Original Research

Journal of Intensive Care Medicine

1-5
Maximum Norepinephrine Dosage Within ª The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
24 Hours as an Indicator of Refractory DOI: 10.1177/0885066619860736
journals.sagepub.com/home/jic
Septic Shock: A Retrospective Study

Daisuke Kasugai, MD1 , Akihiko Hirakawa, MD, PhD2,

Masuyuki Ozaki, MD, PhD1, Kazuki Nishida, MD3 , Takao Ikeda, MD4,
Kunihiko Takahashi, PhD3 , Shigeyuki Matsui, PhD3,
and Norimichi Uenishi, MD4

Abstract
Background: The management of refractory septic shock remains a major challenge in critical care and its early indicators are not
fully understood. We hypothesized that the maximum norepinephrine dosage within 24 hours of intensive care unit (ICU) admission
may be a useful indicator of early mortality in patients with septic shock. Methods: In this retrospective single-center observational
study, patients with septic shock admitted to the emergency ICU of an academic medical center between April 2011 and March
2017 were included. Individuals with cardiac arrest and those with do-not-resuscitate orders before admission were excluded.
We analyzed if the maximum norepinephrine dosage within 24 hours of ICU admission (MD24) was associated with 7-day
mortality. Results: Among 152 patients with septic shock, 20 (15%) did not survive by day 7. The receiver operating charac-
teristic curve analysis for predicting 7-day mortality revealed a cutoff of MD24 of 0.6 mg/kg/min (sensitivity 47%, specificity 93%). In
the multivariable regression analysis, a higher MD24 was significantly associated with 7-day mortality (odds ratio: 7.20; 95%
confidence interval [CI]: 2.02-25.7; P ¼ .002) but not with 30-day mortality. Using the inverse probability of treatment weighting
method in a propensity scoring analysis, a higher MD24 was significantly associated with 7-day (hazard ratio [HR]: 8.9; 95% CI: 3.2-
25.0; P < .001) and 30-day mortality (HR: 2.7; 95% CI: 1.2-5.8; P ¼ .012). Conclusions: An MD24 0.6 mg/kg/min was significantly
associated with 7-day mortality in patients with septic shock and may therefore be a useful indicator of refractory septic shock.

Keywords
septic shock, sepsis, norepinephrine, intensive care unit, 7-day mortality

Background useful for the management of refractory shock, considering the

favorable 28-day survival in their study compared to previous
Sepsis is commonly encountered in intensive care units (ICUs)
reports, while they underscored that the maximum vasopressor
worldwide.1 Despite improvements in treatment strategies over
the past decade,2,3 it remains the leading cause of the death in
patients admitted to noncardiac ICUs.1 Refractory shock is one
of the major causes of death in patients with sepsis, accounting 1
Department of Emergency and Critical Care, Nagoya University Graduate
for 40% of deaths.4 To improve the management of septic School of Medicine, Nagoya, Japan
2
shock, early recognition of refractory shock is necessary. Department of Disaster and Traumatology, Fujita Health University,
Toyoake, Japan
The definition of refractory septic shock varies from persis- 3
Department of Biostatistics, Nagoya University Graduate School of Medicine,
tent shock in spite of optimized therapy5,6 to a vasopressor Nagoya, Japan
4
requirement of >0.5 to 1 mg/kg/min of norepinephrine or equiv- Department of Emergency and General Internal Medicine, Fujita Health
alent,7,8 and no consensus about its criteria has been reached University Hospital, Toyoake, Japan
yet. The effect of the norepinephrine dosage on mortality in Received May 10, 2019. Received revised June 5, 2019. Accepted June 11,
patients with septic shock is controversial.9 Several studies 2019.
have described a subgroup of septic shock receiving high-
dose vasopressors (HDVs).10-12 Some authors reported that Corresponding Author:
Daisuke Kasugai, Department of Emergency and Critical Care, Nagoya
patients with septic shock who received HDVs showed almost University Graduate School of Medicine, Tsurumai-cho 64, Syowa-ku,
100% mortality.10,11 Auchet et al analyzed a cohort of 106 Nagoya, Aichi 8560, Japan.
patients receiving HDVs and concluded that HDVs may be Email: [email protected]
2 Journal of Intensive Care Medicine XX(X)

dose should not be used as the sole factor to assess prognosis.12 The primary end point was 7-day mortality; this was chosen
Since these studies did not include all patients with septic to analyze cause-specific mortality due to refractory shock.4
shock,10-12 a better understanding of the role of catecholamine The secondary end point was 30-day mortality.
dosage as an indicator for refractory shock remains to be estab-
lished. Several articles reported the usefulness of cumulative
dose of vasopressors in the prediction of mortality in septic Statistical Analysis
shock,12,13 while its clinical application is currently limited due Continuous variables are expressed as means + standard
to the difficulty in calculation especially in the early phase. deviations or medians and interquartile ranges (25th-75th per-
Conrad et al reported that the hemodynamic response after centiles), as appropriate, and were compared using the Student
6 mg/kg/min of phenylephrine administration can be used to t test or Mann-Whitney U test, respectively. Categorical vari-
predict refractory septic shock.14 However, as the clinical ables are shown as numbers (%) and compared using the Fisher
application of phenylephrine in patients with septic shock is exact test. Receiver operating characteristic (ROC) curve anal-
limited according to current guidelines,2 another indicator of ysis was performed to assess the optimal cutoff of the MD24.
refractory shock is needed. We hypothesized that the maximum The optimal cutoff was defined as the point of the maximum
norepinephrine dosage within 24 hours of ICU admission sum of sensitivity and specificity results. The MD24 was then
(MD24) may be a useful early indicator of refractory shock. converted into categorical variables using this cutoff.
The associations between the variables of interest and out-
comes were assessed using a multivariable logistic regression
Methods model. Clinically and biologically plausible variables, namely
the MD24, an SOFA score 14, patient age, serum lactate
Study Setting and Population level, steroid and vasopressin use, and RRT were considered
In this retrospective single-center observational study, we in the model. Based on a previous study, the SOFA score was
included patients with septic shock admitted to the emergency converted into a categorical variable with a cutoff of 14.14 The
ICU of the Fujita Health University Hospital between April effect of the MD24 on mortality was assessed using a Cox
2011 and March 2017. The Fujita Health University Hospital proportional hazards model with and without inverse probabil-
is a tertiary medical center with 1435 beds, including 9 ICUs. ity of treatment weighting (IPTW), using propensity score
The emergency ICU contains 10 beds to provide care for those adjustment. The R (version 3.4.3) and EZR software (version
admitted from the emergency department. 1.36) were used for all statistical analyses (Saitama Medical
Eligible patients were patients with septic shock aged Center, Jichi Medical University, Saitama, Japan).16
>18 years. Septic shock was defined according to the Sepsis-3
definition.15 Individuals with cardiac arrest and those with do-
not-resuscitate orders before ICU admission were excluded.
Results
Baseline Patient Characteristics
Initial Management of Septic Shock
Of 177 eligible patients, 9 and 16 patients were excluded for
Patients with septic shock were treated according to the current
having do-not-resuscitate orders and due to cardiac arrest
guidelines for the management of septic shock.2 After adequate
before ICU admission, respectively. Thus, 152 patients were
fluid resuscitation (administration of more than 30 mL/kg of
included in the final analysis.
crystalloid) that was initiated upon arrival at the hospital,
Table 1 shows the baseline characteristics of the patients.
norepinephrine was titrated to obtain a mean arterial
The median SOFA score was 11 (9-13), and the median MD 24
pressure >65 mm Hg. Vasopressin 0.02 to 0.03 units/min
was 0.24 (0.16-0.38) mg/kg/min. Vasopressin and steroid were
and hydrocortisone 200 mg/d were administered when
administered in 58 (38%) and 67 (44%) cases, respectively.
considered appropriate by the attending physician at the
Twenty (15%) patients died within 1 week and 48 (32%) within
emergency ICU.
1 month. Patients not surviving by day 7 had a higher MD24,
higher APACHE II and SOFA scores, a higher incidence of
Variables of Interest and Outcome Measurements pneumonia, and a higher rate of vasopressin use.
Data were retrospectively collected from electronic medical
charts. The variables of interest included patient characteristics
(age, sex, body mass index, and focus of infection), the sequen-
Receiver Operating Characteristic Analysis
tial organ failure assessment (SOFA) score and Acute Physiol- The ROC curve of the MD24 to predict 7-day mortality is
ogy And Chronic Health Evaluation (APACHE) II scores on the shown in Figure 1. The area under the curve of the ROC was
day of ICU admission, laboratory data at the time of ICU admis- 0.777 (95% confidence interval [CI]: 0.661-0.894). The opti-
sion, the MD24, and adjunctive therapy (vasopressin, steroid, mal cutoff value of the MD24 to predict 7-day mortality was
renal replacement therapy [RRT]). The MD24 was calculated 0.6 mg/kg/min, with a sensitivity of 47% and specificity of
by dividing norepinephrine dosage by actual body weight. 93%. We used this cutoff for all further analyses.
Kasugai et al 3

Table 1. Patient Characteristics.a

Characteristic Total, n ¼ 152 Survivorsb, n ¼ 132 Nonsurvivorsb, n ¼ 20 P Value

Age, years 69 (14) 69 (15) 69 (14) .986

Male sex, n (%) 91 (60) 79 (60) 12 (60) >.999
Body weight, kg 52.6 (45.0-2.3) 52.9 (45.9-63.2) 49.3 (41.7-58.2) .149
Focus of infection, n (%)
Lung 44 (29) 31 (23) 13 (65) <.001
Urinary tract 25 (16) 25 (19) 0 (0) .046
Skin and soft tissue 12 (8) 11 (8) 1 (5) >.999
Abdomen 24 (16) 24 (18) 0 (0) .044
Others 47 (31) 41 (31) 6 (39) >.999
APACHE II score 25 (8.4) 24 (8.0) 31.5 (8.2) <.001
SOFA score 11 (9-13) 10 (8-13) 14 (11-14.5) <.001
Creatinine, mg/dL 1.8 (1.2-3.0) 1.7 (1.1-2.5) 2.2 (1.3-3.9) .11
Platelet count, 103/mL 12.6 (6.3-19) 13.2 (6.2-19) 8.8 (7.1-18.9) .715
Bilirubin, mg/dL 1.0 (0.6-1.9) 1.0 (0.6-1.6) 1.0 (0.5-2.2) .532
P-F ratio 242 (152-393) 251 (153-393) 180 (106-385) .191
GCS 12 (5.5-15) 12 (6-15) 7 (3-13) .018
Urine output, mL/d 625 (182-1383) 695 (207-1455) 328 (69-540) .042
WBC,  103/mL 5.675 12.500 20.225 13.6 (7.6-21.1) 5.5 (2.7-14.9) .005
Hematocrit, % 31.1 (25.7-36.5) 31.3 (26.8-36.6) 27.3 (24.0-34.4) .132
FDP, mg/mL 15.1 (9.1-35.9) 15.6 (8.8-35.7) 13.6 (10.8-56.1) .82
CRP, mg/dL 16.9 (9.0-25.6) 16.8 (9.3-25.4) 19.9 (5.6-25.3) .787
PCT, ng/mL 27.2 (6.0-74.0) 26.5 (6.0-71.7) 27.9 (6.1-99.8) .798
Lactate, mmol/L 4.4 (3.1-7.1) 4.3 (3.1-6.5) 5.3 (4.3-10.3) .011
MD24, mg/kg/min 0.24 (0.16-0.38) 0.23 (0.14-0.36) 0.5 (0.27-0.71) <.001
Vasopressin use, n (%) 57 (38) 43 (33) 14 (70) .002
Steroid use, n (%) 72 (47) 59 (45) 13 (65) .099
RRT, n (%) 80 (53) 66 (50) 14 (70) .337
Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; CRP, C-reactive protein; FDP, fibrin degradation products; GCS, Glasgow Coma Scale;
MD24, maximum norepinephrine dosage within 24 hours of intensive care unit (ICU) admission; PCT, procalcitonin; P-F ratio, PaO2-FiO2 ratio; RRT, renal
replacement therapy; SOFA, sequential organ failure assessment; WBC, white blood cell count.
a
Data are presented as means (standard deviations), medians and interquartile ranges (25%-75% percentile), or absolute frequencies and percentages.
b
Survival was based on the 7-day outcome.

P ¼ .008) were independently associated with 7-day mortality.

In contrast, the MD24 did not show a significant association
with 30-day mortality.
The adjusted estimated survival curves are depicted in
Figure 2. The hazard ratios for 7- and 30-day mortality
based on the MD24, adjusted by IPTW using propensity
scoring, were 8.9 (95% CI: 3.2-25.0, P < .001) and 2.7
(95% CI: 1.2-5.8, P ¼ .034) respectively (Table 3), suggest-
ing that an MD24 0.6 mg/kg/min was associated with an
increased risk of death; this association was more obvious in
the first 7 days.

Discussion
Figure 1. Receiver operating characteristic (ROC) curve using MD24
for predicting 7-day mortality. MD24, maximum dosage of norepi-
Refractory shock is a major challenge in the management of
nephrine within 24 hours of intensive care unit (ICU) admission. septic shock. 4 Although several case-series studies of
patients with refractory septic shock described its mortal-
ity,10-12 little evidence on its indicators exist. To the best of
Association Between the MD24 and Mortality our knowledge, the present study is the first to show that an
Table 2 shows the results of the multivariable logistic regres- MD24 0.6 mg/kg/min was significantly associated with
sion analysis. The MD24 (odds ratio [OR]: 6.98; 95% CI: 1.94- early mortality in all patients with septic shock included
25.2; P ¼ .003) and SOFA score (OR: 5.25; 95% CI: 1.54-17.8; in the study population.
4 Journal of Intensive Care Medicine XX(X)

Table 2. Multivariable Logistic Regression Analysis. multifactorial,4 and the time of death differs by cause. Death
due to refractory shock occurs during the first 7 days of the
Variable OR 95% CI P Value
syndrome.4 Our findings suggest that, while the MD24 is asso-
Prediction of 7-day mortality ciated with the severity of the shock, its effect on long-term
MD24 0.6 mg/kg/min 7.20 2.02-25.7 .002 mortality may be limited. Therefore, the subgroup of patients
SOFA score 14 4.87 1.53-15.5 .007 with refractory shock may have a favorable outcome if they
Vasopressin use 2.15 0.61-7.55 .233 somehow survived beyond shock phase. This may explain the
Age 0.99 0.95-1.03 .677
reason for the difference in 28-day mortality between previous
Prediction of 30-day mortality
MD24 0.6 mg/kg/min 1.58 0.52-4.85 .422 reports that analyzed patients with septic shock who received
SOFA score 14 3.08 1.34-7.08 .008 HDVs.10-12 Our findings also suggest that, in the management
Vasopressin use 2.49 1.11-5.55 .026 of refractory shock, focusing on the short-term outcome
Age 1.00 0.97-1.02 .875 (ie, 7-day mortality) may be more valuable than focusing on
the 28-day or longer-term outcome. In any sense, we should
Abbreviations: CI, confidence interval; MD24, maximum norepinephrine
dosage within 24 hours of intensives care unit (ICU) admission; OR, odds ratio; consider detailed hemodynamic monitoring and additional
SOFA, sequential organ failure assessment. therapeutic approaches for those require higher norepinephrine
especially in acute phase.

MD24 Cutoff in Refractory Septic Shock

We utilized a threshold of 0.6 mg/kg/min of norepinephrine
based on the ROC curve analysis. This is consistent with a
study by Conrad et al 14 that reported that 6 mg/kg/min of
phenylephrine (an equivalent of 0.6 mg/kg/min of norepinephr-
ine) predicted refractory shock. Based on this high specificity,
this threshold may be a useful indicator for refractory septic
shock. On the other hand, the sensitivity for early death in our
study group was below 50%. Despite the low sensitivity, and
given the low rate of early death in patients who received the
lower MD24, it may be safe to increase the norepinephrine
dosage to 0.6 mg/kg/min without considering early fatal conse-
quences. The low sensitivity may partly be due to not prospec-
Figure 2. Adjusted estimated survival curves. ICU indicates intensive tively evaluating the required dose of norepinephrine as an
care unit. indicator of refractory shock or due to the nonprotocolized
administration of vasopressin. Vasopressin administration is
Table 3. Hazard Ratios for Mortality Based on an MD24 0.6 mg/kg/ known to be associated with a reduction in norepinephrine
min.
requirement. 7 In this study, vasopressin administration
HR 95% CI P Value may have resulted in a lower requirement for norepinephrine
in patients with refractory shock and thereby may have
Crude HR for 7-day mortality 7.9 3.2-19.8 <.001 increased the number of patients in the false-negative group
Adjusted HR for 7-day mortalitya 8.9 3.2-25.0 <.001
(ie, those who die due to refractory shock with norepinephrine
Crude HR for 30-day mortality 2.7 1.3-5.5 .004
Adjusted HR for 30-day mortalitya 2.7 1.2-5.8 .012 <0.6 mg/kg/min).
As the management of hemodynamic shock was not proto-
Abbreviations: CI, confidence interval; HR, hazard ratio; IPTW, Inverse prob- colized in this study, the MD24 cutoff of 0.6 mg/kg/min
ability of treatment weighting; MD24, maximum norepinephrine dosage within
requires external validation in a future prospective study using
24 hours of intensive care unit (ICU) admission; SOFA, sequential organ failure
assessment. detailed protocols of catecholamine administration.
a
IPTW using propensity scoring was used for the adjustment. Variables used to
calculate the propensity score included age, lactate, the SOFA score, steroid
use, vasopressin use, and renal replacement therapy. Limitations
This study has several limitations, including the retrospective
Effect of the MD24 on Mortality study design. Although the norepinephrine dosage was
We found that the MD24 was significantly associated with increased according to the severity of the septic shock, it was
7-day but not with 30-day mortality in the multivariate analy- adjusted based on decision-making by clinicians and not based
sis. After propensity scoring adjustment, it was associated with on protocols. Therefore, further protocol-based validation of
an increased risk of death at 7 and at 30 days, but we initially our results is needed. In addition, due to the retrospective study
demonstrated that this association was more evident during the design, we cannot exclude that the higher norepinephrine
earlier phase. The cause of death in patients with sepsis is dosage itself worsened the 7-day outcome. Second, the effect
Kasugai et al 5

of the focus of infection on survival was not evaluated due to (J-SSCG 2016). J Intensive Care. 2018;6:7. doi:10.1186/
the small sample size. Specifically, the incidence of pneumonia s40560-017-0270-8.
was higher in nonsurvivors. Thus, respiratory failure may wor- 3. Stevenson EK, Rubenstein AR, Radin GT, Wiener RS, Walkey
sen the severity of the shock and may be associated with a AJ. Two decades of mortality trends among patients with severe
higher rate of early death. Finally, steroids and vasopressin sepsis: a comparative meta-analysis. Crit Care Med. 2014;42(3):
were administered in the study population. The effect of these 625-631. doi:10.1097/CCM.0000000000000026.
confounders should be minimized in a future prospective study 4. Moskowitz A, Omar Y, Chase M, et al. Reasons for death in
evaluating if the norepinephrine dosage can be used as an early patients with sepsis and septic shock. J Crit Care. 2017;38:
indicator of refractory septic shock. 284-288. doi:10.1016/j.jcrc.2016.11.036.
5. Nandhabalan P, Ioannou N, Meadows C, Wyncoll D. Refractory
Conclusions septic shock: our pragmatic approach. Crit Care. 2018;22(1):215.
doi:10.1186/s13054-018-2144-4.
Maximum dose of norepinephrine within 24 hours 0.6 mg/kg/ 6. Buckley MS, MacLaren R. Concomitant vasopressin and hydro-
min was significantly associated with 7-day mortality in cortisone therapy on short-term hemodynamic effects and vaso-
patients with septic shock and may therefore be a useful indi-
pressor requirements in refractory septic shock. J Crit Care. 2017;
cator of refractory septic shock.
42:6-11. doi:10.1016/j.jcrc.2017.06.016.
Authors’ Note 7. Bassi E, Park M, Azevedo LCP. Therapeutic strategies for high-
dose vasopressor-dependent shock. Crit Care Res Prac. 2013;
Daisuke Kasugai and Norimichi Uenishi conceived and designed this
study. Daisuke Kasugai and Takao Ikeda contributed to the acquisition 2013:654708.
of the data. Kazuki Nishida, Kunihiko Takahashi, and Shigeyuki Mat- 8. Brand DA, Patrick PA, Berger JT, et al. Intensity of vasopressor
sui contributed to the data analysis. Akihiko Hirakawa, Masuyuki therapy for septic shock and the risk of In-hospital death. J Pain
Ozaki, and Norimichi Uenishi assisted with the interpretation of the Symptom Manage. 2017;53(5):938-943. doi:10.1016/j.jpainsym-
data. Daisuke Kasugai was responsible for drafting, editing, and sub- man.2016.12.333.
mission of the manuscript. All authors reviewed and revised the manu- 9. Yamamura H, Kawazoe Y, Miyamoto K, Yamamoto T, Ohta Y,
script as well as approved the final manuscript. The data sets used and/ Morimoto T. Effect of norepinephrine dosage on mortality
or analyzed during the current study are available from the corre- in patients with septic shock. J Intensive Care. 2018;6:12.
sponding author on reasonable request. The institutional review board
doi:10.1186/s40560-018-0280 -1.
of Fujita health university approved this study. As this was a retro-
spective study, the need for patient consent was waived. 10. Jenkins CR, Gomersall CD, Leung P, Joynt GM. Outcome of
patients receiving high dose vasopressor therapy: a retrospective
Acknowledgments cohort study. Anaesth Intensive Care. 2009;37(2):286-289.
The authors would like to thank all nurses and physicians of the 11. Döpp-Zemel D, Groeneveld AB. High-dose norepinephrine treat-
emergency ICU of the Fujita Health University Hospital for providing ment: determinants of mortality and futility in critically ill
support during treatment. The authors also thank Misako Hashiguchi, patients. Am J Crit Care. 2013;22(1):22-32. doi:10.4037/
an administrative assistant, for help with data collection. ajcc2013748.
12. Auchet T, Regnier MA, Girerd N, Levy B. Outcome of patients
Declaration of Conflicting Interests with septic shock and high-dose vasopressor therapy. Ann Inten-
The author(s) declared no potential conflicts of interest with respect to sive Care. 2017;7(1):43. doi:10.1186/s13613-017-0261-x.
the research, authorship, and/or publication of this article. 13. Dargent A, Nguyen M, Fournel I, et al. Vasopressor cumulative
dose requirement and risk of early death during septic shock:
Funding
an analysis from the EPISS cohort. Shock. 2018;49(6):625-630.
The author(s) received no financial support for the research, author-
doi:10.1097/SHK.0000000000001022.
ship, and/or publication of this article.
14. Conrad M, Perez P, Thivilier C, Levy B. Early prediction of
ORCID iD norepinephrine dependency and refractory septic shock with mul-
Daisuke Kasugai https://orcid.org/0000-0002-8692-3003 timodal approach of vascular failure. J Crit Care. 2015;30(4):
Kazuki Nishida https://orcid.org/0000-0003-0367-8557 739-743. doi:10.1016/j.jcrc.2015.03.029.
Kunihiko Takahashi https://orcid.org/0000-0003-2387-7772 15. Singer M, Deutschman CS, Seymour CW, et al. The third inter-
national consensus definitions for sepsis and septic shock
References (Sepsis-3). JAMA. 2016;315(8):801-810. doi:10.1016/S0140-
1. Florian BM, Sachin Y, Derek CA. Epidemiology of severe sepsis. 6736(07)61344-0.
Virulence. 2014;5(1):4-11. doi:10.4161/viru.27372. 16. Kanda Y. Investigation of the freely available easy-to-use soft-
2. Nishida O, Ogura H, Egi M, et al. The Japanese clinical practice ware ‘EZR’ for medical statistics. Bone Marrow Transplant.
guidelines for management of sepsis and septic shock 2016 2013;48(3):452-458. doi:10.1038/bmt.2012.244.