INFECTIOUS DISEASE/ORIGINAL RESEARCH

Emergency Department Management of Sepsis Patients:

A Randomized, Goal-Oriented, Noninvasive
Sepsis Trial
Win Sen Kuan, MBBS, MRCSEd(A&E)*; Irwani Ibrahim, MBBS, FRCSEd(A&E); Benjamin S. H. Leong, MBBS, MRCSEd(A&E);
Swati Jain, MD; Qingshu Lu, PhD; Yin Bun Cheung, PhD; Malcolm Mahadevan, MBBS, FRCSEd(A&E)
*Corresponding Author. E-mail: win_sen_kuan@nuhs.edu.sg.

Study objective: The noninvasive cardiac output monitor and passive leg-raising maneuver has been shown to be
reasonably accurate in predicting fluid responsiveness in critically ill patients. We examine whether using a noninvasive
protocol would result in more rapid lactate clearance after 3 hours in patients with severe sepsis and septic shock in the
emergency department.

Methods: In this open-label randomized controlled trial, 122 adult patients with sepsis and serum lactate concentration
of greater than or equal to 3.0 mmol/L were randomized to receive usual care or intravenous fluid bolus administration
guided by measurements of change of stroke volume index, using the noninvasive cardiac output monitor after
passive leg-raising maneuver. The primary outcome was lactate clearance of more than 20% at 3 hours. Secondary
outcomes included mortality, length of hospital and ICU stay, and total hospital cost. Analysis was intention to treat.

Results: Similar proportions of patients in the randomized intervention group (70.5%; N¼61) versus control group
(73.8%; N¼61) achieved the primary outcome, with a relative risk of 0.96 (95% confidence interval [CI] 0.77 to 1.19).
Secondary outcomes were similar in both groups (P>.05 for all comparisons). Hospital mortality occurred in 6 patients
(9.8%) each in the intervention and control groups on or before 28 days (relative risk¼1.00; 95% CI 0.34 to 2.93).
Among a subgroup of patients with underlying fluid overload states, those in the intervention group tended to receive
clinically significantly more intravenous fluids at 3 hours (difference¼975 mL; 95% CI –450 to 1,725 mL) and attained
better lactate clearance (difference¼19.7%; 95% CI –34.6% to 60.2%) compared with the control group, with shorter
hospital lengths of stay (difference¼–4.5 days; 95% CI –9.5 to 2.5 days).

Conclusion: Protocol-based fluid resuscitation of patients with severe sepsis and septic shock with the noninvasive
cardiac output monitor and passive leg-raising maneuver did not result in better outcomes compared with usual care.
Future studies to demonstrate the use of the noninvasive protocol-based care in patients with preexisting fluid overload
states may be warranted. [Ann Emerg Med. 2015;-:1-12.]

Please see page XX for the Editor’s Capsule Summary of this article.

0196-0644/$-see front matter

Copyright © 2015 by the American College of Emergency Physicians.
http://dx.doi.org/10.1016/j.annemergmed.2015.09.010

INTRODUCTION necessity of central venous pressure monitoring to assess

Background
The global incidence and mortality from severe sepsis
and septic shock remain high.1-3 The Surviving Sepsis
Campaign4 is an initiative to improve mortality from sepsis,
incorporating some principles from the landmark article on
early goal-directed therapy.5 Various resuscitation bundles
have since been formulated and practiced to treat severe
sepsis.6 However, the exact targets of early, quantitative
resuscitation of severe sepsis and septic shock of the
landmark study have been challenged, particularly by 3
recent large multicenter trials.7-9 The generalizability of
early goal-directed therapy is also questioned.10 The
fluid status and blood transfusion also lacks evidence.11-13
Drawbacks include the invasive nature and complications
of central catheter insertion, and the significant amount
of physician and nursing time required to set up the
equipment in the emergency department (ED).14
Lactate levels in the blood have been shown to be a key
indicator of global organ hypoperfusion and shock.15 It has
also been well documented to be a good marker for the
severity and outcome of sepsis; likewise, for the success of
resuscitation.16,17 In addition, if there is a failure of lactate
clearance in the process of resuscitation, a much worse
outcome ensues.18 Studies have shown that the speed with

Volume -, no. - : - 2015 Annals of Emergency Medicine 1

Emergency Department Management of Sepsis Patients
Editor’s Capsule Summary
What is already known on this topic
Structured quantitative resuscitation has been
associated with improved outcomes in patients with
severe sepsis and septic shock. However, structured
noninvasive measurements as goals of resuscitation
have not been studied extensively.
What question this study addressed
Whether a noninvasive measurement of change
stroke volume using a noninvasive cardiac output
monitor and passive leg raise would lead to more
rapid clearance of lactate compared with usual care.
What this study adds to our knowledge
Noninvasive cardiac output monitoring led to no
improvement in lactate clearance compared with
usual care; however, clinically important differences
in amount of fluid administered and lactate clearance
percentage were observed in patients with underlying
volume overload.
How this is relevant to clinical practice
This study informs clinical practice that clinician
judgment and usual care in guiding resuscitation is
equivalent to quantitative noninvasive physiologic
measurements. Future studies of resuscitation should
focus on difficult clinical scenarios such as volume
overload.
which lactate is cleared is an independent prognostic factor
for improved mortality.19-21 As such, targeting lactate
clearance as a surrogate outcome in clinical studies is a valid
endpoint.
Improving cardiac output and end-organ perfusion by
increasing preload through the administration of fluid
boluses remains a cornerstone for resuscitation of patients
with sepsis. However, ascertaining the volume status and
fluid responsiveness of critically ill patients remains a
challenge for physicians practicing in the ED, especially in
patients with impaired cardiac function. There are few data
to show the incidence of preload responsiveness in patients
presenting to the ED. Excessive fluid administration
could have deleterious effects on the individual, such as
interstitial edema and organ dysfunction.22 Hence, there
is a need for noninvasive and uncomplicated alternatives
to advanced hemodynamic assessment in the ED.23
The noninvasive cardiac output monitor (Cheetah
Medical, Tel Aviv, Israel) is capable of providing
2 Annals of Emergency Medicine
Kuan et al
continuous dynamic measurement of hemodynamic
parameters by using bioreactance technology (Appendix
E1, available online at www.annemergmed.com). Clinical
validation studies have shown that the noninvasive cardiac
output monitor has acceptable accuracy.24,25 Noninvasive
monitoring of hemodynamic parameters can be coupled
with the passive leg-raising maneuver, which results in
mobilization of fluid from the venous capacitance system of
the lower extremities to the core, resulting in an effective
volume challenge of 250 to 300 mL (Appendix E2,
available online at www.annemergmed.com). The ability of
passive leg raising to assess fluid responsiveness has been
well demonstrated.26-30
Importance
To our knowledge, no randomized trials have reported
a resuscitation protocol that examines the use of a
noninvasive hemodynamic monitoring device such as the
noninvasive cardiac output monitor in sepsis resuscitation
particularly relating to the outcome of such a strategy. The
use of noninvasive hemodynamic protocols may allow
treating physicians in the ED to be able to quickly attain
hemodynamic optimization and resultant faster lactate
clearance, which could translate into better outcomes for
patients.18,19
Goal of This Investigation
Our goal in this study was to determine whether the use
of noninvasive hemodynamic optimization combined with
passive leg-raising maneuver resulted in early clearance of
lactate in patients with severe sepsis and septic shock in
the ED.
MATERIALS AND METHODS
Study Design and Setting
This was a randomized, open-label, single-center
study enrolling patients with severe sepsis or septic shock
from the ED of National University Hospital, a tertiary
university-affiliated hospital in Singapore. The study period
was during 22 months, from February 2012 to November
2013. Patients or their legally acceptable representatives
provided written informed consent, and all protocols were
approved by the Domain Specific Review Board, National
Healthcare Group, Singapore (DSRB 2011/00286).
Selection of Participants
Eligible patients were those who had serum lactate level
greater than or equal to 3 mmol/L with suspected infection
and fulfilled at least 2 systemic inflammatory response
syndrome criteria: (1) body temperature greater than
Volume -, no. - : - 2015
Kuan et al
38.3 C (100.9 F) or less than 36.0 C (96.8 F); (2) pulse
rate greater than 90 beats/min; (3) respiratory rate greater
than 20 breaths/min or PaCO2 less than 32 mm Hg; and
(4) WBC count greater than 12.0109/L or less than
4.0109/L. Three of the 4 systemic inflammatory response
syndrome criteria (ie, temperature, respiratory rate, and
pulse rate) were obtained as part of triage. In general, if any
patient fulfilled the temperature criterion and at least 1
other criterion with suspected infection, he or she would
have had blood drawn for lactate. Some patients who did
not fulfill the abovementioned criteria may have been
included according to the WBC count of the systemic
inflammatory response syndrome criteria.
Exclusion criteria were known pregnancy, younger
than 21 years, prisoners, do-not-attempt resuscitation
status, known severe aortic insufficiency or severe anatomic
abnormalities of the thoracic aorta, requirement for
immediate surgery, previous above- or below-knee
amputation, inability to do passive leg raising (eg, ankylosis
of hip joint, severe sciatica), and treating physician deemed
aggressive care unsuitable. Additional exclusion criteria
were individuals who received or were receiving b-agonist
therapy (intravenous, oral, nebulized, or inhaled) within
the last 4 hours, those who received 200 mL of crystalloids
out-of-hospital, those unable to provide informed consent,
and a primary diagnosis of trauma, burns, active seizures,
acute cerebral vascular accident, acute coronary syndrome,
status asthmaticus, major cardiac arrhythmias, active
gastrointestinal hemorrhage, seizure, or drug overdose.
All eligible patients were randomized 1:1 to the control
or the intervention group. Blocked randomization, with
randomly permutated block sizes of 4 and 6, was performed
by investigators or research coordinators at the site through
a Web randomization system at the Singapore Clinical
Research Institute, which was not involved in patient
enrollment and assessment. Before patient enrollment, a
randomization list was generated with a validated SAS
program (SAS Institute, Inc., Cary, NC) and then
uploaded to the randomization system. After hospital
admission, all patients continued to have standard care in
the ward or ICU. Clinicians in the ward and ICU were
blinded to the patients’ enrollment in either group. Data
were collected until hospital discharge or death, which-
ever occurred earlier.
Two to 3 research coordinators were available on
weekdays from 8 AM to 10 PM and occasionally on
weekends and night shifts. They screened patients for
eligibility, assessed for inclusion and exclusion criteria,
obtained informed consent from the patients or legally
acceptable representatives, and randomized the patients
into the study. They advised the managing team of
Volume -, no. - : - 2015
Emergency Department Management of Sepsis Patients
physicians and nurses in setting up the noninvasive cardiac
output monitor and performing the passive leg-raising
maneuver according to the protocol. Patients were enrolled
and randomized as soon as informed consent was obtained.
The protocol was continued in the ED for up to 3 hours
or until patients were transferred to the ward or ICU,
whichever occurred sooner.
Interventions
The management entailed a systematic approach in a
stepwise manner optimizing cardiac preload, afterload, and
contractility, with the aim of improving cardiac pump
function and oxygen delivery to the tissues. This included
assessment of fluid responsiveness with the noninvasive
cardiac output monitor and passive leg-raising maneuver
before administration of fluid boluses (Appendices E1 and
E2, available online at www.annemergmed.com). The
passive leg-raising maneuver was performed by first placing
a patient with noninvasive cardiac output monitoring in
a semirecumbent position and obtaining hemodynamic
parameters averaged during 3 minutes. Subsequently, both
lower limbs were raised passively at least 45 degrees from
the horizontal, with the torso supine. The hemodynamic
parameters obtained in this position were also averaged
during the next 3 minutes.
Fluid responsiveness was defined as a stroke volume
index increase of more than 10%. A stroke volume index
increase of more than 20% mandated the administration of
a 1,000-mL intravenous crystalloid bolus to the patient,
whereas an increase greater than 10% but less than or
equal to 20% triggered the administration of a 500-mL
intravenous crystalloid bolus. A change in stroke volume
index less than or equal to 10% and a mean arterial
pressure (MAP) target of greater than or equal to 65 mm
Hg and less than 90 mm Hg were used as hemodynamic
goals. In patients for whom stroke volume index less than
or equal to 10% was achieved, but not the MAP values,
vasoactive agent treatment (first dopamine and then
norepinephrine if required) was initiated to achieve the
MAP targets. Assessment of fluid responsiveness with
passive leg-raising maneuver was performed 10 minutes
after completion of each fluid bolus and every 30 minutes
for patients who were deemed not fluid responsive initially.
The control group was managed under usual care at
the clinician’s discretion in accordance with current best
practice that involved the administration of intravenous
fluids, vasopressors, and inotropes according to the clinical
judgment of the attending physicians managing the patient
to attain a hemodynamic goal of MAP greater than or
equal to 65 mm Hg. Patients in the control group did
not undergo the passive leg-raising maneuver. Research
Annals of Emergency Medicine 3
Emergency Department Management of Sepsis Patients
coordinators charted relevant clinical information for the
study and did not participate in the treatment of patients in
the control group.
In both groups, a central venous catheter could be
inserted should the need for norepinephrine arise.
Outcome Measures
The primary endpoint was lactate clearance greater than
20% at 3 hours after the commencement of resuscitation.
Secondary endpoints included hospital, ICU, and 28-day
mortality, proportion of patients requiring vasoactive
agents, length of hospital and ICU stay, and total hospital
cost. Hospital mortality was defined as death that occurred
inhospital, including ED, ICU, and wards, whereas 28-day
mortality was death that occurred fewer than or equal to 28
days from admission in the hospital. Hemodynamic goals
were considered achieved at 3 hours in the intervention
group if the change in stroke volume index was less than or
equal to 10% and the MAP was greater than or equal to 65
mm Hg and less than 90 mm Hg. The hemodynamic goal
in the control group was MAP greater than or equal to
65 mm Hg. Length of hospital stay was defined as time
from admission to discharge from the hospital or to death
if the patient died before discharge.
We had decided a priori to compare patients with
preexisting fluid overload conditions such as congestive
heart failure and renal failure in a subgroup analysis. This
subgroup was determined from the patient’s history of
congestive cardiac failure (based on previous admission
records or echocardiographic evaluation for heart failure
with reduced ejection fraction of 50%) or end-stage renal
failure requiring hemodialysis or peritoneal dialysis. They
were not randomized separately.
Primary Data Analysis
Proportion of patients achieving the primary endpoint
was tabulated by treatment group as randomized and
compared with Fisher’s exact test. Relative risk (RR) of
achieving the primary endpoint was reported together with
its 95% confidence interval (CI). Log-binomial regression
was performed to adjust for prespecified covariates: age, sex,
disease severity, Charlson comorbidity index, and
physiologic scores.31,32 Continuous outcomes, unless
otherwise stated, were compared between treatment groups
with 2-sample t test. Highly skewed continuous outcomes,
such as length of hospital stay and cost, were compared
with the Mann-Whitney U test. Data analysis was
performed with SAS (version 9.3; SAS Institute, Inc.) and
was made according to the intention-to-treat principle.
Categorical variables are reported in frequency and
percentages and continuous variables in mean (SD). For
4 Annals of Emergency Medicine
Kuan et al
continuous variables that were highly skewed, median
(interquartile range [IQR]) is reported instead. As-treated
analysis was conducted for safety data.
According to our previous prospective 18 months of
data in the study of patients with severe sepsis and septic
shock at our hospital between July 2008 and December
2009, 45% of patients had lactate clearance greater
than 20% between 2 and 6 hours.33 We assumed this
proportion (45%) of patients had lactate clearance greater
than 20% at 3 hours after commencement of fluid
resuscitation. One hundred forty patients (70 in each
group) were required to detect an increase of 25%
(intervention 70% versus control 45%) in proportion of
patients with lactate clearance greater than 20% with a
power of 80% at a type I error of 5%.
RESULTS
Characteristics of Study Subjects
A total of 122 patients were enrolled from the ED of
National University Hospital, Singapore. Patients were
randomized to either control (N¼61) or intervention
group (N¼61) (Figure 1). All patients who were assigned
to the intervention group adhered strictly to the protocol
(Figure 2), apart from 3 patients (4.9%) assigned to the
intervention group who switched to the control group at
the decision of treating physicians. All efficacy analyses,
however, were performed on an intention-to-treat basis.
Baseline characteristics were similar between the 2
groups (Table 1). There were slight imbalances in age, sex,
and physiologic scores, with patients in the intervention
group tending to be older, to be women, and to have
increased severity of illness. However, these differences did
not reach statistical significance. The baseline vital signs
and laboratory parameters were similar between the 2
groups, except for a higher baseline MAP in the
intervention group compared with the control group
(P¼.04) (Table 1).
Compliance with the protocol in the intervention group
was high (95.1%). The following are the descriptions of the
3 patients who switched from the intervention to the
control group: (1) 1 patient with pneumonia had fluid
boluses administered despite not being fluid responsive on
the noninvasive cardiac output monitor and passive leg-
raising maneuver, later developed pulmonary edema, and
subsequently died from severe pneumonia and septic shock
after 3 days; (2) 1 patient with pneumonia was deemed
fluid responsive according to the protocol but had fluid
boluses withheld in the ED because of underlying
congestive heart failure and renal failure, was admitted to
the hospital for 20 days, and subsequently was discharged;
Volume -, no. - : - 2015
Kuan et al
Figure 1. Enrollment flowchart indicating numbers of patients screened for eligibility, enrolled, allocated to each group, switched
over, and analyzed in the study. SIRS, Systemic Inflammatory Response Syndrome.
(3) 1 patient was considered fluid responsive but also had
fluid boluses ceased and stayed in the ICU for 13 days
before dying of pneumonia.
Main Results
In this time-sensitive trial, the median time from
measurement of first lactate level to randomization (ie,
time taken to obtain informed consent) was 9 minutes
(IQR 4 to 18 minutes) in the intervention group and 11
minutes (IQR 5 to 19 minutes) in the control group
(difference¼–2.0; 95% CI –5.0 to 2.0 minutes). The
majority of patients (88.5%) had the passive leg-raising
maneuver performed 2 to 4 times (range 1 to 6) within
the 3 hours. The median time for the performance of the
first passive leg-raising maneuver from randomization was
25 minutes (IQR 15 to 32 minutes). There were 48
patients (78.7%) who were fluid responsive (52.5%
with change in stroke volume index >20% and 26.2%
Volume -, no. - : - 2015
Emergency Department Management of Sepsis Patients
with change >10% and 20%) on the initial passive
leg-raising maneuver in the intervention group. Fewer
patients (60.6%) remained fluid responsive by the time
the second passive leg-raising maneuver was performed
at a median time of 76.5 minutes (IQR 54.5 to 93.0
minutes) after randomization.
For the primary outcome, a total of 43 patients
(70.5%) in the intervention group and 45 (73.8%) in
the control group achieved a lactate clearance of greater
than 20% at 3 hours, with an RR of 0.96 (95% CI 0.77
to 1.19) (Table 2). The RR did not materially change
after adjustment for prespecified covariates. The mean
lactate clearance (ie, percentage of lactate reduction)
was 32.1% in the intervention group and 39.9% in
the control group (difference¼–7.9%; 95% CI –21.1%
to 5.3%).
A total of 19 patients (15.6%; 9 intervention, 10
control) had a serum lactate level increase at 3 hours
Annals of Emergency Medicine 5
Emergency Department Management of Sepsis Patients
Figure 2. Protocol in the intervention group. DSVI, Change in stroke volume index; IV, intravenous; PLR, passive leg raising.
from the initial level. Of these patients, 6 (31.6%) died on
or before day 28 compared with 6 (5.8%) deaths in 103
patients who had lactate clearance (RR¼5.42; 95% CI
1.95 to 15.04). Similarly, of 34 patients who had a lactate
clearance of less than 20% at 3 hours, 8 (23.5%) died on
or before day 28 compared with 4 (4.5%) deaths in 88
patients who achieved a lactate clearance of greater than
20% at 3 hours (RR¼5.18; 95% CI 1.67 to 16.07).
Hospital mortality occurred in 8 patients (13.1%) in the
intervention group, 6 deaths (9.8%) of which happened on
or before 28 days, whereas in the control group the number
was 6 patients (9.8%), all of whom died on or before 28
days of enrollment (RR of 28-day mortality¼1.00; 95% CI
0.34 to 2.93) (Table 3). The median length of stay in the
hospital was 7.0 days in both groups (difference¼0 days;
95% CI –3.0 to 3.0 days). The lengths of stay in the ED
and ICU were comparable between groups (Table 3). The
median total hospital cost was Singapore $7,597 in the
intervention group and Singapore $9,298 in the control
group (difference¼Singapore $1,701; 95% CI
Singapore $8,306 to Singapore $3,197). Analysis of
safety data indicated comparable results in nosocomial
infection, pulmonary edema, and acute kidney injury
(P>.60 for all indicators) (Table E1, available online at
www.annemergmed.com).
The cumulative volume of crystalloids administered was
larger, although not statistically significant, in the
intervention than in the control group at 1, 2, and 3 hours
after commencement of treatment (Table 4). The mean
cumulative volume increased by approximately 700 mL/
6 Annals of Emergency Medicine
Kuan et al
hour for the first 3 hours in each group, attaining values
of approximately 2,000, 4,000, and 5,500 mL at 3, 24,
and 72 hours, respectively. Central venous catheters were
inserted in only 4 patients (3.3%), and 18 patients (14.8%)
required intubation in the ED.
We performed a planned subgroup analysis to test
our original hypothesis in patients with preexisting fluid
overload states such as congestive heart failure and renal
failure (Table 5). There were a total of 17 patients (13.9%)
with congestive cardiac failure or renal failure, 9 in the
intervention group and 8 in the control group. A larger
volume of cumulative fluid was administered at 3 hours in
the intervention group (median¼1,250 mL) than in the
control group (median¼275 mL) (difference¼975 mL;
95% CI –450 to 1,725 mL). The median lactate clearance
was higher at 3 hours in the intervention group (31.2%
versus 11.5%; difference¼19.7%; 95% CI –34.6% to
60.2%), which corresponded to a larger proportion of
patients achieving lactate clearance of greater than 20% at 3
hours (66.7% versus 50.0%; RR¼1.3; 95% CI 0.6 to 3.1).
One death occurred in the control group and none in the
intervention group. The hospital length of stay was shorter
in the intervention group (median¼8 days) than in the
control group (median¼12.5 days) (difference¼–4.5 days;
95% CI –9.5 to 2.5 days).
LIMITATIONS
There are several limitations to our study. First, the
target sample size of 140 patients was not achieved because
Volume -, no. - : - 2015
Kuan et al Emergency Department Management of Sepsis Patients

Table 1. Demographic and baseline disease characteristics, vital Table 2. Lactate clearance.
signs, and laboratory parameters (values are mean [SD] unless Intervention Control Effect Size
indicated otherwise). Variable (N[61) (N[61) (95% CI)
Variable Intervention (N[61) Control (N[61) Arterial lactate, mean Mean difference*
Age, y 66.6 (14.9) 64.3 (14.6) (SD), mmol/L
Female sex, No. (%) 30 (49.2) 21 (34.4) 0h 4.8 (2.0) 4.8 (2.0) –0.1 (–0.8 to 0.7)
Ethnicity, No. (%) 3h 3.2 (2.3) 2.8 (1.8) 0.4 (–0.4 to 1.1)
Chinese 41 (67.2) 42 (68.9) After 3 h† 3.0 (2.4) 2.9 (1.8) 0.1 (–0.7 to 0.9)
Malay 13 (21.3) 11 (18.0) Lactate clearance at 3 h, 32.1 (40.3) 39.9 (33.1) –7.9 (–21.1 to 5.3)
Indian 4 (6.6) 6 (9.8) mean (SD), %
Other 3 (4.9) 2 (3.3) RR‡
Disease severity, Lactate clearance at 3 h 43 (70.5) 45 (73.8) 0.96 (0.77 to 1.19)
No. (%) >20%, No. (%)
Severe sepsis 49 (80.3) 50 (82.0) Hemodynamic goals 41 (67.2) 50 (82.0) 0.82 (0.66 to 1.01)
Septic shock 12 (19.7) 11 (18.0) achieved within 3 h,
Charlson comorbidity 1.7 (2.0) 1.8 (2.3) No. (%)
index* *Mean difference was calculated as intervention group–control group.
Congestive heart failure, 8 (13.1)†‡ 6 (9.8)§ †
Assessment of arterial lactate after 3 hours was conducted for 48 and 49 patients in
No. (%) the intervention and control groups, respectively.
End-stage renal failure, 2 (3.3)† 2 (3.3) ‡
RR was calculated for intervention group versus control group.
No. (%)
Liver cirrhosis, No. (%) 3 (4.9)‡ 3 (4.9)§
Immunosuppressive 5 (8.2)‡ 4 (6.6) of manpower constraints that prohibited continual
state, No. (%)jj
Chronic obstructive lung 0 2 (3.3)
recruitment beyond 122 patients. Although the study did
disease/bronchiectasis, not achieve the calculated target sample size, the 95% CI
No. (%) (0.77 to 1.19) of the RR obtained for the primary outcome
MPM0, %{ 32.9 (19.0) 28.6 (15.5)
SOFA# 3.6 (3.0) 3.2 (2.1)
is relatively narrow, depicting adequate precision and power
MEDS** 6.2 (3.2) 5.5 (3.3) of the study.34 We also analyzed the results of the primary
Arterial lactate, mmol/L 4.8 (2.0) 4.8 (2.0) endpoint under the most extreme scenario. We assumed
Temperature,  C 38.0 (1.3) 38.1 (1.1) that 9 of 18 patients who were to be recruited were
Respiratory rate, breaths/ 24.2 (6.4) 23.8 (6.8)
min randomized to the intervention group and the rest to the
Pulse rate, beats/min 106.7 (19.5) 110.4 (20.7) control group. Under the most extreme scenario in favor of
Systolic arterial pressure, 128.8 (28.7) 120.9 (25.9) the intervention group, all 9 patients in the intervention
mm Hg
MAP, mm Hg 91.3 (19.2) 84.5 (16.7)
group would have achieved the primary endpoint and all 9
pH 7.42 (0.09) 7.44 (0.09) patients in the control group would not have. Combining
PaCO2, mm Hg 28.4 (7.3) 28.7 (9.9) with the data from our 122 recruited patients gives a
WBC count, 109/L 12.2 (6.8) 12.6 (7.2)
Hemoglobin, g/dL 13.0 (2.5) 13.2 (2.3)
proportion of patients achieving the primary endpoint of
Platelets, 109/L 211.0 (97.0) 226.8 (94.5) 74% (52/70) and 64% (45/70) in the intervention and
Creatinine, mmol/L 131.5 (103.1) 156.2 (129.0) control groups, respectively (P¼.27). The converse is also
MPM0, Mortality probability model on admission; SOFA, Sequential Organ Failure true if we assumed the extreme scenario to be in favor of
Assessment; MEDS, Mortality in Emergency Department Sepsis. the control group (P¼.07). The implication is that our
*Charlson comorbidity index, a prediction of 10-year mortality with a range of
comorbid conditions, with higher scores corresponding to higher risk. study conclusion would not have changed even if we had
†
One patient in the intervention group had congestive heart failure and end-stage met the original enrollment goal. Therefore, we maintain
renal failure.
‡
One patient in the intervention group had congestive heart failure, liver cirrhosis, and confidence in the conclusion that there was no difference in
an immunosuppressive state.
§
lactate clearance between the 2 groups.
One patient in the control group had congestive heart failure and liver cirrhosis.
jj
Immunosuppressive state includes patients receiving long-term steroids,
Second, blinding of study personnel and patients was
immunosuppressive agents for autoimmune diseases, or active chemotherapy, or not possible. However, care was taken to minimize bias by
those with advanced HIV infection.
{
MPM0, an ICU severity-of-illness scoring system, with higher scores on admission
complying with the study protocol systematically in the
corresponding to higher prediction for mortality. Components of the score are age, intervention group, and fluid administration occurred
pulse rate, systolic blood pressure, Glasgow Coma Scale score, emergency admission, through the largest-bore intravenous cannulae available
cardiopulmonary resuscitation before admission, cancer, chronic renal failure,
possibility of infection, previous ICU admission within 6 months, and surgical service for both groups of patients. We also ensured that all
at ICU admission.
#
included patients did not receive more than 200 mL of
SOFA, an ICU severity-of-illness scoring system, with higher scores corresponding to
higher prediction for mortality. crystalloids out-of-hospital through verification with the
**MEDS, an emergency clinical prediction rule, with higher scores corresponding to paramedics if the patient was transported to the ED by
higher predicted 28-day mortality rate.
ambulance.

Volume -, no. - : - 2015 Annals of Emergency Medicine 7

Emergency Department Management of Sepsis Patients Kuan et al

Table 3. Mortality, length of hospital stay, and cost.

Variable Intervention (N[61) Control (N[61) Effect Size (95% CI)
Mortality RR*
Hospital mortality, No. (%) 8 (13.1) 6 (9.8) 1.33 (0.49 to 3.61)
ICU mortality, No. (%)† 6 (27.3) 4 (17.4) 1.57 (0.51 to 4.82)
28-day mortality, No. (%) 6 (9.8) 6 (9.8) 1.00 (0.34 to 2.93)
Length of stay, median (IQR) Median difference‡
ED, h 4.0 (3.5 to 5.0) 4.0 (3.5 to 5.0) 0 (–0.5 to 0.5)
ICU, h†§ 96.0 (72.0 to 120.0) 96.0 (60.0 to 144.0) 0 (–35.0 to 42.0)
Hospital, days 7.0 (3.0 to 12.0) 7.0 (4.0 to 14.0) 0 (–3.0 to 3.0)
Total hospital charge, median (IQR), 7,597 (2,588 to 14,989) 9,298 (4,686 to 14,981) –1,701 (–8,306 to 3,197)
Singapore $
*RR was calculated for intervention group versus control group.
†
ICU admission occurred for 22 and 23 patients in the intervention and control groups, respectively.
‡
Median difference was calculated as intervention group–control group.
§
Length of stay in the ICU was summarized for patients admitted to the ICU.

Third, given the relatively high number of excluded included in this study, thereby contributing to a lower
patients (Figure 1) and lack of 24-hour recruitment, the mortality. Notwithstanding, the included population of
sample may not be representative of all the eligible patients patients represents the typical case mix of patients with
at the study site. There was relatively low mortality and a sepsis in the ED who could benefit from early interventions
smaller proportion of patients who were admitted to the and possible avoidance of ICU admission.
ICU or who received vasopressors in this study compared Fourth, reassessment of patients 3 hours into the
with other sepsis studies. We postulate that the lack of resuscitation is brief compared with 6-hour protocols
24-hour recruitment and our stringent inclusion and published previously. The shorter protocol was selected in
exclusion criteria that precluded participation of many accordance with the knowledge of the mean ED length
severely ill patients and those with significant vasculopathy of stay of approximately 4 hours in critically ill patients in
from this study may limit the generalizability of the results. our ED, thus allowing the entire protocol to be completed
That we included patients with a serum lactate level of at within the ED. This approach has been supported by
least 3.0 mmol/L compared with 4.0 mmol/L may reflect experimental evidence that suggested lactate clearance of
the lower baseline risk and better prognosis of patients at least 10% at a minimum of 2 hours after initiation
Table 4. Treatments.
Intervention Control Effect Size
Variable (N[61) (N[61) (95% CI)
Cumulative volume of crystalloids, median Median diff*
(IQR), mL
1h 1,000 (500 to 1,000) 500 (250 to 1,000) 500 (–100 to 500)
2h 1,500 (1,100 to 2,000) 1,100 (600 to 2,100) 400 (–100 to 600)
3h 2,100 (1,600 to 2,550) 1,600 (1,000 to 2,700) 500 (–100 to 1,000)
24 h 3,550 (2,850 to 4,800) 3,400 (2,100 to 5,600) 150 (–1,300 to 900)
72 h 5,100 (3,500 to 6,800) 5,400 (3,250 to 8,200) –300 (–1,600 to 1,350)
Other treatments, No. (%) RR†
Dopamine 10 (16.4) 6 (9.8) 1.7 (0.7 to 4.3)
Norepinephrine 3 (4.9) 3 (4.9) 1.0 (0.2 to 4.8)
Dobutamine 2 (3.3) 2 (3.3) 1.0 (0.2 to 6.9)
Glyceryl trinitrate 3 (4.9) 0 NA
Steroids 1 (1.6) 3 (4.9) 0.3 (0.1 to 3.1)
Blood transfusion 0 0 NA
Antibiotics use 55 (90.2) 51 (83.6) 1.1 (0.9 to 1.2)
Central venous catheter insertion in ED 3 (4.9) 1 (1.6) 3.0 (0.3 to 28.0)
Intubation 10 (16.4) 8 (13.1) 1.3 (0.5 to 3.0)
Time from first lactate measured to antibiotics Median diff*
use,‡ median (IQR), min 57.0 (41.0 to 109.0) 61.0 (33.0 to 93.0) –4.0 (–24.0 to 29.5)
*Mean difference was calculated as intervention group–control group.
†
RR was calculated for intervention group versus control group.
‡
Time was calculated for 55 and 51 patients in the intervention and control groups, respectively, who received antibiotics.

8 Annals of Emergency Medicine Volume -, no. - : - 2015

Kuan et al Emergency Department Management of Sepsis Patients

Table 5. Treatments and outcomes in the subgroup of patients with preexisting fluid overload conditions (values are median [IQR] unless
indicated otherwise).
Variable Intervention (N[9) Control (N[8) Effect Size (95% CI)
Cumulative volume of crystalloids, mL Median diff*
1h 500 (450 to 1,000) 0 (0 to 300) 500 (0 to 1,050)
2h 1,000 (650 to 1,500) 150 (0 to 600) 850 (–50 to 1,500)
3h 1,250 (1,100 to 1,500) 275 (50 to 1,050) 975 (–450 to 1,725)
24 h 1,700 (1,200 to 2,800) 900 (350 to 2,625) 800 (–2,000 to 2,300)
72 h 2,300 (2,000 to 3,050) 2,475 (600 to 6,150) –175 (–5,400 to 2,350)
Other treatments, No. (%) Risk diff*
Dopamine 2 (22.2) 0 22.2 (–13.9 to 54.7)
Norepinephrine 2 (22.2) 0 22.2 (–13.9 to 54.7)
Dobutamine 2 (22.2) 0 22.2 (–13.9 to 54.7)
Glyceryl trinitrate 1 (11.1) 0 11.1 (–22.6 to 43.5)
Steroids 0 0 NA
Blood transfusion 0 0 NA
Antibiotics use 9 (100.0) 7 (87.5) 12.5 (–19.1 to 47.1)
Central venous catheter insertion in ED 2 (22.2) 0 22.2 (–13.9 to 54.7)
Intubation 2 (22.2) 0 22.2 (–13.9 to 54.7)
Median diff*
Time from first lactate measured to 81.0 (42.0 to 126.0) 73.0 (31.0 to 137.0) 8.0 (–81.0 to 95.0)
antibiotics use, min
Arterial lactate, mmol/L
0h 3.8 (3.2 to 5.1) 4.1 (3.4 to 6.0) –0.3 (–2.8 to 1.6)
3h 3.4 (1.9 to 3.6) 3.8 (2.9 to 4.5) –0.4 (–1.9 to 1.5)
After 3 h† 2.5 (1.6 to 3.1) 3.2 (2.2 to 3.9) –0.7 (–2.1 to 0.9)
Lactate clearance at 3 h, % 31.2 (7.8 to 50.1) 11.5 (–12.5 to 48.7) 19.7 (–34.6 to 60.2)
RR‡
Lactate clearance at 3 h >20%, No. (%) 6 (66.7) 4 (50.0) 1.3 (0.6 to 3.1)
Hemodynamic goals achieved within 3 6 (66.7) 8 (100.0) 0.7 (0.4 to 1.1)
h, No. (%)
Median diff*
Length of hospital stay, days 8.0 (4.0 to 12.0) 12.5 (9.5 to 14.5) –4.5 (–9.5 to 2.5)
Total hospital charge, Singapore $ 11,398 (3,867 to 13,456) 13,290 (8,338 to 17,828) –1,892 (–12,323 to 5,119)
*Median and risk differences were calculated as intervention group–control group.
†
Assessment of arterial lactate level after 3 hours was conducted for 6 and 7 patients in the intervention and control groups, respectively.
‡
RR was calculated for intervention group versus control group.

of resuscitation as a method to assess initial response to
resuscitation in severe sepsis.35 We recognize that some
patients may require more than 3 hours of initial ED
management before improvements can be appreciated.
DISCUSSION
In this open-label randomized controlled study, we
report similar outcomes between the intervention group
comprising a noninvasive fluid resuscitation protocol using
the noninvasive cardiac output monitor and passive leg-
raising maneuver compared with the control group of usual
care according to best practice. This conclusion is in line
with those of 3 large multicenter trials that have been
published recently concluding that usual care did not
perform worse than early goal-directed therapy or protocol-
based resuscitation.7-9 However, we believe usual care in
the current era, in which early recognition of a septic
patient is coupled with early fluid resuscitation and
antibiotics, has led to significantly better outcomes,
differing greatly from more than a decade ago.6,36 There
is a suggestion that the temporal trend resulting from
learning effect and familiarity in managing patients with
severe sepsis was reflected in more than 70% of the control
group’s achieving greater than 20% lactate clearance at 3
hours compared with local data obtained in 2009 and
used for the sample size calculation of this study. This effect
likely resulted in no differences in fluid administration
observed between the 2 groups. Therefore, no conclusions
about the efficacy or lack of efficacy of using fluid
resuscitation to drive lactate clearance can be made in
this study. Nevertheless, the ability to obtain reasonably
accurate and, more important, informative trends of
hemodynamic parameters within 10 minutes noninvasively
appeals to the fast-paced environment of critical care in
the ED.
The response to passive leg raising using stroke volume
variation and pulse pressure variation has been described in
intubated ICU patients.37-39 However, in our study we

Volume -, no. - : - 2015 Annals of Emergency Medicine 9

Emergency Department Management of Sepsis Patients
chose changes in stroke volume index and cardiac index,
measured noninvasively in spontaneously breathing
patients because this was more representative of a typical
septic ED patient and avoided the well-known
complication of invasive monitoring such as infection and
limb ischemia.40 Various studies on the accuracy of the
noninvasive cardiac output monitor to detect fluid
responsiveness using passive leg raising have shown positive
outcomes.41-43
Elevation of lactate concentrations in patients recruited
may not have been entirely due to the lack of delivery of
oxygen to the tissues and organs (type A lactic acidosis).
Numerous causes of type B lactic acidosis by underlying
diseases (eg, liver cirrhosis), drugs (eg, metformin, b-
agonists, salicylates, acetaminophen), and toxins (eg,
alcohol) may have led to elevation of lactate concentrations
in our study population.44 In this study, there were 7
patients with liver cirrhosis; 4 (2 in Child-Pugh class A and
2 in class B) in the intervention group and 3 (2 in class
A and 1 in class B) in the control group. Aerobic glycolysis
to rapidly generate adenosine triphosphate in the
hyperdynamic stage of sepsis, stimulation of Naþ-Kþ-
ATPase activity, and inhibition of mitochondrial activity
can also lead to hyperlactatemia.35,44 Drawbacks in using
lactate clearance as a target for resuscitation have been
debated, including that as many as 30% of patients with
septic shock do not manifest hyperlactatemia, where
results of this study would have been irrelevant.45
Despite this, there is evidence to suggest that lactate
clearance and normalization of lactate in the first 6 hours
of resuscitation is a good predictor of mortality, although
the optimal margin and timing of clearance remain
to be elucidated. 46 In this study with a threshold
lactate clearance of 20% at 3 hours, mortality was
4.5% compared with 23.5% for patients who did not
achieve this threshold.
We sought to answer the question of whether greater
precision in optimizing fluid resuscitation with the use
of noninvasive hemodynamic monitors such as the
noninvasive cardiac output monitor would result in faster
lactate clearance. Apart from MAP, the use of passive leg
raising coupled with cardiac output measurements such as
stroke volume index in ED resuscitation has yet to be
established as efficacious. Our study did not show any
benefit in terms of our primary or secondary endpoints.
Previous and ongoing educational efforts by sepsis
champions in the ED to improve sepsis care have likely
led to heightened awareness and lower threshold for
administering initial fluid boluses to a septic patient
compared with that in the past.33 One of the measurable
key performance indices of our ED is to administer fluid
10 Annals of Emergency Medicine
Kuan et al
boluses to patients with severe pneumonia, an effect that
may have cascaded to other septic patients.47 For the
broad population of patients with sepsis in the ED,
clinical acumen of physicians and adherence to strong,
pragmatic, evidence-based practices are sufficient in
optimization of fluid resuscitation in the ED. In light of
our results, a simplified protocol incorporating the severe
sepsis care bundle can be implemented to include the
use of passive leg raising and a noninvasive cardiac
output monitor in selected circumstances. Our results on
cumulative volume of crystalloids administered (Table 4)
have demonstrated greater precision through smaller IQRs
in the intervention group. The purpose of a simplified
protocol would be to reduce the heterogeneity in practice
among emergency physicians with various levels of
experience.
However, the application of the noninvasive cardiac
output monitor was particularly useful in the fluid
resuscitation of patients with underlying fluid overload
states such as congestive heart failure and renal failure.
Results from this group of patients were consistent with our
initial hypothesis that using a noninvasive hemodynamic
optimization strategy would result in early lactate clearance,
with consequent lower morbidity and mortality. The
median lactate clearance and volume of fluid administered
was higher, albeit without any statistical significance,
probably because of the small number of patients, and
should be the focus of a larger study. The differences in the
amount of fluid administered and outcomes between these
2 groups in the subgroup analysis likely arose not from the
more aggressive fluid resuscitation in the intervention
group but rather the fear of causing overt fluid overload in
the control group, as evidenced by comparing the volume
of fluid administered at 3 hours with that used for the
entire cohort. We postulate that the noninvasive cardiac
output monitor helped guide the emergency physicians and
instill confidence in them to be more aggressive with fluid
resuscitation for these patients with congestive heart failure
and renal failure. This observation is hypothesis-generating
and merits further research for this subgroup.
In conclusion, our main findings were that protocol-
based care using a noninvasive cardiac output monitor
and passive leg-raising maneuver did not result in any
significant differences in lactate clearance at 3 hours,
fluid administration, mortality, hospital length of
stay, and total hospital cost compared with usual care.
Future studies involving the subgroup of patients with
congestive cardiac failure and renal failure may yield
more promising use of the noninvasive cardiac output
monitor according to our preliminary results for this
subgroup.
Volume -, no. - : - 2015
Kuan et al
Supervising editor: Alan E. Jones, MD
Author affiliations: From the Emergency Medicine Department,
National University Hospital, National University Health System,
Singapore and the Department of Surgery, Yong Loo Lin School of
Medicine, National University of Singapore, Singapore (Kuan,
Ibrahim, Leong, Mahadevan); the Centre for Quantitative Medicine,
Duke–NUS Graduate Medicine School, Singapore (Kuan, Jain, Lu,
Cheung); the Singapore Clinical Research Institute, Singapore (Lu);
and the Department of International Health, University of Tampere,
Tampere, Finland (Cheung).
Author contributions: WSK, II, BSHL, YBC, and MM conceived the
study, designed the trial, and obtained research funding. WSK, SJ,
and QL supervised the conduct of the trial and data collection and
managed the data, including quality control. QL and YBC provided
statistical advice on study design and analyzed the data. WSK, QL,
and MM drafted the article, and all authors contributed
substantially to its revision. WSK takes responsibility for the paper
as a whole.
Funding and support: By Annals policy, all authors are required to
disclose any and all commercial, financial, and other relationships
in any way related to the subject of this article as per ICMJE conflict
of interest guidelines (see www.icmje.org). The authors have stated
that no such relationships exist and provided the following details:
This work was supported by the National University Health System
Clinician Research Grant (FY2011) and the National Medical
Research Council New Investigator Grant (NIG11may049) to Dr.
Kuan. Dr. Kuan reports receiving grant support from the National
Medical Research Council Transition Award. Dr. Lu’s institution
(Singapore Clinical Research Institute) received funding from the
National University Health System Clinician Research Grant
(FY2011) in statistical analysis and preparation of the article.
Publication dates: Received for publication April 12, 2015.
Revisions received June 26, 2015, and August 31, 2015. Accepted
for publication September 2, 2015.
Clinical trial registration number: NCT01453270
Cheetah Medical did not participate in the design of the study or
the decision to submit this article for publication.
REFERENCES
1. Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J
Med. 2013;369:840-851.
2. Martin GS. Sepsis, severe sepsis and septic shock: changes in
incidence, pathogens and outcomes. Expert Rev Anti Infect Ther.
2012;10:701-706.
3. Adhikari NK, Fowler RA, Bhagwanjee S, et al. Critical care and the
global burden of critical illness in adults. Lancet. 2010;376:
1339-1346.
4. Levy MM, Dellinger RP, Townsend SR, et al. The Surviving Sepsis
Campaign: results of an international guideline-based performance
improvement program targeting severe sepsis. Crit Care Med.
2010;38:367-374.
5. 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:
1368-1377.
6. 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.
Volume -, no. - : - 2015
Emergency Department Management of Sepsis Patients
7. Yealy DM, Kellum JA, Huang DT, et al. A randomized trial of protocol-
based care for early septic shock. N Engl J Med. 2014;370:
1683-1693.
8. Peake SL, Delaney A, Bailey M, et al. Goal-directed resuscitation for
patients with early septic shock. N Engl J Med. 2014;371:1496-1506.
9. Mouncey PR, Osborn TM, Power GS, et al. Trial of early, goal-directed
resuscitation for septic shock. N Engl J Med. 2015;372:1301-1311.
10. Jones A, Kline J. Use of goal-directed therapy for severe sepsis shock in
academic emergency departments. Crit Care Med. 2005;33:
1888-1889.
11. Marik PE, Varon J. Early goal-directed therapy: on terminal life support?
Am J Emerg Med. 2010;28:243-245.
12. Marik PE, Baram M, Vahid B. Does central venous pressure predict
fluid responsiveness? a systematic review of the literature and the tale
of seven mares. Chest. 2008;134:172-178.
13. Marik PE, Cavallazzi R. Does the central venous pressure predict fluid
responsiveness? an updated meta-analysis and a plea for some
common sense. Crit Care Med. 2013;41:1774-1781.
14. Ballard DW, Reed ME, Rauchwerger AS, et al. Emergency physician
perspectives on central venous catheterization in the emergency
department: a survey-based study. Acad Emerg Med. 2014;21:
623-630.
15. Howell MD, Donnino M, Clardy P, et al. Occult hypoperfusion and
mortality in patients with suspected infection. Intensive Care Med.
2007;33:1892-1899.
16. Mikkelsen ME, Miltiades AN, Gaieski DF, et al. Serum lactate is
associated with mortality in severe sepsis independent of organ failure
and shock. Crit Care Med. 2009;37:1670-1677.
17. Zhang Z, Xu X. Lactate clearance is a useful biomarker for the
prediction of all-cause mortality in critically ill patients: a systematic
review and meta-analysis. Crit Care Med. 2014;42:2118-2125.
18. Nguyen HB, Kuan WS, Batech M, et al. Outcome effectiveness of the
severe sepsis resuscitation bundle with addition of lactate clearance
as a bundle item: a multi-national evaluation. Crit Care. 2011;15:
R229.
19. Bakker J, Gris P, Coffernils M, et al. Serial blood lactate levels can
predict the development of multiple organ failure following septic
shock. Am J Surg. 1996;171:221-226.
20. Nguyen HB, Rivers EP, Knoblich BP, et al. Early lactate clearance is
associated with improved outcome in severe sepsis and septic shock.
Crit Care Med. 2004;32:1637-1642.
21. Arnold RC, Shapiro NI, Jones AE, et al. Multicenter study of early lactate
clearance as a determinant of survival in patients with presumed
sepsis. Shock. 2009;32:35-39.
22. Wiedemann HP, Wheeler AP, Bernard GR, et al. Comparison of two
fluid-management strategies in acute lung injury. N Engl J Med.
2006;354:2564-2575.
23. Melot J, Sebbane M, Dingemans G, et al. Use of indicators of fluid
responsiveness in septic shock: a survey in public emergency
departments. Ann Fr Anesth Reanim. 2012;31:583-590.
24. Squara P, Denjean D, Estagnasie P, et al. Noninvasive cardiac output
monitoring (NICOM): a clinical validation. Intensive Care Med.
2007;33:1191-1194.
25. Raval NY, Squara P, Cleman M, et al. Multicenter evaluation of
noninvasive cardiac output measurement by bioreactance technique.
J Clin Monit Comput. 2008;22:113-119.
26. Monnet X, Rienzo M, Osman D, et al. Passive leg raising predicts
fluid responsiveness in the critically ill. Crit Care Med. 2006;34:
1402-1407.
27. Monnet X, Teboul JL. Passive leg raising. Intensive Care Med.
2008;34:659-663.
28. Teboul JL, Monnet X. Prediction of volume responsiveness in critically
ill patients with spontaneous breathing activity. Curr Opin Crit Care.
2008;14:334-339.
29. Thiel SW, Kollef MH, Isakow W. Non-invasive stroke volume
measurement and passive leg raising predict volume responsiveness
Annals of Emergency Medicine 11
Emergency Department Management of Sepsis Patients
in medical ICU patients: an observational cohort study. Crit Care.
2009;13:R111.
30. Boulain T, Achard JM, Teboul JL, et al. Changes in BP induced by
passive leg raising predict response to fluid loading in critically ill
patients. Chest. 2002;121:1245-1252.
31. Montiel-Jarquín A, Láscarez-Lagunas I, Sánchez-Gasca C, et al. Lactate
clearance is a prognostic factor in patients on shock state. Eur J Gen
Med. 2012;9:98-103.
32. Liu V, Morehouse JW, Soule J, et al. Fluid volume, lactate values, and
mortality in sepsis patients with intermediate lactate values. Ann Am
Thorac Soc. 2013;10:466-473.
33. Kuan WS, Mahadevan M, Tan JH, et al. Feasibility of introduction and
implementation of the Surviving Sepsis Campaign bundle in a
Singapore emergency department. Eur J Emerg Med. 2013;20:
344-349.
34. Schulz KF, Grimes DA. Sample size calculations in randomised trials:
mandatory and mystical. Lancet. 2005;365:1348-1353.
35. Kraut JA, Madias NE. Lactic acidosis. N Engl J Med. 2014;371:
2309-2319.
36. Coz Yataco A. Protocol-based care for early septic shock. N Engl J Med.
2014;371:384-385.
37. Geerts B, de Wilde R, Aarts L, et al. Pulse contour analysis to assess
hemodynamic response to passive leg raising. J Cardiothorac Vasc
Anesth. 2011;25:48-52.
38. Préau S, Saulnier F, Dewavrin F, et al. Passive leg raising is
predictive of fluid responsiveness in spontaneously breathing
patients with severe sepsis or acute pancreatitis. Crit Care Med.
2010;38:819-825.
12 Annals of Emergency Medicine
Kuan et al
39. Cavallaro F, Sandroni C, Marano C, et al. Diagnostic accuracy of
passive leg raising for prediction of fluid responsiveness in adults:
systematic review and meta-analysis of clinical studies. Intensive Care
Med. 2010;36:1475-1483.
40. Belda FJ, Aguilar G, Teboul JL, et al. Complications related to less-
invasive haemodynamic monitoring. Br J Anaesth. 2011;106:
482-486.
41. Benomar B, Ouattara A, Estagnasie P, et al. Fluid responsiveness
predicted by noninvasive bioreactance-based passive leg raise test.
Intensive Care Med. 2010;36:1875-1881.
42. Lamia B, Molano L, Declercq P, et al. Non invasive prediction of volume
responsiveness using bioreactance in hemodynamically unstable
patients with spontaneously breathing activity. Am J Respir Crit Care
Med. 2010;181:A4540.
43. Duus N, Shogilev DJ, Skibsted S, et al. The reliability and validity of
passive leg raise and fluid bolus to assess fluid responsiveness in
spontaneously breathing emergency department patients. J Crit Care.
2015;30: 217.e1-5.
44. Vernon C, Letourneau JL. Lactic acidosis: recognition, kinetics, and
associated prognosis. Crit Care Clin. 2010;26:255-283.
45. Jones AE. Lactate clearance for assessing response to resuscitation in
severe sepsis. Acad Emerg Med. 2013;20:844-847.
46. Puskarich MA, Trzeciak S, Shapiro NI, et al. Whole blood lactate
kinetics in patients undergoing quantitative resuscitation for severe
sepsis and septic shock. Chest. 2013;143:1548-1553.
47. Lim HF, Phua J, Mukhopadhyay A, et al. IDSA/ATS minor criteria aid pre-
intensive care unit resuscitation in severe community-acquired
pneumonia. Eur Respir J. 2014;43:852-862.
Volume -, no. - : - 2015
Kuan et al Emergency Department Management of Sepsis Patients

APPENDIX E1*
*Pictures courtesy of Cheetah Medical.

Figure E1. The Cheetah noninvasive cardiac output monitor.

Figure E2. Placement of noninvasive cardiac output monitor electrodes on the torso.

Volume -, no. - : - 2015 Annals of Emergency Medicine 12.e1

Emergency Department Management of Sepsis Patients Kuan et al

Figure E3. Hemodynamic parameters obtained before and after the PLR challenge.

12.e2 Annals of Emergency Medicine Volume -, no. - : - 2015

Kuan et al Emergency Department Management of Sepsis Patients

APPENDIX E2
Passive leg-raising maneuver

PLR was performed by first placing a patient with

noninvasive cardiac output monitoring in a semirecumbent
position and obtaining hemodynamic parameters averaged
during 3 minutes. Subsequently, the patient was placed
with both lower limbs raised passively at least 45
degrees from the horizontal and the torso supine. The
hemodynamic parameters obtained in this position were
also averaged during the next 3 minutes.

Table E1. Inhospital complications.

Variable Intervention (N[58) Control* (N[64) RR (95% CI)†
‡
Pulmonary edema, No. (%)
Yes 4 (6.9) 4 (6.3) 1.10 (0.29–4.21)
No 54 (93.1) 60 (93.8)
Acute kidney injury, No. (%)§
Yes 18 (31.0) 18 (28.1) 1.10 (0.64–1.91)
No 40 (69.0) 46 (71.9)
Nosocomial infection, No. (%)jj
Yes 2 (3.4) 1 (1.6) 2.21 (0.21–23.70)
No 56 (96.6) 63 (98.4)
Line complications, No. (%){
No 58 (100.0) 64 (100.0) NA
*Data from 3 patients who switched from the intervention group to the control group were analyzed as control group data.
†
RR was calculated for intervention group versus control group.
‡
Pulmonary edema was clinically or radiologically determined by the inpatient treating physicians, who were blinded to the patients’ enrollment group during their evaluation of the
patient in the ward or ICU.
§
Acute kidney injury was defined according to the KDIGO Clinical Practice Guideline for Acute Kidney Injury.1
jj
Nosocomial infections (bloodstream infection, pneumonia, urinary tract infection, or surgical site infection) were determined by the inpatient treating physicians, who were
blinded to the patients’ enrollment group during their evaluation of the patient in the ward or ICU.
{
Line complications were determined by the presence of the following: (1) arterial catheters: permanent ischemic damage to the limb, local infection, pseudoaneurysm,
arteriovenous fistula, peripheral neuropathy from nerve damage, and central embolization; and (2) central venous catheter: bleeding, inadvertent arterial puncture, arrhythmia, air
embolism, thoracic duct injury, pneumothorax, hemothorax, delayed infection at insertion site, venous thrombosis, catheter migration or embolization, myocardial perforation, and
nerve injury.

REFERENCE
1. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney
Injury Work Group. KDIGO clinical practice guideline for acute kidney
injury. Kidney Int Suppl. 2012;2:1-138.

Volume -, no. - : - 2015 Annals of Emergency Medicine 12.e3