[Abstract in Chinese]

Sepsis is a common cause of hospital admission worldwide. The annual incidence of sepsis has been reported to be approximately 300 to 1031 per 100 000 population in the US, and is increasing.1 In-hospital mortality, however, decreased from 35% in 2000 to <20% in 2013.2 Numerous studies have been performed or are ongoing in this field. The following discussion provides an update on the change to sepsis definition, three recent trials on protocolised early goal-directed therapy (EGDT), and individual components of sepsis management.

In 1991, sepsis was defined as fulfilling two or more than two systemic inflammatory response syndrome (SIRS) criteria in the presence of infection (Table 1).3 Many seriously infected patients (eg old or immunocompromised), however, are unable to mount a SIRS. Using SIRS criteria to define severe sepsis will miss one in eight otherwise similar patients with substantial mortality.2 In addition, the mortality risk has been shown to increase linearly with each additional SIRS criterion and there is no transition point noted at a threshold of two SIRS criteria.

To acknowledge the above shortcomings, the Sepsis-3 (Third International Consensus Definitions for Sepsis and Septic Shock) in 2016 redefined sepsis as “life-threatening organ dysfunction caused by a dysregulated host response to infection”.4 Organ dysfunction is identified as “an acute change in the total Sequential Organ Failure Assessment (SOFA) score of ≥2 points” (Table 1). In addition, a quick SOFA (qSOFA) score was introduced for bedside screening. Meeting two or more than two qSOFA criteria (respiratory rate ≥22/min, altered mentation with a Glasgow Coma Scale score of <15, systolic blood pressure ≤100 mm Hg) should prompt consideration of possible infection in undiagnosed patients, investigation for organ dysfunction that defines sepsis in infected patients, and initiation of sepsis management where appropriate.4 Despite these recent changes in definition, clinicians should maintain a high index of suspicion and always consider sepsis as a possible explanation/diagnosis when faced with new-onset organ dysfunction in a patient.

Apart from clinical assessment, various serum biomarkers have been studied for their potential role in early diagnosis of sepsis. C-reactive protein is commonly used but it has a low specificity for sepsis. Procalcitonin (PCT) is a prohormone of calcitonin that is released into the circulation in response to severe systemic inflammation due to bacterial infection. A recent meta-analysis showed its clinical value in diagnosing sepsis in critically ill patients (an area under the receiver operating characteristic curve of 0.85).5 The meta-analysis, however, was limited by substantial heterogeneity across different studies, a wide range of cut-offs used, absence of a true reference diagnostic standard, and potential publication bias. It is noteworthy that PCT can be falsely elevated in inflammation due to other causes, such as trauma, rhabdomyolysis, surgery, severe pancreatitis, autoimmune disorders, cardiogenic shock, and following prolonged resuscitation. It cannot be solely relied on to discriminate sepsis from other causes of inflammation, but a plasma PCT level of ≥0.5 ng/mL is a helpful adjunct when interpreted along with additional clinical information and serial monitoring might have a role in guiding subsequent antibiotic treatment (see below). In case of doubt, it is advisable to initiate treatment for sepsis early, and adjust subsequent management and antibiotics according to the patient’s clinical progress, results of investigations, and possibly serial PCT monitoring.

In 2001, Rivers et al6 randomised 263 patients with severe sepsis or septic shock in an emergency department (ED) to EGDT or usual care. The sequential goals of EGDT were central venous pressure (CVP) achieved by fluid resuscitation, mean arterial pressure (MAP) with vasopressors, and central venous oxygen saturation (ScvO₂) with red cell transfusion and dobutamine. The result was an absolute reduction in in-hospital mortality of 16%.6 The protocol was incorporated into the surviving sepsis campaign (SSC) 2004, 2008, and the 2012 Guidelines.7

Over the years, concerns have remained about the external validity of the original trial, haemodynamic goals, use of CVP and ScvO₂ monitoring, blood transfusion, dobutamine, and resultant higher costs of its implementation. Subsequently, three large-scale multicentre randomised controlled trials were published in 2014-2015: the ProCESS (Protocolized Care for Early Septic Shock) trial,8 the ARISE (Australasian Resuscitation in Sepsis Evaluation) trial,9 and the ProMISe (Protocolised Management in Sepsis) trial.10 All three studies were negative; there was no survival benefit using protocolised care compared with usual care (Table 2).6 8 9 10

A lack of survival benefit of EGDT in the latest three trials may be the result of improved sepsis management since the original trial: nearly all patients received antibiotics within 6 hours of presentation, and a significant amount of fluid was already administered before randomisation (Table 2). Treatment in the usual care groups was guided by clinical assessment of volume and perfusion status, and achieved similar mean and systolic blood pressures at the end of the intervention period. These trials demonstrated that CVP and ScvO₂ goals confer no additional benefit for sepsis survival.

In response to the new evidence, the SSC Guidelines updated its 6-hour bundle in April 2015, and recommended reassessment in the event of persistent arterial hypotension with either physical examination or “two of the following (measure CVP, measure ScvO₂, bedside cardiovascular ultrasound, dynamic assessment of fluid responsiveness with passive leg raise or fluid challenge)”.11

Source control includes drainage of any infected fluid collection, debridement of infected solid tissue, and removal of infected foreign bodies or devices. It should best be achieved within 12 hours of identification by imaging and/or diagnostic sampling of the infection foci. Minimally invasive intervention including percutaneous and endoscopic treatments should be considered, but surgery is indicated if control remains inadequate or if there is diagnostic uncertainty. Damage-control surgery for life-threatening peritonitis is associated with improved outcomes.12 It involves an abbreviated initial laparotomy for haemorrhage and contamination control, followed by resuscitation before the final definitive repair and abdominal closure.

Delay in antimicrobial treatment is associated with increased mortality, adverse clinical outcome, and longer intensive care unit (ICU) and hospital stay.13 Effective intravenous antimicrobials should be initiated as soon as possible after recognition and within 1 hour for both sepsis and septic shock.14 One study showed that each hour delay reduces survival by 7.6% in the first 6 hours following the onset of septic shock.15 Although a recent meta-analysis pooling data from 11 observational studies showed no survival benefit of early antibiotic therapy,16 it failed to analyse all eligible studies and lacked microbiological considerations. Early therapy remains logical, especially in patients with severe infections, although the optimum time frame for administration remains unknown.17 Delay in administration can occur anywhere, from ED triage, making a diagnosis, antibiotic order, drug dispensary to the actual administration, and should be addressed with clinical, administrative, and logistics measures to improve timeliness of treatment.

The choice of initial empirical antimicrobials should be broad enough to cover the likely pathogens, while also taking into account recent culture results, host factors, and susceptibility patterns. The EPIC II (Extended Prevalence of Infection in Intensive Care) study showed that in Asia, the common infective sources are the respiratory system, abdomen, blood stream, and renal/urinary tract, while the commonest organisms are Streptococcus pneumoniae, vancomycin-sensitive Enterococcus, Klebsiella spp, Pseudomonas spp, and Acinetobacter spp.18 Local microbiologists can regularly provide antibiotic sensitivity patterns for reference.

Combination therapy, defined as administration of two or more different classes of antimicrobials with different mechanisms of action, has the advantages of broadening the spectrum of coverage, and possible additive or synergistic effects on pathogens. Meta-analyses showed that combination antibiotic therapy improves survival in the most severely ill patients with septic shock, but may be detrimental to low-risk patients and increases nephrotoxicity.19 20 Once the causative pathogens and their susceptibility patterns are known, de-escalation of antimicrobial therapy should follow to prevent development of resistance, as well as reduce drug toxicity and cost. Discontinuing antibiotics can be considered when PCT is ≤0.5 ng/mL or serial monitoring shows a decline of ≥80% of its peak value.21

Appropriate antibiotic dosing to achieve effective bacterial killing while preventing toxicity and emergence of resistance is also important. This is particularly relevant in critically ill patients with substantial pharmacokinetic variability (Table 3), who are infected by pathogens with higher minimum inhibitory concentrations (MIC). Individualised dosing adjustment requires knowledge of pharmacokinetic targets and MIC for the organism. Readers can refer to published reviews for specific recommendations.22 23 In general, changes to the volume of distribution will affect initial dosing, while changes in drug clearance will affect the maintenance dose. Killing by time-dependent antibiotics (eg β-lactams) correlates with the time fraction when serum drug concentration exceeds MIC. This can be achieved by frequent dosing and use of continuous infusion. Conversely, concentration-dependent killing (eg aminoglycosides) correlates with the ratio of peak drug concentration to MIC. A higher dosage with extended dosing intervals will maximise killing while minimising toxicity. Due to the complexity and variability of various factors at play, therapeutic drug monitoring has been advocated in the critically ill patients but remains to be available universally.22 23 24 Therapeutic drug monitoring involves direct measurement of serum antibiotic concentrations and timely comparison with a therapeutic target to facilitate adjustments by the clinician or any dosing software.

Choice of non-blood product can be broadly divided into crystalloids and colloids. Crystalloids include normal saline or balanced solutions (eg lactated Ringer’s solution [B Braun, US], Hartmann’s solution [Fresenius Kabi, Australia], Plasma-Lyte 148 [Baxter, US]). Colloids include natural human albumin and semi-synthetic solutions (gelatin-like Gelofusine [B Braun, US] or Haemaccel [Sanofi, France], dextran, and hydroxyethyl starch). Normal saline has a high chloride content and may produce hyperchloraemic acidosis and renal vasoconstriction. Balanced solutions minimise these side-effects by using lactate or acetate as buffers. Weak evidence suggests that balanced solutions compared with normal saline reduce acute kidney injury (AKI), the need for renal replacement therapy (RRT), and mortality in sepsis.25 26 Nonetheless, the recent SPLIT trial did not find a difference in AKI among a heterogeneous group of ICU patients who received a balanced crystalloid or saline, although the study recruited predominantly postoperative patients at low risk who received small doses (median, 2 L) of fluid.27

Colloids theoretically maintain a higher oncotic pressure and hence intravascular volume but the CRISTAL (Colloids Versus Crystalloids for the Resuscitation of the Critically Ill) trial found no mortality difference among ICU patients with hypovolaemic shock who were resuscitated with either colloids or crystalloids.28 Dextran has ceased to be a resuscitation fluid due to its high anaphylactoid potential, impact on platelet aggregation with resultant bleeding complications, and interference with erythrocyte cross-matching. Gelatins have the highest risk among the colloids for anaphylactoid reaction and the lowest intravascular persistence due to their rapid urinary excretion. Hydroxyethyl starch is not advisable for acute volume resuscitation because it deposits in the kidneys, liver, skin and other tissues, and is associated with increased mortality, AKI, new-onset hepatic failure, and higher incidences of pruritus and rash.29 30 31 32 Concerning albumin, the SAFE (Saline versus Albumin Fluid Evaluation) trial33 demonstrated no survival benefit among a general ICU population when either 4% albumin or normal saline was used for fluid resuscitation, but predefined subgroup analysis suggested a trend towards improved survival in patients with severe sepsis who received albumin solution. A decade later, however, the ALBIOS (Albumin Italian Outcome Sepsis) study of patients with severe sepsis could not confirm a survival benefit when albumin was used in addition to crystalloids compared with crystalloids alone to maintain a serum albumin level of ≥30 g/L, although there was a small haemodynamic advantage and post-hoc subgroup analysis showed a significantly lower 90-day mortality in patients with septic shock who received albumin.34

In summary, crystalloids (possibly balanced solutions) remain the initial fluid of choice in the resuscitation of sepsis. Routine use of albumin is not warranted given its higher cost, but it may be considered in patients with septic shock who do not respond to crystalloid. There is no evidence that gelatins are more beneficial, and dextran and hydroxyethyl starch should be avoided.

Both under- and over-hydration can be harmful.35 It is therefore recommended that 30 mL/kg crystalloid be given with reassessment of fluid responsiveness (defined as >10%-15% increase in stroke volume in response to volume administration) or tissue perfusion afterwards. Static indices like CVP and pulmonary capillary wedge pressure are not good indicators of fluid responsiveness, and are not recommended for use to guide fluid therapy. Dynamic indices obtained by inducing a change in the preload and monitoring the corresponding change in cardiac output (CO) or its derivatives should be used instead (Table 4).

An arterial waveform pulse pressure variation (PPV) of >13%, induced by heart-lung interactions during mechanical ventilation, predicts fluid responsiveness with a high degree of accuracy in controlled settings.36 37 Its accuracy, however, is lowered by arrhythmia, spontaneous breathing activity, low-tidal-volume ventilation (<8 mL/kg ideal body weight), low heart-rate-to-respiratory-rate ratio (<3.6), and right ventricular dysfunction (peak systolic velocity of tricuspid annulus <0.15 m/s). Raised intra-abdominal pressure (IAP) exaggerates PPV, and one study found that PPV/IAP of <1.41 could identify false-positive patients.38 These confounders limit the application of PPV in routine clinical practice, in particular during protective ventilation for acute respiratory distress syndrome (ARDS).

A virtual fluid challenge test with passive leg raising can avoid the above caveats as it does not rely on mechanical ventilation to induce changes in preload and, unlike other methods, has been validated in patients with breathing efforts and arrhythmia. When coupled with CO monitoring, passive leg raising has an excellent predictive accuracy.39 Accuracy drops when arterial pulse pressure instead of CO is used, as well as in patients with raised IAP of ≥16 mm Hg.40

A mini-fluid challenge (100 mL infused rapidly over 1 minute) is an alternative method and limits cumulative positive fluid balance in non-responders. Unlike PPV, it remains accurate at times of low-tidal-volume ventilation.41

Respiratory variation of the inferior vena cava diameter is another accurate marker of fluid responsiveness in patients who are mechanically ventilated,42 but its use in spontaneously breathing patients is more controversial.

Guidelines of SSC recommend maintaining MAP at a minimal of 65 mm Hg.14 The SEPSISPAM (Sepsis and Mean Arterial Pressure) Investigators studied 776 septic shock patients, and found that targeting a MAP of 80-85 mm Hg rather than 65-70 mm Hg did not result in any difference in 28- or 90-day mortality.43 A vasopressor should be considered when MAP of ≥65 mm Hg cannot be maintained despite adequate fluid resuscitation.

For the choice of vasopressors, studies that compared norepinephrine (with or without additional dobutamine in patients with low CO) with epinephrine found no difference in all-cause 28-day mortality,44 or the time to achievement of a clinician-prescribed MAP goal.45 Epinephrine use was associated with significant tachycardia and lactic acidosis that did not affect haemodynamic stabilisation or survival. The hyperlactataemia represents exaggerated aerobic glycolysis instead of ongoing tissue hypoxia, but potentially interferes with interpretation of serial serum lactate measurements. The 2008 VASST (Vasopressin and Septic Shock Trial)46 randomised 779 septic shock patients to receive either norepinephrine alone or norepinephrine plus low-dose vasopressin (0.03 U/min), and found no difference in all-cause 28-day mortality. The SOAP II Trial randomised 1679 patients with shock to receive either norepinephrine or dopamine, and found no difference in all-cause 28-day mortality but a significantly higher rate of arrhythmias in the dopamine group (the number needed to harm was 9). In their subgroup analysis, dopamine was associated with higher mortality in cardiogenic shock, but not septic and hypovolaemic shock.47

Norepinephrine is therefore recommended as the first-line vasopressor for septic shock. In refractory hypotension, epinephrine or low-dose vasopressin (0.03 units/min) may be added. Dopamine should be avoided except in highly selected patients who are bradycardic and at low risk of tachyarrhythmias.

The optimal goal for sepsis resuscitation remains unknown. While under resuscitation is detrimental, achieving supranormal targets has also been shown to cause harm.48 The MAP (perfusion pressure) of ≥65 mm Hg and urine output of ≥0.5 mL/kg/h are the recommended targets. The EMShockNet Trial showed that there was no difference in hospital mortality when using lactate clearance (>10%) or ScvO₂ (>70%) as goals of early sepsis resuscitation.49 Hyperlactataemia in sepsis, however, can result from increased production driven by endogenous or exogenous epinephrine-stimulated aerobic glycolysis, endotoxin inhibition of pyruvate dehydrogenase, and decreased lactate metabolism due to liver and renal dysfunction. Thus, persistent hyperlactataemia does not necessarily indicate anaerobic metabolism and tissue hypoxia, and should not be solely relied on to guide therapy that aims to boost oxygen delivery in patients who are otherwise clinically improving. Conversely, normalisation of serum lactate is reassuring as it is associated with reduced hospital mortality in critically ill patients.50

In the original EGDT protocols, once ScvO₂ drops below 70%, blood transfusion to achieve a haematocrit level of ≥30% was recommended to boost oxygen delivery. The 1999 TRICC (Transfusion Requirements in Critical Care) trial demonstrated lower rates of in-hospital mortality with a restrictive rather than liberal transfusion strategy. This trial, however, excluded septic shock patients. The 2014 TRISS (Transfusion Requirements in Septic Shock) trial randomised 998 septic shock patients to either a liberal blood transfusion strategy with a transfusion threshold of haemoglobin of ≤90 g/L or a restrictive strategy with a threshold of ≤70 g/L.51 Mortality at 90 days, rate of ischaemic events, and use of life support were similar. A transfusion threshold of 70 g/L is therefore recommended. For patients with ongoing acute coronary syndrome or chronic cardiovascular disease, targeting a higher haemoglobin level of 100 g/L might be beneficial but remains to be proven.52

Glucocorticoids have anti-inflammatory and immunosuppressive effects. Despite the positive Annane Trial in 2002, the subsequently larger multicentre CORTICUS (Corticosteroid Therapy of Septic Shock) Trial53 was negative. It randomised 499 patients with septic shock to receive 6-hourly 50-mg hydrocortisone or placebo, with the dose tapered over 11 days. Hydrocortisone did not improve 28-day survival in patients with septic shock, and should not be routinely used for septic shock before adequate fluid resuscitation and vasopressor therapy.14 If used in refractory shock, early administration within 9 hours of commencement of vasopressor is advised.54

Tight glycaemic control (blood glucose, 4.4-6.1 mmol/L) was once commonly practised after the 2001 Leuven Surgical Trial. In 2009, the NICE-SUGAR (Normoglycemia in Intensive Care Evaluation—Surviving Using Glucose Algorithm Regulation) study randomised 6104 ICU patients, and showed that intensive glucose control (4.5-6.0 mmol/L) increased mortality compared with a target of <10 mmol/L.55 Post-hoc analysis further demonstrated an association between hypoglycaemia and an increased risk of death in a dose-response relationship. This association was strongest for death from distributive, including septic shock.56 Guidelines of SSC recommend targeting an upper blood glucose level of <10 mmol/L to reduce the risk of hypoglycaemia.14

The optimal timing of RRT in the absence of overt life-threatening complications (severe metabolic acidosis, hyperkalaemia, and/or fluid overload) is uncertain. Prior studies as well as the recent ELAIN (Early vs Late Initiation of Renal Replacement Therapy in Critically Ill Patients With Acute Kidney Injury57) and AKIKI (Artificial Kidney Initiation in Kidney Injury58) trials have yielded contradictory results, partly because of the heterogeneous definitions of ‘early’ and ‘late’ initiation of RRT. It is hoped that the upcoming IDEAL-ICU (Initiation of Dialysis Early versus Delayed in Intensive Care Unit59) and STARRT-AKI (Standard versus accelerated initiation of renal replacement therapy in acute kidney injury60) trials will provide more evidence on the subject. Regarding the intensity of renal support in critically ill patients with AKI, an effluent rate of 25 mL/kg/h is considered adequate and high-volume haemofiltration is not superior.61 62 Survival benefit of blood purification strategies has yet to be proven.

Of note, ARDS is a frequent complication of sepsis. Optimal ventilatory support prevents further lung injury and the resultant biotrauma from cytokine release. A lung protective strategy with low tidal volume (6 mL/kg ideal body weight) remains the cornerstone of treatment.63 A higher positive end-expiratory pressure should be reserved for patients with moderate-to-severe ARDS as defined by the latest Berlin definition.64 Early (intubated for <36 hours) and sustained (≥16 consecutive hours per day) prone positioning in moderate-to-severe ARDS has proven survival advantage when practised in conjunction with lung protective ventilation.65

Optimal sepsis management involves both refinement of clinical interventions and administrative logistics for the timeliness of their delivery. Early recognition of sepsis, timely source control, prompt and effective antibiotic administration at the right dose, immediate fluid resuscitation as guided by bedside reassessment, and dynamic indices of fluid responsiveness remain the mainstay of sepsis management.

All authors have disclosed no conflicts of interest.

1. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care Med 2013;41:1167-74. Crossref

2. Kaukonen KM, Bailey M, Pilcher D, Cooper DJ, Bellomo R. Systemic inflammatory response syndrome criteria in defining severe sepsis. N Engl J Med 2015;372:1629-38. Crossref

3. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992;101:1644-55. Crossref

4. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801-10. Crossref

5. Wacker C, Prkno A, Brunkhorst FM, Schlattmann P. Procalcitonin as a diagnostic marker for sepsis: a systematic review and meta-analysis. Lancet Infect Dis 2013;13:426-35. Crossref

6. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368-77. Crossref

7. Dellinger RP, Levy MM, Rhodes A, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013;41:580-637. Crossref

8. ProCESS Investigators, Yealy DM, Kellum JA, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med 2014;370:1683-93. Crossref

9. ARISE Investigators; ANZICS Clinical Trial Group, Peake SL, Delany A, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med 2014;371:1496-506. Crossref

10. Mouncey PA, Osborn TM, Power GS, et al. Trial of early, goal-directed resuscitation for septic shock. N Engl J Med 2015;372:1301-11. Crossref

11. Surviving Sepsis Campaign. Updated bundles in response to new evidence. Available from: http://www.survivingsepsis.org/SiteCollectionDocuments/SSC_Bundle.pdf. Accessed 5 Jun 2016.

12. Waibel BH, Rotondo MF. Damage control for intra-abdominal sepsis. Surg Clin North Am 2012;92:243-57. Crossref

13. Ferrer R, Martin-Loeches I, Phillips G, et al. Empirical antibiotic treatment reduces mortality in severe sepsis and septic shock from the first hour: results from a guideline-based performance improvement program. Crit Care Med 2014;42:1749-55. Crossref

14. Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Crit Care Med 2017;45:486-552. Crossref

15. Kumar A, Roberts D, Wood KE, 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;34:1589-96. Crossref

16. Sterling SA, Miller WR, Pryor J, Puskarich MA, Jones AE. The impact of timing of antibiotics on outcomes in severe sepsis and septic shock: a systematic review and meta-analysis. Crit Care Med 2015;43:1907-15. Crossref

17. Almeida M, Ribeiro O, Aragão I, et al. Differences in compliance with Surviving Sepsis Campaign recommendations according to hospital entrance time: day versus night. Crit Care 2013;17:R79. Crossref

18. Vincent JL, Rello J, Marshall J, et al. International study of the prevalence and outcomes of infection in intensive care units. JAMA 2009;302:2323-9. Crossref

19. Kumar A, Safdar N, Kethireddy S, Chateau D. A survival benefit of combination antibiotic therapy for serious infections associated with sepsis and septic shock is contingent only on the risk of death: a meta-analytic/meta-regression study. Crit Care Med 2010;38:1651-64. Crossref

20. Paul M, Benuri-Silbiger I, Soares-Weiser K, Leibovici L. Beta lactam monotherapy versus beta lactam-aminoglycoside combination therapy for sepsis in immunocompetent patients: systematic review and meta-analysis of randomised trials. BMJ 2004;328:668. Crossref

21. de Jong E, van Oers JA, Beishuizen A, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis 2016;16:819-27. Crossref

22. Jamal JA, Economou CJ, Lipman J, Roberts JA. Improving antibiotic dosing in special situations in the ICU: burns, renal replacement therapy and extracorporeal membrane oxygenation. Curr Opin Crit Care 2012;18:460-71. Crossref

23. Choi G, Gomersall CD, Tian Q, Joynt GM, Li AM, Lipman J. Principles of antibacterial dosing in continuous renal replacement therapy. Blood Purif 2010;30:195-212. Crossref

24. Roberts JA, Abdul-Aziz MH, Lipman J, et al. Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect Dis 2014;14:498-509. Crossref

25. Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA 2012;308:1566-72. Crossref

26. Rochwerg B, Alhazzani W, Sindi A, et al. Fluid resuscitation in sepsis: a systematic review and network meta-analysis. Ann Intern Med 2014;161:347-55. Crossref

27. Young P, Bailey M, Beasley R, et al. 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 2015;314:1701-10. Crossref

28. Annane D, Siami S, Jaber S, et al. Effects of fluid resuscitation with colloids vs crystalloids on mortality in critically ill patients presenting with hypovolemic shock: the CRISTAL randomized trial. JAMA 2013;310:1809-17. Crossref

29. Brunkhorst FM, Engel C, Bloos F, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125-39. Crossref

30. Perner A, Haase N, Guttormsen AB, et al. Hydroxyethyl starch 130/0.42 versus Ringer’s acetate in severe sepsis. N Engl J Med 2012;367:124-34. Crossref

31. Myburgh JA, Finfer S, Bellomo R, et al. Hydroxyethyl starch or saline for fluid resuscitation in intensive care. N Engl J Med 2012;367:1901-11. Crossref

32. Guidet B, Martinet O, Boulain T, et al. Assessment of hemodynamic efficacy and safety of 6% hydroxyethylstarch 130/0.4 vs. 0.9% NaCl fluid replacement in patients with severe sepsis: the CRYSTMAS study. Crit Care 2012;16:R94. Crossref

33. Finfer S, Bellomo R, Boyce N, et al. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004;350:2247-56. Crossref

34. Caironi P, Tognoni G, Masson S, et al. Albumin replacement in patients with severe sepsis or septic shock. N Engl J Med 2014;370:1412-21. Crossref

35. Boyd JH, Forbes J, Nakada TA, 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;39:259-65. Crossref

36. Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med 2009;37:2642-7. Crossref

37. Michard F, Boussat S, Chemla D, et al. Relation between respiratory changes in arterial pulse pressure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med 2000;162:134-8. Crossref

38. Royer P, Bendjelid K, Valentino R, Résière D, Chabartier C, Mehdaoui H. Influence of intra-abdominal pressure on the specificity of pulse pressure variations to predict fluid responsiveness. J Trauma Acute Care Surg 2015;78:994-9. Crossref

39. Monnet X, Marik P, Teboul JL. Passive leg raising for predicting fluid responsiveness: a systematic review and meta-analysis. Intensive Care Med 2016;42:1935-47. Crossref

40. Mahjoub Y, Touzeau J, Airapetian N, et al. The passive leg-raising maneuver cannot accurately predict fluid responsiveness in patients with intra-abdominal hypertension. Crit Care Med 2010;38:1824-9. Crossref

41. Mallat J, Meddour M, Durville E, et al. Decrease in pulse pressure and stroke volume variations after mini-fluid challenge accurately predicts fluid responsiveness. Br J Anaesth 2015;115:449-56. Crossref

42. Barbier C, Loubières Y, Schmit C, et al. Respiratory changes in inferior vena cava diameter are helpful in predicting fluid responsiveness in ventilated septic patients. Intensive Care Med 2004;30:1740-6. Crossref

43. Asfar P, Meziani F, Hamel JF, et al. High versus low blood-pressure target in patients with septic shock. N Engl J Med 2014;370:1583-93. Crossref

44. Annane D, Vignon P, Renault A, et al. Norepinephrine plus dobutamine versus epinephrine alone for management of septic shock: A randomised trial. Lancet 2007;370:676-84. Crossref

45. Myburgh JA, Higgins A, Jovanovska A, et al. A comparison of epinephrine and norepinephrine in critically ill patients. Intensive Care Med 2008;34:2226-34. Crossref

46. Russell JA, Walley KR, Singer J, et al. Vasopressin versus norepinephrine infusion in patients with septic shock. N Engl J Med 2008;358:877-87. Crossref

47. De Backer D, Biston P, Devriendt J, et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med 2010;362:779-89. Crossref

48. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med 1994;330:1717-22. Crossref

49. Jones AE, Shapiro NI, Trzeciak S, et al. Lactate clearance vs central venous oxygen saturation as goals of early sepsis therapy: a randomized clinical trial. JAMA 2010;303:739-46. Crossref

50. Nichol A, Bailey M, Egi M, et al. Dynamic lactate indices as predictors of outcome in critically ill patients. Crit Care 2011;15:R242. Crossref

51. Holst LB, Haase N, Wetterslev J, et al. Lower versus higher hemoglobin threshold for transfusion in septic shock. N Engl J Med 2014;371:1381-91. Crossref

52. Docherty AB, O’Donnell R, Brunskill S, et al. Effect of restrictive versus liberal transfusion strategies on outcomes in patients with cardiovascular disease in a non-cardiac surgery setting: systematic review and meta-analysis. BMJ 2016;352:i1351. Crossref

53. Sprung CL, Annane D, Keh D, et al. Hydrocortisone therapy for patients with septic shock. N Engl J Med 2008;358:111-24. Crossref

54. Katsenos CS, Antonopoulou AN, Apostolidou EN, et al. Early administration of hydrocortisone replacement after the advent of septic shock: impact on survival and immune response. Crit Care Med 2014;42:1651-7. Crossref

55. NICE-SUGAR Study Investigators, Finfer S, Chittock DR, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283-97. Crossref

56. NICE-SUGAR Study Investigators, Finfer S, Liu B, et al. Hypoglycemia and risk of death in critically ill patients. N Engl J Med 2012;367:1108-18. Crossref

57. Zarbock A, Kellum JA, Schmidt C, 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;315:2190-9. Crossref

58. Gaudry S, Hajage D, Schortgen F, et al. Initiation strategies for renal-replacement therapy in the Intensive Care Unit. N Engl J Med 2016;375:122-33. Crossref

59. Barbar SD, Binquet C, Monchi M, Bruyère R, Quenot JP. Impact on mortality of the timing of renal replacement therapy in patients with severe acute kidney injury in septic shock: the IDEAL-ICU study (initiation of dialysis early versus delayed in the intensive care unit): study protocol for a randomized controlled trial. Trials 2014;15:270. Crossref

60. Wald R, Adhikari NK, Smith OM, et al. Comparison of standard and accelerated initiation of renal replacement therapy in acute kidney injury. Kidney Int 2015;88:897-904. Crossref

61. RENAL Replacement Therapy Study Investigators, Bellomo R, Cass A, et al. Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med 2009;361:1627-38. Crossref

62. Joannes-Boyau O, Honoré PM, Perez P, et al. High-volume versus standard-volume haemofiltration for septic shock patients with acute kidney injury (IVOIRE study): a multicentre randomized controlled trial. Intensive Care Med 2013;39:1535-46. Crossref

63. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000;342:1301-8. Crossref

64. Briel M, Meade M, Mercat A, 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;303:865-73. Crossref

65. Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med 2013;368:2159-68. Crossref