YJCRC-53276; No of Pages 7

Journal of Critical Care xxx (2019) xxx

Contents lists available at ScienceDirect

Journal of Critical Care

journal homepage: www.journals.elsevier.com/journal-of-critical-care

Arterial vs venous lactate: Correlation and predictive value of mortality of

patients with sepsis during early resuscitation phase
Ata Mahmoodpoor a, Kamran Shadvar a, Sarvin Sanaie b, Samad E.J. Golzari a, Rukma Parthvi c,
Hadi Hamishehkar d, Nader D. Nader e,⁎
a
Department of Anesthesiology & Critical Care Medicine, Faculty of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran
b
Lung Disease and Tuberculosis Research Center, Tabriz University of Medical Sciences, Tabriz, Iran
c
Department of Pulmonary, Critical Care and Sleep Medicine, State University of New York at Buffalo, Buffalo, NY, USA
d
Department of Clinical Pharmacy, Tabriz University of Medical Sciences, Tabriz, Iran
e
Department of Anesthesiology, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA

a r t i c l e i n f o a b s t r a c t

Available online xxxx Purpose: To compare the lactate concentrations obtained from venous to those obtained from arterial blood in
predicting hospital mortality of patients with sepsis and septic shock. To also assess lactate clearance as predictor
for mortality.
Keywords: Methods: 100 patients with septic shock were prospectively enrolled. Serum was sampled at baseline and after
Arterial 6 h of resuscitation from arterial and venous lines. Demographic, severity indices, hemodynamic measures as
Venous well as lactate clearance levels were noted. Data were analyzed for bias and precision.
Blood Results: There was correlation between venous and arterial lactate concentrations at the baseline (R = 0.68) and
Serum lactate
at the 6-hour time point (R = 0.95). Venous concentrations were consistently higher than those obtained from
Correlation
an arterial access by 0.684 mg/dL. Further, arterial lactate level N 3.2 mmol/L and clearance of b20% were consid-
Level of agreement
Mortality ered the cutoff for the mortality risk. While only 8% of the patients with no risk died, all 20 patients who had lac-
tate level N 3.2 mmol/L and clearance of b20% died within the hospital.
Conclusion: Our data suggests a strong correlation between arterial and peripheral venous the lactate levels and
in the initial phase of resuscitation in septic shock patients we can use venous lactate level as biomarker instead
of arterial lactate level. The study also showed that combining lactate levels and its clearance is a reliable predic-
tor of mortality in sepsis.
Published by Elsevier Inc.

1. Introduction
Recent data suggested that severe sepsis and septic shock cause high
mortality. Therefore, septic shock is an important health care problem
and represents an economic burden. The recent Sepsis-3 definition de-
fined sepsis as a dysregulated immune response to infection which re-
sults in organ dysfunction, and septic shock as septic patients who
need vasopressor for keeping a mean arterial pressure N65 mmHg and
lactate N2 mmol/L [1]. During sepsis and circulatory failure an increasing
lactate concentration is the result of a complex condition consisting of
anaerobic and aerobic metabolism through Na/K ATPase channels and
a decrease in lactate consumption [2].
⁎ Corresponding author at: Anesthesiology, UB-Anesthesia, 77 Goodell Street, Suite 550,
Buffalo, NY 14203, USA.
E-mail addresses: sanaies@tbzmed.ac.ir (S. Sanaie), rukmapar@buffalo.edu (R. Parthvi),
nnader@buffalo.edu (N.D. Nader).
Many markers measured in critically ill patients have been used to-
wards the diagnostic and prognostic evaluation for the cost of therapy,
monitoring of adequacy of treatment and its timing [3]. Although it is
unlikely that one marker to possess both diagnostic and prognostic
properties, some markers may be more predictive of a poor outcome
in clinical situations. Lactate measurement is a core principal in the
early diagnosis and appropriate management of sepsis. Basic lactate
concentration and lactate clearance are good markers of hypoperfusion,
organ dysfunction and mortality in septic critically ill patients [4,5]. Lac-
tate clearance monitoring during initial 6 h of resuscitation in septic
shock patients has a good correlation with mortality but its relation
after 6 h of resuscitation is unclear. The peripheral arterial lactate con-
centration has been considered the standard for lactate measurement.
As arterial sampling may not be feasible in all patients, in this study
we compared the correlation between arterial and venous blood lactate
concentrations in adult patients with septic shock who were admitted
to an intensive care unit. We hypothesize that venous concentrations

https://doi.org/10.1016/j.jcrc.2019.05.019
0883-9441/Published by Elsevier Inc.

Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
2 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx
of lactate accurately reflect its arterial concentrations and may be used
to monitor tissue perfusion in critical care settings.
2. Patients and methods
The study design, protocol and the informed consent forms were
reviewed and approved by the institutional review committee on re-
search ethics of Tabriz University of Medical Science on 7 January
2017 under approval number 5D-989801. After obtaining informed
consent from patients or their next of kin, 100 patients with the diagno-
sis of sepsis or septic shock which occurred during the initial 6 h of re-
suscitation were enrolled in this cross-sectional study. Inclusion
criteria were patients with sepsis/septic shock who were diagnosed
during the initial 6 h of resuscitation. Diagnostic criteria for sepsis/septic
shock was based on the new Sepsis-3 definition. Exclusion criteria were
contraindication for arterial puncture and patients b18 years of age.
After diagnosis of sepsis/septic shock a radial arterial line and central
venous access were introduced for all patients. One set of arterial and
venous samples were taken at the baseline from each patient. The ve-
nous samples were taken from central line access hence a tourniquet
was not needed. Patient management in the intensive care unit (ICU)
was started to achieve international guideline targets during the
first 6 and 24 h after admission [6]. After an initial fluid bolus of
20–30 mL/kg, a fluid challenge technique was implemented as long as
hemodynamic improvement was witnessed. Norepinephrine was
started as the first choice of vasopressor to maintain mean arterial pres-
sure (MAP) N 65 mmHg. If a second vasopressor drug was needed, Vaso-
pressin at a dose of 0.03 IU/min was added to the regimen. In the
presence of myocardial dysfunction, defined by low cardiac output, ele-
vated cardiac filling pressures, or ongoing signs of hypoperfusion de-
spite achieving adequate MAP or intravascular volume, inotropic
support was provided by dobutamine or milrinone infusion. If the target
MAP was not achieved with vasopressor therapy, a short term of low
dose hydrocortisone was added. A target hemoglobin level of 7–9 g/dL
was anticipated in all patients except those with ischemic coronary ar-
tery disease in whom the target hemoglobin was set at 10 g/dL. A
targeted glycemic control was performed with insulin regular infusion
to maintain blood glucose levels between 150 and 180 mg/dL. The nutri-
tional risk assessment was performed using online calculated nutri-
tional risk in critically ill (NUTRIC) scores which was originally
described by Heyland et al [7]. Caloric intake was managed via enteral
nutrition unless the patients were unable to tolerate enteral feeding or
needed a high dose vasopressor. If this was the case, parenteral nutrition
protocols were incorporated based on a NUTRIC score of N5 in the first
48 h after the diagnosis of sepsis. All patients received deep vein and
stress ulcer prophylaxis and underwent mechanical ventilation based
on low tidal volume strategy. Another set of arterial and venous
serum samples were taken after completion of the initial 6-hour resus-
citation period.
Demographic characteristics, acute physiologic assessment and
chronic health evaluation (APACHE-II) scores, the sequential organ fail-
ure assessment (SOFA) scores, central venous saturations of oxygen
(SCVO2), MAP, and lactate clearance levels were noted for all patients.
Glomerular filtration rate was calculated using Chronic Kidney Disease
Epidemiology Collaboration (CKD-EPI) equation and was expressed as
mL/min /1.73 m2 [8]. Values below 60 mL/min/1.73 m2 were considered
significant drops in kidney function.
3. Statistical analysis and sample size determination
A study of patients was planned and their values of venous against
arterial concentrations of serum lactate were regressed. Prior data indi-
cated that the standard deviation (SD) of the venous lactate concentra-
tions was 2.01 while the SD of regression errors for the arterial
concentration was 1.931. Online sample size calculator was available
at the website http://www.sample-size.net/correlation-sample-size of
Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
the Institute of Translational Research, University of California in San
Francisco (UCSF). If the true slope of the line obtained by regressing ve-
nous against arterial concentrations was 0.7, we needed to study 22 to
be able to detect this level of correlation with a two-tailed alpha of
0.05 and beta error of 10%. All patient data were analyzed with SPSS
25.0 statistical package (IBM Inc., Chicago, IL). Chi-square analysis was
used for evaluation of correlation between mortality and lactate clear-
ance. Quantitative variables were analyzed with independent sample
t-tests where the numerical variables had a normal distribution. Pear-
son correlation coefficient was used for the relation between venous
and arterial lactate levels. The correlation between venous and arterial
blood samples after adjustment with confounding factors were per-
formed by partial correlation test. Bland-Altman plots were graphed
to estimate the bias (mean) ± precision (1.96 x SD) of [arterial – ve-
nous] concentrations of serum lactate in mg/dL. ANCOVA testing was
used for comparison of arterial lactate level between survivors and
non-survivors.
4. Results
Fifty-seven men and 43 women with septic shock within the initial
6 h of resuscitation were enrolled in this study. The average age of the
patients was 63 ± 11 years old. Demographic and clinical characteristics
of the patients are shown in Table 1. The average serum lactate concen-
trations from the arterial samples at the time of enrollment was 3.0 ±
0.6 mg/dL, while it was 3.7 ± 0.8 mg/dL in those samples obtained
from central venous catheters. Serum lactate concentration significantly
decreased after 6 h of resuscitation both in arterial (mean difference 0.86
with 95% confidence interval of 0.74–0.98 mg/dL) and venous (mean
difference 0.98 with 95% confidence interval of 0.81–1.15 mg/dL)
samples.
Although there was a direct linear correlation between the arterial
concentrations of serum lactate and those obtained from a venous ac-
cess (R2 = 0.587; P b .001), serum lactate levels in venous samples
were consistently higher than those that were simultaneously obtained
from an arterial access (Bias of 0.684 ± Precision of 0.956 mg/dL) at the
time of enrollment (Fig. 1). The level of agreement was even higher in
samples that were obtained after 6 h of resuscitation. Similarly, there
was a stronger linearity between the venous and arterial concentrations
of lactate in the serum (R2 = 0.887; P b .00) with a bias of 0.563 and pre-
cision of 0.785 mg/dL as the venous sampling yielded to comparatively
higher concentrations of serum lactate following 6 h of resuscitation
(Fig. 2).
Receiver-operator characteristics (ROC) analyses were performed
for arterial and venous samples of serum lactate concentrations on ad-
mission, as well as the lactate clearance, so called lactate clearance
from both arterial and venous bloods. The area under the curve (AUC)
of lactate concentrations on admission was higher if it was measured
in blood samples obtained from an artery (0.85 ± 0.05) than those ob-
tained from a venous access (0.74 ± 0.06, P b .001) (Fig. 3A). The cutoff
Table 1
Demographic and baseline characteristics of patients.
Characteristic
Age (years) 63 ± 11
Gender 57 Male
43 Female
Sequential organ failure assessment (SOFA) 14.7 ± 2.1
Acute Physiology and Chronic Health Evaluation (APACHE-II) 30.4 ± 4.5
Lactate level at time of admission (mg/dL)
Venous 3.7 ± 0.8
Arterial 3.0 ± 0.6
Lactate level after 6 h (mg/dL)
Venous 2.7 ± 1.2
Arterial 2.2 ± 1.1
ScvO2 at time of admission (Percent) 65 ± 3
ScvO2 at after 6 h (Percent) 70 ± 4
 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx 3

Fig. 1. On the left Bland-Altman plot that compares arterial and venous concentrations of serum lactate. Bias and precision for this comparison is depicted on the plot. On the right scatter
plot demonstrates the linear correlation between the two sampling methods at the time of admission.

for arterial lactate that predicted ICU mortality was 3.2 mg/dL, while the
cutoff for the venous lactate was 4.0 mg/dL. The sensitivity and specific-
ity for arterial concentration of 3.2 mg/dl was 66% and 94% respectively.
The predictive values both arterial and venous lactate removal within
6 h of resuscitation were significantly higher than the baseline concen-
trations of lactate from an artery at the time of admission (AUCs 0.89 ±
0.04 and 0.85 ± 0.05, respectively) (Fig. 3A). The cutoff value for lactate
clearance that reliably predicted ICU mortality was calculated to be 20%
within 6 h of active resuscitation.
In order to improve the value of serum lactate measurement as bio-
marker in predicting ICU mortality, we combined the admission base-
line concentrations of lactate ≥3.2 mg/dL in arterial samples with and
the arterial lactate clearance of b20% as risk factors for ICU mortality. Pa-
tients were grouped as Group 1 if their admission concentration of the
arterial lactate was b3.2 and had an arterial lactate clearance of ≥20%;
as Group 2 if only one of the two criteria was met or Group 3 if they
met both risk criteria meaning admission lactate levels ≥3.2 with b20%
lactate clearance. The sensitivity and specificity of this model for mortal-
ity prediction were 88% and 95% respectively with a predictive value of
91% based on ROC curve analysis (Fig. 3B).
Table 2 depicts all outcome variables according to this group. Strik-
ingly, all of the 20 patients who had both initial lactate ≥3.2 mg/dL
and lactate removal of b20% after 6 h died. Among the 17 patients in
Group 2, 10 (59%) died, while from 63 patients (33 men and 30
women), patients in group 1 who lacked any of lactate risk, only 5 pa-
tients (8%) died.
5. Discussion
This study evaluated the correlation between arterial and venous
lactate levels during initial resuscitation of patients with septic shock
admitted to an intensive care unit. Results of the study showed that ve-
nous lactate samples can be used instead of arterial samples during ini-
tial resuscitation hours in septic patients. However, arterial
concentration of lactate has a better predictive value compared to its ve-
nous concentration. We further demonstrated that combining the arte-
rial lactate level and the rate of lactate clearance over the period of time
provide the best predictor of outcome during early resuscitation of the
patients suffering from septic shock.
Nichol et al. in their study in 5047 critically ill patients showed that
measurement of dynamic lactate (lactate concentration over time) is

Fig. 2. On the left Bland-Altman plot that compares arterial and venous concentrations of serum lactate. Bias and precision for this comparison is depicted on the plot. On the right scatter
plot demonstrates the linear correlation between the two sampling methods after six hours of resuscitation.

Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
4 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx

Fig. 3. Receiver-operator characteristics curves are plotted to predict hospital mortality by arterial and venous concentrations of serum lactate (A) and its clearance from the blood after 6
hours of resuscitation (B). The value of combined arterial lactate risk in predicting hospital mortality is depicted (C). The AUC for this combined model was 0.91.

the best prognostic factor in septic shock patients [9]. However, arterial threatening distal blood flow (especially in hand) and finally carries
sampling has some disadvantages as it needs greater technical skills, is more cost [10]. On the other hand, peripheral venous samples are usu-
difficult in hypotensive patients, is more invasive, holds a risk of ally influenced by changes in local perfusion and may be falsely

Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx 5

Table 2
Clinical and laboratory information of patients with or without having the lactate risk factor that included either admission arterial concentration of lactate ≥3.2 mg /dL or b 20% decrease
on arterial lactates concentration following 6 h of resuscitation. Group 1 had none of these risk factors, while Group 2 patients had only one risk factor and Group 3 had both factors.

Group 1 Group 2 Group 3 P-Value

N = 63 N = 17 N = 20

Gender
Male 33 (52%) 10 (59%) 14 (70%) 0.377
Female 30 (48%) 7 (41%) 6 (30%)
Age (year) 64 ± 11 61 ± 14 61 ± 12 0.516
Acute Physiology and Chronic Health Evaluation (APACHE-II) 28.4 ± 3.5 31.9 ± 4.2 35.4 ± 2.9 b0.001
Sequential organ failure assessment (SOFA) 14.1 ± 2.1 15.2 ± 2.1 16.5 ± 1.3 b0.001
Blood Urea Nitrogen (mg/dL) 24 ± 4 26 ± 5 29 ± 5 b0.001
Admission levels
Serum Creatinine (mg/dL) 1.4 ± 0.1 1.4 ± 0.2 1.5 ± 0.3 0.004
Admission levels
Glomerular Filtration Rate 47 ± 9 46 ± 10 45 ± 13 0.838
(mL/min/1.73 m2)
Mean Arterial Pressure (mmHg) 67 ± 5 64 ± 6 63 ± 5 0.005
on Admission
Mean Arterial Pressure (mmHg) 71 ± 4 69 ± 6 66 ± 6 b0.001
After 6 h resuscitation
Central Venous Saturation of O2 66 ± 3 65 ± 4 63 ± 3 0.002
on Admission (Percent)
Central Venous Saturation of O2 72 ± 4 69 ± 4 68 ± 4 b0.001
After 6 h resuscitation (Percent)
Mortality with the Intensive Care Unit 5 (8%) 10 (59%) 20 (100%) b0.001

elevated, especially when a tourniquet has been applied to the limb
from which a sample is taken [11].
Gallagher et al. in one study compared peripheral venous lactate
levels with arterial levels in 74 patients and showed a mean difference
of 0.22 mmol/L with 95% confidence intervals of −1.3 to 1.7 mmol/L
[12]. Bloom et al. in a meta-analysis evaluated the role of venous
blood gas monitoring in emergency department and showed that
there is a poor correlation between arterial and venous blood lactate,
however if lactate concentration is normal, the arterial concentration
is likely to be in normal range [13]. There are a few trials about the cor-
relation of arterial and venous blood lactate levels in literature; there is
one trial that compared them in children with septic shock, which
showed a good correlation between venous and arterial lactate in con-
centrations below 2 mmol/L [14].
Samaraweera et al. showed that a blood lactate level of b2 mmol/L
can be used as a surrogate measurement of arterial lactate in children
with septic shock which is similar to our study, but in values
N2 mmol/L arterial samples should be taken to obtain accurate results
[14]. The hemodynamic compromise in septic patients could be the rea-
son to the difference between venous and arterial lactate concentrations
in higher levels of lactate. Another reason for this difference can be due
to the time difference between two samples which was almost one hour
and makes the conclusion unreliable. They also did not report whether
or not to use a tourniquet, which can affect the lactate level. Nascente
and coworkers showed that central vein lactate had a good agreement
with arterial lactate concentration but peripheral vein lactate tends to
overestimate the arterial lactate. This may be due to the metabolic ab-
normalities or hypoperfusion in septic patients [15]. They recom-
mended arterial lactate sampling as the first choice in septic patients
and if it is not available they recommended central venous lactate sam-
pling. In their results there was a great discrepancy between measure-
ments from central route vs. peripheral ones which creates a
possibility of laboratory error or delay in sampling analysis as well as
the possibility of misplacing the tests. Zhou et al. in their letter men-
tioned that arterial and venous lactate concentration cannot be equiva-
lent and could be mixed together because they were two distinctly
different measuring methods which could generate two different re-
sults based on the same patient at the same time point [16]. Bloom
et al. showed a great bias between venous and arterial lactate measure-
ment and found a poor agreement between them especially with the
cut off level of 2 mmol/L in the emergency department [2]. A cumulative
analysis of major studies that have shown correlation between arterial
and venous lactate in adults and pediatric population with sepsis is
depicted in Table 3 [17-28].
Despite a strong correlation between the arterial and venous con-
centrations of serum lactate, there is no universal agreement on using
venous sampling in lieu of arterial sampling. The proponents of using
venous sampling argue for the ease of venous access and its availability
even in the emergency room setting at the time of admission even be-
fore transferring the patient to the ICU and placement of an arterial
line. These clinicians generally refer to the literature that provide an ev-
idence for a good agreement between the arterial-venous agreement of
serum lactate [29]. We have shown here that this agreement is stronger
when the serum concentrations of lactate is below 2 mmol/L. Unfortu-
nately, the correlation between the arterial lactate and venous lactate
decreases as its concentrations in the serum rise. This will pose a big
limitation in using venous sampling for serum lactate measurement
upon admission, as the critically ill patients with higher acuity in more
server condition generally will have higher concentrations of serum lac-
tate where the A-V correlation is the weakest. However, a serial mea-
surement of serum lactate will adequately guide the direction of the
medical management by monitoring the trend of lactate concentration
regardless of its sampling site.
There are some concerns about selection bias; additionally, the au-
thors did not mention the time difference between samples of venous
and arteries [17,19,30]. There are also some studies that reported high
venous lactate as a prognostic factor for mortality in emergency depart-
ment or trauma patients. Mikami et al. evaluated the correlation be-
tween arterial and venous lactate measurement and found a possibly
clinically useful equation to calculate arterial from venous Lac data (ar-
terial Lac = −0.259 + v-Lac × 0.996) [31]. This equation is only affected
by paO2 and other variables do not affect this relation. Kim et al. showed
similar results to ours in patients admitted to an intensive care unit in
many clinical contexts regarding pH, paCO2, bicarbonate and lactate
[21]. Our results also showed a good agreement between mortality
and lactate clearance (arterial and venous), so we can use it as a prog-
nostic factor in septic critically ill patients during initial resuscitation
period.
Using lactate levels for predicting mortality has been a topic of de-
bate for quite some time now. An Australian study of a tertiary ICU cen-
ter has shown that though using lactate levels and its clearance may
serve as biomarker in sepsis, but it may require other parameters to
be included in the assessment [21]. Though the lactate level may be a re-
liable indicator of tissue perfusion, measuring lactate clearance every

Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
6 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx

Table 3
Major studies that show significant correlation between arterial and venous lactate levels.

Author/year of study Type of study Number of patients Recommendation

Correlation of arterial and venous lactate in adult population with sepsis

Theerawit et al [28].
Theerawit et al [29].
Contenti et al [20].
Marti & Machado [24]
Kim et al [23].
Réminiac et al [27].
Pattharanitima et al [25].
Weil et al [30].
Nascente et al [16].
Prospective, cross-sectional study 63 paired samples/
Prospective, cross-sectional study 73 pair-samples
Prospective 103 patients
Prospective, cross-sectional study 26 patients/107 paired
samples
Prospective 34 patients 151 paired
samples
Retrospective study 188 patients/673 samples
Prospective, cross-sectional study 30 patients
Prospective 12 patients/50 paired samples
Cross-sectional 32 patients/ 238 paired
samples
VL cannot replace AL however VL ≥ 4.5 can predict AL ≥ 4
mmol/L
VL can be used in place of AL b4mml/l.
VBL is not as effective as ALB
Femoral VL may be used as AL substitute.
Other variables can be used as AL equivalent
VL can substitute AL when drawn within 30 min
Central VL may be used in place on Al when it is b10 mmol/L
Very high correlation found between AL and VL
Central VL may replace Al but not peripheral VL.

Correlation of arterial and venous lactate in pediatric population with sepsis

Phumeetham et al [26]. Prospective, cross-sectional study 48 patients Very high correlation found between AL and VL
Samarraweera et al [15]. Retrospective 60 patients VL ≤ 2 mmol/may be used but AL required for higher values.
Fernandez Sarmiento et al Longitudinal Retrospective observational 42 patients Good correlation was found between AL and VL
[21]. study

1–2 h has shown to be associated with improved outcomes in septic pa-
tients regardless of the initial lactate level [32]. Hence in recent times a
greater impetus is placed on lactate clearance [33-35]. One study has
shown that the clearance of lactate was the strongest predictor of mor-
tality during early resuscitation in sepsis in emergency department [36].
Similar results were seen in critical care settings [9,37]. Since there are
various mechanisms and kinetics involved in lactate levels in stable pa-
tients, it may not be of value in hemodynamically stable patients [20]. In
our study we combined the initial lactate and it clearance and found
100% mortality in patients with higher lactate level and poor clearance
as shown in Table 4.
Masyuk et al. evaluated delta-lactate (ΔLac) 24 h after admission
(Δ24Lac) to an ICU in critically ill patients [38]. Inclusion criterion was
a lactate level at admission above 2.0 mmol/L. According to them
Δ24Lac might constitute an independent, easily available and important
parameter for risk stratification in the critically ill. Masyuk et al. in their
study showed that ΔLac ≤19% during 24 h was further associated with
increased long-term mortality (HR 2.22; 95%CI 1.90–2.61; p b .001; re-
gardless of admission diagnosis which remained after correction for
APACHE II scores (HR 1.53; 95%CI 1.27–1.85; p b .001), SAPS II scores
(HR 1.46; 95%CI 1.21–1.76; p b .001). [39] We can also do it for APACHE.
6. Limitation of study
This is a single center study which is performed only in surgical sep-
tic patients, but for generalizing the results we require future trials with
more heterogenous critically ill septic patients. We also didn't define a
cut off value for our lactate levels. The evidence suggests that an analysis
of venous lactate levels provides a useful screen. Venous and arterial lac-
tates agree reasonably at all values, but the agreement is highest at nor-
mal values. More work is required to ascertain the lactate in order to
confidently use venous values as a screen for critical illness. Further
work is also needed to ascertain the value of the venous lactate analysis
in assessing response to treatment.
Table 4
Predicting mortality with serum levels of Lactate in arterial blood.
Risk groups Group 1 (Low) Group 2 (Medium) Group 3 (High)
Lactate level b 3.2 mmol/L and b 3.2 mmol/L or N 3.2 mmol/L and
Lactate clearance ≥ 20% ≥ 20% b 20%
Mortality 8% 59% 100%
7. Conclusion
Our data suggest that different variables have only minimal impact
on the relationship between arterial and peripheral venous samples,
and in the initial phase of resuscitation in septic shock patients we can
use venous lactate measurement instead of arterial lactate measure-
ment. Additionally, when using lactate as a biomarker, both initial lac-
tate level and its clearance should be used to predict prognosis in
critically ill septic patients.
Conflict of interests
No funding was requested.
No conflict of interest was declared by the authors.
Acknowledgement
AM contributed literature search, data collection, study design, anal-
ysis of data, manuscript preparation, and review of manuscript; KS con-
tributed to literature search, clinical care and review of manuscript; SS
contributed to literature search and manuscript preparation; MESG
contributed to data collection and manuscript preparation; RP contrib-
uted to literature search and manuscript preparation; HH contributed
to study design and data collection and NDN contributed to study de-
sign, literature search, data analysis & interpretation and critical review
of the manuscript.
References
[1] Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al.
The third international consensus definitions for sepsis and septic shock (Sepsis-3).
JAMA 2016;315(8):801–10.
[2] Bloom B, Pott J, Freund Y, Grundlingh J, Harris T. The agreement between abnormal
venous lactate and arterial lactate in the ED: a retrospective chart review. Am J
Emerg Med 2014;32(6):596–600.
[3] Bakker J, Nijsten MW, Jansen TC. Clinical use of lactate monitoring in critically ill pa-
tients. Ann Intensive Care 2013;3(1):12.
[4] Jones AE, Shapiro NI, Trzeciak S, Arnold RC, Claremont HA, Kline JA, et al. Lactate
clearance vs central venous oxygen saturation as goals of early sepsis therapy: a ran-
domized clinical trial. JAMA 2010;303(8):739–46.
[5] Mahmoodpoor A, Shadvar K, Saghaleini SH, Koleini E, Hamishehkar H, Ostadi Z, et al.
Which one is a better predictor of ICU mortality in septic patients? Comparison be-
tween serial serum lactate concentrations and its removal rate. J Crit Care 2018;44:
51–6.
[6] Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving
sepsis campaign: international guidelines for management of severe sepsis and sep-
tic shock: 2012. Crit Care Med 2013;41(2):580–637.

Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
 A. Mahmoodpoor et al. / Journal of Critical Care xxx (2019) xxx
[7] Heyland DK, Dhaliwal R, Jiang X, Day AG. Identifying critically ill patients who ben-
efit the most from nutrition therapy: the development and initial validation of a
novel risk assessment tool. Crit Care 2011;15(6):R268.
[8] Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro III AF, Feldman HI, et al. A new
equation to estimate glomerular filtration rate. Ann Intern Med 2009;150(9):
604–12.
[9] Nichol A, Bailey M, Egi M, Pettila V, French C, Stachowski E, et al. Dynamic lactate in-
dices as predictors of outcome in critically ill patients. Crit Care 2011;15(5):R242.
[10] Adams II J, Hazard P. Comparison of blood lactate concentrations in arterial and pe-
ripheral venous blood. Crit Care Med 1988;16(9):913–4.
[11] Benzon HT, Toleikis JR, Meagher LL, Shapiro BA, Ts'ao CH, Avram MJ. Changes in ve-
nous blood lactate, venous blood gases, and somatosensory evoked potentials after
tourniquet application. Anesthesiology 1988;69(5):677–82.
[12] Gallagher EJ, Rodriguez K, Touger M. Agreement between peripheral venous and ar-
terial lactate levels. Ann Emerg Med 1997;29(4):479–83.
[13] Bloom BM, Grundlingh J, Bestwick JP, Harris T. The role of venous blood gas in the
emergency department: a systematic review and meta-analysis. Eur J Emerg Med
2014;21(2):81–8.
[14] Samaraweera SA, Gibbons B, Gour A, Sedgwick P. Arterial versus venous lactate: a
measure of sepsis in children. Eur J Pediatr 2017;176(8):1055–60.
[15] Nascente AP, Assuncao M, Guedes CJ, Freitas FG, Mazza BF, Jackiu M, et al. Compar-
ison of lactate values obtained from different sites and their clinical significance in
patients with severe sepsis. Sao Paulo Med J 2011;129(1):11–6.
[16] Zhou X, Ye Y, Tian F, Wu F. Lactate levels in arterial and venous blood may be corre-
lated but not equivalent. J Crit Care 2017;40:267–8.
[17] Callaway DW, Shapiro NI, Donnino MW, Baker C, Rosen CL. Serum lactate and base
deficit as predictors of mortality in normotensive elderly blunt trauma patients. J
Trauma 2009;66(4):1040–4.
[18] Contenti J, Corraze H, Lemoel F, Levraut J. Effectiveness of arterial, venous, and cap-
illary blood lactate as a sepsis triage tool in ED patients. Am J Emerg Med 2015;33
(2):167–72.
[19] Fernandez Sarmiento J, Araque P, Yepes M, Mulett H, Tovar X, Rodriguez F. Correla-
tion between arterial lactate and central venous lactate in children with sepsis. Crit
Care Res Pract 2016;2016:7839739.
[20] Jansen TC, van Bommel J, Bakker J. Blood lactate monitoring in critically ill patients: a
systematic health technology assessment. Crit Care Med 2009;37(10):2827–39.
[21] Kim BR, Park SJ, Shin HS, Jung YS, Rim H. Correlation between peripheral venous and
arterial blood gas measurements in patients admitted to the intensive care unit: a
single-center study. Kidney Res Clin Pract 2013;32(1):32–8.
[22] Marti YN, Machado FR. Use of femoral vein catheters for the assessment of perfusion
parameters. Rev Bras Ter Intensiva 2013;25(2):168–74.
[23] Pattharanitima P, Tongyoo S, Ratanarat R, Wilachone W, Poompichet A, Permpikul C.
Correlation of arterial, central venous and capillary lactate levels in septic shock pa-
tients. J Med Assoc Thai 2011;94(Suppl. 1):S175–80.
Please cite this article as: A. Mahmoodpoor, K. Shadvar, S. Sanaie, et al., Arterial vs venous lactate: Correlation and predictive value of mortality of
patients with sepsis du..., Journal of Critical Care, https://doi.org/10.1016/j.jcrc.2019.05.019
7
[24] Phumeetham S, Kaowchaweerattanachart N, Law S, Chanthong P, Pratumvinit B.
Close correlation between arterial and central venous lactate concentrations of chil-
dren in shock: a cross-sectional study. Clin Chim Acta 2017;472:86–9.
[25] Reminiac F, Saint-Etienne C, Runge I, Aye DY, Benzekri-Lefevre D, Mathonnet A, et al.
Are central venous lactate and arterial lactate interchangeable? A human retrospec-
tive study. Anesth Analg 2012;115(3):605–10.
[26] Theerawit P, Na Petvicham C. Correlation between arterial lactate and venous lactate
in patients with sepsis and septic shock. Crit Care 2014;18(Suppl. 1):P177.
[27] Theerawit P, Na Petvicharn C, Tangsujaritvijit V, Sutherasan Y. The correlation be-
tween arterial lactate and venous lactate in patients with sepsis and septic shock. J
Intensive Care Med 2018;33(2):116–20.
[28] Weil MH, Michaels S, Rackow EC. Comparison of blood lactate concentrations in cen-
tral venous, pulmonary artery, and arterial blood. Crit Care Med 1987;15(5):489–90.
[29] van Tienhoven AJ, van Beers CAJ, Siegert CEH. Agreement between arterial and pe-
ripheral venous lactate levels in the ED: a systematic review. Am J Emerg Med
2019;37(4):746–50.
[30] Trzeciak S, Dellinger RP, Chansky ME, Arnold RC, Schorr C, Milcarek B, et al. Serum
lactate as a predictor of mortality in patients with infection. Intensive Care Med
2007;33(6):970–7.
[31] Mikami A, Ohde S, Deshpande GA, Mochizuki T, Otani N, Ishimatsu S. Can we predict
arterial lactate from venous lactate in the ED? Am J Emerg Med 2013;31(7):
1118–20.
[32] Moran JL, Santamaria J. Reconsidering lactate as a sepsis risk biomarker. PLoS One
2017;12(10):e0185320.
[33] Nguyen HB, Rivers EP, Knoblich BP, Jacobsen G, Muzzin A, Ressler JA, et al. Early lac-
tate clearance is associated with improved outcome in severe sepsis and septic
shock. Crit Care Med 2004;32(8):1637–42.
[34] 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;345(19):1368–77.
[35] Vincent JL, Quintairos ESA, Couto Jr L, Taccone FS. The value of blood lactate kinetics
in critically ill patients: a systematic review. Crit Care 2016;20(1):257.
[36] Puskarich MA, Trzeciak S, Shapiro NI, Albers AB, Heffner AC, Kline JA, et al. Whole
blood lactate kinetics in patients undergoing quantitative resuscitation for severe
sepsis and septic shock. Chest 2013;143(6):1548–53.
[37] Jansen TC, van Bommel J, Schoonderbeek FJ, Sleeswijk Visser SJ, van der Klooster JM,
Lima AP, et al. Early lactate-guided therapy in intensive care unit patients: a multi-
center, open-label, randomized controlled trial. Am J Respir Crit Care Med 2010;
182(6):752–61.
[38] Masyuk M, Wernly B, Lichtenauer M, Franz M, Kabisch B, Muessig JM, et al. Prognos-
tic relevance of serum lactate kinetics in critically ill patients. Intensive Care Med
2019;45(1):55–61.