An increase in the aortic diastolic pressure is believed to improve coronary blood flow during chest compressions and improve the overall likelihood of successful resuscitation.

However, adrenaline may exert deleterious effects during CPR through its beta-adrenergic receptor-mediated effects. Beta stimulation leads to increased myocardial oxygen demand, tachycardia, and tachyarrhythmias. The increase in oxygen demand may be particularly harmful in cardiac arrest provoked by ischemic heart disease. Besides, adrenaline is known to cause thrombogenesis and platelet activation, adversely affecting blood flow to vital organs, including the brain, and may lead to poor neurological outcomes.

Three randomized controlled studies (RCTs) have been carried out to evaluate the efficacy of adrenaline during CPR. Patients who suffered out of hospital cardiac arrest and received advanced cardiac life support by the Oslo Emergency Medical System were randomized to receive intravenous medication or no intravenous access.(4) Hospital admission with ROSC was significantly higher with intravenous medication; however, survival to hospital discharge, the primary outcome, was no different between groups. Furthermore, there was no difference in survival to a favorable neurological outcome nor the 1-y survival. Adrenaline was used in 79% of patients randomized to receive intravenous medication in this study. Patients who received adrenaline had a higher rate of ROSC; however, long term survival and functional recovery were significantly lower.(5)Another placebo-controlled RCT among out of hospital cardiac arrests attended by paramedics in Western Australia revealed very similar findings.(6) Pre-hospital ROSC was significantly higher with adrenaline administration; however, survival to hospital discharge was no different with adrenaline compared to placebo. A large, propensity-matched observational study was conducted in Japan to assess the effect of adrenaline in out of hospital cardiac arrest. The findings of this study were similar to the RCTs mentioned above. There was a significantly higher incidence of ROSC, with adrenaline use; however, survival to 1 month and neurological recovery were significantly lower.(7) These studies clearly demonstrate that though adrenaline may facilitate ROSC, survival to hospital discharge and neurological outcomes are unchanged or even worse.

The most recent and the largest randomized controlled trial that tested the efficacy of adrenaline in cardiac arrest was conducted by the ambulance services of the National Health Services in the UK.(8) As in previous studies, survival to hospital discharge was significantly higher with adrenaline compared to placebo (23.8 vs. 8%). Survival at 30 days, the primary outcome which the study was powered for, was poor in both arms, but significantly higher with adrenaline (3.2 vs. 2.4%). The number of patients who needed to be treated with adrenaline to save one life was 112. A favorable neurological outcome at hospital discharge was no different between groups; importantly, patients with a poor neurological outcome (a score of 4 or 5 on the Modified Rankin Scale) was much higher with adrenaline use (31.0% vs. 17.8%). This study strengthens the evidence that pre-hospital adrenaline administration as currently recommended may improve survival at the expense of poor neurological outcomes. The median time to administration of the study drug was about 21 minutes; hence the finding of this study are not extrapolatable to in-hospital cardiac arrests.

A 3-phase model has been proposed to postulate the physiological changes during CPR.(9) This model involves an early electrical phase extending from the onset of cardiac arrest to approximately 4 min, during which defibrillation and compressions are crucial. This is followed by the circulatory phase, between 4–10 min after cardiac arrest, during which interventions that improve tissue oxygen delivery are important, including adrenaline and continued compressions. Finally, the metabolic phase ensues after 10 min. Adrenaline may worsen outcomes during this phase by increasing the oxygen demand, thus worsening myocardial ischemia, and may also lead to thrombogenesis and an increase in lactate levels. The observational study from Osaka seems to support this hypothesis.(7)

Does the time-honoured dose of 1 mg of adrenaline, based mainly on older animal models need to be modified? Intuitively, it would seem to be a mega-dose, especially if one considers the physiological effect of such a large bolus. Perhaps a smaller dose, delivered as an infusion, may be more appropriate and needs investigation.(10) Besides, a dosing interval of longer than 5 min may be more optimal.(11)


  • Adrenaline improves coronary blood flow by increasing the aortic diastolic pressure; however, it increases the myocardial oxygen demand. Platelet activation and propensity to thrombogenesis may lead to poor outcomes.
  • There is robust evidence that adrenaline administration facilitates ROSC; however, neurological outcomes are unchanged or even worse.
  • Adrenaline administration later during CPR may not be effective.
  • Future research should focus on the timing and a lower dose of adrenaline, possibly as an infusion compared to bolus doses.



  1. Crile G, Dolley DH. An experimental research into the resuscitation of dogs killed by anesthetics and asphyxia. J Exp Med. 1906 Dec 21;8(6):713–25.
  2. Redding JS. Resuscitation from ventricular fibrillation. Drug therapy. JAMA J Am Med Assoc. 1968 Jan 22;203(4):255–60.
  3. Neumar Robert W., Shuster Michael, Callaway Clifton W., Gent Lana M., Atkins Dianne L., Bhanji Farhan, et al. Part 1: Executive Summary. Circulation. 2015 Nov 3;132(18_suppl_2):S315–67.
  4. Olasveengen TM, Sunde K, Brunborg C, Thowsen J, Steen PA, Wik L. Intravenous Drug Administration During Out-of-Hospital Cardiac Arrest. :8.
  5. Olasveengen TM, Wik L, Sunde K, Steen PA. Outcome when adrenaline (epinephrine) was actually given vs. not given – post hoc analysis of a randomized clinical trial. Resuscitation. 2012 Mar;83(3):327–32.
  6. Jacobs IG, Finn JC, Jelinek GA, Oxer HF, Thompson PL. Effect of adrenaline on survival in out-of-hospital cardiac arrest: A randomised double-blind placebo-controlled trial. Resuscitation. 2011 Sep;82(9):1138–43.
  7. Hagihara A, Hasegawa M, Abe T, Nagata T, Wakata Y, Miyazaki S. Prehospital epinephrine use and survival among patients with out-of-hospital cardiac arrest. JAMA. 2012 Mar 21;307(11):1161–8.
  8. Perkins GD, Ji C, Deakin CD, Quinn T, Nolan JP, Scomparin C, et al. A Randomized Trial of Epinephrine in Out-of-Hospital Cardiac Arrest. N Engl J Med. 2018 Aug 23;379(8):711–21.
  9. Weisfeldt ML, Becker LB. Resuscitation After Cardiac Arrest: A 3-Phase Time-Sensitive Model. JAMA. 2002 Dec 18;288(23):3035.
  10. Arrich J, Sterz F, Herkner H, Testori C, Behringer W. Total epinephrine dose during asystole and pulseless electrical activity cardiac arrests is associated with unfavourable functional outcome and increased in-hospital mortality. Resuscitation. 2012 Mar;83(3):333–7.
  11. Warren SA, Huszti E, Bradley SM, Chan PS, Bryson CL, Fitzpatrick AL, et al. Adrenaline (epinephrine) dosing period and survival after in-hospital cardiac arrest: a retrospective review of prospectively collected data. Resuscitation. 2014 Mar;85(3):350–8.









Controversies in feeding the critically ill…


Nutritional support is one of the key elements of care in critically ill patients. Providing adequate nutrition to patients with multiorgan failure can pose several challenges. Many new concepts have emerged over the years that have enabled optimization of the nutritional strategy. I will address key issues related to feeding the critically ill in this brief review.

What is the preferred route for nutritional support?

Although the enteral route is generally preferred in most critically ill patients, parenteral nutrition may be equally efficacious and may not be associated with worse outcomes. In a multicenter study in the UK, enteral nutrition was compared with parenteral nutrition administered within 36 hours of ICU admission and continued for 5 days. (1) The 30-day mortality, which was the primary outcome, was not significantly different between the enteral and parenteral routes of administration. Unlike older studies, no increase in infective complications was observed with parenteral nutrition. The 90-day mortality, one of the secondary outcomes, was also similar between groups. However, considering the ease of administration and the lower cost of care, the enteral route would still be preferred in most clinical situations.

How early should nutritional support be initiated?

Initiation of enteral nutrition within 24–48 hours of initiation of mechanical ventilation is appropriate in critically ill patients. Early commencement of enteral nutrition may preserve enterocyte function, reduce the incidence of infective complications, and may prevent stress ulcers. In a retrospective observational study, initiation of enteral nutrition within 48 hours of commencement of mechanical ventilation was associated with a significantly lower ICU and hospital mortality. (2) The mortality was lowest in the sickest quartiles of patients. Lewis et al. performed a meta-analysis of 11 randomized controlled trials that compared enteral feeding within 24 hours with a variable period of nil by mouth management after gastrointestinal surgery. In six studies, feeding was by the oral route, while it was directly into the small bowel in five studies. Early enteral feeding significantly reduced the incidence of infectious complications. Furthermore, there was a reduced risk of anastomotic leak, surgical site infections, pneumonia, intra-abdominal infections, and mortality with early enteral feeding. These studies offer strong evidence that early enteral nutrition is feasible and may be associated with improved clinical outcomes in critically ill patients, especially in the more severely ill.

What is an appropriate nutritional dose?

Contrary to widely held belief, an early, full nutritional dose compared to a low-calorie restrictive feeding strategy may be associated with adverse outcomes. During critical illness, extensive damage occurs to cellular organelles and protein aggregates, leading to organ failure. Recovery involves the process of autophagy to clear and repair the damage. Early full nutritional support may inhibit autophagy and worsen clinical outcomes. (3) Administration of calories of up to 70% of the resting energy expenditure has been shown to reduce mortality. A higher calorie dose was associated with increased mortality, longer ICU stay, and duration of mechanical ventilation. (4) Even small volumes of enteral feed may exert beneficial effects including preservation of gut epithelium, prevention of translocation of bacteria from the gut and enhanced immune function. In a randomized controlled study of trophic compared to full nutrition in patients with acute lung injury, no difference was observed in the number of ventilator-free days to day 28, or the 90-day mortality. (5)

What route for enteral nutrition?

There is a largely theoretical, increased risk of aspiration and pneumonia with the gastric delivery of enteral nutrition. Post-pyloric feeding may reduce the gastric residual volume; however, no clear advantage has been shown in clinical outcomes, including duration of ventilation, ICU stay, or mortality. Besides, it is also unclear whether post-pyloric feeding reduces the incidence of ventilator-associated pneumonia. Post-pyloric feeds may be appropriate in patients with intolerance to gastric feeds with a high risk of aspiration.

Supplemental parenteral nutrition

If enteral nutrition alone cannot provide caloric requirement, would supplementation with parenteral nutrition help? In the EPANIC study, patients were randomized to receive early supplementation within the first 48 hours of ICU admission to later initiation of parenteral nutrition, after 8 days. (6) Clinical outcomes were more favorable with late supplementation, including reduced ICU and hospital stay, fewer infective complications, reduced duration of renal replacement therapy, and lower health care costs. In a later study, parenteral nutrition was supplemented to meet calorie requirements within the first 4–8 days of ICU, compared to continued enteral nutrition. (7) Nosocomial infections were significantly lower with parenteral nutritional supplementation. Based on overall evidence, it may be inappropriate to commence supplemental parenteral nutrition within the first 5–8 days on ICU.(8)

Should we withhold enteral nutrition in shock?

In patients who are hemodynamically unstable and need support with vaso-active gastrointestinal hypoperfusion may occur. Arterial and venous flow are in opposite directions in the intestinal villi. This results in the transfer of oxygen from the artery to the vein in a countercurrent fashion, with a progressively lower oxygen content from the base to the apex of the villi. The apices of the villi may thus be prone to ischemia in low-flow states. Furthermore, a splanchnic “steal” phenomenon may occur in shocked patients, compromising oxygen delivery to vital organs. Non-occlusive bowel necrosis has also been reported in critically ill patients who were administered post-pyloric feeds. Presumably, direct feeding distal to the stomach may lead to distension of the small intestine and increased intraluminal pressure, predisposing to ischemia. Most guidelines recommend primary focus on resuscitative measures in unstable patients and commencement of low-volume, trophic feeds once reasonable hemodynamic stability has been achieved. (9)

Do we need to diligently measure gastric residual volumes and titrate feeds accordingly?

Contrary to conventional wisdom, gastric residual volumes are poor indicators of feeding tolerance. Intolerance to feeds is more clearly discernible by the presence of gastric distension, vomiting, or passive regurgitation. There is no evidence to support a threshold for gastric residual volume that may indicate inadequate absorption of feeds. In a randomized controlled study, continued administration of feeds without measurement of residual volumes did not increase the incidence of ventilator-associated pneumonia; besides, it enabled better achievement of nutritional targets. There was no significant difference in the incidence of other infectious complications, duration of ventilation, ICU stay, or mortality. (10)

Is a mandatory period of fasting required for intubated patients who undergo specific interventions?

In many critically ill patients, multiple interventions may be required that mandate fasting based on conventional wisdom. As a result, a substantial number of patients may fail to achieve adequate nutrition due to frequent interruption of enteral nutrition. In many clinical situations, it may not make intuitive sense to withhold feeds, when the planned procedure is unlikely to require airway manipulation that may pose a risk of aspiration. A shortened duration of 45 minutes of fasting did not lead to complications during bedside tracheostomy. (11) Uninterrupted delivery of enteral nutrition throughout the perioperative period has also been shown to be safe in burns patients who undergo serial debridement. (12)


  • Early initiation of enteral nutrition is preferred in most critically ill patients. Commencement of hypocaloric feeds may be appropriate in most situations. Trophic feeds of 20 ml/h is generally well tolerated and may carry non-nutritional benefits.
  • Supplemental parenteral nutrition may be considered when nutritional targets are not achieved after 5–8 days.
  • Enteral nutrition may be commenced in patients who are in shock, once the dose of vaso-active drugs has stabilized.
  • Titration of feeds based on measurement of gastric residual volumes may be unnecessary in most patients.
  • Mandatory duration of fasting recommended in anesthetic guidelines may need modification in patients who are intubated and mechanically ventilated.



  1. Harvey SE, Parrott F, Harrison DA, Bear DE, Segaran E, Beale R, et al. Trial of the Route of Early Nutritional Support in Critically Ill Adults. N Engl J Med. 2014 Oct 30;371(18):1673–84.
  2. Artinian V, Krayem H, DiGiovine B. Effects of Early Enteral Feeding on the Outcome of Critically Ill Mechanically Ventilated Medical Patients. Chest. 2006 Apr;129(4):960–7.
  3. Van Dyck L, Casaer MP, Gunst J. Autophagy and Its Implications Against Early Full Nutrition Support in Critical Illness. Nutr Clin Pract. 2018 Jun;33(3):339–47.
  4. Zusman O, Theilla M, Cohen J, Kagan I, Bendavid I, Singer P. Resting energy expenditure, calorie and protein consumption in critically ill patients: a retrospective cohort study. Crit Care [Internet]. 2016 Dec [cited 2019 Jan 9];20(1). Available from:
  5. The National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Rice TW, Wheeler AP, Thompson BT, Steingrub J, Hite RD, et al. Initial Trophic vs Full Enteral Feeding in Patients With Acute Lung Injury: The EDEN Randomized Trial. JAMA J Am Med Assoc. 2012 Feb 22;307(8):795–803.
  6. Casaer MP, Mesotten D, Hermans G, Wouters PJ, Schetz M, Meyfroidt G, et al. Early versus Late Parenteral Nutrition in Critically Ill Adults. N Engl J Med. 2011 Aug 11;365(6):506–17.
  7. Heidegger CP, Berger MM, Graf S, Zingg W, Darmon P, Costanza MC, et al. Optimisation of energy provision with supplemental parenteral nutrition in critically ill patients: a randomised controlled clinical trial. The Lancet. 2013 Feb;381(9864):385–93.
  8. Bost RB, Tjan DH, van Zanten AR. Timing of (supplemental) parenteral nutrition in critically ill patients: a systematic review. Ann Intensive Care [Internet]. 2014 Dec [cited 2019 Jan 9];4(1). Available from:
  9. Dhaliwal R, Cahill N, Lemieux M, Heyland DK. The Canadian Critical Care Nutrition Guidelines in 2013. Nutr Clin Pract. 2014 Feb 1;29(1):29–43.
  10. Reignier J, Mercier E, Le Gouge A, Boulain T, Desachy A, Bellec F, et al. Effect of not monitoring residual gastric volume on risk of ventilator-associated pneumonia in adults receiving mechanical ventilation and early enteral feeding: a randomized controlled trial. JAMA. 2013 Jan 16;309(3):249–56.
  11. Gonik N, Tassler A, Ow TJ, Smith RV, Shuaib S, Cohen HW, et al. Randomized Controlled Trial Assessing the Feasibility of Shortened Fasts in Intubated ICU Patients Undergoing Tracheotomy. Otolaryngol-Head Neck Surg. 2016 Jan;154(1):87–93.
  12. McElroy LM, Codner PA, Brasel KJ. A Pilot Study to Explore the Safety of Perioperative Postpyloric Enteral Nutrition. Nutr Clin Pract. 2012 Dec;27(6):777–80.









Tracheostomy: how do you get the timing right?


One of the earliest references in history to a procedure that vaguely resembles a tracheostomy is alluded to in the Rig Veda, with the description of the healing of a throat incision.The Egyptians, ancient pioneers of “modern” medicine as we know it today, were past masters of many different surgical procedures in their time and may have documented the first ever tracheostomy in history.  Alexander the Great, the historical hero that he was, did not hesitate to stab the windpipe of one of his fellow soldiers, who was choking on a piece of bone.1

In more recent times, the debate on what is a good time to perform a tracheostomy continues to captivate intensivists. To say that the evidence has been contradictory would perhaps be an understatement.

The putative benefits of tracheostomy include ease of tracheal suctioning and patient mobilization, reduction or earlier cessation of sedation, and facilitation of speech and oral intake. Besides, even more importantly, a shorter tube may help with spontaneous breathing due to a lower airflow resistance and reduced work of breathing, enabling earlier weaning and liberation from mechanical ventilation. One of the early randomized controlled trials (RCT) compared tracheostomy within the first 48 hours of ventilation to a later tracheostomy between days 14–16.This study revealed marked benefits of an early tracheostomy, including reduced mortality, a lower incidence of pneumonia, and less duration on ventilation and in the ICU. However, the findings of many later studies have been more sobering. Several meta-analysts have also thrown in their weight trying to find an answer to this vexing question. Griffiths et al., in their meta-analysis of five RCTs, found no significant difference in mortality or the incidence of pneumonia. However, an early tracheostomy resulted in a reduced duration of mechanical ventilation and stay in the ICU.3 However, Wang et al. did not find any reduction in the duration of mechanical ventilation nor in the length of ICU or hospital stay with an early tracheostomy.A Cochrane review of more recent studies revealed a lower mortality at the longest time of follow-up with early tracheostomy; however, the authors cautioned that high-quality information is not available on the subject, and their findings may only be “suggestive”.5

The TracMan study is the largest, multi-center, RCT that has been conducted to evaluate the possible benefit of an early tracheostomy.Critically ill patients from 70 general intensive care units across the UK were eligible if they were within 4 days of ICU admission, and, would require at least 7 days of mechanical ventilation according to clinician judgment. An early tracheostomy was performed within the first 4 days of admission to the critical care unit; patients randomized to the control group were subjected to a tracheostomy after 10 days if it was still considered necessary by the clinician.

In the early group, 84.6% of patients underwent a tracheostomy as planned; however, among patients randomized to the late group, only 45% received a tracheostomy. Eighty-nine (19.6%) of patients in the late group were discharged from the ICU by day 10, while 78 (17.2%) were still in ICU, but off ventilator support. Thus, 167 out of the 448 patients who were allocated to the late group could be weaned off ventilation and extubated by day 10. This staggering statistic suggests that clinicians could easily misjudge the duration of ventilation and the requirement for tracheostomy. All-cause mortality at 30 days, the primary outcome for which the study was powered for, did not differ significantly between groups. The secondary outcomes, including survival rates at ICU and hospital discharge, and at 1- and 2-year follow-up were also not significantly different. Furthermore, the duration of ventilation and the duration of ICU stay were similar between groups. The sole advantage observed with an early tracheostomy was less use of sedation among 30-day survivors. The targeted sample size could not be achieved as the recruitment rate slowed down over time.  Besides, the TracMan trial included only 5% of patients with a primary neurological illness, a subgroup of patients who might probably benefit from an early tracheostomy.

At the end of the day, what should our timing strategy be in patients who, according to our judgment may require long-term ventilator support?

In my opinion, there are relatively few patients who might benefit from a tracheostomy within the first 3–5 days of ventilation; perhaps patients with neurological illnesses may be able to be liberated from mechanical ventilation early and cared for in a ward with an early tracheostomy. This may be important, especially when the cost of care is an important consideration. It is also important to consider the clinical situation while contemplating tracheostomy. If reducing the level of ventilator support seems unlikely, there may not be much point in performing a tracheostomy. This applies to patients who may require high levels of PEEP, FiO2 or inspiratory pressures. Attempting a tracheostomy in this setting may lead to worsening of gas exchange and is more likely to cause harm. A tracheostomy should perhaps also be deferred if the prospect of a reasonably meaningful recovery seems unlikely.


  1. Colice GL (1994) Historical background. In: Tobin MJ (ed) Principles and practice of mechanical ventilation. McGraw-Hill, New York, pp 1-37
  2. Rumbak MJ, Newton M, Truncale T, Schwartz SW, Adams JW, Hazard PB. A prospective, randomized, study comparing early percutaneous dilational tracheotomy to prolonged translaryngeal intubation (delayed tracheotomy) in critically ill medical patients. Crit Care Med. 2004 Aug;32(8):1689–94.
  3. Griffiths J, Barber VS, Morgan L, Young JD. Systematic review and meta-analysis of studies of the timing of tracheostomy in adult patients undergoing artificial ventilation. BMJ. 2005;330(7502):1243.
  4. Wang F, Wu Y, Bo L, Lou J, Zhu J, Chen F, Li J, Deng X. The timing of tracheotomy in critically ill patients undergoing mechanical ventilation: a systematic review and meta-analysis of randomized controlled trials. Chest. 2011 Dec;140(6):1456-65.
  5. Gomes Silva BN, Andriolo RB, Saconato H, Atallah AN, Valente O. Early versus late tracheostomy for critically ill patients. Cochrane Database Syst Rev. 2012 Mar 14;3:CD007271. doi: 10.1002/14651858
  6. Young D, Harrison DA, Cuthbertson BH, Rowan K; TracMan Collaborators. Effect of early vs late tracheostomy placement on survival in patients receiving mechanical ventilation: the TracMan randomized trial. JAMA. 2013 May 22;309(20):2121-9


Aerosolized antibiotics for ventilator-associated pneumonia

Ventilator-associated pneumonia (VAP) caused by multidrug-resistant bacteria continues to be a major cause of morbidity and mortality in our ICUs. We have a limited choice of antibiotics to combat the resistant bacterial flora prevalent in many units. Besides, most systemically administered antibiotics fail to attain therapeutic concentrations in the lung. This has led many clinicians to resort to aerosolized antibiotics, often as an adjuvant to systemic therapy in the treatment of VAP. The use of inhaled antibiotics is based on sound rationale, with the possibility of delivering a high concentration of the drug to the target site. Furthermore, the emergence of resistant organisms may also be reduced with preservation of the gut flora. An added advantage may be to cut down the duration of systemic antibiotics, and perhaps, even use inhaled antibiotics are monotherapy.

Inhalational antibiotic therapy is not a new therapeutic modality; a 1975 study of ICU patients used polymyxin B in the atomized form or by endotracheal instillation in intubated patients as prophylaxis against pneumonia. (1) Predictably, such unlimited, universal, largely prophylactic therapy resulted in a high level of polymyxin resistance and high mortality.

Does inhalational therapy really work?

Although theoretically rational, there is a paucity of evidence with the use of aerosolized antibiotic therapy regarding efficacy, appropriate dosing, and method of administration.  Most of the evidence is from retrospective observational studies with no large randomized controlled trial that has addressed the efficacy of inhalational therapy. Zampieri et al. performed a meta-analysis comparing the combination of intravenous and aerosolized antibiotics with intravenous administration alone in the treatment of VAP. (2) The antibiotics used included gentamicin, amikacin, tobramycin, ceftazidime, and colistin. The likelihood of clinical cure was significantly higher with combined therapy; however, there was no significant difference in microbiological cure rates, ICU or hospital mortality, ventilation days, and ICU length of stay. In another meta-analysis, intravenous colistin was compared with adjunctive inhalational colistin therapy. This study reported a significantly higher clinical response, microbiological eradication, and infection-related mortality. (3) There was no difference in overall mortality or nephrotoxicity.

A retrospective cohort study compared intravenous colistin alone with combination therapy in patients with microbiologically proven VAP. (4) Cure rates were significantly higher with combination therapy; however, no difference was noted in all-cause ICU or hospital mortality. Similar findings were also observed in a retrospective case-control study; (5) Acinetobacter baumanni was the most common organism, followed by Klebsiella pneumoniae, and Pseudomonas aeruginosa in this study. A higher clinical cure was observed with combination therapy; however, there was no difference in mortality or pathogen eradication. There are few randomized controlled studies that have compared inhalational to intravenous therapy. The results have been mixed, with reports of reduction in the clinical pulmonary infection score (CPIS), (6) and microbiological cure with unchanged clinical outcomes. (7) A recent study using a combination of aerosolized amikacin and fosfomycin revealed no difference in CPIS, clinical cure, or mortality. (8)

Do we use inhalational therapy considering the mixed evidence?

Although there is no firm evidence of benefit yet, inhaled antibiotics seem to improve clinical and microbiological cure rates. Importantly, inhalational therapy has not led to adverse outcomes among available studies. Considering the often dismal outcomes from multi-drug resistant VAP and the relative paucity of effective antibiotics, it may be reasonable to administer aerosolized therapy, especially if conventional intravenous therapy proves ineffective.


The dose of aerosolized antibiotics varies widely between different studies. However, based on the limited data available, the following doses may be reasonable for VAP or ventilator-associated tracheobronchitis.

Colistin:  150 mg (5 million units of colistimethate sodium) BD

Gentamicin and tobramycin: 300 mg BID

Amikacin: 400 mg BID

Points to remember when administering aerosolized colistin

  • In India, the commonly available formulation is colistimethate sodium (polymyxin E), which is a prodrug. Do remember the conversion formula:

80 mg colistimethate sodium (prodrug) = 30 mg colistin base activity = approximately 1,000,000 (1 million) units of colistimethate sodium

  • Use normal or half normal saline for dilution in a total volume of 3–5 ml; administer the solution immediately after reconstitution. Although jet and ultrasonic nebulizers are commonly used, vibrating plate nebulizers may be preferable.
  • If the patient is prone to bronchospasm, pre-treatment with nebulized salbutamol is recommended.
  • The ventilator filter at the expiratory end of the circuit may get clogged with aerosolization; some guidelines recommend replacing the filter after every dose.



  1. Feeley TW, du Moulin GC, Hedley-Whyte J, Bushnell LS, Gilbert JP, Feingold DS. Aerosol Polymyxin and Pneumonia in Seriously Ill Patients. N Engl J Med. 1975 Sep 4;293(10):471–5.
  2. Zampieri FG, Nassar Jr AP, Gusmao-Flores D, Taniguchi LU, Torres A, Ranzani OT. Nebulized antibiotics for ventilator-associated pneumonia: a systematic review and meta-analysis. Crit Care [Internet]. 2015 Dec [cited 2018 Dec 29];19(1). Available from:
  3. Valachis A, Samonis G, Kofteridis DP. The Role of Aerosolized Colistin in the Treatment of Ventilator-associated Pneumonia: A Systematic Review and Metaanalysis*. Crit Care Med. 2015 Mar 1;43(3):527–33.
  4. Korbila IP, Michalopoulos A, Rafailidis PI, Nikita D, Samonis G, Falagas ME. Inhaled colistin as adjunctive therapy to intravenous colistin for the treatment of microbiologically documented ventilator-associated pneumonia: a comparative cohort study. Clin Microbiol Infect. 2010 Aug 1;16(8):1230–6.
  5. Kofteridis DP, Alexopoulou C, Valachis A, Maraki S, Dimopoulou D, Georgopoulos D, et al. Aerosolized plus Intravenous Colistin versus Intravenous Colistin Alone for the Treatment of Ventilator-Associated Pneumonia: A Matched Case-Control Study. Clin Infect Dis. 2010 Dec 1;51(11):1238–44.
  6. Palmer LB, Smaldone GC, Chen JJ, Baram D, Duan T, Monteforte M, et al. Aerosolized antibiotics and ventilator-associated tracheobronchitis in the intensive care unit*. Crit Care Med. 2008 Jul 1;36(7):2008–13.
  7. Rattanaumpawan P, Lorsutthitham J, Ungprasert P, Angkasekwinai N, Thamlikitkul V. Randomized controlled trial of nebulized colistimethate sodium as adjunctive therapy of ventilator-associated pneumonia caused by Gram-negative bacteria. J Antimicrob Chemother. 2010 Dec 1;65(12):2645–9.
  8. Kollef MH, Ricard J-D, Roux D, Francois B, Ischaki E, Rozgonyi Z, et al. A Randomized Trial of the Amikacin Fosfomycin Inhalation System for the Adjunctive Therapy of Gram-Negative Ventilator-Associated Pneumonia: IASIS Trial. Chest. 2017 Jun 1;151(6):1239–46.











Early extubation followed by immediate noninvasive ventilation vs. standard extubation in hypoxemic patients: a randomized clinical trial (1)

Background: Non-invasive ventilation (NIV) has been established to be an effective modality to facilitate extubation in the presence of hypercapnia, especially in patients with chronic obstructive pulmonary disease, cardiogenic pulmonary edema, and following abdominal surgery.(2) However, NIV use to expedite liberation from invasive mechanical ventilation (iMV) in non-hypercapnic patients with hypoxemic respiratory failure has not been adequately investigated.  The present study was conducted to evaluate the efficacy of NIV in facilitating liberation from mechanical ventilation among non-hypercapnic patients with hypoxemic respiratory failure.

Setting: The study was conducted over a 3-y period in six Chinese and three Italian intensive care units of academic centers in both countries.

Population: Adult patients who were on mechanical ventilation for more than 48 h were eligible if they had (1) a P/F ratio of 200–300 on an FiO2of 0.6 or less on pressure support ventilation with a total applied pressure of  25 cm of H2O or less, with a PEEP of 8–13 cm of H2O; (2) a respiratory rate of 30/min or less; (3) PaCO2of 50 mm Hg or less and a pH of 7.35 or more; (4) a tidal volume of less than 8 ml/kg of ideal body weight, (5) a normal GCS, and (6) a temperature of less than 38.5 C. Patients had an adequate cough with requirement for endotracheal suctioning of less than two times per hour. After exclusion of 1129/1259 eligible patients for various reasons, including hemodynamic instability, vasoactive agent use, life-threatening arrhythmias, sepsis, two or more organ failures, and BMI of > 30 kg/cm2, 130 patients were randomized.

Intervention: Patients were extubated and commenced on NIV using the same settings on pressure support mode at the time of extubation. NIV pressures were weaned down according to a protocol. Briefly, this involved increasing the PEEP and waiting if the P/F was less than 225; once the P/F was more than 225, the PEEP and inspiratory pressure levels were weaned down. NIV was ceased when the P/F ratio was more than 250 mm Hg at a PEEP of 8 cm H2O and PS of 10 cm H2O. Following this, patients were put on a ventimask at FiO2of 0.35 to maintain pH ≥7.35, PaCO2 ≤50 mmHg, P/F ratio≥200 mmHg and respiratory rate of ≥ 30/min.

Control: Invasive ventilation was continued using the same protocol-based, stepwise reduction in inspiratory pressure and PEEP levels used to wean down NIV support in the intervention arm. Prophylactic NIV could be used soon after extubation for a maximum duration of 12 hours at the discretion of the treating physician.

In both groups, respiratory failure requiring reintubation, NIV, or non-invasive CPAP within 48 hours of unassisted breathing was considered as “treatment failure”.

Primary outcomes: The co-primary outcomes evaluated in the study were (1) the duration of iMV and (2) the duration of ICU stay. The duration of iMV was significantly less with early extubation to NIV [5.5 (4.0–9.0) vs. 4.0 (3.0–7.0) days; p = 0.004). The duration of ICU stay was not significantly different between groups [9.0 (6.5–12.5) vs. 8.0 (6.0–12.0) days, p = 0.259]. Surgical patients seemed to benefit most from NIV-facilitated early extubation.

Secondary outcomes: The incidence of treatment failure, serious adverse events, and requirement for tracheostomy were not significantly different between groups; the ICU and hospital mortality were also not significantly different. The incidence of ventilator-associated pneumonia and tracheobronchitis, use of sedatives, and hospital length of stay were significantly lower with early extubation to NIV. On Kaplan Meir analysis, the total duration of iMV and NIV combined was not significantly different between groups.

Comments: The study was conducted on patients with P/F ratios of 200–300, on pressure support ventilation. Conventional practice in most settings among such patients would be to expedite weaning, using a short spontaneous breathing trial and consider extubation if successful.(3,4) It is likely that extubation may have been unnecessarily delayed in most patients in the control group who underwent continued iMV and weaning using a stepwise protocolized approach. The similar duration of ICU length of stay, in spite of a minimal difference of approximately 1.5 days of iMV, would also support the possibility that most patients in the control arm were also ready for earlier extubation. The shorter duration of hospital stay in the treatment arm is hard to explain, considering that the duration of ICU stay was similar. Almost 90% of eligible patients were excluded for various reasons, which questions the validity of the findings in the real world. Besides, in an unblinded study, wherein observers are aware of group allocation, performance and detections biases are likely, that could have affected the final results. The study was conducted across multiple centers in only two countries which may limit generalizability.

My take: In our practice, a P/F ratio of 200–300 on any level of pressure support ventilation would be an indication for a short spontaneous breathing trial and extubation if the patient is able to sustain. We would use NIV post extubation in selected patients, based on clinical judgment. I strongly feel that this study just shows that a complex weaning protocol may delay extubation in patients who are otherwise ready for liberation from invasive mechanical ventilation

What is your weaning and extubation policy?  Please offer your valuable comments.


  1. Vaschetto R, Longhini F, Persona P, Ori C, Stefani G, Liu S, et al. Early extubation followed by immediate noninvasive ventilation vs. standard extubation in hypoxemic patients: a randomized clinical trial. Intensive Care Med [Internet]. 2018 Dec 10 [cited 2018 Dec 27]; Available from:
  2. Official ERS/ATS clinical practice guidelines: noninvasive ventilation for acute respiratory failure | European Respiratory Society [Internet]. [cited 2018 Dec 27]. Available from:
  3. Perkins GD, Mistry D, Gates S, Gao F, Snelson C, Hart N, et al. Effect of Protocolized Weaning With Early Extubation to Noninvasive Ventilation vs Invasive Weaning on Time to Liberation From Mechanical Ventilation Among Patients With Respiratory Failure: The Breathe Randomized Clinical Trial. JAMA. 2018 Nov 13;320(18):1881.
  4. Nava S, Gregoretti C, Fanfulla F, Squadrone E, Grassi M, Carlucci A, et al. Noninvasive ventilation to prevent respiratory failure after extubation in high-risk patients: Crit Care Med. 2005 Nov;33(11):2465–70.


“Early” antibiotics: absolute sine qua non or unjustified paranoia?

There is increasing emphasis by regulatory bodies and expert group guidelines to administer antibiotics expeditiously once an infection is suspected. The surviving sepsis campaign proposes a “1-h bundle” comprising of a slew of measures, including antibiotic administration. Unarguably, antibiotic therapy should not be delayed in patients who are truly septic; however, would a tight timeframe lead to injudicious administration of antibiotics in patients who may not require them, including those with non-infective illnesses? Undoubtedly, the widespread use of broad-spectrum antibiotics has resulted in the genesis of multidrug-resistant organisms in the community, almost leading to a global crisis. Besides, there is a significant risk of inducing resistant organisms, including fungi, in the individual patient due to selective pressure.

One of the early studies that caused understandable paranoia suggested that every hour of delay in antibiotic administration resulted in an increase in mortality by 7.6% in the first 6 hours after the onset of septic shock. (1) There was no documented infection in a substantial number of patients (22.1%) in this study. No information was available on the timing of source control, which clearly takes precedence in septic patients. Three patient cohorts of “approximately” 150 each were retrospectively studied, over a 15-yr period between 1989 to 2004. Other retrospective studies have also arrived at similar conclusions with a stepwise increase in mortality with delay in initiation of antibiotic therapy. (2)

However, prospective studies that address this important question of antibiotic timing have a different tale to tell. In a before-after study of patients admitted to a surgical ICU, antibiotic therapy commenced soon after suspicion of infection was compared with a conservative approach wherein antibiotics were commenced after microbiological or other objective evidence of infection was obtained. The conservative approach was more often associated with appropriate initial therapy, and resulted in a significantly lower all-cause mortality. Importantly, a significantly lower mortality was also observed in patients who required vasoactive drugs for a mean arterial pressure of less than 60 mm Hg. (3)

In 715 consecutive patients with septic shock who presented to an emergency department, there was no significant increase in 28-d mortality for up to 5 h of delay in antibiotic administration after the onset of shock. However, failure to achieve initial goals of resuscitation, the SOFA score, and lactate levels were associated with mortality on multivariate analysis. (4) Puskarich et al. observed similar findings; there was no increase in mortality for up to 6 h of delay in antibiotic administration following the diagnosis of septic shock, in patients who received a structured, early resuscitation protocol. (5)

In a randomized controlled study, patients suspected to have sepsis were administered empirical ceftriaxone by ambulance personnel or offered usual care. The median time of receiving antibiotics was 26 min before arrival to the emergency department in the intervention group compared to 70 min after arrival in the control group. The 28-d mortality was not significantly different between groups. This study included relatively few patients (3.8%) with septic shock; however, in less seriously ill patients, a delayed approach did not lead to adverse outcomes. (6)

In mechanically ventilated patients in the ICU, diagnosis of ventilator-associated pneumonia (VAP) can be difficult; radiographic infiltrates are notoriously non-specific and fever may occur due to non-infectious causes. This is a typical situation in which clinicians may feel the pressure to initiate antibiotic therapy even if there is a low index of suspicion. Besides, if there is a perceived lack of response to initial therapy, subsequent antibiotic jugglery may well ensue, with an exponential increase in superinfection with multidrug-resistant organisms. If septic shock is present, an expeditious approach is justified; however, in less severely ill patients, waiting for more objective evidence including gram stain or culture results may perhaps be more appropriate.

More recent, prospective studies suggest that initial resuscitation and attempts to identify the likely source, with treatment directed towards the most likely organisms, keeping in mind the local microbial environment may be a more appropriate approach. The importance of adequate source control cannot be overemphasized. In fact, some infections may require source control alone, such as an infected central venous catheter that leads to sepsis. Extravagant, broad-spectrum antibiotic use without adequate evaluation, in an attempt to reduce “delay”, is likely to contribute to the ever-growing list of superbugs in the community; besides, at the individual patient level, resistant pathogens are likely to freely proliferate. It may also be pointed out that in many of the studies that address delay, the time lag between the actual onset of sepsis and diagnosis is unknown; interpretation of delay when time zero is obscure may be fraught with misperceptions.

Clearly, we need to strike a balance here!  It would probably pay to apply some considered thought and get the early resuscitation going before you throw the most powerful weed killer at the presumed invasion by a bacterial army.



  1. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock*: Crit Care Med. 2006 Jun;34(6):1589–96.
  2. Liu VX, Fielding-Singh V, Greene JD, Baker JM, Iwashyna TJ, Bhattacharya J, et al. The Timing of Early Antibiotics and Hospital Mortality in Sepsis. Am J Respir Crit Care Med. 2017 Oct;196(7):856–63.
  3. Hranjec T, Rosenberger LH, Swenson B, Metzger R, Flohr TR, Politano AD, et al. Aggressive versus conservative initiation of antimicrobial treatment in critically ill surgical patients with suspected intensive-care-unit-acquired infection: a quasi-experimental, before and after observational cohort study. Lancet Infect Dis. 2012 Oct;12(10):774–80.
  4. Ryoo SM, Kim WY, Sohn CH, Seo DW, Oh BJ, Lim KS, et al. Prognostic Value of Timing of Antibiotic Administration in Patients With Septic Shock Treated With Early Quantitative Resuscitation. Am J Med Sci. 2015 Apr;349(4):328–33.
  5. Puskarich MA, Trzeciak S, Shapiro NI, Arnold RC, Horton JM, Studnek JR, et al. Association between timing of antibiotic administration and mortality from septic shock in patients treated with a quantitative resuscitation protocol*: Crit Care Med. 2011 Sep;39(9):2066–71.
  6. Alam N, Oskam E, Stassen PM, Exter P van, van de Ven PM, Haak HR, et al. Prehospital antibiotics in the ambulance for sepsis: a multicentre, open label, randomised trial. Lancet Respir Med. 2018 Jan;6(1):40–50.



Albumin infusion in the critically ill: are we wiser today?


Commercial preparations of albumin have been in use from the 1940s. The first protein to be extracted from human plasma, it was extensively used in the battles of World War II and subsequently, in civilian practice. A major controversy erupted and continues to surround the use of albumin since the publication of the systematic review by the Cochrane Injuries Group in 1998.The authors performed a meta-analysis on a mixed group of patients including surgical, trauma, sepsis, and burns and demonstrated an overall significant increase of mortality with the use of albumin, compared to alternative fluids. The study was limited by a vastly heterogeneous group of patients; besides, albumin was compared to several different types of fluids. Furthermore, many of the studies were several decades old, with practices that may not have been contemporaneous. Predictably, this report drew sharp reactions from the medical community and from the mainstream media. There was a considerable decline in the usage of albumin solutions in the following years, especially in UK intensive care units.

Albumin may seem close to being the ideal intravenous fluid to fill the intravascular compartment based on the Starling hypothesis, considering its colloid osmotic pressure and molecular size. Let us examine if this assumption is really true.

According to the Starling hypothesis, fluid filtration occurs across the arterial end of capillaries to the interstitium across a hydrostatic pressure gradient. On the venous side, the reverse movement was proposed to occur, from the interstitium back into the capillaries, driven by the higher colloid osmotic pressure of the plasma. However, it has been clearly established that there is no fluid reabsorption into the capillaries as believed previously; the filtered fluid is cleared by the lymphatic system.According to the new approach, the endothelium is lined by a layer of glycoproteins and proteoglycans, constituting the glycocalyx (Fig. 1). The glycocalyx layer is bound to proteins, mainly, albumin, and constitutes a barrier to the movement of fluid out of the capillaries along the hydrostatic gradient. There is a sub-glycocalyx space at the gap between the endothelial cells, where fluid movement occurs; this space is devoid of protein, and hence, cannot exert a colloid osmotic pressure. Hence, contrary to the Starling hypothesis, fluid movement cannot occur from the interstitium back into the capillaries.


Fig. 1: The glycocalyx layer lines the inside of the capillary endothelium and acts as a barrier to fluid filtration. Filtration occurs through the gap (colored white) between endothelial cells (colored brown). The capillary lumen is separated from the interstitium at the gaps by the glycocalyx (colored green) and the subglycocalyx space (arrow). The subglycocalyx space is devoid of protein, and hence, cannot facilitate reabsorption of filtered fluid back into the capillary lumen

The glycocalyx layer gets disrupted in critical illness, including sepsis, trauma, and postoperative patients. Once the glycocalyx breaks down, colloidal solutions, including albumin can filter through the capillary endothelium and distribute within the interstitial space. This is why, contrary to conventional wisdom, the volume of fluid required to fill the intravascular compartment is nearly equal with crystalloids and colloids, including albumin, in contrast to the predicted 3:1 ratio. Once the glycocalyx disintegrates, both types of fluid leak out, regardless of the colloid osmotic pressure or molecular size. Albumin may have other putative benefits including anti-inflammatory and free radical scavenging effects, which may be beneficial in critically ill patients.

Do these purported physiological advantages of albumin translate to clinical benefits? The Saline versus Albumin Fluid Evaluation (SAFE) Study compared the administration of 4% albumin to normal saline in 6,977 patients in a randomized controlled trial.There was no difference overall in the 28-day mortality, requirement for ventilator support and renal replacement therapy. The duration of ICU and hospital stays were also similar. On subgroup analysis, there was a non-significant trend towards reduced 28-day mortality in patients with severe sepsis. On the contrary, the mortality risk was higher in the subgroup of trauma patients; the risk of death was mainly due to the increased mortality among patients with traumatic brain injury who were administered albumin. The SAFE study was primarily designed to test the hypothesis that the use of 4% albumin does not increase 28-day mortality compared to normal saline. The sample size was calculated based on an assumed mortality of 15% and to test a 3% difference in mortality between groups. The study mortality was nearly 21% in both arms. Although not meant to demonstrate a significant difference in clinical outcomes, the sample size was adequate to detect any clinical outcome benefit with albumin use.

The Albumin Italian Outcome Sepsis (ALBIOS) study was conducted to specifically investigate outcome benefits with the use of albumin in patients with severe sepsis, considering its anti-inflammatory and free radical scavenging properties. Along with crystalloids, 20% albumin was administered in the study arm, targeting a serum albumin level of 30 g/L; the control group received crystalloids alone.No difference was observed in the 28 or 90-day mortality; however, there was a statistically significant, one day difference in the duration of vasopressor support. On non-predefined subgroup analysis, there was a trend towards improved survival at 90 days in the albumin group.

The Early Albumin Resuscitation during Septic Shock study (EARSS) evaluated the use of 100 ml of 20% albumin, administered 8 hourly for the first 72 hours after the diagnosis of septic shock in a multicentre, randomized, placebo-controlled study in France.The preliminary findings, reported only in an abstract form, did not reveal any improvement in 28-day survival with the use of 20% albumin. The full report of this study is still awaited.

It may seem that the theoretical advantages of albumin use as an intravenous colloid do not translate to improved clinical outcomes in a general critically ill population. There is no evidence so far that it may improve outcomes in severe sepsis. However, it may offer clinical benefit in specific subgroups of patients, including in spontaneous bacterial peritonitis,as replacement fluid during abdominal paracentesis,and as combination therapy, with terlipressin in Type I hepato-renal syndrome.8

The bottom line

  • Transcapillary leak of fluid occurs due to disruption of the glycocalyx layer in critically ill patients. Colloid solutions including albumin, are distributed into the interstitial space similar to crystalloids.
  • Albumin-based resuscitation use does not improve clinical outcomes in a general critically ill population.
  • Therapeutic use of albumin targeting a serum level of 30/L does not improve outcomes in septic patients.
  • Specific subgroups of patients with liver disease may benefit from albumin use.



  1. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. BMJ 317, 235–240 (1998).
  2. Adamson, R. H. et al.Oncotic pressures opposing filtration across non-fenestrated rat microvessels. J. Physiol. 557, 889–907 (2004).
  3. A Comparison of Albumin and Saline for Fluid Resuscitation in the Intensive Care Unit. N. Engl. J. Med.10 (2004).
  4. Caironi, P. et al.Albumin Replacement in Patients with Severe Sepsis or Septic Shock. N. Engl. J. Med. 370, 1412–1421 (2014).
  5. Chapentier, Mira. Efficacy and tolerance of hyperoncotic albumin administration in septic shock patients: the EARSS study [abstract]. Intensive care medicine, 37(Supplement 2): S115-438. 2011; 37, 438
  6. Salerno, F., Navickis, R. J. & Wilkes, M. M. Albumin Infusion Improves Outcomes of Patients With Spontaneous Bacterial Peritonitis: A Meta-analysis of Randomized Trials. Clin. Gastroenterol. Hepatol. 11, 123-130.e1 (2013).
  7. Bernardi, M., Maggioli, C. & Zaccherini, G. Human Albumin in the Management of Complications of Liver Cirrhosis. in Annual Update in Intensive Care and Emergency Medicine 2012(ed. Vincent, J.-L.) 421–430 (Springer Berlin Heidelberg, 2012). doi:10.1007/978-3-642-25716-2_39
  8. Italian Association for the Study of the Liver (AISF). AISF position paper on nonalcoholic fatty liver disease (NAFLD): Updates and future directions. Dig. Liver Dis. Off. J. Ital. Soc. Gastroenterol. Ital. Assoc. Study Liver 49, 471–483 (2017).





Salt or sugar for the swollen brain?


Consider the following case scenario: A 42-year-old man is brought to the Emergency Department following a car crash. He has a GCS of 6 and is intubated and ventilated. The CT scan shows acute subdural hemorrhage with a midline shift of 3 cm and cerebral edema. He needs urgent evacuation of the subdural hematoma; in the meanwhile, you consider osmotherapy to reduce the intracranial pressure. Would you choose 20% mannitol or hypertonic saline?

Nearly a century ago, Weed and Mckibben serendipitously demonstrated a reduction in ICP following intravenous administration of hypertonic saline in cats.Following this seminal study, clinicians were quick to adopt this practice in their patients; a variety of osmotic agents were used, including glycerol and urea. From the 1960s, mannitol became firmly entrenched in clinical practice as standard osmotic therapy. This practice continued until there was an upsurge of interest with the use of hypertonic saline in the past two decades.

Hypertonic saline has several potential advantages over mannitol. It expands the intravascular compartment and may better maintain cardiac output and blood pressure; in contrast, mannitol may lead to excessive diuresis and hypovolemia. Mannitol may cause renal dysfunction in higher doses; besides, it may lead to a hyperglycemic, hyperosmolar state and cause encephalopathy. Other theoretical advantages of hypertonic saline include immunomodulatory and anti-inflammatory effects. It may also prevent the accumulation of the excitatory amino acid, glutamine, and prevent neuronal damage.2

In the light of these putative advantages, how does hypertonic saline compare with mannitol in the real word? Several randomized controlled studies have been performed that compared mannitol with boluses of hypertonic saline in different strengths to evaluate the effect on ICP. While some studies have demonstrated a more profound and sustained ICP reduction with hypertonic saline,3,4 others have not shown a significant difference.5,6 Importantly, no study has demonstrated a difference in clinical outcomes, including mortality. In the most exhaustive meta-analysis that compared both treatments, including 11 randomized controlled trials, no difference was seen in the degree of ICP reduction or on mortality in patients with severe traumatic brain injury.7

Hypertonic saline is available in solutions of different strengths, including 3%, 7.5%, and 23.4%. The bolus volume required to achieve the desired osmolality varies from 30 ml for 23.4% saline to 150 ml for 3%. It is interesting to note that nearly all the randomized controlled trials with hypertonic saline used it as intermittent boluses in response to raised ICP. Many of us use continuous infusions of hypertonic saline, often without ICP monitoring. It remains unclear whether continuous infusions are as effective in reducing ICP compared to bolus doses. This is particularly relevant when the intracranial pressure is not monitored. A sodium level of 145–150 mmol/l is often recommended when hypertonic saline is used; however, this target level may be little more than just a ballpark number. The brain trauma foundation (BTF) continues to recommend mannitol for ICP reduction, considering the lack of firm evidence to support the use of hypertonic saline.8

Coming back to our patient, our practice until a couple of years ago would have been to use hypertonic saline, as a combination of boluses and a continuous infusion targeting a sodium level of around 150 mmol/l. The turnaround time to obtain sodium levels and diligent titration to reasonably precise levels often turned out to be time and labor-intensive. Several recent studies suggest that the difference in ICP reduction is at best, modest, and clinical outcomes including mortality are unchanged with the use of hypertonic saline, compared to mannitol. Hence, we have gone back to our previous practice of using mannitol, unless there is a concern with inducing hemodynamic instability or renal dysfunction. In spite of the many theoretic advantages, hypertonic saline may cause significant volume overload in patients with poor heart function; besides, the possibility of central pontine demyelination cannot be entirely ruled out with a rapid increase in sodium levels.

At the end of the day, osmotic therapy is at best, only intended to buy time before a definitive intervention is carried out, as in our gentleman with the acute subdural hematoma. A transient, modest difference in ICP, even if it may assume statistical significance, may not translate to a perceptible change in clinical outcomes. What would be your choice of osmotic agent in the case we discussed?


  1. Weed LH, Mckibben PS. Pressure changes in the cerebrospinal fluid following intravenous injection of solutions of various concentrations. Am J Physiol 48, 512–30 (1919).
  2. Relationship between excitatory amino acid release and outcome after severe human head injury. – PubMed – NCBI. Available at: (Accessed: 15th December 2018)
  3. Ichai, C. et al.Sodium lactate versus mannitol in the treatment of intracranial hypertensive episodes in severe traumatic brain-injured patients. Intensive Care Med. 35, 471–479 (2009).
  4. Battison, C., Andrews, P. J. D., Graham, C. & Petty, T. Randomized, controlled trial on the effect of a 20% mannitol solution and a 7.5% saline/6% dextran solution on increased intracranial pressure after brain injury. Crit. Care Med. 33, 196–202; discussion 257-258 (2005).
  5. Cottenceau, V. et al.Comparison of effects of equiosmolar doses of mannitol and hypertonic saline on cerebral blood flow and metabolism in traumatic brain injury. J. Neurotrauma 28, 2003–2012 (2011).
  6. Sakellaridis, N. et al.Comparison of mannitol and hypertonic saline in the treatment of severe brain injuries. J. Neurosurg. 114, 545–548 (2011).
  7. Berger-Pelleiter, E., Émond, M., Lauzier, F., Shields, J.-F. & Turgeon, A. F. Hypertonic saline in severe traumatic brain injury: a systematic review and meta-analysis of randomized controlled trials. CJEM 18, 112–120 (2016).
  8. Carney, N. et al.Guidelines for the Management of Severe Traumatic Brain Injury. 4th edition 244 (2016)


Lactate in Sepsis: Much Maligned, but Not Quite the Evil Devil!

In the 2018 iteration of the Surviving Sepsis Guidelines, a 1-h bundle is recommended for expeditious resuscitation and management of severe sepsis. Serial lactate measurements are advocated to guide resuscitation with the aim to normalize levels. High lactate levels are considered to indicate tissue hypoperfusion in sepsis.1

Glucose metabolism

Aerobic pathway

Let us consider how lactate is generated. In the presence of an adequate supply of oxygen, aerobic metabolism occurs. Glucose is first converted to pyruvate, yielding two molecules of ATP. The pyruvate is then converted by pyruvate dehydrogenase to acetyl CoA (Fig. 1, red arrow). The acetyl CoA enters the tricarboxylic acid cycle (TCA) resulting in the formation of 36 molecules of ATP. If there is a lack of oxygen, the anaerobic pathway takes over. In this pathway, the first step, namely, the conversion of glucose to pyruvate occurs as previously, because oxygen is not required for this process. However, the next step, namely, the conversion of pyruvate to acetyl CoA, and subsequent entry into the TCA cannot happen in the absence of oxygen.


Fig. 1 Glucose metabolism. Red arrow: aerobic pathway through the TCA cycle. Green arrow: Due to enhanced aerobic glycolysis,  pyruvate accumulation occurs as the TCA pathway becomes saturated. The pyruvate is converted to lactate by lactate dehydrogenase 

Anaerobic pathway

What happens to the pyruvate that accumulates consequent to the failure of its conversion to acetyl CoA due to a lack of oxygen? Another enzyme, lactate dehydrogenase, takes over and converts the pyruvate to lactate (Fig. 1, green arrow). The lactate is then transported to the liver, where it is utilized in three different ways (Fig. 2): (1) Conversion to pyruvate by the same enzyme, lactate dehydrogenase, by a reverse process. The pyruvate generated by this process is converted to glucose by gluconeogenesis. The glucose thus produced goes back into circulation. The conversion of two molecules of pyruvate to one molecule of glucose consumes six molecules of ATP and hence is an energy expending process. (2) Conversion to glycogen, which may be converted to glucose for later use, and (3) Conversion to pyruvate, which re-enters the TCA under aerobic conditions to generate ATP. The excess lactate is also transported to the myocardium, where it is converted to pyruvate and enters the TCA cycle to yield ATP for the myocardial cells. Metabolism in the myocardial cells is always aerobic.


Fig. 2 The lactate generated from pyruvate (green arrow, Fig. 1) is transported to the liver. It may be converted to glucose (gluconeogenesis),  to glycogen, or to pyruvate in the liver  

Lactate levels in sepsis

Why do lactate levels rise in sepsis? Is it because of reduced tissue oxygen delivery, and anaerobic metabolism of glucose? Tissue hypoxia due to impaired oxygen delivery is highly unlikely is sepsis. In most cases, there is preserved or increased cardiac output with adequate oxygen delivery. Lactate levels rise in sepsis, through entirely different mechanisms that are unrelated to tissue hypoxia. In sepsis, there is excessive stimulation of beta-2 receptors. This leads to accelerated aerobic metabolism of glucose. Increased aerobic metabolism of glucose leads to high levels of pyruvate. However, the pyruvate dehydrogenase pathway through to the TCA cycle gets overwhelmed due to the excessively high pyruvate levels. Consequently, the excess pyruvate gets converted to lactate by the lactate dehydrogenase enzyme. Hence, with an increase in aerobic metabolism of glucose, the lactate levels rise hand in hand. However, lactate production through this mechanism is not due to tissue hypoxia; it is entirely due to excessive aerobic metabolism of glucose, stimulated by beta 2 receptor-mediated release of epinephrine, a characteristic feature of sepsis. In fact, failure of lactate levels to rise in response to epinephrine infusion in shocked patients has been associated with increased mortality.Other mechanisms of lactate generation are also seen in sepsis, including excessive production in the lung due to ARDS and hepatic dysfunction leading to reduced utilization of lactate.

It has been clearly established that high lactate levels indicate more severe illness. Furthermore, increased lactate levels predict mortality.Hyperlactatemia without hypotension has been associated with significantly higher mortality compared to patients with hypotension and normal lactate levels.Be that as it may, will targeting resuscitative interventions with the lactate level as a therapeutic goal be appropriate? First, if the rise in lactate levels is unlikely to be due to oxygen debt, it would seem illogical to attempt to increase oxygen delivery with an eye on the lactate level. If a patient has improving blood pressures, urine output, and acid-base status, it defies logic to assume that continued measures aimed specifically at lactate levels would be beneficial.

The use of the term “lactate clearance” may also be flawed. Clearance is the volume of plasma from which a substance is completely removed per unit time. In a shocked patient, the lactate levels are more likely to fall due to reduced production; not from removal by metabolism. A falling lactate level is definitely more likely to be associated with survival; however, the rate of fall in lactate levels that may be associated with a better outcome is hard to define.5

The bottom line

  • Lactate levels rise in sepsis mainly because of catecholamine-mediated enhanced aerobic glycolysis.
  • Oxygen debt is uncommon in septic shock; hence “bundles” that target increased oxygen delivery are unlikely to help.
  • Lactate is an important metabolic fuel in shock. It is a substrate for the production of glucose and glycogen in the liver, and a source of aerobic generation of ATP for the myocardium.
  • A high lactate level is a strong predictor of mortality.
  • A fall in lactate levels carries a good prognosis; however, resuscitative interventions that target lactate levels are unlikely to be beneficial.
  • It is important not to view lactate levels in isolation; a holistic approach that addresses all the vital parameters that govern the circulatory status is essential.



  1. Levy, M. M., Evans, L. E. & Rhodes, A. The Surviving Sepsis Campaign Bundle: 2018 update. Intensive Care Med.44, 925–928 (2018).
  2. Wutrich, Y. et al.Early increase in arterial lactate concentration under epinephrine infusion is associated with a better prognosis during shock. Shock Augusta Ga34, 4–9 (2010).
  3. Serum lactate as a predictor of mortality in patients with infection. – PubMed – NCBI. Available at: (Accessed: 12th December 2018)
  4. Mortality is Greater in Septic Patients With Hyperlactatemia Than With Refractory Hypotension. – PubMed – NCBI. Available at: (Accessed: 11th December 2018)
  5. Vincent, J.-L., Quintairos e Silva, A., Couto, L. & Taccone, F. S. The value of blood lactate kinetics in critically ill patients: a systematic review. Crit. Care20, (2016).







Noninvasive Ventilation in Severe Community-Acquired Pneumonia​: To Do or Not to Do, That Is the Question!


Invasive mechanical ventilation may be complicated by ventilator-associated lung injury, ventilator-associated pneumonia, the need for sedation and muscle paralysis, and the possibility of airway-related problems. Noninvasive ventilation (NIV) is widely used by clinicians in community-acquired pneumonia (CAP), in the hope of avoiding intubation thereby improving clinical outcomes. Although the efficacy of NIV in patients with chronic obstructive airways disease (COPD) with pneumonia is reasonably well established, the evidence for its use in non-COPD patients is less clear. We are currently faced with an epidemic of viral pneumonia in many parts of India; would it be appropriate to offer a trial of NIV to some of these patients?

In an early observational study of patients with CAP and acute respiratory failure, a high intubation rate (66%) was observed, although there was an initial improvement in the PaO2/FiO2ratios.A large, registry-based cohort study was carried out during the 2009 H1N1 pandemic from 35 intensive care units in Argentina. NIV use carried a survival of 43/64 patients (67%) in this study, significantly higher than those who underwent invasive ventilation.The authors speculated that the improved survival observed with NIV may have been due to a lower severity of illness at baseline. In our series of 31 patients during the 2009 H1N1 pandemic, noninvasive ventilation was initially carried out in eight patients; four patients failed and subsequently underwent invasive ventilation.Several early studies suggest worse outcomes with NIV in acute hypoxemic respiratory failure, including CAP.More recent studies, some using a helmet device, suggest benefit with NIV use.5,6 Elderly, immunocompromised patients with CAP were shown to have a better 90-d survival with NIV use in a retrospective cohort study.7

Against the background of this fairly dodgy evidence, where do we stand in relation to NIV use in severe CAP with acute hypoxemic respiratory failure? Unfortunately, there are no robust, controlled studies that can guide us in deciding whether to resort to NIV as an initial line of therapy. Should we give a trial of NIV or just carry on with intubation and invasive ventilation in patients who present with CAP and acute hypoxemic respiratory failure? Are there reliable predictors of NIV failure that can help decision making?

In a prospective observational study of 64 patients with CAP, NIV was successful in 28 (44%). On multivariate analysis, a decrease in the oxygenation index (mean airway pressure × FIO2 × 100/PaO2) by 1.2 or more and an increase in the PaO2/FiO2ratio by 42.2 or more predicted NIV success. Furthermore, on ROC curve analysis, a pH of more than 7.38 and a respiratory rate of less than 27/min were also predictive of successful NIV use.In another prospective study of 184 patients with CAP, NIV failed more often in patients without previous cardiac or respiratory disease compared to those who had comorbidities. Radiological worsening and higher SOFA scores also predicted NIV failure. After an NIV trial of 1 h, an increase in heart rate, a fall in the PaO2/FiO2ratio, and lower bicarbonate levels were independent predictors of failure.In another retrospective cohort study of 209 patients, NIV failure occurred in 90 of 117 patients (77%); patients who failed NIV had a significantly higher mortality compared to those who succeeded. On multivariate analysis, higher APACHE scores and the requirement for vasopressors at 2 h were associated with NIV failure.10

I am sure you would agree that the situation is far from clear-cut. We have been mortified by lack of success with NIV in a handful of patients in our unit during this flu season. With the limited observational evidence we have, I would highlight the following points to ponder on before embarking upon an NIV trial in patients with CAP and acute hypoxemic respiratory failure.

  1. It is crucial to identify patients who are likely to benefit from NIV.
  2. Patients who have no significant comorbidities are more likely to failNIV (counterintuitive as it may seem).
  3. The higher the baseline severity of illness, the lower the likelihood of NIV success.
  4. Failure to improve physiological parameters, including the respiratory rate, P/F ratio, oxygenation index, and bicarbonate levels within the first few hours is likely to result in failure.
  5. Radiological worsening in the first 24 hours calls for invasive ventilation.
  6. It is important to recognize early signs of failure and resort to invasive ventilation expeditiously.


Please feel free to offer your valuable comments and input.


  1. Jolliet, P., Abajo, B., Pasquina, P. & Chevrolet, J. C. Non-invasive pressure support ventilation in severe community-acquired pneumonia. Intensive Care Med.27, 812–821 (2001).
  2. Estenssoro, E. et al.Pandemic 2009 influenza A in Argentina: a study of 337 patients on mechanical ventilation. Am. J. Respir. Crit. Care Med.182, 41–48 (2010).
  3. Chacko, J., Gagan, B., Ashok, E., Radha, M. & Hemanth, H. V. Critically ill patients with 2009 H1N1 infection in an Indian ICU. Indian J. Crit. Care Med. Peer-Rev. Off. Publ. Indian Soc. Crit. Care Med.14, 77–82 (2010).
  4. Honrubia, T. et al.Noninvasive vs conventional mechanical ventilation in acute respiratory failure: a multicenter, randomized controlled trial. Chest128, 3916–3924 (2005).
  5. Cosentini, R. et al.Helmet continuous positive airway pressure vs oxygen therapy to improve oxygenation in community-acquired pneumonia: a randomized, controlled trial. Chest138, 114–120 (2010).
  6. Brambilla, A. M. et al.Non-invasive positive pressure ventilation in pneumonia outside Intensive Care Unit: An Italian multicenter observational study. Eur. J. Intern. Med.(2018). doi:10.1016/j.ejim.2018.09.025
  7. Johnson, C. S., Frei, C. R., Metersky, M. L., Anzueto, A. R. & Mortensen, E. M. Non-invasive mechanical ventilation and mortality in elderly immunocompromised patients hospitalized with pneumonia: a retrospective cohort study. BMC Pulm. Med.14, (2014).
  8. Carron, M., Freo, U., Zorzi, M. & Ori, C. Predictors of failure of noninvasive ventilation in patients with severe community-acquired pneumonia. J. Crit. Care25, 540.e9–14 (2010).
  9. Carrillo, A. et al.Non-invasive ventilation in community-acquired pneumonia and severe acute respiratory failure. Intensive Care Med.38, 458–466 (2012).
  10. Murad, A., Li, P. Z., Dial, S. & Shahin, J. The role of noninvasive positive pressure ventilation in community-acquired pneumonia. J. Crit. Care30, 49–54 (2015).




Is normal saline really worth its salt​?


The quest for suitable intravenous fluids began with the cholera pandemic of the 1830s. The deadly disease was characterized by repeated evacuation of large volumes of a rice-water-like fluid leading to severe dehydration among those afflicted. It spread from India to South East Asia and the Middle East, then towards Russia and the rest of Europe killing thousands of people along the way. Latta, a Scottish physician, concerned by the loss of huge volumes of fluid from the body, suggested replenishment with a “salt” solution. This solution was administered rectally first, and later, intravenously, with a profound effect on those affected. Several decades later, a Dutch scientist, Hamburger, suggested that the concentration of salt in the human body was 0.9%. He proposed that an identical concentration of salt would be “normal” and physiologically appropriate. From this early, humble beginning, “normal” saline percolated down generations of physicians and became deeply entrenched into contemporary clinical practice. Yes, normal saline has been, by far, the most commonly prescribed intravenous fluid for nearly two centuries.

Is normal saline really physiological as Hamburger believed?

The sodium and chloride levels of normal saline are markedly different from serum levels. Normal saline is acidic (pH: 5.4) and contains 154 mmol/L each of sodium and chloride. Compared to serum levels, the Na+level in normal saline is approximately 10% higher, and even more importantly, the chloride level is almost 50% higher.

Does normal saline cause acidosis? According to Stewart’s hypothesis, the pH of any fluid is dependent only on three factors, including the PCOlevel, the content of non-volatile acid (albumin and phosphate), and most importantly, the strong ion difference (SID). SID is the difference between the fully dissociated cations and anions. Thus, SID = Strong cations (Na+, K+, Ca++, and Mg++) – Strong anions (Cl, lactate, ketoacids, and organic anions). The normal SID is 40 mmol/L; a lower SID which may occur with a relative increase in the strong anions, leading to a lower pH (acidosis). The pH increases when the SID is higher (alkalosis). When normal saline is infused, both Na+and the Cllevels rise. However, the increase in Clis much higher than the increase in Na+. This is because the Cllevel in normal saline is almost 50% higher than the serum level, compared to the minimal difference in Na+levels. Hence, the infusion of normal saline results in a relative increase in the Cllevels with a reduction in the SID, leading to acidosis.

Normal saline and the kidneys

The excessive Clcontent of the fluid filtered through the glomeruli is sensed by the macula densa, a group of specialized cells located at the junction of the ascending limb of the loop of Henle and the distal convoluted tubule. The Clsignal is transmitted to the afferent arteriole, which is in close contact with the macula densa. This leads to vasoconstriction of the afferent arteriole, a drop in the perfusion pressure in the glomeruli, and reduced glomerular filtration (Fig. 1).


Fig. 1: The macula densa senses a high concentration of chloride in the filtered fluid and signals vasoconstriction in the afferent arteriole, thus reducing the glomerular filtration rate (From: Li, Heng, Shi-ren Sun, John Q. Yap, Jiang-hua Chen, and Qi Qian. 2016. 0.9% Saline Is Neither Normal nor Physiological. Journal of Zhejiang University-SCIENCE B 17(3): 181–187)

Normal saline and potassium levels

Many clinicians believe that it may be safer to infuse normal saline compared to Ringer Lactate in patients with high potassium levels. This is based on the assumption that, Ringer Lactate, with a K+ concentration of 4.0 mmol/L, may exacerbate hyperkalemia. However, this assumption is largely incorrect; the acidosis arising from normal saline infusion leads to a shift of Kfrom the intracellular to the extracellular compartment with an increase in the serum K+ levels. Several clinical studies have confirmed higher potassium levels with the infusion of normal saline compared to Ringer’s lactate (Weinberg et al. 2017).

Does the use of normal saline as a resuscitation fluid lead to clinical harm?

It is somewhat bewildering that the ubiquitous use of normal saline remained unquestioned for more than a century. Beginning from the 1980s, concerns were raised regarding the high Clcontent in normal saline and the possibility of consequent renal injury. Clinical investigation into possible harm commenced much later, with several observational studies suggesting a detrimental effect (Raghunathan et al. 2015; Shaw et al. 2012).

In a sequential pilot study that compared 6 months each of a chloride-liberal vs. chloride-restricted intravenous fluid, the incidence of acute kidney injury (AKI) and the need for renal replacement therapy were significantly lower during the chloride-restricted period (Yunos et al. 2012). The SPLIT study was a double-blind, cluster-randomized, crossover study that compared 0.9% saline with Plasma-Lyte 148 (Young et al. 2015). There was no significant difference in the incidence of AKI, the need for renal replacement therapy, and in-hospital mortality between groups. However, this study was aimed primarily to evaluate the feasibility and assess sample size for future studies. The study was meant to be conducted over a specific period of time with no fixed sample size. No power calculation was made considering the lack of previous randomized controlled trials.

Two large randomized controlled trials were conducted at the Vanderbilt University Medical Centre comparing normal saline Vs balanced crystalloids (Ringer’s solution or Plasma Lyte A) and published earlier this year (Self et al. 2018; Semler et al. 2018). The SMART study was carried out in five intensive care units. Patients were randomly assigned to receive one of the two types of fluid on alternate months. The primary composite outcome was one or more major kidney events during 30 days (MAKE-30) of follow-up. The MAKE-30 criteria included mortality, the requirement for renal replacement therapy, and a rise in creatinine to twice the baseline or more in 30 days. The composite outcome was significantly less with balanced crystalloids compared to normal saline (14.3 vs 15.4%; p = 0.04). The SALT-ED study was conducted on non-critically ill patients presenting to the emergency department and admitted to the wards. Using a similar design, this study also revealed a more favorable composite outcome (4.7 vs. 5.6%; p = 0.01). Subgroup analysis suggested that the impact of using balanced crystalloids was more pronounced in patients who received larger volumes of fluid and those with sepsis.

The question of whether balanced crystalloids is preferable to normal saline may be far from unequivocally answered as yet. However, we know today that the centuries-old practice of unrestricted use of normal saline may lead to unfavorable clinical outcomes. Balanced crystalloids, including Ringer’s lactate and Plasma Lyte, offer possible safer alternatives, especially if large volume resuscitation is required. It is also pertinent to point out that a perceptible outcome difference based on the choice of fluid may be largely confined to the most severely ill patients. An 8800-patient randomized controlled trial comparing normal saline vs. Plasma Lyte 148 (the PLUS study) is currently in progress across multiple ICUs in Australia and New Zealand and may add substantially to our body of knowledge. In our practice, we confine to Ringer Lactate for most of our patients who need fluid resuscitation. The findings of SMART and SALT-ED have only strengthened our bias.


Raghunathan, Karthik, Anthony Bonavia, Brian H. Nathanson, et al. 2015.Association between Initial Fluid Choice and Subsequent In-Hospital Mortality during the Resuscitation of Adults with Septic Shock. Anesthesiology 123(6): 1385–1393.

Self, Wesley H., Matthew W. Semler, Jonathan P. Wanderer, et al. 2018.   Balanced Crystalloids versus Saline in Noncritically Ill Adults. The New England Journal of Medicine 378(9): 819–828.

Semler, Matthew W., Wesley H. Self, Jonathan P. Wanderer, et al. 2018.   Balanced Crystalloids versus Saline in Critically Ill Adults. The New England Journal of Medicine 378(9): 829–839.

Shaw, Andrew D., Sean M. Bagshaw, Stuart L. Goldstein, et al. 2012.        Major Complications, Mortality, and Resource Utilization after Open Abdominal Surgery: 0.9% Saline Compared to Plasma-Lyte. Annals of Surgery 255(5): 821–829.

Weinberg, L., L. Harris, R. Bellomo, et al. 2017.      Effects of Intraoperative and Early Postoperative Normal Saline or Plasma-Lyte 148® on Hyperkalaemia in Deceased Donor Renal Transplantation: A Double-Blind Randomized Trial. British Journal of Anaesthesia 119(4): 606–615.

Young, Paul, Michael Bailey, Richard Beasley, et al. 2015.Effect of a Buffered Crystalloid Solution vs Saline on Acute Kidney Injury Among Patients in the Intensive Care Unit: The SPLIT Randomized Clinical Trial. JAMA 314(16): 1701–1710.

Yunos, Nor’azim Mohd, Rinaldo Bellomo, Colin Hegarty, et al. 2012.        Association between a Chloride-Liberal vs Chloride-Restrictive Intravenous Fluid Administration Strategy and Kidney Injury in Critically Ill Adults. JAMA 308(15): 1566–1572.





The hypoxic drive – an urban legend?

It is not unusual to see physicians painstakingly titrate oxygen flows with elaborate precision, especially in CO2 retaining patients with chronic obstructive pulmonary disease (COPD). I have occasionally come across novice ICU trainees cease oxygen therapy in dyspneic patients ostensibly to stimulate the hypoxic drive. The driving principle behind this line of thinking is the tradition-borne belief that supplemental oxygen may inhibit the hypoxic drive and lead to hypoventilation and rise in PCO2 levels. How important is the contribution of the hypoxic drive in patients with respiratory failure, especially among those with hypercapnia? Does supplemental oxygen really suppress ventilation in spontaneously breathing patients with COPD and lead to a rise in CO2 levels?

Aubier et al. studied 22 spontaneously breathing patients with acute exacerbation of COPD.They measured minute ventilation and performed arterial blood gases analysis, initially on room air, and subsequently while breathing pure oxygen through a Douglas bag. The minute ventilation dropped in the first few minutes but recovered rapidly to near baseline levels at approximately 10 minutes after oxygen administration. However, the CO2 levels continued to rise (Fig. 1). Clearly, hypoventilation was not the mechanism behind the rise in PCOin these patients.


Fig. 1 Minute ventilation (VE) decreased transiently after oxygen administration but recovered to baseline levels. The PCO2 levels continued to rise even after ventilation had recovered (From Abdo WF, Heunks LMA: Oxygen-induced hypercapnia in COPD: myths and facts. Critical Care 2012; 16:323)

Hypoxic pulmonary vasoconstriction and effect on PCO2levels

If oxygen therapy does not cause hypoventilation, what may be the mechanism behind the rise in CO2 levels? In chronic lung disease, blood flow through the lung is redistributed by hypoxia-induced vasoconstriction of the pulmonary vasculature. Blood vessels in regions of the lung that are poorly ventilated become constricted to match perfusion to ventilation and prevent shunting (flow of blood through the lungs without gas exchange). Supplemental oxygen results in an increase in the oxygen level within alveoli that are poorly ventilated with the abolition of hypoxia-induced vasoconstriction. This leads to a redistribution of blood flow from the relatively better ventilated to less ventilated regions of the lung. In effect, this results in an increase in perfusion relative to ventilation (low V/Q) in under-ventilated areas of the lung while the reverse occurs in the relatively well-ventilated areas of the lung (increase in alveolar dead space). Both these mechanisms may contribute to the rise in the alveolar and arterial COlevels.

The Haldane effect

About 10% of CO2 is carried in the blood as carbamino compounds, especially in combination with hemoglobin, as carbaminohemoglobin. Hemoglobin combines more avidly with CO2 when it is in the deoxygenated form. Thus, in the lung, as the hemoglobin takes up oxygen, it gives up the COby the Haldane effect. The CO2 is normally flushed out by ventilation. However, in a damaged lung, when the ventilation is inadequate, CO2 cannot be washed out adequately. The extra CO2 released from hemoglobin dissolves in the plasma with a rise in the PCO2 levels.

The bottom line

It is abundantly clear that the rise in CO2 levels seen in COPD patients with oxygen therapy is unrelated to the abolition of the “hypoxic drive” as was conventionally believed. The inhibition of ventilation is transient and insignificant; the PCO2 levels rise through other mechanisms. Theories and hypotheses apart, the bottom line is that supplemental oxygen should never be denied to hypoxic patients. The chances of death due to uncorrected hypoxia is overwhelmingly higher than possible harm from the administration of supplemental oxygen. Needless to add, most patients, regardless of their COlevels, do not need an oxygen saturation of more than 92%. Increasing the FiO2 aiming for much higher levels of oxygenation is not appropriate under most circumstances; less is more, so sayeth the wise intensivist!


  1. Aubier M, Murciano D, Milic-Emili J, Touaty E, Daghfous J, Pariente R, et al. Effects of the administration of O2 on ventilation and blood gases in patients with chronic obstructive pulmonary disease during acute respiratory failure. Am Rev Respir Dis 1980; 122:747-754.




Peripheral Venous Cannulation Under Ultrasonographic Guidance

Most of us routinely insert central venous catheters under real-time ultrasound guidance. The technique is time-tested, and there is robust evidence that it is safer and more reliable compared to the landmark-based approach. However, peripheral venous access can, at times, be even more challenging in critically ill patients. Quite often, access may be difficult due to thrombophlebitis or edema. It is not unusual for poorly experienced operators to struggle with the insertion of peripheral IV lines causing avoidable pain and discomfort to the patient. Would ultrasonography enable visualization of veins that are invisible to the naked eye and enable ease of insertion?

The importance of using an effective tourniquet cannot be overemphasized for ultrasound-guided cannulation of peripheral veins. Many monitors have a built-in, sustained, cuff inflation mechanism on the NIBP module to enable venipuncture. The other option is to inflate a manual blood pressure cuff to slightly lower than the systolic blood pressure. A high-frequency linear probe is most suited to visualize peripheral veins. The initial scan may be on the short axis view just distal to the cubital fossa, where the cannula will not get bent or displaced with arm movement. If veins are not readily visible at this location, move the probe distally along the forearm.

Once you see a vessel of reasonable size, you need to ensure that it is, in fact, a vein. This may occasionally be more difficult than it may seem. If the inflation pressure is too high,  arteries may be compressed on minimal probe pressure, making the distinction difficult. Pulse wave Doppler helps to distinguish between arteries and veins but may not be absolutely confirmatory if the inflation pressure is set close to the systolic pressure. However, a good way to differentiate is using what I call the “deflation test.” When the cuff is deflated, the artery becomes more prominent, while the vein collapses.

I would strongly recommend cannulation in the long axis view, using an in-plane technique. This enables precise guidance; besides, it is possible to visualize the cannula from the site of puncture all along the soft tissue through to the vein. The stylet of the cannula projects beyond the plastic tip by about a millimeter. With the in-plane technique, it is possible to ensure that the cannula is also within the vein (not just the stylet) and guide it along under vision, without transfixing the vein. This prevents the possibility of the stylet being inside the vein while the cannula lies outside.

The first step is to search for a suitable vein in the short axis view (Fig 1).


Fig 1. Short axis view. A: Artery; V: Vein

Once you identify an optimally sized vein, the probe is turned around to obtain a long axis view of the target vein (Fig 2). A sufficiently good length of the vein, without tortuosity, should be visible to enable ease of insertion.


Fig 2. Long axis view of the vein

In practice, the cannula is inserted just distal to the probe in the long axis to enable an in-plane technique (Fig 3).insertion.jpg

Fig 3. The direction of insertion of the cannula on the long axis view

Puncture the skin and look for movement on the ultrasonographic image; pass the cannula a millimeter at a time and follow it right from under the skin through to the vein (Fig 4).


Fig 4. The cannula (arrow) is seen within the vein on the long axis

Do ensure that you do not lose sight of the cannula at any time. At times, the cannula may seem to be in the vein, when it actually lies outside. This can be prevented by ensuring indentation of the wall of the vein as it makes contact.

How do you confirm that the cannula is in the vein? You could view the cannula inside the lumen on the long axis. However, the definitive method is by performing a bubble test with 10-20 ml of saline (Fig 5).


Fig 5. The bubble test. Opacification is seen within the right ventricle (RV)

The bubble test is carried out by rapid injection of 20-30 ml of saline into the IV line while watching for echocardiographic opacification of the right atrium and the ventricle. Do not forget to record the clip; it helps to review the loop if the opacification is not clear-cut on the initial view.



Is Doppler-based calculation of pulmonary artery pressure valid in critically ill patients on mechanical ventilation?

Pulmonary artery pressure (PAP) is an important parameter in mechanically ventilated patients. In cardiology practice, the pulmonary artery systolic pressure (PASP) is measured by transthoracic echocardiography by continuous wave Doppler interrogation of the tricuspid regurgitation (TR) jet. The measurement is based on the following equations:

Tricuspid pressure gradient (Right ventricular systolic pressure – right atrial pressure) =4V(where V = maximal velocity of the TR jet)

Right ventricular systolic pressure (RVSP) = Tricuspid gradient + CVP. The RVSP is assumed to be identical to the pulmonary artery systolic pressure (PASP).

If the CVP is not measured, conventionally, 10 mm Hg is added to the tricuspid gradient to obtain the PASP. This method of measurement of PASP has been validated in spontaneously breathing patients in cardiology practice. However, the question remains, would it be valid in patients on positive pressure ventilation and PEEP? The evidence so far has been inconclusive.

In a single-center study from France, published online first in the Critical Care Medicine, the investigators assessed the reliability of Doppler-derived PAP measurements compared to invasive measurements using a pulmonary artery catheter in mechanically ventilated patients (Mercado et al., 2018). The PASP was calculated by two formulas (1) PASP = tricuspid pressure gradient + CVP and (2) PASP = tricuspid pressure gradient + 10 mm Hg. The mean PAP was calculated (1) using the isovolemic relaxation time and (2) using the Chemla equation: 0.61 × PASP + 2. The Doppler-based calculations were compared with direct measurements from a pulmonary artery catheter.

The PASP calculated by adding the CVP to the tricuspid pressure gradient correlated best with invasive PASP measured using a pulmonary artery catheter, with a Spearman correlation coefficient of 0.87. When pulmonary artery hypertension was defined as a mean PAP more than 25 mm Hg, a PASP of more than 39 mm Hg, derived using this method, revealed 100% sensitivity and specificity. The mean PAP calculated using the Chemla equation showed a similar correlation (0.87) with invasive measurements. A cut-off value of 26 mm Hg revealed a sensitivity and specificity of 100% for the diagnosis of pulmonary hypertension, defined as mean PAP > 25 mm Hg by invasive measurement. When the PASP was calculated by adding 10 mm Hg to the tricuspid pressure gradient, the correlation coefficient was 0.79 on comparison with invasive measurements from a pulmonary artery catheter.

This study validates TR jet-based pulmonary artery pressure measurements in critically ill patients, who are on mechanical ventilation. The most precise method, with the best correlation, was by addition of the CVP to the tricuspid gradient. However, if the central venous pressure is not available, adding 10 mm Hg to the tricuspid gradient would be a close approximation. This study also showed that contrary to widely held belief, the maximal or minimal diameter of the inferior vena cava nor the variation in diameter bore no correlation with invasively measured CVP.

This study firmly establishes the precision of continuous-wave Doppler-derived assessment of pulmonary artery pressures in mechanically ventilated patients in the intensive care unit. However, a tricuspid regurgitation jet may not be evident in some patients to enable measurement of the gradient. In the present study, the authors report that a measurable jet was present in 60% of patients. Besides, an optimal 4-chamber view, with proper alignment of the Doppler cursor with the axis of the regurgitant flow may not be possible to obtain in some patients.


Mercado, P., Maizel, J., Beyls, C., Kontar, L., Orde, S., Huang, S., Slama, M. (2018). Reassessment of the Accuracy of Cardiac Doppler Pulmonary Artery Pressure Measurements in Ventilated ICU Patients: A Simultaneous Doppler-Catheterization Study. Critical Care Medicine, Online First.





Do Procalcitonin Levels Help Guide Antibiotic Therapy in Acute Exacerbation of Chronic Obstructive Airway Disease?


There has been a growing interest regarding the utility of procalcitonin to guide appropriate initiation and duration of antibiotic therapy in critically ill patients. Two randomized controlled studies in critically ill patients suspected to have bacterial infections arrived at disparate conclusions (De Jong et al. 2016; Bouadma et al. 2010).

In patients presenting with acute exacerbation of chronic obstructive pulmonary disease (COPD), it is often difficult to clinically ascertain whether bacterial infection is the precipitating factor. Previous studies that evaluated the utility of procalcitonin in this setting involved less severely ill patients, with relatively few who required treatment in the intensive care unit. Against this background, Daubin et al. carried out a multicentre randomized controlled study in the intensive care unit of 11 hospitals in France to investigate the utility of procalcitonin-guided antibiotic therapy in acute exacerbations of COPD (Daubin et al. 2018).

The authors hypothesized that (1) procalcitonin-guided treatment would reduce antibiotic exposure and (2) mortality will not be different with a lower antibiotic exposure. To demonstrate non-inferiority, a cut off of 12% was considered as the excess mortality permissible with procalcitonin guidance.

Procalcitonin levels were measured in all patients at inclusion, 6 hours later, and on days 1, 3, and 5 after inclusion. In the intervention group, the initiation and duration of antibiotic therapy followed an algorithm based on procalcitonin levels. In the control group, antibiotic therapy was commenced and continued for an appropriate duration based on clinician judgment.

The primary endpoint was mortality at 3 months. At the end of this period, 30 patients (20%) died in the procalcitonin arm compared to 21 (14%) in the control arm, with a confidence interval of −0.3 to 13.5 %. As the upper limit of the confidence interval exceeded the pre-determined cut off of 12% mortality, the study failed to establish non-inferiority of procalcitonin-guided therapy. Mortality with procalcitonin-guided therapy was significantly higher in patients who were not on antibiotics at inclusion. Besides, there was no difference in secondary endpoints including the requirement for vasopressor or dialytic support, the incidence of acute respiratory distress syndrome, ICU-acquired pneumonia or other infections, and multiorgan failure. The duration ventilation, ICU and hospital stay were also not significantly different between groups.

In contrast to previous studies in acute exacerbation of COPD that showed reduced antibiotic usage with procalcitonin guidance, the present study included patients with a relatively higher severity of illness; 87% of patients required non-invasive or invasive mechanical ventilation. Vasopressor support was required in 17.2% of patients, while 5.6% of patients underwent dialysis. The number of patients who received antibiotics was significantly less in the procalcitonin group throughout the first 6 days of the study. This may suggest that procalcitonin levels may have been falsely low and failed to identify infection in some patients.

There were more patients on home oxygen and non-invasive ventilation in the procalcitonin group, which may indicate a more severe illness that may have contributed to mortality. There is no data available on the time to commencement of antibiotics in either group, which may also have affected outcomes.

The bottom line is that in critically ill patients with acute exacerbation of COPD, procalcitonin guidance may fail to identify patients who may have bacterial infection as the precipitating cause. Early antibiotics based on clinical judgment may be more appropriate in such patients. I must confess that I am not a procalcitonin fan; the results of this study is no surprise to me. Our practice of initiation and continuance of antibiotic treatment based on clinician judgment shall remain unchanged.


  1. Bouadma, Lila, Charles-Edouard Luyt, Florence Tubach, Christophe Cracco, Antonio Alvarez, Carole Schwebel, Frédérique Schortgen, et al. 2010. “Use of Procalcitonin to Reduce Patients’ Exposure to Antibiotics in Intensive Care Units (PRORATA Trial): A Multicentre Randomised Controlled Trial.” Lancet (London, England)375 (9713): 463–74.
  2. Daubin, Cédric, Xavier Valette, Fabrice Thiollière, Jean-Paul Mira, Pascal Hazera, Djillali Annane, Vincent Labbe, et al. 2018. “Procalcitonin Algorithm to Guide Initial Antibiotic Therapy in Acute Exacerbations of COPD Admitted to the ICU: A Randomized Multicenter Study.” Intensive Care Medicine44 (4): 428–37.
  3. Jong, Evelien de, Jos A. van Oers, Albertus Beishuizen, Piet Vos, Wytze J. Vermeijden, Lenneke E. Haas, Bert G. Loef, et al. 2016. “Efficacy and Safety of Procalcitonin Guidance in Reducing the Duration of Antibiotic Treatment in Critically Ill Patients: A Randomised, Controlled, Open-Label Trial.” The Lancet. Infectious Diseases16 (7): 819–27.





Journal Critique

Effect of Thiamine Administration on Lactate Clearance and Mortality in Patients With Septic Shock

Woolum JA. Crit Care Med 2018; 46:1747–1752

doi: 10.1097/CCM.0000000000003311


Clinical Question: Does the administration of thiamine lead to more rapid lactate clearance and improved clinical outcomes in patients with septic shock?

Background: Septic shock is characterized by a hypermetabolic state that resembles thiamine deficiency. Thiamine deficiency is common in critically ill patients. A previous pilot randomized controlled trial had shown significantly lower lactate levels and improved mortality over time in patients with septic shock who were thiamine deficient.

Design: Retrospective, matched cohort study, based on data collected from electronic medical records. Regression analysis was performed with mortality as competing event (if the patient died with a lactate level of more than 2 mmol/l, clearance was considered not achieved). Three models were constructed: (1) with lactate levels alone, (2) after adjustment for age, sex, and race, and (3) with age, sex, race, and other likely factors that influence mortality and lactate clearance. A Cox proportional hazards model was constructed along the same lines for 28-day mortality.

Setting: A single academic center in the US. The study covered a 4-year period between January 1, 2013, and January 1, 2017.

Population: An electronic medical database was queried based on the diagnostic code for septic shock according to 9thor 10thedition of the International Classification of Diseases (ICD).

Inclusion criteria:

  • Patients who were coded as septic shock on the electronic medical database
  • 18 years and older
  • Admission to medical or surgical services

Exclusion criteria:

  • Less than 18 years of age
  • Septic shock not present at admission

After validation using the Sepsis-3 criteria, 1049 patients were included out of the 2270 patients who were initially screened. Out of this cohort, 123 patients who received thiamine were matched with 246 patients who did not.

Intervention: Intravenous administration of thiamine in any dose within the first 24 hours of hospital admission.

Control: Patients who received thiamine were matched with a cohort who did not receive thiamine in a 1:2 ratio.


Primary outcome: Patients who were administered intravenous thiamine in the first 24 hours of hospital admission had more rapid lactate clearance. All three regression models revealed improved lactate clearance with thiamine administration. The subdistribution hazard ratios in the three models ranged between 1.292–1.339. The effect of thiamine on lactate clearance was significantly more in female patients on a gender-based interaction model.

Secondary outcomes: Thiamine was found to significantly reduce 28-day mortality on the three Cox’s proportional hazards models. Similar to lactate clearance, the benefit was more evident in female patients. There was no significant difference in other secondary outcomes including change in SOFA scores on day 5 compared to baseline, vasopressor-free days, ventilator-free days, ICU-free days, incidence of AKI, and the requirement for renal replacement therapy.

Authors’ conclusions:

In patients with septic shock, intravenous thiamine, administered within 24 hours of ICU admission resulted in more rapid lactate clearance and a significantly reduced 28-day mortality.


  • This is the largest study so far to evaluate the effect of thiamine on lactate clearance and mortality in patients with patients with septic shock.
  • Matching was carried out between patients who had thiamine and those who did not.
  • Regression analysis was performed with mortality as the competing event.


  • This a retrospective observational study based on data derived from electronic medical records. Although matching was carried out between patients who received thiamine and those who did not, there may be confounders that have not been accounted for.
  • The dose of thiamine was variable and ranged between 100–500 mg per day.
  • Thiamine levels were not measured; it is unclear whether the benefit may be related to thiamine deficiency.
  • Lactate measurements were presumably carried out at random intervals; this may have led to miscalculation of time to lactate clearance.
  • The mortality of the cohort was high compared to contemporaneous studies (54%) which the authors attribute to a relatively large number of patients with cirrhosis.

My take:

This study adds to the growing body of evidence that intravenous thiamine may improve outcomes in patients with sepsis and septic shock. However, no prospective, controlled trial has evaluated the effect of thiamine in sepsis. In my opinion, given that harmful effects are unlikely, intravenous thiamine may be considered in critically ill patients with sepsis.








Hemodyanamic monitoring of the future

Several years ago, when the dinosaurs among us were in training, we used to look upon each pulmonary artery catheter that we inserted with a sense of pride and fulfillment​. Thankfully, the era of inflating balloons with the pulmonary artery, measurement of wedge pressures, and serial cold saline injections to measure cardiac output seem to be drawing to a close. Arterial lines are often inserted more for convenience than for precisely titrated therapeutic interventions. There is increasing realization that the central venous pressure may be just a ballpark number that may not reflect the preload to the heart that it is assumed to represent. Perhaps we need to look at technological refinement that will enable us to obtain maximal information with minimal invasion.

Non-invasive, continuous blood pressure measurements with display of an arterial waveform is possible today by the finger clamp and the applanation techniques. The former uses a finger cuff with inflation-deflation cycles to maintain the finger volume constant, while the counterpressure applied is reconstructed to obtain arterial blood pressures (Fig. 1). In the applanation technique, a miniaturized transducer placed on the surface exerts pressure on an artery and enables direct pressure measurement. Much work needs to be done to validate measurements using these techniques; however, it seems likely that reliable, continuous, non-invasive blood pressure measurements may be extensively available in the near future.


Fig. 1 The finger clamp method of continous blood pressure measurement

Lack of adequate windows, especially in ventilated patients can be a frustrating experience for the intensivist by the bedside. There are major limitations to performing repeated transesophageal echocardiographic examinations in critically ill patients. However, a miniaturized probe, no bigger than a nasogastric tube, has been introduced that can be placed in the esophagus for repeated examination over a period of 72 hours. Three standard views are utilized, including the mid-esophageal  4-chamber view, the transgastric short axis view, and the superior vena caval view to enable assessment of volume responsiveness, size and function of the ventricles, assessment of valvular function, and detection of pericardial effusion (Fig. 2)


Fig. 2 Hemodynamic transesophageal echocardiography using a miniaturized probe

Despite the debate surrounding the use of lactate levels as a guide to the efficacy of resuscitation, it is well-established that persisting, high lactate levels are associated with increased mortality. Continuous, real-time measurement of lactate levels may help assess the trajectory of illness, especially in critically ill, septic patients. Using a 5-lumen central venous catheter, with saline infusion as a microdialysate, continuous measurement of lactate levels can be carried out, with values displayed in a graphic format (Fig. 3). The direction of the lactate curve may help guide therapy and assess progress in septic patients.


Fig. 3 Continous lactate monitoring by microdialysis

The time to embrace non- or less invasive techniques of hemodynamic monitoring is probably overdue. Technological refinement, would, hopefully allow us in the near future to extract maximal information from our patients to tailor appropriate intervention with the least risk of damage.


Corticosteroids in H1N1 pneumonia – damned if you do, and damned if you don’t?


We are in the middle of yet another H1N1 epidemic in India. Karnataka has been particularly affected, with several new cases being reported every day. Several deaths have been reported so far, and the toll is likely to mount in the days to come. The current epidemic shares several common features with the global pandemic of 2009, with predominance in young adults and dense lung infiltrates leading to severe problems with gas exchange. Several rescue interventions have been resorted to, including the use of extracorporeal membrane oxygenation with a view to tide over the crisis until natural healing occurs.

Against this background, we reconsider the use of corticosteroids – considered savior by many and maligned by others in many different clinical situations. Several anecdotal reports of dramatic improvement were noted with corticosteroid therapy during the 2009 epidemic; subsequently, many observational studies have been published with no clear-cut answers; however, some studies suggest worse outcomes with the use of corticosteroids. To my knowledge, there has been no robust, prospective controlled study that has addressed this question. The World Health Organization is reasonably categorical in its H1N1 treatment guidelines, suggesting that “Patients who have severe or progressive clinical illness, including viral pneumonitis, respiratory failure, and ARDS due to influenza virus infection, should not be given systemic corticosteroids unless indicated for other reasons or as part of an approved research protocol”.

I tried to pool the limited data available so far, through a PubMed search using a combination of search terms including “corticosteroids and H1N1”, “steroids and H1N1”, “methylprednisonlone and H1N1”, “hydrocortisone and H1N1”, and “dexamethasone and H1N1”. I retrieved nine studies that analyzed mortality as the endpoint. I chose mortality for the longest period reported in each study as the outcome. This is what I found.

Fig 1. Pooled data on mortality from nine observational studies with the use of corticosteroids in H1N1 infection

Screen Shot 2018-10-14 at 5.55.41 PM 

The pooled data from these observational studies seem to suggest that the use of corticosteroids use in H1N1 infection may lead to increased mortality. Two previous meta-analyses have also reported similar results. (1,2)

However, how much importance do we attach pooled data in a clinical situation fraught with poor outcomes with conventional measures? We have also seen a dramatic improvement on the odd occasion with steroids when we were with our backs to the wall. Several important questions remain unanswered especially when data is pooled across heterogeneous patient populations. Specifically, would steroids be helpful in 1) the most severe forms of the disease 2) would the timing matter – early vs. late? 3) Is there a preferred corticosteroid preparation (methylprednisolone vs. hydrocortisone)? 4) Would corticosteroids improve outcomes in severe ARDS due to H1N1 infection?

Unfortunately, these questions are difficult to answer; it is an onerous if not impossible task to prospectively study specific patient populations who are likely to benefit from corticosteroid administration. Perhaps, when faced with similar difficult clinical situations when a clear-cut answer is not forthcoming, we should continue to have equipoise and keep an open mind.

Have you been taking your flu booster shots?


  1. Zhang Y, Sun W, Svendsen ER, Tang S, MacIntyre RC, Yang P, et al. Do corticosteroids reduce the mortality of influenza A (H1N1) infection? A meta-analysis. Crit Care. 2015;19(1):46.
  2. Rodrigo C, Leonardi-Bee J, Nguyen-Van-Tam JS, Lim WS. Effect of Corticosteroid Therapy on Influenza-Related Mortality: A Systematic Review and Meta-analysis. J Infect Dis. 2015 Jul 15;212(2):183–94.


Contentious Use of Corticosteroids in The Critically Ill


There is a long-drawn-out history with the use of corticosteroids in septic shock. In the 1980s, methylprednisolone was used in industrial strengths as a short course treatment, with predictably poor results.[1]After several studies that suggested poor outcomes in septic shock, the use of corticosteroids slowly faded away. However, in the 1990s, there was a rekindling of interest with the use of corticosteroids in lower, more physiological doses, as replacement therapy, considering the possibility of “relative adrenal insufficiency” in septic shock. Three adequately powered randomized controlled trials have been published with the use of “physiological” dose corticosteroids in septic shock (Annane et al.[2], Corticus[3], and ADRENAL[4]). A pooled analysis of these three studies does not demonstrate improved survival with the use of corticosteroids in septic shock.


Fig 1. A pooled analysis of studies on mortality with the use of corticosteroids in septic shock 

However, earlier shock reversal seems likely with the use of corticosteroids, as evidenced in most of the studies. The ADRENAL trial also revealed marginally lower ventilation days with corticosteroid use (6 vs. 7 days) with the initial episode of mechanical ventilation; however, there was no difference between groups with days alive and free of ventilation.

Severe acute respiratory distress syndrome (ARDS) may follow several acute illnesses, including sepsis, trauma, and acute pancreatitis. Corticosteroids are often used in patients who continue to remain hypoxic after optimization of mechanical ventilation. Meduri et al. carried out two RCTs,[5],[6]with 1:2 randomization. Both studies seemed to favor the use of corticosteroids, with improved lung injury scores, improved oxygenation, and less time on ventilation. However, the results of the ARDSnet study of 180 patients with ARDS was different.[7] Although steroid use was associated with improved P/F ratios and other parameters of respiratory physiology, there was no difference in the 60 or 180-day mortality. A pooled analysis of these three trials and a recent RCT does not show any survival advantage with the use of corticosteroids in ARDS

fig 2.jpg

Fig 2. A pooled analysis of studies on mortality with the use of corticosteroids in ARDS

Are corticosteroids beneficial in community-acquired pneumonia (CAP)? One of the early studies on critically ill patients revealed improved P/F ratios, earlier resolution of shock, shorter hospital stay, and improved mortality.[8]Subsequently, there have been several small RCTs that evaluated the possible beneficial effect of corticosteroids in CAP[9],[10],[11],[12],[13],[14]. Most of these studies have been performed on patients with a low severity of illness, and low mortality. A pooled analysis of all studies carried out after 2005 suggests a mortality benefit with the use of corticosteroids (Fig 3). 

fig 3

Fig 3. A pooled analysis of studies on mortality with the use of corticosteroids in community-acquired pneumonia 

However, several questions remain unanswered. Viral pneumonias are notorious to lead to a severe disease with profound impairment of oxygenation on occasions. We are seeing a resurgence of severe H1N1 pneumonia after several years in India. Would corticosteroids be of benefit in these patients? There are no robust data available to guide us in this situation. The limited evidence available so far seems to suggest that corticosteroids may have either have no effect or even be harmful in viral pneumonias.


[1] Veterans Administration Systemic Sepsis Cooperative Study Group, Effect of high-dose glucocorticoid therapy on mortality in patients with clinical signs of systemic sepsis.  N Engl J Med (1987);317659- 665

[2] Annane D1Sébille VCharpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock.JAMA. 2002 Aug 21;288(7):862-71.

[3] Sprung CL, Annane D, Keh D, et al. Hydrocortisone therapy for patients with septic shock. New England Journal of Medicine. 2008 Jan 10;358(2):111.

[4] Venkatesh B, Finfer S, Cohen J, Rajbhandari D, Arabi Y, Bellomo R, Billot L, Correa M, Glass P, Harward M, Joyce C. Adjunctive glucocorticoid therapy in patients with septic shock. New England Journal of Medicine. 2018 Mar 1;378(9):797-808.

[5] Meduri GU, Headley AS, Golden E, Carson SJ, Umberger RA, Kelso T, Tolley EA. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: a randomized controlled trial. Jama. 1998 Jul 8;280(2):159-65.

[6] Meduri GU, Golden E, Freire AX, et al. Methylprednisolone infusion in early severe ARDS: results of a randomized controlled trial. Chest. 2007 Apr 1;131(4):954-63.

[7] National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Efficacy and safety of corticosteroids for persistent acute respiratory distress syndrome. New England Journal of Medicine. 2006 Apr 20;354(16):1671-84.

[8] Confalonieri M, Urbino R, Potena A, et al.  Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study. Am J Respir Crit Care Med. 2005 Feb 1;171(3):242-8.

[9] Blum CA, Nigro N, Briel M, et al. Adjunct prednisone therapy for patients with community-acquired pneumonia: a multicentre, double-blind, randomised, placebo-controlled trial. The Lancet. 2015 Apr 18;385(9977):1511-8.

[10] Fernández-Serrano S, Dorca J, Garcia-Vidal C, et al. Effect of corticosteroids on the clinical course of community-acquired pneumonia: a randomized controlled trial. Critical Care. 2011 Apr;15(2):R96.

[11] Meijvis SC, Hardeman H, Remmelts HH, et al. Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. The Lancet. 2011 Jun 11;377(9782):2023-30.

[12] Sabry NA, Omar EE. Corticosteroids and ICU course of community acquired pneumonia in Egyptian settings. Pharmacology & Pharmacy. 2011 Apr 25;2(02):73.

[13] Snijders D, Daniels JM, de Graaff CS, et al. Efficacy of corticosteroids in community-acquired pneumonia: a randomized double-blinded clinical trial. American journal of respiratory and critical care medicine. 2010 May 1;181(9):975-82.

[14] Torres A, Sibila O, Ferrer M, et al. Effect of corticosteroids on treatment failure among hospitalized patients with severe community-acquired pneumonia and high inflammatory response: a randomized clinical trial. Jama. 2015 Feb 17;313(7):677-86.


Extracorporeal Membrane Oxygenation (ECMO) for Acute Respiratory Failure – the EOLIA Study


Extracorporeal membrane oxygenation (ECMO) is being increasingly used in acute respiratory failure. It is employed as a rescue intervention when conventional measures including titration of PEEP and prone positioning fail to achieve the desired effect. Historically, two randomized controlled trials (RCTs) had failed to demonstrate efficacy; however, these studies were performed several decades ago, when ECMO techniques were less refined. The CESAR study, performed many years later, demonstrated a significant improvement in the primary outcome of death or disability at six months in ECMO treated-patients. This study used devices with roller pumps, in contrast to the centrifugal pumps that are currently the preferred technique.

Combes et al., in their RCT, compared ECMO with conventional care in severely hypoxic patients with acute respiratory distress syndrome (ARDS). Patients in the control group could crossover to ECMO in case of severe, refractory hypoxemia. They found no significant difference in the primary endpoint of 60-day mortality. The secondary endpoint was treatment failure at 60 days: mortality in the ECMO group, and mortality or crossover to ECMO in the control group. The secondary endpoint was significantly more favorable with ECMO.

It may not be prudent to reject ECMO therapy in acute respiratory failure based on the findings of this study. To begin with, the investigators assumed improved survival by 20% with ECMO. This was based on two previously published studies- the PEEP study by Mercat et al. and the CESAR trial. However, such a large effect size entailed a small sample size, increasing the likelihood of a type II error. Clearly, it would have been unethical to decline ECMO to patients who were dying of hypoxia; however, crossover from the control to the intervention arm also makes the results of the study difficult to interpret.

Thirty-five patients who developed refractory hypoxemia in the control group were crossed over to receive ECMO. At the time of crossover, the median P/F ratio was 51, and the median Saowas 77%. Many of these patients were on the verge of severe cardiovascular failure; nine patients suffered cardiac arrest prior to initiation of ECMO and seven underwent veno-arterial ECMO. None of these patients may be expected to survive with continued conventional care; however, the use of ECMO resulted in 60-day survival in 15 of 35 (43%) patients. Another weakness of this study is the large number (72%) of potentially eligible patients who were excluded; 166 (16%) were excluded because they were already on ECMO. Blinding is not feasible in a study of ECMO; however, investigator bias cannot be excluded.

Perhaps the inclusion criteria for ECMO initiation also need to be considered. In our practice, we would probably not consider ECMO in a patient with a P/F ratio of 80 for six hours, particularly if there is an improving trend. Nor would I be too keen with a PCO2 of 60 mm Hg and pH of 7.25 for 6 hours. Only 62% of patients who underwent ECMO were prone ventilated; in our practice, we would attempt prone ventilation almost always before we consider ECMO.

At the end of the day, it may well be difficult to definitely prove the efficacy of ECMO against conventional care in a randomized controlled trial. First, it may be unethical to withhold ECMO in patients who are dying of hypoxia; approval for such a trial may be denied by most ethics committees. Second, to generate a sufficient sample size to demonstrate a clear effect in patients with refractory hypoxemia may take an inordinately long period. This multicentric study took nearly six years to recruit 249 patients. Therefore, I feel, given the current level of evidence, it may be appropriate to initiate ECMO based more on clinical judgment and local feasibility.





The Sepsis Scenario in India

The World Sepsis Day is held on the 13th of September every year. We held a meeting of local intensive care physicians in Bangalore the other day to mark the occasion. It offered us an opportunity to reflect upon where we stand and the progress we have made over the years in battling this deadly affliction, that kills approximately 6 million people across the globe every year. It is sobering to realize that this number may well be an underestimate because most epidemiological studies do not include “middle” and “low income” countries. While the number of hospital admissions for myocardial infarction and stroke have decreased over the years, there has been a steady increase in patients who require hospitalization for sepsis. Several magic bullets have been tried and fallen by the wayside in our quest to combat sepsis.

In contrast to global epidemiology, there are several infectious diseases that are peculiar to and widely prevalent in India, including malaria, dengue, leptospirosis, typhoid, and tuberculosis. We have a paucity of country-specific epidemiological information on the incidence of and outcomes from sepsis-related illnesses from India. Multicentric studies from critical care units across the country would definitely offer us a plethora of information on the types of diseases, complications, and outcomes in critically ill patients with sepsis, from an Indian perspective. I am sure it will pave the way to enable us to deliver improved care to our patients.

Do we need to combine meropenem with colistin in multidrug-resistant infections?

Colistin is often used as combination therapy in multidrug-resistant infections. The antibiotics used in combination with colistin include meropenem, rifampicin, and minocycline. Combination therapy is favored for several theoretical reasons. Colistin levels in the lung have often resulted in subtherapeutic levels in animal models. Heteroresistance, a phenomenon by which subsets of bacteria may be resistant, even though in vitro testing suggests otherwise, may occur with colistin monotherapy. Heteroresistance may lead to the proliferation of a fully resistant strain during the course of treatment. Antibiotic synergism has also been proposed to explain the benefits of combination therapy with colistin. Furthermore, some studies suggest poor rates of clinical cure with colistin monotherapy. (1)

Paul et al. in a multicenter randomized controlled trial, compared combination therapy with colistin and meropenem with monotherapy using colistin alone in carbapenem-resistant gram-negative infections. (2) A total of 406 patients were included, with 198 in the monotherapy group and 208 with combination therapy. The groups were well matched at baseline; 65% of patients were ventilated, while 18% required hemodynamic support, and 6.4% required renal replacement therapy. The median SOFA score was 6, with an overall mortality of 44%. The most common infection was ventilator-associated pneumonia (VAP), followed by bacteremia, and urosepsis. About 36% of infections were ICU-acquired. The most common pathogen, by far, was Acinetobacter Baumannii.

The primary outcome was “clinical success” of therapy at 14 days, which required all the following criteria: 1. Survival; 2. Systolic blood pressure > 90 mm Hg without vasopressors; 3. Improved or stable SOFA score; 4. Stable or improved P/F ratio in patients with VAP; 5. No growth in blood culture on day 14 for patients with bacteremia. Clinical failure rates were high and similar in both groups; failure of therapy was seen in 156/198 (79%) of patients with monotherapy and 152/208 (73%) with combination therapy. There was no difference in all-cause mortality between groups at 14 and 28 days. There was a lower incidence of AKI on day 14 with combination therapy, in patients with “injury” and “failure” on the RIFLE classification. Clinical failure was lower in ventilator-associated pneumonia, hospital-acquired pneumonia, and bloodstream infection, although the difference was not statistically significant.

This study aimed to answer an important question that we ask in patients with multidrug-resistant infections with in vitro sensitivity to colistin and resistance to meropenem. Hitherto, the general policy in most units has been to combine both, although there was no robust evidence to support the efficacy of such a combination. This study had several limitations, including the use of a composite outcome with short-term (14-day) mortality as one of the components. The overwhelming majority of infections were caused by A.Baumannii, which naturally raises a question about applicability to other organisms. There was considerable heterogeneity in the site of infections. Would the results apply to specific infections such as VAP, given that colistin has poor penetration into lung tissue? Therapeutic drug monitoring was not performed; it may have been interesting to know if adequate drug levels were achieved, especially with colistin.

However, based on this study, I feel that we should strongly consider using colistin as monotherapy if in vitro resistance to meropenem is noted. Antibiotic synergism is at best, hypothetical, especially when resistance has been demonstrated in vitro. Furthermore, in our setting, the manifold escalation of the cost of care with combination therapy should also be a deterrent. The findings of a retrospective observational study from an oncological unit in India, that showed no change in crude mortality with a carbapenem-colistin combination compared to colistin alone, lends further support to monotherapy. (3)


  1. Parchem NL, Bauer KA, Cook CH, Mangino JE, Jones CD, Porter K, et al. Colistin combination therapy improves microbiologic cure in critically ill patients with multi-drug resistant gram-negative pneumonia. Eur J Clin Microbiol Infect Dis. 2016 Sep;35(9):1433–9.
  2. Paul M, Daikos GL, Durante-Mangoni E, Yahav D, Carmeli Y, Benattar YD, et al. Colistin alone versus colistin plus meropenem for treatment of severe infections caused by carbapenem-resistant Gram-negative bacteria: an open-label, randomised controlled trial. Lancet Infect Dis. 2018 Apr;18(4):391–400.
  3. Ghafur A, Devarajan V, Raja T, Easow J, Raja MA, Sreenivas S, et al. Monotherapy versus combination therapy against nonbacteremic carbapenem-resistant gram-negative infections: a retrospective observational study. Indian J Crit Care Med. 2017 Dec 1;21(12):825.

The BICAR study – does bicarbonate therapy help in metabolic acidosis?

In the BICAR-ICU study Jaber et al. randomized critically ill patients with metabolic acidosis with pH less than 7.2 and bicarbonate less than 22 mmol/L to receive 4.2% bicarbonate, targeting a pH of 7.3. They compared outcomes with a control group that did not receive bicarbonate.

Three hundred and eighty-nine patients were enrolled, with 195 in the bicarbonate group and 194 in the control group. The primary outcome, a composite of death at 28 days and at least a single organ failure on day 7, was not significantly different. However, the composite outcome was significantly lower in a pre-specified subgroup of patients with an AKIN score of 2 or 3. Besides, the 28-day survival and the presence of at least a single organ dysfunction on day 7 (the individual components of the composite primary outcome) were also significantly lower in this subgroup. The requirement for renal replacement therapy (RRT), based on pre-specified criteria, was also significantly lower in patients who received bicarbonate therapy.

This is indeed an interesting addition to the scarce body of knowledge that addresses the benefit of bicarbonate therapy in metabolic acidosis due to causes other than loss of alkali from the body. Conventional wisdom advocates treatment of the underlying cause in lactic acidosis, in contrast to correction of pH to a pre-set level. The generally held view is that administration of bicarbonate could lead to a release of CO2 and worsening of intracellular acidosis. Does this study challenge tradition-borne practice and suggest a benefit from bicarbonate therapy to correct the pH in metabolic acidosis?

There are several obvious limitations to appropriate interpretation of this study, some of them pointed out by the authors themselves. Composite primary outcomes have often been problematic to interpret; this study is no exception. How does 28-day survival add up with the presence of a single organ failure of day 7? Among the eligible patients, 58% were excluded; a significant proportion (20%) because they had already received bicarbonate therapy. The overall mortality of the cohort was 49%, largely explained by the high baseline severity, with a SAPS of 60. Only 60% of patients who received bicarbonate attained the set pH target of 7.3 by 48 hours. If the intention of correction of acidosis was not achieved in the majority of patients what may be the putative mechanism by which bicarbonate exerts a beneficial effect? The other major limitation of this study is that 47/194 (24%) of patients in the control group were administered bicarbonate therapy, leading to a major problem with the interpretation of intention to treat analysis.

It is interesting to note that overall, the requirement for RRT (based on prespecified criteria) was lower in the bicarbonate-treated group. The requirement for RRT was also lower in the subgroup of patients with an AKIN score of 2-3. Do we conclude that the improved 28-day survival among bicarbonate-treated patients in this subgroup may have been due to adverse outcomes resulting from RRT in the control group? This seems unlikely. The study was unblinded, which could also have led to bias; no placebo was employed because administration of a significant volume of any type of placebo fluid may have led to electrolyte abnormalities. Another point that I would raise is the difference in the composite outcome that was assumed to calculate power. I cannot recall a major study in recent times that assumed such a big difference (15%) in the primary outcome. The larger the assumed outcome difference, the smaller the sample size and the higher the propensity for a type I error.

Overall, I feel it is an interesting study which suggests improved outcomes in critically ill patients with AKI and metabolic acidosis from bicarbonate therapy. I do feel that there may be several critical care physicians in India who follow a similar approach in the hope of delaying RRT or avoiding it altogether. Perhaps we should consider a controlled study that specifically addresses patients with severe AKI and metabolic acidosis to seek a possible outcome benefit. I feel, the primary outcome of such a study should be 90-day mortality, which seems to be a more appropriate benchmark in this stone age.




Platelet transfusion in Dengue Fever

Following the monsoon rains, we see several cases of Dengue in our ICUs. Many of these patients develop severe thrombocytopenia, with the counts often dropping below 20,000. I feel most clinicians would strongly consider prophylactic platelet transfusion (without any evidence of clinical bleeding) when the count drops to between 10–20,000. However, there is a reasonably sound body of knowledge which suggests that prophylactic platelet transfusions using arbitrary thresholds may not have any favorable effect. There are several retrospective studies and two randomized controlled studies that have clearly shown that there is no reduction in the incidence of bleeding with platelet transfusion using such arbitrary thresholds. Indeed, even the platelet counts do not seem to rise significantly following transfusion. In fact, there is a possibility of potential harm with transfusion of platelets in this manner.

In two pediatric studies, it has been shown that ADAMTS-13 levels may be relatively low in Dengue,  compared to Von Willebrand factor (VWF) levels. This may result in increased platelet adhesion to VWF multimers, and endothelial sequestration. Sequestrated platelets may lead to impaired microcirculatory flow and organ dysfunction (through a pathophysiological mechanism similar to TTP). Hence, it is possible that prophylactic platelet transfusion may cause harm by increased endothelial sequestration and worsening organ function. Furthermore, the harmful effects of transfusion, including transfusion-associated lung injury (TRALI) and fluid overload may have an adverse impact on clinical outcomes.

How do we offset the possible harm from platelet transfusion in Dengue? Clearly, if the ADAMTS-13 levels are low, there may be a compelling reason to replenish it using fresh frozen plasma, prior to transfusion of platelets. Cryo-reduced plasma, which is FFP from which cryoprecipitate has been removed is another rich source of ADAMTS-13. Recombinant ADAMTS-13 is also currently available.

Although based on a well-founded hypothesis, no clinical studies have been done to test the benefit of replenishing ADAMTS-13 levels in Dengue prior to platelet transfusions. I strongly feel we could undertake a multicentric study to assess clinical outcomes with such an intervention. The clinical outcomes to consider may include a rise in platelet counts to a sustained level and clinical bleeding with and without ADAMTS-13 supplementation with any of the aforementioned products.