Original Research

Journal of Intensive Care Medicine

1-9
Effects of Inotropes on the Mortality ª The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
in Patients With Septic Shock DOI: 10.1177/0885066619892218
journals.sagepub.com/home/jic

Ryota Sato, MD1 , Nobuhiro Ariyoshi, MD2, Daisuke Hasegawa, MD3,

Erin Crossey, MD, PhD1, Natsumi Hamahata, MD1, Takuma Ishihara, MSc4,
Michitaka Nasu, MD5, and Gehan Devendra, MD, FCCP6

Abstract
Background: Although surviving sepsis campaign guidelines recommend the use of inotropes in the presence of myocardial
dysfunction, the effects of inotropes, including epinephrine, dobutamine, and milrinone, on in-hospital mortality in patients with
septic shock remains unclear. Materials and Methods: We conducted an international,2-center, retrospective cohort study.
The Cox proportional hazards regression model with time-varying covariates was used to investigate whether epinephrine,
milrinone, or dobutamine reduces in-hospital mortality in patients with septic shock. Sensitivity analysis was performed using
propensity score matching. The primary outcome was in-hospital mortality. The secondary outcome included atrial fibrillation
(Afib) with a rapid ventricular response (RVR) in the intensive care unit (ICU) and ICU-free days. Results: A total of 417 patients
with septic shock were included, 72 (17.3%) of whom received inotropes. The use of epinephrine and dobutamine was associated
with significantly higher in-hospital mortality (epinephrine, hazard ratio [HR]: 4.79, 95% confidence interval [CI]: 2.12-10.82,
P ¼ .001; dobutamine, HR: 2.53, 95% CI: 1.30-4.95, P ¼ .046). The effects of epinephrine and dobutamine were time- and dose-
dependent. The use of milrinone was not associated with increased mortality (HR: 1.07, 95% CI: 0.42-2.68, P ¼ .345). The use of
epinephrine, dobutamine, and milrinone was associated with significantly increased odds of Afib with RVR (epinephrine, odds ratio
[OR]: 3.88, 95% CI: 1.11-13.61, P ¼ .034; dobutamine, OR: 3.95, 95% CI: 1.14-13.76; and milrinone, OR: 3.77, 95% CI: 1.05-13.59).
On the other hand, the use of epinephrine, dobutamine, and milrinone was not associated with less ICU-free days (epinephrine,
adjusted OR: 0.30, 95% CI: 0.09-1.01, P ¼ .053; dobutamine, adjusted OR: 0.91, 95% CI: 0.29-2.84; and milrinone, adjusted OR:
0.60, 95% CI: 0.19-1.87). Conclusion: The present study showed that the use of epinephrine and dobutamine was associated
with significantly increased in-hospital mortality in patients with septic shock. These effects were both time- and dose-dependent.
On the other hand, the use of milrinone was not associated with increased in-hospital mortality.

Keywords
sepsis, septic shock, inotropes, cardiogenic shock, epinephrine, dobutamine, milrinone

Introduction
Myocardial dysfunction is a relatively common complication
of septic shock. 1 Since early goal-directed therapy was
reported, dobutamine has been considered to be an essential
component, especially in the treatment of septic patients with
myocardial dysfunction.2 Currently, surviving sepsis campaign
guidelines (SSCG) suggest the use of dobutamine in the pres-
ence of myocardial dysfunction indicated by elevated cardiac
filling pressures and low cardiac output or ongoing signs of
hypoperfusion, despite achieving adequate intravascular vol-
ume and mean arterial pressure.3 Surviving sepsis campaign
guidelines also suggest milrinone, a phosphodiesterase inhibi-
tor, as an alternative inotropic agent.3 In addition to these med-
ications, epinephrine has frequently been used as an inotropic
agent in cases involving sepsis. However, while these inotropic
agents lead to an increase in cardiac index (CI) and splanchnic
blood flow, 4,5 there are no convincing data supporting
1
Department of Internal Medicine, John A. Burns School of Medicine,
University of Hawaii at Manoa, Honolulu, HI, USA
2
Hospitalist Program, The Queen’s Medical Center, Honolulu, HI, USA
3
Department of Anesthesiology and Critical Care Medicine, Fujita Health
University School of Medicine, Toyoake, Japan
4
Innovative and Clinical Research Promotion Center, Gifu University Hospital,
Gifu, Japan
5
Department of Emergency and Critical Care, Urasoe General Hospital,
Okinawa, Japan
6
Department of Pulmonary and Critical Care, John A. Burns School of
Medicine, University of Hawaii at Manoa, Honolulu, HI, USA
Received August 13, 2019. Received revised October 07, 2019. Accepted
November 13, 2019.
Corresponding Author:
Ryota Sato, Department of Internal Medicine, John A. Burns School of Medi-
cine, University of Hawaii at Manoa,1356 Lusitana Street, 7th Floor, Honolulu,
HI 96813, USA.
Email: st051035@gmail.com
2 Journal of Intensive Care Medicine XX(X)
a morbidity or mortality benefit with the use of inotropes in
septic shock, even with myocardial dysfunction.6 Therefore,
we investigated the effects of epinephrine, dobutamine, and
milrinone on in-hospital mortality in patients with septic shock.
Materials and Methods
We conducted an international, 2-center, retrospective cohort
study comprising adult patients with septic shock who were
admitted to the medical intensive care unit (MICU) in The
Queen’s Medical Center (QMC) and Urasoe General Hospital
(UGH) between January 1, 2015, and December 31, 2017. The
QMC is a teaching hospital with 505 acute and 28 subacute
care beds. Urasoe General Hospital is an acute care teaching
hospital with 334 beds. Among patients who were admitted to
the MICU in QMC and UGH, we enrolled those with septic
shock. The exclusion criteria were patients who died within the
first 24 hours and patients who did not undergo a transthoracic
echocardiogram in the MICU. The investigators gathered all
clinical data from electronic medical records.
Sepsis is defined as a life-threatening organ dysfunction
caused by dysregulated host response to infection. Organ dys-
function was identified as an acute change in total sequential
organ failure assessment (SOFA) score 2 points consequent
to the infection. The baseline SOFA score was assumed to be
zero in patients not known to have preexisting organ dysfunc-
tion.7 Septic shock was defined as a vasopressor requirement to
maintain a mean arterial blood pressure of 65 mm Hg and a
serum lactate level >2 mmol/L in the absence of hypovolemia.
The primary outcome was in-hospital mortality according to
the use of inotropes, including dobutamine, epinephrine, and
milrinone. The secondary outcomes were the incidence of new
onset atrial fibrillation (Afib) with rapid ventricular response
(RVR) and ICU-free days. In patients who received inotropes,
we counted only Afib with RVR that occurred subsequent to
initiation of inotropes as new-onset Afib with RVR. Intensive
care unit-free days are a composite outcome combining mor-
tality and length of ICU stay. The number of ICU-free days
was calculated as 28—the length of ICU stay during the first
28 days. All patients who died before 28 days were counted as
having zero ICU-free days on the basis that those patients
should be counted as having the worst possible outcome.8
We performed statistical analysis using R software version
3.2.1 for Mac (R Development Core Team). The patients’ char-
acteristics were summarized by a median with 25th and 75th
percentiles for continuous variables. Frequencies and percent-
age were shown for categorical variables.
In the primary analysis, we analyzed the use of dobutamine,
epinephrine, and milrinone as a time-varying exposure vari-
able. To assess the effects of using inotropes on mortality,
we used a time-varying Cox proportional hazards multivariable
regression model with adjustment for the following a priori
covariates that were selected because of their clinical rele-
vance: age,9 sex,10 body mass index,11 lactic acid on admission
to the ICU,12 history of heart failure, chronic kidney disease,13
liver disease,14 diabetes,15 neoplastic diseases,16 history of
stroke,17 history of structural lung disease,17 history of baseline
Afib, history of myocardial infarction requiring revasculariza-
tion, immunocompromised status,18 left ventricular ejection
fraction (LVEF) on admission, E/e0 ratio on admission, highest
rate of norepinephrine infusion in the first 24 hours, utilization
of vasopressin in the first 24 hours, utilization of hydrocorti-
sone in the first 24 hours, institution of mechanical ventilation
in the first 24 hours,19 and Acute Physiology and Chronic
Health Evaluation II score.20 Missing covariates were imputed
using multiple imputation methods.21 In addition, we incorpo-
rated 2-way interaction terms as “dobutamine*epinephrine,”
“dobutamine*milrinone,” and “epinephrine*dobutamine” in
the regression model to assess the effect modification. We also
conducted sensitivity analysis using propensity score matching
(PSM), where the propensity score was computed as the prob-
ability of using at least 1 inotrope as a function of the identical
variables for adjustment in the primary analysis. The Cox
regression including time-varying variables of the use of each
inotrope was performed after PSM. We validated the reliability
of the model internally with the bootstrap method.21 In total,
150 sets of bootstrap samples were generated by resampling the
original data, and the optimism was estimated to determine the
degree of overfitting.21 An optimism value <0.2 indicated no
evidence of overfitting. Time-varying Kaplan-Meier curves
were obtained using Simon and Makuch’s method22 with a
significance test using Mantel-Byar estimators.23 We also con-
firmed the effect on mortality using dose as an exposure vari-
able instead of whether inotropes were used. We assessed
nonlinear associations between each dose of inotropes and
mortality by including nonlinear cubic splines in the Cox
regression model.
In the secondary analysis, we defined the exposure as a
time-fixed variable in which each inotrope was used within
24 hours of admission to the ICU. Because ICU-free days were
not normally distributed, we used a proportional odds logistic
regression model with adjustment for covariates, identical to
the primary analysis, to assess the relationship between ICU-
free days and each inotrope. Furthermore, we used a binomial
logistic regression model to assess the effect of inotropes on
Afib with RVR.
A P value of <.05 was considered statistically significant in
all analyses. All authors had full access to all study data and
analyses. This study was approved by the Research and
Institutional Review Committee of the QMC (RA-2018-024)
and UGH.
Results
A total of 8426 patients were admitted to the MICU in QMC or
UGH between January 1, 2015, and December 31, 2017. Of
these, we identified 520 with septic shock and enrolled 417, as
shown in Figure 1. Norepinephrine was administered in all 417
patients. Of these, 188 patients (45.1%) received vasopressin
(0.03 U/min) and 222 patients (53.2%) received hydrocortisone
(50 mg every 6 hours). Seventy-two patients (17.3%) received
inotropic agents, including epinephrine (33 patients, 7.9%),
Sato et al
Figure 1. Selection of the patients. A total of 8426 patients were admitted to the MICU at QMC or UGH between January 1, 2015,
and December 31, 2017. Of these, we identified 520 with septic shock and enrolled 417. Inotropic agents, including epinephrine,
dobutamine, and milrinone, were administered in 72 patients.
dobutamine (30 patients, 7.2%), and milrinone (25 patients,
6.0%). Of these, 8 patients (1.9%) received both epinephrine
and milrinone, 5 patients (1.2%) received both epinephrine and
dobutamine, and 3 patients (0.7%) received both milrinone and
dobutamine. Epinephrine, milrinone, and dobutamine were ini-
tiated within 24 hours from admission in 7 patients (21.2%), 7
patients (28.0%), and 8 patients (26.7%), respectively. The
study population included 237 males (56.8%), with the median
age being 66 years. The characteristics of the patients are
shown in Table 1 and Supplemental Table 1.
Among the patients enrolled, 123 (29.5%) had respiratory
infections, 85 (20.4%) had soft tissue or bone infections, 75
(18.0%) had intra-abdominal infections, 65 (15.6%) had urin-
ary tract infections, 42 (10.1%) had a blood stream infection, 4
(1.0%) had a central nervous system infection, and 23 (5.5%)
had other infections. In addition, 116 (27.8%) had Gram-
positive cocci bacterial infection, 164 (39.3%) had Gram-
negative rods bacterial infection, 29 (7.0%) had other
pathogen-induced infection, and 108 (25.9%) had a culture-
negative infection, as shown in Table 2. Of 116 patients who
had Gram-positive cocci bacterial infection, 20 (4.8%) had
methicillin-resistant Staphylococcus aureus infection. Of 164
patients who had Gram-negative rods bacterial infection, 19
(4.6%) had Pseudomonas aeruginosa infection and 9 (2.2%)
had extended spectrum b-lactamase producing Escherichia coli
or Klebsiella pneumoniae infections. A total of 25 (75.8%)
patients in the epinephrine group who received epinephrine,
17 (56.7%) in the dobutamine group, 13 (52.0%) in the
3
milrinone group, and 95 (27.5%) in the non-inotropes group
died in the hospital.
In the time-varying multivariable Cox regression analysis,
patients who received epinephrine and dobutamine had signif-
icantly higher in-hospital mortality (epinephrine, hazard
ratio [HR]: 4.79, 95% confidence interval [CI]: [2.12-10.82],
P ¼ .001 and dobutamine, HR: 2.53, 95% CI: [1.30-4.95], P ¼
.046). On the other hand, the use of milrinone was not
associated with higher in-hospital mortality (HR: 1.07,
95% CI: [0.42-2.68], P ¼ .345), as shown in Table 3. The
“dobutamine*epinephrine,” “dobutamine* milrinone,” and
“epinephrine* dobutamine” interactions were nonsignificant
(P ¼ .359, P ¼ .098, and P ¼ .845, respectively). In patients
with lactate 4 mmol/L, both the use of epinephrine and the
use of dobutamine were associated with significantly higher
in-hospital mortality (epinephrine, HR: 7.53, 95% CI: [3.38-
16.75], P < .001; dobutamine, HR: 2.27, 95% CI: [1.04-4.96],
P ¼ .041) while milrinone was not (HR: 0.64, 95% CI: [0.29-
1.39], P ¼ .259), as shown in Supplemental Table 2. In patients
with lactate < 4 mmol/L, the use or epinephrine, milrinone, or
dobutamine was not associated with higher in-hospital mortal-
ity (epinephrine, HR: 2.50, 95% CI: [0.17-36.30], P ¼ .503;
milrinone, HR: 3.43, 95% CI: [0.57-20.73], P ¼ .179; dobuta-
mine, HR: 1.10, 95% CI: [0.17-6.98], P ¼ .920).
Analysis with a PSM data set showed similar results as
shown in Table 4 (epinephrine, HR: 2.61, 95% CI: [1.24-
5.51], P ¼ .012; dobutamine, HR: 1.66, 95% CI: [0.77-3.60],
P ¼ .199; milrinone, HR: 0.885, 95% CI: [0.44-1.79],
4 Journal of Intensive Care Medicine XX(X)

Table 1. Patient Characteristics.

Demographics of the Patients Combined, N ¼ 417 Inotropes (þ), N ¼ 72 Inotropes (), N ¼ 345 P Value

M F M F M F

Sex 237 (56.8%) 180 (43.2%) 42 (58.3%) 30 (41.7%) 195 (56.5%) 150 (43.5%) .778

Age 66 (55-75) 64 (54-72) 66 (56-76) .084

BMI 26.5 (22.0-31.9) 26.5(23.1-31.9) 26.3 (21.9-31.9) .52
History of HF 167 (40.0%) 41 (56.9%) 126 (36.5%) .001
History of CKD 166 (39.8%) 25 (34.7%) 141 (40.9%) .332
History of CVA 59 (14.2%) 4 (5.6%) 55 (16.0%) .021
History of myocardial infarction 48 (11.5%) 5 (6.9%) 43 (12.5%) .182
History of liver disease 65 (15.6%) 12 (16.7%) 53 (15.4%) .781
History of structural lung disease 76 (18.3%) 13 (18.1%) 63 (18.3%) .959
History of diabetes 189 (45.5%) 29 (40.3%) 160 (46.6%) .324
History of cancer 70 (16.8%) 9 (12.5%) 61 (17.7%) .285
Baseline Afib 65 (15.6%) 12 (16.7%) 53 (15.4%) .781
E/e0 13.9 (10.4-20.0) 16.2 (10.3-21.7) 13.3 (10.4-19.2) .3
Immunosuppression 72 (17.3%) 8 (11.1%) 64 (18.6%) .129
APACHE II score 23 (19-27) 26 (21-29) 23 (19-27) <.001
Lactic acid on admission 4.0 (2.7-6.4) 5.4 (4-8.8) 3.6 (2.6-5.9) <.001
LVEF on admission 50% (35%-58%) 30% (20%-53%) 55% (40%-60%) <.001
Highest rate of NE in the first 24 hours (mg/kg/min) 0.18 (0.10-0.34) 0.40 (0.20-0.60) 0.17 (0.10-0.27) <.001
Use of vasopressin 188 (45.1%) 53 (73.6%) 135 (39.1%) <.001
Use of hydrocortisone 222 (53.2%) 64 (88.9%) 158 (45.8%) <.001
MV in the first 24 hours 215 (51.6%) 60 (83.3%) 155 (44.9%) <.001
Abbreviations: Afib, atrial fibrillation; APACHE II score, Acute Physiology and Chronic Health Evaluation II score; BMI, body mass index; CKD, chronic kidney
disease; CVA, cerebrovascular accident; F, female; HF, heart failure; LVEF, left ventricular ejection fraction; M, male; MV, mechanical ventilation;
NE, norepinephrine.

Table 2. Sources and Pathogens of Septic Shock Among Patients Table 3. Hazard Ratio for Mortality Among Patients Who Received
Who Received Inotropes and Those Who Did Not. Inotropes and Those Who Did Not.

Inotropes (þ), Inotropes (), Variable HRa 95% CI P Value

N ¼ 72 N ¼ 345
Milrinone, N ¼ 25 1.07 0.42-2.68 .345
The infectious source Epinephrine, N ¼ 33 4.79 2.12-10.82 .001
Respiratory 23 (31.9%) 100 (29.0%) Dobutamine, N ¼ 30 2.53 1.30-4.95 .046
Skin, soft tissue, and bone 19 (26.4%) 66 (19.1%)
Intra-abdominal 9 (12.5%) 66 (19.1%) Abbreviations: CI, confidence interval; HR, hazard ratio.
a
Hazard ratio adjusted for age, sex, body mass index, lactic acid on admission
Urinary tract 8 (11.1%) 57 (16.5%)
to the intensive care unit, history of heart failure, chronic kidney disease, liver
Blood stream infection 9 (12.5%) 33 (9.6%) disease, diabetes, neoplastic diseases, stroke, structural lung disease, and
Central nervous system 0 (0%) 4 (1.2%) myocardial infarction, immunocompromised status, baseline atrial fibrillation,
Others 4 (5.6%) 19 (5.5%) left ventricular ejection fraction on admission, E/e0 on admission, highest rate
The pathogen of norepinephrine infusion in the first 24 hours, whether vasopressin was used
Gram-positive cocci 20 (27.8%) 96 (27.8%) in the first 24 hours, whether hydrocortisone was used in the first 24 hours,
MRSA 4 (5.6%) 16 (4.6%) whether mechanical ventilation was required in the first 24 hours, and Acute
Gram negative rods 32 (44.4%) 132 (38.3%) Physiology and Chronic Health Evaluation II score.
Pseudomonas aeruginosa 5 (6.9%) 14 (4.1%)
ESBL producing Escherichia coli or 2 (2.8%) 7 (2.0%)
Table 4. Hazard Ratio for Mortality Among Patients Who Received
Klebsiella pneumoniae
Inotropes and Those Who Did Not for PSM Data Set.
Others 4 (5.6%) 25 (7.2%)
Unknown 16 (22.2%) 92 (26.7%) Variable HRa 95% CI P Value
Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus; ESBL, Milrinone 0.885 0.44-1.79 .734
extended spectrum beta-lactamases.
Epinephrine 2.61 1.24-5.51 .012
Dobutamine 1.66 0.77-3.60 .199
P ¼ .734). The bootstrap validation, which assesses the extent
Abbreviations: CI, confidence interval; HR, hazard ratio; PSM, propensity score
to which a model may be overfit to the development data, matching.
a
showed that slope shrinkage (optimism) was 0.198, indicating Hazard ratio were obtained using time-varying Cox proportional hazards
no evidence of overfitting. As described by the Kaplan-Meier model for PSM data set.
Sato et al
Figure 2. The effects of inotropes were time-dependent.
curve in Figure 2, the effects of inotropes were time-dependent.
The effects of epinephrine and dobutamine on in-hospital
mortality were dose-dependent (1-5 mg/kg/min; P < .001 and
P ¼ .002, respectively), while that of milrinone was not
(P ¼ .245). The effect of dobutamine seemed to be constant
once it became >5 mg/kg/min, as shown in Figure 3.
The use of epinephrine, milrinone, and dobutamine was
associated with a significantly higher risk of Afib with RVR
(epinephrine, odds ratio [OR]: 3.88, 95% CI: [1.11-13.61],
P ¼ .034; milrinone, OR: 3.77, 95% CI: [1.05- 13.59], P ¼
.043; and dobutamine, OR: 3.95, 95% CI: [1.14- 13.76], P ¼
.031), as shown in Table 5. The use of epinephrine, milrinone,
or dobutamine was not associated with fewer ICU-free days
(epinephrine, adjusted OR: 0.30, 95% CI: [0.09-1.01], P ¼
.053; milrinone, adjusted OR: 0.60, 95% CI: [0.19-1.87], P ¼
.376; and dobutamine, adjusted OR: 0.91, 95% CI: [0.29-2.84],
P ¼ .873), as shown in Table 5.
Discussion
In the present study, we investigated whether the use of ino-
tropes in patients with septic shock is associated with higher in-
hospital mortality. Multivariable Cox proportional hazards
regression analysis revealed that the use of epinephrine and
dobutamine in septic shock was associated with significantly
5
higher in-hospital mortality and a higher risk of Afib with
RVR. The use of milrinone was not found to be significant.
In the present study, there were no clear criteria to initiate
inotropes. Therefore, whether inotropes were used depended
greatly on the physician preference based on clinical informa-
tion such as LVEF. While current SSCG suggests the use of
inotropes in the presence of myocardial dysfunction, there has
not been sufficient evidence to support this strategy. Previous
studies have shown that maintaining cardiac index higher than
normal is not beneficial in critically ill patients.24,25 A rando-
mized controlled trial showed no benefit in the mortality and
prevention of organ dysfunction in septic patients with the use
of levosimendan, which is a relatively new inotropic agent.26
Although levosimendan was not used in the present study, the
results suggest that the use of inotropes does not necessarily
improve mortality in septic patients. Even in patients with
severe heart failure, a meta-analysis showed that dobutamine
did not improve mortality. 27 Furthermore, a propensity-
matched retrospective observational analysis of over 65 000
patients from the Acute Decompensated Heart Failure National
Registry revealed that the use of inotropes doubled in-hospital
mortality.28 All these studies suggest that maintaining cardiac
index may not be associated with or might even increase mor-
tality, while inotropic agents are known to lead to an increase in
cardiac index and splanchnic blood flow.3,5
6 Journal of Intensive Care Medicine XX(X)

Figure 3. The effects of epinephrine and dobutamine on in-hospital mortality were dose-dependent (1-5 mg/kg/min; P < .001 and P ¼ .002,
respectively), while that of milrinone was not (P ¼ .245). The effect of dobutamine seemed to be constant once it became >5 mg/kg/min.

The main purpose of the use of inotropic agents is to
improve tissue oxygen delivery, which depends greatly on
microcirculation, which is typically impaired in sepsis. There-
fore, regarding sepsis, tissue oxygen delivery may not depend
on cardiac output, and an increase in cardiac output may not
always mean improvement of microcirculatory abnormal-
ities,29 as it is well-known that b-adrenergic stimulation can
cause lactic acidosis.30 In addition, sepsis can cause sepsis-
induced cardiomyopathy, which is temporary myocardial dys-
function, and it is known that sepsis-induced cardiomyopathy is
not associated with worse outcomes and may even be associ-
ated with better outcomes.31 Vieillard Baron et al reported that
septic patients in a hyperkinetic state had significantly higher
mortality than those in a hypokinetic state, or with normal
output.32 In addition, a meta-analysis failed to show a signifi-
cant difference between septic patients with left ventricular
dysfunction or dilatation and those without.33 Therefore,
reduced left ventricular function does not seem to be a prog-
nostic factor or an effective therapeutic target.
On the other hand, in experimental environments, epinephr-
ine has been shown to increase the contractility of myocardial
fibers to a lesser degree than norepinephrine, despite a higher
energy cost.34 Furthermore, a previous study revealed that the
use of dopamine that has a strong b-1 adrenergic effect and that
epinephrine is associated with an increased risk of death com-
pared with norepinephrine, even in cases of cardiogenic shock,
which is characterized by low cardiac index.35 Another recent
study revealed that the use of epinephrine was associated with
refractory shock compared with norepinephrine in patients with
cardiogenic shock.36 In a recent meta-analysis, the use of epi-
nephrine for the management of cardiogenic shock was asso-
ciated with a 3-fold increase of death.37 In the present study,
both epinephrine and dobutamine were associated with signif-
icantly higher in-hospital mortality especially in the high-risk
population (lactate 4 mmol/L) while milrinone was not.
In addition, a randomized controlled trial reported that a
short-acting selective b-1 blocker improved survival in patients
with septic shock,38 suggesting that a b-1 adrenergic effect may
be associated with worse outcomes. Another randomized con-
trolled trial described that the combination therapy with milri-
none and a short-acting selective b-1 blocker improved cardiac
function and 28-day survival rate in septic patients.39
On the other hand, Nguyen et al reported that the use of
dobutamine as a second vasoactive agent was associated with
Sato et al 7

Table 5. Multivariable Analysis of Inotropes Associated With ICU- patients who received inotropes and those who did not, and our
Free Days and Atrial Fibrillation With Rapid Ventricular Response.a results might reflect these differences. However, we used suf-
Adjusted Odds
ficient variables as well as a PSM analysis as an assessment of
Outcome Inotrope Ratiob 95% CI P Value sensitivity and this resulted in very similar result. Therefore,
we believe that this risk was minimized. Third, epinephrine can
ICU-free days Milrinone 0.60 0.19-1.87 .376 be used as a second vasopressor with another inotropic agent.
Epinephrine 0.30 0.09-1.01 .053 However, we incorporated 2-way interaction terms as
Dobutamine 0.91 0.29-2.84 .873
“dobutamine*epinephrine,” “dobutamine*milrinone,” and
Outcome Inotrope Odds Ratioc 95% CI P Value “epinephrine*dobutamine” in the regression model to assess
effect modification. These 2-way interactions were nonsignifi-
Afib with RVR Milrinone 3.77 1.05-13.59 .043 cant, therefore, we believe that the risk of interaction among
Epinephrine 3.88 1.11-13.61 .034
each inotrope was minimized. To verify our findings, a pro-
Dobutamine 3.95 1.14-13.76 .031
spective randomized controlled trial is warranted.
Abbreviation: Afib, atrial fibrillation; CI, confidence interval; ICU, intensive care
unit; RVR, rapid ventricular response.
a
Each model was adjusted for age, sex, body mass index, lactic acid on admis- Conclusion
sion to the ICU, history of heart failure, chronic kidney disease, liver disease,
diabetes, neoplastic diseases, stroke, structural lung disease, and myocardial The present study showed that the use of epinephrine and
infarction, immunocompromised status, baseline Afib, left ventricular ejection dobutamine was associated with significantly increased in-
fraction on admission, E/e0 on admission, highest rate of norepinephrine infu- hospital mortality in patients with septic shock. These effects
sion in the first 24 hours, whether vasopressin was used in the first 24 hours,
whether hydrocortisone was used in the first 24 hours, whether mechanical were both time- and dose-dependent. On the other hand, the use
ventilation was required in the first 24 hours and Acute Physiology and of milrinone was not associated with increased in-hospital
b
Chronic Health Evaluation II score. mortality.
Adjusted odds ratios were obtained using proportional logistic regression
model.
c
Odds ratios were obtained using binomial logistic regression model. Authors’ Note
This study was approved by the Research and Institutional Review
Committee of QMC (RA-2018-024) and UGH (2019A004). The data
lower mortality compared to dopamine, phenylephrine, and set generated and analyzed during the current study are available from
vasopressin.40 However, dobutamine is an inotrope rather than the corresponding author on reasonable request. Ryota Sato, Nobuhiro
a vasopressor and there is a possibility that the dobutamine Ariyoshi, and Gehan Devendra were responsible for conception of the
group might be hemodynamically more stable given the nature article and drafted and revised the manuscript. Daisuke Hasegawa,
of the retrospective study. In fact, the severity score and pre- Erin Crossey, Natsumi Hamahata and Michitaka Nasu contributed
dicted mortality were significantly lower among the dobuta- substantially to the data collection, and draft the manuscript. Takuma
mine group and this could have affected the result. Therefore, Ishihara contributed substantially to the study design, data analysis,
and interpretation.
further prospective investigations are warranted.
The present study revealed a significant association
Declaration of Conflicting Interests
between the use of epinephrine and the occurrence of AFib
with RVR. On the other hand, it is well-known that sepsis The author(s) declared no potential conflicts of interest with respect to
complicated by AFib has significantly worse mortality com- the research, authorship, and/or publication of this article.
pared with no AFib.41
These findings suggest that the use of inotropes, especially Funding
excessive b-1 stimulation, leads to potential risks in septic The author(s) received no financial support for the research, author-
patients, and the risk of excessive b-1 stimulation is often ship, and/or publication of this article.
underestimated. Most importantly, no prospective trial has sup-
ported the use of epinephrine or dobutamine over norepinephr- ORCID iD
ine in cases of myocardial dysfunction, or even in cardiogenic Ryota Sato, MD https://orcid.org/0000-0001-8806-2852
shock, even though these inotropes have been widely used all
over the world. Improvements in surrogate markers such as Supplemental Material
cardiac index with inotropes do not indicate improvements of Supplemental material for this article is available online.
patient-centered outcomes such as mortality.
Our study has several limitations. First, as this was a retro- References
spective study, there may have been a selection bias. In fact, the 1. Sato R, Kuriyama A, Takada T, Nasu M, Luthe SK. Prevalence
highest dose of norepinephrine in the first 24 hours was rela- and risk factors of sepsis-induced cardiomyopathy: a retrospective
tively low. However, we enrolled all consecutive patients with cohort study. Medicine. 2016;95(39):e5031.
septic shock during the study period; therefore, we believe that 2. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy
the risk of selection bias was minimized. Second, there are in the treatment of severe sepsis and septic shock. N Engl J Med.
significant differences in severity and cardiac histories between 2001;345(19):1368-1377.
8 Journal of Intensive Care Medicine XX(X)
3. Rhodes A, Evans LE, Alhazzani W, et al. Surviving sepsis cam-
paign: International Guidelines for Management of Sepsis and
Septic Shock: 2016. Intensive Care Med. 2017;43(3):304-377.
4. Reyad AR, Elkharboutly W, Wahba A, Elmorshedi M, Hasaneen
NA. Effect of intraoperative dobutamine on splanchnic tissue
perfusion and outcome after Whipple surgery. J Crit Care.
2013;28(4):531.e537-515.
5. Barton P, Garcia J, Kouatli A, et al. Hemodynamic effects of IV
milrinone lactate in pediatric patients with septic shock. A pro-
spective, double-blinded, randomized, placebo-controlled, inter-
ventional study. Chest. 1996;109(5):1302-1312.
6. Sato R, Nasu M. Time to re-think the use of dobutamine in sepsis.
J Intensive Care. 2017;5(1):65.
7. Singer M, Deutschman CS, Seymour CW, et al. The third inter-
national consensus definitions for sepsis and septic shock (Sepsis-
3). JAMA. 2016;315(8):801-810.
8. Young P, Hodgson C, Dulhunty J, et al. End points for phase II
trials in intensive care: recommendations from the Australian and
New Zealand clinical trials group consensus panel meeting. Crit
Care Resusc. 2012;14(3):211-215.
9. Lagu T, Rothberg MB, Shieh MS, Pekow PS, Steingrub JS,
Lindenauer PK. Hospitalizations, costs, and outcomes of severe
sepsis in the United States 2003 to 2007. Crit Care Med. 2012;
40(3):754-761.
10. Asai K, Hiki N, Mimura Y, Ogawa T, Unou K, Kaminishi M.
Gender differences in cytokine secretion by human peripheral
blood mononuclear cells: role of estrogen in modulating LPS-
induced cytokine secretion in an ex vivo septic model. Shock.
2001;16(5):340-343.
11. Pepper DJ, Sun J, Welsh J, Cui X, Suffredini AF, Eichacker PQ.
Increased body mass index and adjusted mortality in ICU patients
with sepsis or septic shock: a systematic review and meta-analy-
sis. Crit Care. 2016;20(1):181.
12. Weil MH, Afifi AA. Experimental and clinical studies on lactate
and pyruvate as indicators of the severity of acute circulatory
failure (shock). Circulation. 1970;41(6):989-1001.
13. Doi K, Leelahavanichkul A, Hu X, et al. Pre-existing renal disease
promotes sepsis-induced acute kidney injury and worsens out-
come. Kidney Int. 2008;74(8):1017-1025.
14. Foreman MG, Mannino DM, Moss M. Cirrhosis as a risk factor
for sepsis and death: analysis of the National Hospital Discharge
Survey. Chest. 2003;124(3):1016-1020.
15. de Miguel-Yanes JM, Mendez-Bailon M, Jimenez-Garcia R, Her-
nandez-Barrera V, Perez-Farinos N, Lopez-de-Andres A. Trends
in sepsis incidence and outcomes among people with or without
type 2 diabetes mellitus in Spain (2008-2012). Diabetes Res Clin
Pract. 2015;110(3):266-275.
16. Abou Dagher G, El Khuri C, Chehadeh AA, et al. Are patients
with cancer with sepsis and bacteraemia at a higher risk of
mortality? A retrospective chart review of patients presenting
to a tertiary care centre in Lebanon. BMJ open. 2017;7(3):
e013502.
17. Kempker JA, Kramer MR, Waller LA, Martin GS. Risk factors for
septicemia deaths and disparities in a longitudinal US cohort.
Open Forum Infect Dis. 2018;5(12):ofy305.
18. Jamme M, Daviaud F, Charpentier J, et al. Time course of septic
shock in immunocompromised and nonimmunocompromised
patients. Crit Care Med. 2017;45(12):2031-2039.
19. Annane D, Aegerter P, Jars-Guincestre MC, Guidet BCUB-Réa
Network. Current epidemiology of septic shock: the CUB-Rea
network. Am J Respir Crit Care Med. 2003;168(2):165-172.
20. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE
II: a severity of disease classification system. Crit Care Med.
1985;13(10):818-829.
21. Harrell FE. Regression Modeling Strategies: With Applications to
Linear Models, Logistic Regression, and Survival Analysis. New
York, NY: Springer; 2001.
22. Simon R, Makuch RW. A non-parametric graphical representa-
tion of the relationship between survival and the occurrence of an
event: application to responder versus non-responder bias. Stat
Med. 1984;3(1):35-44.
23. Mantel NBD. Evaluation of response-time data involving transi-
ent states: an illustration using heart-transplant data. J Am Stat
Assoc. 1974;69:81-88.
24. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson
D. Elevation of systemic oxygen delivery in the treatment of
critically ill patients. N Engl J Med. 1994;330(24):1717-1722.
25. Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented
hemodynamic therapy in critically ill patients. N Engl J Med.
1995;333(16):1025-1032.
26. Gordon AC, Perkins GD, Singer M, et al. Levosimendan for the
prevention of acute organ dysfunction in sepsis. N Engl J Med.
2016;375(17):1638-1648.
27. Tacon CL, McCaffrey J, Delaney A. Dobutamine for patients with
severe heart failure: a systematic review and meta-analysis of
randomised controlled trials. Intensive Care Med. 2012;38(3):
359-367.
28. Abraham WT, Adams KF, Fonarow GC, et al. In-hospital mor-
tality in patients with acute decompensated heart failure requiring
intravenous vasoactive medications: an analysis from the acute
decompensated heart failure national registry (ADHERE). J Am
Coll Cardiol. 2005;46(1):57-64.
29. De Backer D, Creteur J, Preiser JC, Dubois MJ, Vincent JL.
Microvascular blood flow is altered in patients with sepsis. Am
J Respir Crit Care Med. 2002;166(1):98-104.
30. Day NP, Phu NH, Bethell DP, et al. The effects of dopamine and
adrenaline infusions on acid-base balance and systemic haemo-
dynamics in severe infection. Lancet. 1996;348(9022):219-223.
31. Sato R, Nasu M. A review of sepsis-induced cardiomyopathy.
J Intensive Care. 2015;3:48.
32. Vieillard Baron A, Schmitt JM, Beauchet A, et al. Early preload
adaptation in septic shock? A transesophageal echocardiographic
study. Anesthesiology. 2001;94(3):400-406.
33. Huang SJ, Nalos M, McLean AS. Is early ventricular dysfunc-
tion or dilatation associated with lower mortality rate in adult
severe sepsis and septic shock? A meta-analysis. Crit Care.
2013;17(3):R96.
34. Heubach JF, Ravens U, Kaumann AJ. Epinephrine activates
both Gs and Gi pathways, but norepinephrine activates only
the Gs pathway through human beta2-adrenoceptors
Sato et al
overexpressed in mouse heart. Mol Pharmacol. 2004;65(5):
1313-1322.
35. De Backer D, Biston P, Devriendt J, et al. Comparison of dopa-
mine and norepinephrine in the treatment of shock. N Engl J Med.
2010;362(9):779-789.
36. Levy B, Clere-Jehl R, Legras A, et al. Epinephrine versus nore-
pinephrine for cardiogenic shock after acute myocardial infarc-
tion. J Am Coll Cardiol. 2018;72(2):173-182.
37. Leopold V, Gayat E, Pirracchio R, et al. Epinephrine and short-
term survival in cardiogenic shock: an individual data meta-
analysis of 2583 patients. Intensive Care Med. 2018;44(6):847-856.
38. Morelli A, Ertmer C, Westphal M, et al. Effect of heart rate
control with esmolol on hemodynamic and clinical outcomes in
9
patients with septic shock: a randomized clinical trial. JAMA.
2013;310(16):1683-1691.
39. Wang Z, Wu Q, Nie X, Guo J, Yang C. Combination therapy with
milrinone and esmolol for heart protection in patients with severe
sepsis: a prospective, randomized trial. Clin Drug Investig. 2015;
35(11):707-716.
40. Nguyen HB, Lu S, Possagnoli I, Stokes P. Comparative effective-
ness of second vasoactive agents in septic shock refractory to
norepinephrine. J Intensive Care Med. 2017;32(7):451-459.
41. Kuipers S, Klein Klouwenberg PM, Cremer OL. Incidence,
risk factors and outcomes of new-onset atrial fibrillation in
patients with sepsis: a systematic review. Crit Care. 2014;
18(6):688.