• 
• 

Three Part Question

In [an adult with septic shock receiving vasopressors] is [serum lactate] a [predictor of mortality]?

Clinical Scenario

A 40 year old male presents to the emergency department in septic shock. He is given fluid and vasopressor therapy. Can serial serum lactate measurements be used as a predictor of mortality?

Search Strategy

Studies were then identified by using 3 electronic databases (PubMed, Web of Science and The Cochrane Library) from January 1990 to November 2020.
((septic shock[Title/Abstract]) AND (lactate[Title/Abstract])) AND (vasopressors)) Filters applied: Adult: 19+ years

Search Outcome

15 articles were included in the review from the electronic databases, one of which was found from the reference list.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Omar et al. 2011 South Africa	40 patients admitted to an intensive care unit in tertiary hospital over a period of 6 months. A baseline blood sample was taken for lactate measurement. Adrenaline was the vasopressor used in all patients. Adrenaline was escalated in order to achieve the target mean arterial pressure, at which the dose required was recorded. Blood lactate was then measured every two hours over the next 24 hours. The study quantified a ratio of the change in whole blood lactate concentration to the increase in adrenaline dosage over eight hours as the lactate index (LI). Inclusion criteria: over 18 years and initiated on adrenaline for the treatment of septic shock. Exclusion criteria: known adrenal disease, secondary hypertension, known use of monoamine oxidase inhibitors and dobutamine use.	Prospective cohort study (1b)	Serum lactate levels	Median baseline lactate (0 hour): 2.3 mmol/L Lactate (8 hours): 2.9 mmol/L Peak lactate occurred within eight hours in all participants. The peak lactate correlated with peak adrenaline.	1. Mortality was calculated at ICU discharge so, whether lactate index is predictive of 28-day mortality is unsure. 2. There was a small study size with only 40 patients, all of which were ICU-associated sepsis. Further study would need to include all cases of septic shock. 3. The study was observational hence potential confounders were not able to be controlled for. For example, any fluid resuscitation before study inclusion was not determined.
			In-hospital mortality	Predicted hospital mortality at ICU admission: 44%. The mortality rate at ICU discharge: 52.5% (n=21). Survivors had a significantly higher lactate index than non-survivors.
Wilson et al. 1992 South Africa	15 patients admitted to an intensive care unit in tertiary hospital over a period of 12 months. Inclusion criteria: septic shock (systolic blood pressure (SBP) of 90 mmHg or less and in need for vasopressor support. Patients also had a pulmonary artery catheter in situ. No exclusion criteria were stated. Baseline measurements were taken including serum lactate. The vasopressor used was adrenaline. The infusion was continuous at a dose on 0.025 ug/kg/min and was increased by 0.025 every 20 minutes until a SBP of 120 mmHg was sustained.	Prospective cohort study (1b)	Serum lactate levels	Baseline lactate (0 hour): 2.2 mmol/L Lactate at maximum adrenaline dose: 3.5 mmol/L Change in lactate levels: 1.3 mmol/L (59%) Lactate concentration increased in correlation with incremental adrenaline doses.	1. Lactate levels were only measured in 13 patients 2. Lactate levels were only measured 1 hour after infusion with adrenaline. 3. There was a small study size with only 15 patients 4. Mortality rate was calculated at point of ICU discharge with no patient follow up
			In-hospital mortality rate	In-hospital mortality rate: 66% (n=10).
Day et al. 1996 UK	23 patients admitted to a tertiary hospital over a period of 18 months. Inclusion criteria: septic shock defined by a SBP of less than 80 mmHg and a venous plasma lactate concentration greater than 4.0 mmol/L. Patients also included in the study if they had severe falciparum malaria. 10 patients included had septic shock, 13 included had malaria. 9 patients were randomised to receive adrenaline, and 14 received dopamine. No exclusion criteria were stated.	Randomised control trial (1a)	Lactate trends after adrenaline infusion	Baseline lactate (0 hour): 4.4 mmol/l (P<0.001) Lactate at maximum adrenaline dose: 7.6 mmol/l (P<0.001) Change in lactate levels: 3.2 mmol/L (65%) (P<0.001) The arterial pH fell below pH<7.345 in the adrenaline group in eight (35%) patients (P=0·07).	1. There was a small study size with only 23 patients 2. The study included 13 patients with malaria and 10 who had septic shock 3. There was no follow-up with patient prognosis or mortality rates calculated
			Lactate trends after dopamine infusion	Baseline lactate (0 hour): 5.4 mmol/l (P<0.01) Lactate at maximum dopamine dose: 4.4 mmol/l (P<0.01) Change in lactate levels: -1.0 mmol/L (-19%) (P<0.01) Arterial pH fell below pH<7.345 in two (9%) patients (P=0·07).
			Sepsis-specific in- hospital mortality	Sepsis-specific mortality rate: 30% (n=3)
Menezes et al. 2019 Brazil	106 patients admitted to a tertiary hospital were included between January 2015 to October 2017. Inclusion criteria: Over 18 years old and septic shock defined as sepsis plus a MAP <65 mmHg and a serum lactate above 2.0 mmol/L even after initial volume expansion and requiring vasopressors. Exclusion criteria: cardiac arrhythmias, severe bleeding, and severe hepatopathy and coagulopathy. The patients were divided into 4 groups based on presence or absence of hyperlactatemia (>2.0mmol/L) and based on perfusion index peak values above or below 66%. The mortality rate was compared between groups. Noradrenaline was the vasopressor used. Sample size estimates were also calculated.	Prospective cohort study (1b)	Mortality rate in patients with a lactate (<2.0mmol/L)	Patients with a perfusion index of <66%, 33% died (n=10/30 patients) Patients with a perfusion index of >66%, 38% died (n=10/26 patients)	1. Monocentric study performed in a tertiary hospital with high mortality rate 2. The study limited reducibility by only obtaining one measurement 3. Results do not differentiate the body’s response to shock versus the direct effects of vasopressors. 4. Mortality rates were calculated in-hospital with no follow up.
			Mortality rate in patients with a lactate (>2.0mmol/L)	Patients with a perfusion index of <66%, 37% died (n=9/24 patients) Patients with a perfusion index of >66%, 81% died (n=21/26 patients)
			Baseline lactate	Survivors: 1.7 mmol/L Non-survivors: 2.1 mmol/L
Udy et al. 2019 Australia	1102 patients were included in the Australasian Resuscitation In Sepsis Evaluation (ARISE) trial which studied the utilisation of vasopressors (VP), the association between time to VP and 90-day mortality. Early VP group had VP commenced within 4 hours of Emergency Department arrival. Inclusion criteria: sepsis plus hypotension (SBP <90mmHg) and hypoperfusion (blood lactate >4.0mmol/L). VP included are noradrenaline, adrenaline, metaraminol and vasopressin. No exclusion criteria were stated.	Randomised control trial (1b)	Mortality in early vasopressor group	ICU mortality rate: n=158 (15%) Hospital mortality rate: n=212 (19%) 90 day mortality rate: n=252 (23%)	1. The decision to commence vasopressor therapy was at the practitioner’s discretion which can cause a selection bias 2. Volume of fluid administered was recorded at a constant rate over a time period, however, fluid boluses were more likely to be prescribed in clinical practice 3. It is unsure whether any delay in prescribing vasopressors represented an active clinical decision or is due to guidelines at different institutions.
			Mortality in late vasopressor group	ICU mortality rate: n=74 (8%) Hospital mortality rate: n=125 (11%) 90 day mortality rate: n=166 (15%)
			Vasopressor used	Noradrenaline n=203 (78%) Adrenaline n=28 (11%) Metaraminol n=28 (11%) Vasopressin n=2 (1%)
			Fluid resuscitation	The early VP group received significantly less i.v. fluid (3.0 litres) The late VP group received more i.v. fluid 3.8 litres P Over-resuscitation on patients with persistently high lactate was commonly observed.
Ryoo et al. 2018 Korea	1060 patients were included in this study which was performed between January 2010 to January 2016 at single urban tertiary centre. Noradrenaline was the vasopressor used in this study. Inclusion criteria: sepsis plus hypotension (SBP <90mmHg) and serum lactate greater than or equal to 2 mmol/L and vasopressor use. Exclusion criteria: a DNR in place, were transferred to another hospital, refused informed consent or a repeat lactate were not performed. Serum lactate levels were then measured at baseline and 6 hours from septic shock recognition.	Retrospective cohort study (2a)	Serum lactate levels	Survivors: • 0 hour: 4.1 mmol/L • 6 hour: 2.5 mmol/L • P<0.01 • Maximal noradrenaline dose: 0.16ug/kg/min Non-survivors: • 0 hour: 5.3 mmol/L • 6 hour: 4.6 mmol/L • P<0.01 • Maximal noradrenaline dose: 0.32ug/kg/min	1. The study was at a single institution which prohibits the generalisation of the finding to the whole population 2. Of the 1060 patients included, 376 patients had missing values for hour 6 lactate. The missing results were statistically calculated by a biomedical statistician.
			28 day mortality	28-day mortality rate: 25% n=265 Lactate levels above 2.0 mmol/L were associated with mortality (OR, 1.27 [95% CI, 1.21– 1.34] Lactate clearance was associated with mortality (OR 0.992 [95% CI, 0.989–0.995]) Lactate had a higher prognostic value than lactate clearance (AUC, 0.70 vs 0.65; P<0.01)
Barzegar et al. 2014 Iran	30 patients were included in this study which was performed between November 2012 to April 2014 at a tertiary teaching hospital. Inclusion criteria: patients over 18 years with sepsis plus hypotension (SBP <90mmHg) and serum lactate greater than or equal to 2 mmol/L. Exclusion criteria: more than 12 hours after septic shock diagnosis, prior vasopressin use, heart failure (class III or IV of NYHA), Sodium < 130 mmol/L, pregnancy, poor prognosis (death anticipated within hours). Patients were split in two groups; the first group received noradrenaline (NA) infusion. The second group received the same protocol plus vasopressin (AVP).	Randomised controlled trial (1b)	Serum lactate levels in noradrenaline group	Baseline lactate (0 hour): 3.9 mmol/L Lactate at 24 hours: 3.2 mmol/L Lactate at 48 hours: 1.8 mmol/L Change in lactate levels: 2.1 mmol/L decrease of 18% in 24 hours and 54% by 48 hours.	1. This study had a small sample size 2. The study was not blinded 3. Arterial lactate could not be measured, venous lactate was measured instead 4. Early serum lactate (6 hours) was not available.
			Serum lactate levels in noradrenaline plus vasopressin group	Baseline lactate (0 hour): 4.6 mmol/L Lactate at 24 hours: 2.6 mmol/L Lactate at 48 hours: 1.1 mmol/L Change in lactate levels: 3.5 mmol/L decrease of 43% in 24 hours and 76% by 48 hours
			Lactate clearance	Lactate clearance (LC) after 24 hours were 21% with NA and 46% with AVP. There was a significant correlation between NA dose at 24 and 48 hours and lactate clearance at the same times.
			Mortality	ICU mortality noradrenaline: 53.3% n= 8 ICU mortality protocol + vasopressin: 47.5% n= 7 28-day mortality noradrenaline: 46.7% n= 7
Levy et al. 2010 France	40 ventilated patients with sepsis and a control group of 10 ICU patients were included in the study which was performed at a tertiary hospital. Plasma lactate and pyruvate concentrations were measured every 6 hours via a microdialysis probe which was inserted into the quadriceps muscle. There were two measurable outcomes: 1. Whether muscle produces lactate and pyruvate in sepsis patients 2. Whether muscle lactate production is linked to adrenaline levels and the severity of the patient’s condition. Inclusion criteria: septic shock defined as a MAP >65 mmHg, use of vasopressor therapy and a plasma lactate concentration above 2.0 mmol/L. Vasopressor used was adrenaline. Exclusion criteria: patients who have been previously treated with beta-blockers	Randomised controlled trial (1b)	Serum lactate levels	Baseline lactate levels: 2.0 mmol/L (P<0.01) 12 hour lactate levels: 2.9 mmol/L (P<0.05) Serum lactate levels increased by 49% between baseline to hour 12. Adrenaline correlated with an increa in lactate concentration in patients with sepsis but also in healthy volunteers at rest and during exercise.	1. This study had a small sample size 2. Only 15 out of 40 patients developed septic shock within 48 hours after severe sepsis diagnosis 3. There was no follow up of patients
			Gradient between muscle and arterial levels for lactate	Control group: no gradient between muscle and arterial levels for lactate and pyruvate. Sepsis group: muscle lactate and pyruvate concentrations were greater than the arterial levels (P<0.01) and plasma adrenaline levels were also raised (P<0.05). Septic shock group: had higher musculo- arterial gradients of lactate (2.9 vs. 0.7 mmol/l) (P<0.05) and pyruvate (740 vs. 200 mmol/l) (P<0.05), and higher levels of plasma adrenaline (6.2 vs. 2.5 nmol/l) (P<0.05) when compared to control.
Wutrich et al. 2010 France	100 patients over a period of 4 years Inclusion criteria: A. septic shock defined by a MAP of 65 mmHg or below with no improvement after fluid intervention B. no adrenaline infusion at the time of admission C. arterial lactate measurement performed immediately before adrenaline administration Exclusion criteria: A. received adrenaline before ICU admission B. admitted after a cardiac arrest C. younger than 18 years	Retrospective cohort study (2b)	Serum lactate levels	Survivors: • 0 hour: 3.27 mmol/L (P<0.002) • 4 hour (peak): 7.35 mmol/L (P<0.23) Non-survivors: • 0 hour: 5.72 mmol/L (P<0.002) • 4 hour (peak): 8.64 mmol/L (P<0.23)	1. The study sample had a high mortality rate with patients having comorbidities and a high degree of organ failure at the point of admission. 2. Additional elements that could have changed the outcome including ventilation were not specifically documented. 3. Percentage change in lactate levels were not correctly calculated between baseline and peak (hour 4).
			28-day mortality	28-day mortality rate was 72% n=72 Survivors had a higher percentage change in lactate (124%) compared to non- survivors (51%). At 24 hours, lactate concentration was higher in non-survivors than in survivors.
Levy et al. 2005 France	30 patients admitted to an intensive care unit in a tertiary university hospital were included in this study. Inclusion criteria: Patients aged over 18 years with a MAP below 60 mmHg after fluid administration and hyperlactatemia >2.0 mmol/L. No exclusion criteria was stated in the study. Patients were split into two groups. Group 1 received an infusion of noradrenaline- dobutamine. Group 2 received adrenaline titrated to obtain a MAP of 80 mmHg or above. Measurements taken included arterial and mixed venous gases, lactate and pyruvate blood levels were taken at 0, 1, 6, 12, and 24 hours.	Randomised controlled trial (1b)	Serum lactate levels in adrenaline group	Baseline lactate (0 hour): 3.1 mmol/L Lactate at 6 hours: 5.9 mmol/L (p<0.01) Change in lactate levels: 2.8 mmol/L (90% increase)	1. The clinical consequences of changes in pH is unknown 2. There was no follow-up of patients included in this study which resulting in the loss of ability to predict patient outcomes and survival
			Serum lactate levels in noradrenaline- dobutamine group	Baseline lactate (0 hour): 3.1 mmol/L Lactate at 6 hours: 2.7 mmol/L (p<0.01) Change in lactate levels: 0.4 mmol/L (12% decrease)
			Lactate/pyruvate ratio and pH measurements	Adrenaline group: • Lactate/pyruvate ratio increased from 15.5 to 21 (p<0.01) • pH decreased • Both measurements returned to normal within 24 hours Noradrenaline-dobutamine group: • 13.8 to 14 (p<0.01). Remained unchanged. • pH was normalised within 6 hours (p<0.01). • No statistical difference was found between adrenaline and noradrenaline-dobutamine for systemic hemodynamic measurements.
Avni et al. 2015 Israel	32 peer-reviewed randomized controlled trials (RCT) published between 1989–2012 were included in the analysis. 3,544 patients were recruited. Inclusion criteria: patients aged 18 years or above with septic shock defined by a MAP below 60 mmHg after fluid administration and hyperlactatemia >2.0 mmol/L. The intervention assessed was vasopressor versus a different vasopressor, a combination of vasopressors, placebo or no vasopressor. Vasopressors included: dopamine, norepinephrine, epinephrine, phenylephrine, vasopressin and terlipressin. Exclusion criteria: Studies comparing inotrope agents. Studies that assessed different dosages of the same vasopressors.	Systematic Review and Meta-Analysis (1a)	Mortality	Weighted mean all-cause mortality rate: n= 1,595 (45%) 28-day mortality noradrenaline: n=376 (45%) 28-day mortality dopamine n=450 (50%) Noradrenaline and vasopressin were not associated with a benefit in mortality (RR 0.96, 95% CI 0.86–1.04, n = 16, and RR 1.01, 95% CI 0.88–1.15, n = 8, respectively. There was no statistically significant mortality benefit with noradrenaline over adrenaline (RR 0.96, 95% CI 0.77–1.21, n = 4).	1. Patient follow-up ranged from in-hospital (3 trials) to 90 days (median 28 days) 2. The follow up duration was not stated in 21/32 trials 3. The included trials span a long period, between 1989 and 2012, whereby advances in treatment have been introduced to improve care in ICU. 4. Meta-analysis of differences from baseline for lactate levels was not performed. Only end values were reported.
			Serum lactate levels	Baseline lactate levels are lower with noradrenaline compared to vasopressin/terlipressin by a mean 0.23 mmol/L.
Mahmoud et al. 2012 Egypt	60 patients admitted to an intensive care unit in a tertiary hospital were included in this study. Inclusion criteria: patients aged over 18 years with the diagnosis of septic shock defined by MAP below 60 mmHg after fluid administration and hyperlactatemia >2.0 mmol/L. No exclusion criteria was stated. Patients with MAP <70 mmHg were divided randomly into two equal groups.	Randomised controlled trial (1b)	Serum lactate levels in group 1 (noradrenaline and dobutamine)	Baseline: 2.88 ± 0.49 P<0.81 At 24 hours: 2.76 ± 0.42 P<0.11 At 48 hours: 2.30 ± 0.43 P<0.04 Statistically significant at 48 hours	1. Small sample size 2. Single centre study makes it hard to compare to the population. 3. Data was statistically significant at 48 hours
			Serum lactate levels in group 2 (noradrenaline and adrenaline)	Baseline: 2.91 ± 0.29 P<0.81 At 24 hours: 2.96 ± 0.33 P<0.11 At 48 hours: 2.94 ± 0.28 P<0.04 Statistically significant at 48 hours
			28-day all-cause mortality	Group 1: n=15 (50%) Group 2: n=16 (53%)
Martin et al. 2000 France	97 patients admitted to an intensive care unit in a tertiary university hospital were included in this study. Inclusion criteria: Patients aged over 18 years with septic shock defined by MAP below 60 mmHg after fluid administration and hyperlactatemia >2.0 mmol/L No exclusion criteria was stated. Patients were then divided into groups: High dose dopamine n=40 High dose dopamine + adrenaline n=40 Noradrenaline n=57 Noradrenaline + adrenaline n=10	Prospective cohort study (2b)	In-hospital mortality	Noradrenaline: 62% (n=38) Dopamine: 82% (n=33) All-cause mortality:73% n=71	1. No further lactate levels were stated after baseline. They were then recorded as above or below 4mmol/L 2. The study has a non-randomised, open-label, observational design. So a RCT would be needed to establish if noradrenaline improves mortality of patients with septic shock
			Serum lactate levels	Patients receiving noradrenaline Baseline: 5.4 +/- 1.7 mmol/L Patients receiving dopamine Baseline: 5.7 +/- 1.9 mmol/L
			Variables associated with poor outcome and higher hospital mortality	Admission blood lactate concentration > 4 mmol/L (91% vs. 63%, p < .01; relative risk: 1.60; 95% confidence interval: 1.27–1.84) was independently associated with a higher mortality rate.
Shankar-Hari et al. 2016 International	A systematic review and meta- analysis of observational studies which identified 44 papers published between January 1992 and December 2015 with a total of 166,479 patients. The review aimed to determine clinical criteria currently reported to identify septic shock. Inclusion criteria: adults aged over 18 and in septic shock defined as sepsis-induced hypotension despite fluid resuscitation and the presence of hyperlactatemia or organ dysfunction and requiring vasopressors. No exclusion criteria was stated.	Systematic Review and Meta-Analysis (1a)	All-cause mortality	Septic shock-associated crude mortality was 46.5% (95% CI, 42.7%-50.3%) n= 77,413	1. The systematic review did not assess study quality and was restricted to MEDLINE publications only 2. There is no gold standard diagnostic criteria for septic shock which makes it hard to compare the studies. 3. All studies had missing data that could cause a form of selection bias 4. Numerical lactate levels at baseline and at discharge were not documented 5. Type of vasopressor used was not stated in the study making it hard to differentiate between response from vasopressor use
			Serum lactate level associated 28-day mortality	Serum lactate level >2 mmol/L: n= 8520 (42.3%) 95% CI: [41.2-43.3] Serum lactate level >4 mmol/L: n= 1277 (38.3%) 95% CI: [21.5-55.1]
Hyun Oh et al. 2019 South Korea	1043 patients admitted to an intensive care unit in a tertiary university hospital were included in this study. Inclusion criteria: patients aged over 18 years and meet the revised definition of septic shock being MAP below 60 mmHg after fluid administration and hyperlactatemia >2.0 mmol/L. No exclusion criteria was stated. Patients were divided into a high lactate group with serum lactate > 2 mmol/L and the low lactate group with serum lactate 2 mmol/L or less. Vasopressor used was noradrenaline.	Retrospective cohort study (2b)	Mortality	Total mortality rate: 16.0% (164/1,022) High lactate (n=653) • 7-day mortality: n=67 (10.3%) • 28-day mortality: n=102 (15.6%) • P<0.001 Low lactate (n = 369) • 7-day mortality: n= 6 (1.6%) • 28-day mortality: n=18 (4.9%) • P<0.001 High arterial lactate was significantly associated in univariate analysis with a higher rate of mortality when compared to low lactate levels (p =0.037) unadjusted OR (95% CI) 3.366 (1.073–10.555)	1. 21 patients were excluded due to missing data 2. Baseline SOFA and APACHE II scores were greater in the high lactate group than in the low lactate group 3. Data was collected form a single centre 4. Nosocomial infections were excluded causing selection bias 5. Long-term period study which cannot account for changes in technology, techniques and guidelines. 6. Numerical lactate levels after baseline not recorded.
			Baseline serum lactate levels	High lactate (n = 653) 5.55 ± 3.27 mmol/L P<0.001 Low lactate (n = 369) 1.27 ± 0.44 mmol/L P<0.001

Comment(s)

Six papers found that having a raised serum lactate (above 2.0 mmol/L) to be a predictor of mortality. Three papers were indifferent and six papers found having a raised serum lactate is not a predictor of mortality. Evidence from the papers all suggest that having a high initial lactate after fluid resuscitation but before vasopressor therapy is an independent predictor for mortality. However, studies which used adrenaline 1st line observed that a rise in serum lactate in the first 4 hours after infusion is not a predictor of mortality. Those studies using a noradrenaline, dopamine and/or vasopressin protocol found that having consistently raised serum lactate levels is a predictor of mortality.

Clinical Bottom Line

Due to the mixed evidence, it is difficult to come to a conclusion which is in itself an outcome as it is against the current directive which advises noradrenaline to be superior to adrenaline in patients with septic shock requiring vasopressor therapy. A larger study is needed to address the relationship between lactate and vasopressors and help quantify the extent to which a rise in lactate provides beneficial value when in septic shock with a 28-day follow up. This review could be implemented in practice by aiding the generation of a model which could be used on patients in septic shock receiving vasopressors to help select patients for ICU care.

References

1. Omar et al. The relationship between blood lactate and survival following the use of adrenaline in the treatment of septic shock
2. Wilson et al. Septic shock: does adrenaline have a role as a first-line inotropic agent?
3. Day et al. The effects of dopamine and adrenaline infusions on acid-base balance and systemic haemodynamics in severe infection.
4. Menezes et al. Increase of Perfusion Index During Vascular Occlusion Test is Paradoxically Associated With Higher Mortality in Septic Shock After Fluid Resuscitation: A Prospective Study.
5. Udy et al. Incidence, Patient Characteristics, Mode of Drug Delivery, and Outcomes of Septic ShockPatients Treated With Vasopressors in the Arise Trial.
6. Ryoo et al. Lactate Level Versus Lactate Clearance for Predicting Mortality in Patients With Septic Shock Defined by Sepsis-3.
7. Barzegar et al. The Therapeutic Role of Vasopressin on Improving Lactate Clearance During and After Vasogenic Shock: Microcirculation, Is It The Black Box?
8. Levy et al. Increased muscle-to-serum lactate gradient predicts progression towards septic shock in septic patients.
9. Wutrich et al. Early increase in arterial lactate concentration under epinephrine infusion is associated with a better prognosis during shock.
10. Levy et al. Bench-to-bedside review: Is there a place for epinephrine in septic shock?
11. Avni et al. Vasopressors for the Treatment of Septic Shock: Systematic Review and Meta-Analysis.
12. Mahmoud et al. Norepinephrine supplemented with dobutamine or epinephrine for the cardiovascular support of patients with septic shock.
13. Martin et al. Effect of norepinephrine on the outcome of septic shock.
14. Shankar-Hari et al. Developing a New Definition and Assessing New Clinical Criteria for Septic Shock: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).
15. Hyun Oh et al. Risk factors for mortality in patients with low lactate level and septic shock.