This is essentially a synopsis of the Surviving Sepsis Campaign 2016 guideline, which provides very good, up to date advice on approaching patients with sepsis.
The main points of the SSC are:
Treat sepsis as a medical emergency and start treatment and resuscitation immediately
In the case of sepsis induced hypoperfusion, start with IV fluid resuscitation, aiming for at least 30ml/kg within the first 3 hours
After initial fluid resuscitation, a more careful assessment of the patient’s cardiovascular state is needed to determine the next line of therapy.
Assessing fluid status is challenging, but it is recommended to use dynamic rather than static variables for this e.g. passive leg raising, changes in stroke volume with fluid challenge, variation in stroke volume/pulse pressure with changes in intrathoracic pressure (i.e. with mechanical ventilation).
An initial target for mean arterial pressure should be 65mmHg.
Normalisation of lactate could be used as a target for resuscitation.
The use of the old ‘Early Goal Directed Protocols’ is no longer recommended, although it was noted that there was no evidence of harm from it.
Initial Fluid Resuscitation In the context of sepsis induced organ hypoperfusion or septic shock, aggressive treatment of this complication is important. The initial treatment with IV fluid resuscitation in this case is well established, pretty simple to achieve and has a decent pathophysiological basis. The use of 30ml/kg within 3 hours as the guidance for this has limited evidence, but was a common factor within many of the trials (e.g. PROCESS and ARISE). Initial fluid resuscitation allows some ‘thinking time’ to better assess the patient in more depth.
Cardiovascular Assessment After this 30 ml/kg initial resuscitation, a more detailed assessment of the patient’s cardiovascular state is needed if there is persisting hypoperfusion. This is to assess the nature of the next intervention needed - fluids vs vasopressors vs inotropes. It is clear that a full clinical assessment, drawing together a range of routine clinical variables remains an essential part of this (e.g. heart rate, blood pressure). A common clinical question is whether the patient is ‘fluid deplete’ and thus would benefit from further IV fluids so improve their tissue hypoperfusion. It is increasingly recognised that static haemodynamic markers have little role to play in answering this question, a clear example being central venous pressure. Instead, it is recommended that ‘dynamic haemodynamic markers’ are used to aid with this. Examples include:
passive leg raise, monitoring haemodynamic variables
Changes in systolic pressure, pulse pressure of stroke volume with respiratory variation in the ventilated patient.
Mean Arterial Pressure Target Several studies haven’t shown any detectable difference in a range of relevant outcomes with a higher MAP target than 65 mmHg, although cardiac output has been shown to increase. A subgroup of chronically hypertensive patients in one study did show a reduced need for RRT though when a MAP target of 85 mmHg was used. Higher MAP targets generally resulted in more arrhythmias. Given these benefits with no major detectable harms, an initial target MAP of 65 mmHg is recommended.
Antibiotics should be given as soon as possible in both sepsis and septic shock, and definitely within 1 hour.
Initial antimicrobial therapy should be broad spectrum to cover all possible causing pathogens (including viral and fungal if appropriate).
Antimicrobial spectrum should be narrowed once a pathogen is known or clinical improvement is ongoing.
Procalcitonin is suggested as a biomarker to help guide the duration of antimicrobial therapy in patients with sepsis.
Although there is difficulty in getting good quality data (e.g. what time does sepsis start?) there is clinical evidence that shows that delays in antibiotics worsen outcome, including mortality and organ dysfunction. There are 2 landmark papers by Kumar which demonstrate this well. In one study of patients with septic shock, each hour delay in antibiotics was associated with a 7.6% increase in hospital mortality. We have reviewed another paper on this area here: http://www.rapidsequence.org.uk/blog/effective-antibiotic-therapy-and-sepsis
The choice of antimicrobial had to be effective against the suspected pathogen, or there is still a significant rise in mortality and morbidity. The choice of antimicrobial must be of a spectrum that covers all possible pathogens. The potential pathogens will be determined by a spectrum of features of the patient's history including:
Local pathogen epidemiology, both of pathogens and resistance
The pathogens which cause sepsis are primarily gram negative and gram positive bacteria. Less common presentations such as candidiasis should be considered in certain clinical presentations.
A more detailed review of the selection of antimicrobials will be covered elsewhere.
Procalcitonin Several studies have looked at the role of procalcitonin (PCT) in guiding duration of antibiotics (this biomarker is discussed briefly in the sepsis assessment notes). An antibiotic stewardship algorithm has been composed to provide advice on this:
PCT <0.25 microgram/L - Stopping strongly encouraged PCT decreased by > 80% peak or value 0.25 - 0.5 microgram/L - Stopping encouraged PCT decreased by <80% and >0.5 microgram/L - Continuing antibiotics encouraged Increasing PCT compared to peak and >0.5 microgram/L - Change of antibiotics encouraged
There is evidence that this approach can be used without harm arising to the patients, and even suggestion of a mortality benefit in one study. They note that is still has drawbacks and requires integration with other information.
Crystalloids are the fluid of choice for resuscitation and subsequent fluid management.
There is little to guide the choice of crystalloid, but high chloride load should be avoided.
Albumin solution could be considered in cases where substantial amounts of fluid are needed.
Hydroxyethyl starches (HES) should be avoided.
A ‘fluid challenge’ approach is advocated, whereby fluid is continued if there is cardiovascular benefit
As noted above, fluid therapy remains a cornerstone of resuscitation, but the problems with it are increasingly being recognised. There is also a limited amount of research to guide this therapy. The FEAST trial, primarily looking at children with malaria in Africa questioned this approach, but the applicability of this study to our practice seems limited. There is some evidence that a positive fluid balance is a harmful factor on critical care, and hence the caution with excessive fluids. The SSC recommends that the focus on such fluid balance takes place after the initial resuscitation phase though.
Choice of Fluid There is generally no data to support colloids over crystalloids, and similarly little assessment between different crystalloids. There is a single study which suggested more AKI and RRT in a chloride-liberal fluid regime. As such hyperchloraemia should be avoided, and this may make it easier to avoid 0.9% NaCl with its high chloride load. See our review of a paper here (http://www.rapidsequence.org.uk/blog/chloride-restriction-aki) There is now fairly convincing evidence that shows that HESs result in increased harm to patients when used as a resuscitation fluid and in general they have been withdrawn from clinical care.
Albumin The question on albumins role in resuscitation seems limited by quite heterogenous data. There does seem to be a trend amongst the studies and several meta analyses for a mortality benefit from using albumin as a resuscitation fluid. In general, the SSC felt that the quality of evidence was fairly low, but supported its use alongside crystalloids in the cases of large volumes of fluid being needed.
Noradrenaline should be the first line vasopressor
Vasopressin (up to 0.03 U/min) or adrenaline could be added to this to raise the MAP to target.
Vasopressin (up to 0.03 U/min) could be added to noradrenaline with the aim of reducing the noradrenaline dose.
Dobutamine is recommended as an inotrope in patients with persisting hypoperfusion despite adequate fluid loading and vasopressors.
Low dose dopamine for ‘renal protection’ is not recommended.
In general, vasopressors should be targeted against an end point of perfusion. In the context of worsening hypotension or arrhythmia, it should be stopped or reduced.
Noradrenaline Noradrenaline has its effects primarily through vasoconstriction, with little impact on heart rate or stroke volume. It is more effective at dopamine than reversing hypotension in septic shock, and a recent meta-analysis showed a better mortality than dopamine (RR 0.89) and fewer arrhythmias (RR 0.48). As such noradrenaline is the clear first choice vasopressor.
Adrenaline There is no evidence of any survival difference in using adrenaline instead of noradrenaline in clinical studies. There is some laboratory evidence of reduced splanchnic circulation with its use, and the Beta2 effects cause a lactic production, confusing lactate use.
Vasopressin Vasopressin levels in sepsis have been reported to be lower than would be expected for a shocked state, producing the concept of relative vasopressin deficiency. The evidence for any benefit from it remains unconvincing and it is not recommended as a first line vasopressor. There may be some benefit from adding it to noradrenaline at lower doses (no greater than 0.03 U/min).
Myocardial Dysfunction Some myocardial dysfunction may occur due to sepsis in a subset of patients. The SSC note that cardiac output is often maintained by ventricular dilatation, tachycardia and reduced SVR. As such it can be challenging to identify which patient have impaired cardiac function as a component of their oxygen delivery. As such, use of cardiac output measuring alongside measures of perfusion is recommended to help with this assessment. In these patients with impaired cardiac output, inotropic therapy may be beneficial. Routine use of inotrope to achieve supranormal cardiac function is not recommended.
Dobutamine The first line inotrope is dobutamine once adequate left ventricular filling (fluid status) and MAP have been achieved. The majority of clinical evidence just shows improvement in desired physiological values rather than any outcome data. However, it was a key drug in the EGDT protocols and no notable adverse effects were noted from its use.
Alternative inotropes. The are minimal clinical data on the alternative inotropes in sepsis. Levosimendan increases cardiac myocyte calcium responsiveness and therefore has theoretical benefit when considering the pathophysiology of impaired calcium handling in sepsis. However, it remains expensive and the limited clinical evidence shows no benefit over dobutamine.
If adequate fluid resuscitation and vasopressor therapy do not restore cardiovascular stability, corticosteroids (Hydrocortisone 200 mg/day IV) are recommended.
Routine use of steroids in sepsis is not recommended
The benefit from steroids seems to depend on the responsiveness to fluid and vasopressor therapy. Several RCTs showed reversal of shock with steroids when there had been a failure to respond to initial fluid and vasopressor therapy. However, a large multicentre study (CORTICUS) failed to show any mortality benefit from their use.
The evidence of any benefit from low dose steroids is conflicting, with several different systematic reviews presenting different interpretations. Given this uncertainty, there is a weak recommendation against their routine use.
Random cortisol levels may be useful to detect patients with absolute adrenal deficiency. However, their use in septic patients to detect relative adrenal insufficiency are difficult to interpret and have not been demonstrated to be useful.
The duration of steroid use is not clear, with both fixed (3 or 7 days) and titrated regimes described. The SSC recommend tapering steroids when no longer needed to avoid a rebound effect.
A conservative red blood cell transfusion threshold of 7.0g/dL should be employed for patients with sepsis (in the absence of significant extenuating circumstances).
The use of erythropoietin for anaemia is not recommended.
Transfusion of fresh frozen plasma (FFP) is not recommended in the absence of bleeding or planned invasive procedures.
Platelet transfusion thresholds are unchanged in sepsis - 10 x 10^9/L normally, 20 x 10^9/L if high bleeding risk, or 50 x 10^9/L if bleeding or invasive procedures planned.
The evidence for red blood cell transfusion in sepsis shows no difference in outcome with a liberal versus conservative threshold for transfusion. This is based on one specific major trial, but also the EGDT trials where blood transfusion was part of the algorithm. As such, given the well established risks of transfusion in general, a conservation transfusion threshold is recommended.
There is no real evidence to guide advice on the use of FFP to correct the coagulation abnormalities seen in sepsis. It is noted that transfusion often fails to correct the mild abnormalities seen on coagulation testing, in the absence of complications. Thus there is a weak recommendation to avoid the risks associated with transfusion unless complications are present i.e. bleeding, or bleeding may arise from an invasive procedure.
This is the same for platelet transfusion, and the evidence for this is mainly derived from post chemotherapy thrombocytopenia which likely has a very different pathophysiology. However, the best evidence is to use these same transfusion thresholds (as above)
Good Critical Care
As with all critical illness, attention to the details of good critical care are essential. This includes:
Stress ulcer prophylaxis
The majority of the principles of these domains are unchanged in the context of sepsis.
Vitamin C A very interesting small before and after trial looked at the use of vitamin C, hydrocortisone and thiamine as a treatment for septic shock. Though only a small study (47 patients in each arm) of poor quality, the difference in mortality was huge. In hospital mortality was reduced from 40% to 8% by the intervention, and vasopressor duration was significantly reduced. This is very interesting hypothesis generating stuff and not to change practice but well worth a read
The use of IV immunoglobulins in sepsis is not recommended.
The evidence base for the use of IVIG is beset by data quality issues. Some trials do actually suggest a possible benefit to their use, but they have tended to be small with a high risk of bias. When the SSC examined the largest, highest quality trials there was no evidence of benefit. Further, larger clinical trials are needed though.
The use of recombinant activated protein C (APC) is no longer recommended (and no longer available).
The use of antithrombin is not recommended
There is no guidance on the use of thrombomodulin or heparin.
The coagulopathy in sepsis has been considered a possible target for therapy in sepsis. There was initial promise from in this regard, but the PROWESS-SHOCK trial showed no benefit from its use and it was withdrawn from the market by its manufacturers. Whilst the drop in levels of antithrombin is recognised as an adverse feature in sepsis, provided replacement antithrombin has not been shown to be effective in a phase 3 clinical trial. There was an increased rate of bleeding with its use though. There is a suggestion that thrombomodulin may be beneficial in septic patients with DIC, but a larger RCT is currently assessing this. Similarly, the evidence for heparin remains too limited to provide a recommendation.
Kumar, A. 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(6):1589-1596
Kumar, A. et al. Initiation of Inappropriate Antimicrobial Therapy Results in a Fivefold Reduction of Survival in Human Septic Shock. Chest. 2009. 136:1237-1248
Garnacho-Montero, J et al. Adequate antibiotic admission prior to ICU admission in patients with severe sepsis and septic shock reduces hospital mortality. Crit Care. 2015. 19(1): 302
Povoa, P. Salluh, J.Biomarker-guided antibiotic therapy in adult critically ill patients: a critical review. Ann Intensive Care. 2012. 2:32. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3475044/#B40
Bouadma, L et al. Use of procalcitonin to reduce patients’ exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet. 2010. 375: 463-474
Yunos NM et al. Association between a chloride liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012. 308 (15). 1566-72. (Our review available here: http://www.rapidsequence.org.uk/blog/chloride-restriction-aki)
Marik, P. et al. Hydrocortisone, Vitamin C and Thiamine for the Treatment of Severe Sepsis and Septic Shock: A Retrospective Before-After Study. Chest. 2016. Available at: http://journal.publications.chestnet.org/article.aspx?articleid=2593508
Mathieu, S. Hydrocortisone, Vitamin C and Thiamine for the Treatment of Severe Sepsis and Septic Shock: A Retrospective Before-After Study. Thebottomline.com. 2017 (Review of the paper). Available at: http://www.thebottomline.org.uk/summaries/icm/marik/
Marik, P. The Cure for Sepsis. Critical Care Reviews 2017. Vodcast. Available at: http://www.criticalcarereviews.com/index.php/meetings/485-ccr-meetings/ccr-meeting-2017/2896-the-cure-for-sepsis-with-paul-marik