Journal of Critical Care 29 (2014) 8892

Contents lists available at ScienceDirect

Journal of Critical Care

journal homepage: www.jccjournal.org

Delirium/Depression

Prediction of delirium in critically ill patients with elevated C-reactive protein,

Zhongheng Zhang, MM , Lifei Pan, MB, Hongsheng Deng, MM, Hongying Ni, MM, Xiao Xu, MB
Department of Critical Care Medicine, Jinhua Municipal Central Hospital, Jinhua Hospital of Zhejiang University, Jinhua, Zhejiang Province 321000, P.R. China

a r t i c l e
Keywords:
Delirium
C-reactive protein
Risk factors
Multivariate analysis

i n f o

a b s t r a c t
Background and Purpose: Delirium is thought to be associated with systemic inammatory response.
However, its association with the most widely used inammatory biomarker C-reactive protein (CRP) has not
been well established. We aimed to examine whether CRP on intensive care unit (ICU) entry was associated
with subsequent development of delirium.
Design and Setting: This prospective observational study was conducted in a mixed 24-bed ICU in a tertiary
teaching hospital.
Methods: All patients admitted to the ICU from February 2011 to June 2012 were screened for eligibility.
Demographic data and clinical characteristics of included patients were recorded. Patients were screened for
the presence of delirium by using the tool Confusion Assessment Method for the ICU (CAM-ICU). C-reactive
protein was obtained on ICU entry and 24 hours thereafter. Eligible patients were followed up for 28 days or
until death. Univariate and multivariate analyses were performed to evaluate independent risk factors for
delirium. Clinical outcome included the length of stay (LOS) in the ICU, 28-day mortality, and duration of
mechanical ventilation. Two-tailed P b .05 was considered statistically signicant.
Results: A total of 223 patients were included during study period. In univariate analysis, patients with
delirium showed signicantly higher CRP values than those without (120.5 vs 57.5 mg/L; P = .0001). By
adjusting for confounding variables (including age, sex, Acute Physiology and Chronic Health Evaluation II,
intubation, living alone, physical restraint, alcohol drinking, smoking, type of medical condition, and hospital
LOS before ICU admission) in logistic regression model, CRP remained an independent predictor of delirium
(odds ratio, 1.07; 95% condence interval, 1.01-1.15). As compared with nondelirious patients, those with
delirium showed longer LOS in ICU (13 vs 5 days; P b .001) and duration of mechanical ventilation (6 vs 1
days; P b .001). An increase in CRP greater than 8.1 mg/L within 24 hours was associated with 4-fold increase
in the risk of delirium (odds ratio: 4.47, 95% condence interval, 1.28-15.60).
Conclusion: C-reactive protein measured on ICU entry and its changes within 24 hours are risk indicators of
delirium. Further studies exploring the treatment of delirium according to CRP levels are warranted.
2014 Elsevier Inc. All rights reserved.

1. Introduction
Delirium is a frequently encountered clinical syndrome that is
characterized by altered mental status, inattention, disorganized
thinking, and uctuation of mental status. It is estimated that the
incidence of delirium in the intensive care unit (ICU) ranges between
20% and 80% across different study populations and institutions [1-5].
In analogy to acute kidney injury, delirium is now regarded as a
syndrome of acute brain dysfunction resulting from multiple risk
factors, and it imposes signicant negative impact on patient
outcomes [6,7]. Thus, the prevention and management of delirium
are of vital importance in the treatment of critically ill patients.

The work was performed at Department of Critical Care Medicine, Jinhua

Municipal Central Hospital, Zhejiang, People's Republic of China.
Financial disclosure: There is no conict of interest.
Corresponding author. Tel.: +86 579 82552667.
E-mail address: zh_zhang1984@hotmail.com (Z. Zhang).
0883-9441/$ see front matter 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcrc.2013.09.002

Currently, most studies have investigated the risk factors of delirium,

paving the way for further investigations on the prevention and
treatment of delirium.
Pathophysiologically, delirium is thought to be associated with
inammatory response. Proinammatory cytokines such as interleukin 1, tumor necrosis factor , and interleukin 8 have been associated
with the development of delirium in both ICU and non-ICU patients
[8,9]. However, most of these cytokines are not routinely measured in
clinical settings, and some are only for research purpose. C-reactive
protein (CRP) is a nonspecic inammatory biomarker that rises by
up to 10 000-fold in response to acute stimulus [10]. This biomarker
has already been linked to patient outcome in critical care setting
[11,12]. A recent meta-analysis by our group found that prolonged
elevation of CRP is associated with higher mortality rate [13].
However, the relationship between CRP and delirium has not been
well established. To the best of our knowledge, only several pilot
studies have reported their association, and the results are conicting
[14,15]. Therefore, we performed this study aiming to better dene
the association of CRP and delirium.

Z. Zhang et al. / Journal of Critical Care 29 (2014) 8892

2. Materials and methods

2.1. Study population and design
This prospective, observational study was conducted in a mixed
24-bed ICU in a tertiary teaching hospital from February 2011 to June
2012. The ICU received adult surgical and medical patients as well as
neurosurgical patients and patients of cardiopulmonary bypass (CPB)
surgery. The study protocol was approved by the ethical committee of
Jinhua Municipal Central Hospital. C-reactive protein is routinely
measured in our institution. Informed consent was waived due to the
observational nature of the study. The study was performed according
to the Declaration of Helsinki. Clinical data were collected by clinicians
(ZZ, HD, HN, and XX) and nursing staff (LP). The nursing staff
performed the Confusion Assessment Method for the ICU (CAM-ICU)
assessment, and all of them had completed a 1-week formal training
in the CAM-ICU performance.
All patients admitted to the ICU were screened for potential
eligibility. Eligibility criteria were as follows: (1) patients with
Glasgow Coma Scale greater than 10 or Richmond Agitation Sedation
Scale (RASS) 3 or higher [16]; (2) patients with the 18 years or
older; (3) patients who were expected to stay in ICU for more than 48
hours. Given good evidence that CAM-ICU performs well in patients
with dementia, this subgroup was included in our analysis [17].
Dementia was determined based on the history in medical chart,
which was obtained by asking the patient's family members.
Exclusion criteria are as follows: (1) patients with acute onset
structural brain diseases such as trauma, intracranial hemorrhage,
stroke, and subdural hematoma; (2) patients or their next of kin
signed do-not-resuscitation order; and (3) patients with delirium that
was conrmed by using the tool CAM-ICU at the time of presentation/
ICU admission (see below for detailed descriptions of CAM-ICU).
2.2. Measurement of CRP
C-reactive protein was measured in central laboratory for each
patient on their ICU entry, which is already a clinical routine in our
institution for a decade. The measurement was repeated on 24 hours
after ICU admission. Change in CRP (CRP) was dened as the
difference between these 2 measurements. Plasma CRP was measured
using immunouorescence method (i-CHROMATM) [18]. Briey, a
sandwich immunodetection method was used, such that by mixing
detection buffer with blood specimen in test vial, CRP antigen in the
specimen will bind to the uorescence-labeled detector anti-CRP
antibody in buffer. The amount of CRP can be reected by the signal
intensity of uorescence of detector antibody.

89

and RASS [21,22]. There are 4 features in the CAM-ICU tool: acute
onset or uctuating course, inattention, altered level of consciousness,
and disorganized thinking. Patients with delirium were categorized
into hyperactive (persistent positive RASS values: +1 to + 4 during
all assessments), hypoactive (persistently neutral or negative RASS
values: 0 to 3 during all assessments), and mixed types of delirium
(episode with both positive and negative RASS values) according to
RASS assessments obtained during delirium episodes [23]. Treatment
of delirium included nonpharmacologic and pharmacologic strategies.
The former included timely removal of catheter and physical
restraints, early mobilization, repeated reorientation of patients,
minimization of unnecessary noise/stimuli, and use of scheduled
pain protocol, and the latter included the use of haloperidol.
2.4. Statistical analysis
Quantitative data were expressed as mean SE or median and
interquartile range, as appropriate. Categorical data were expressed as
the number and percentage. Comparisons between delirious and
nondelirious groups were performed using t test or Wilcoxon rank
sum test, as appropriate. 2 Test was used for categorical data.
Conventional logistic regression model was tted to adjust for the
odds ratio (OR) of the development of delirium. All variables that
might interact with delirium and CRP were added to the model. These
variables included age, sex, APACHE II score, tracheal intubation,
physical restraints, living alone, history of drinking and smoking, and
length of hospital stay before ICU admission. The nal model was
assessed for its validity in aspects of interaction, collinearity, and
goodness of t [24]. Interaction term was created for CRP and age
because the impact of CRP on the development of delirium may vary
across different age groups. Collinearity was assessed using variance
ination factor, which is an indicator of how much of the ination of
the SE could be caused by collinearity. A variance ination factor
greater than 10 indicates signicant collinearity. Model t was
assessed by using Hosmer-Lemeshow goodness-of-t test [25]. The
CRP was assessed for its linkage with the development of delirium
by using CRP less than or equal to 9.8 mg/L as the reference group,
and the effect was adjusted for all potential confounders in logistic
regression model, as previously described. Change in CRP values was
classied into quartiles, and patients in the lowest quartile were used
as the reference group for the comparison with groups of other
quartiles. Diagnostic performance of CRP in predicting the development of delirium was assessed using the receiver operating
characteristic curve. All statistical analyses were performed using
the software Stata 11.0 (College Station, TX). Two-tailed P b .05 was
considered statistically signicant.

2.3. Clinical data collection

3. Results
On ICU entry, demographic data and baseline characteristics of
eligible patients including age, sex, Acute Physiology and Chronic
Health Evaluation II (APACHE II) score, type of medical condition,
tracheal intubation, living status, alcohol drinking, cigarette smoking,
and physical restraint were collected. Physical restraint is dened as
any device, material, or equipment attached to a person's body and
which cannot be controlled or easily removed by the person and
which deliberately prevents a person's free body movement to a
position of choice [19]. These variables were considered to be
associated with the development of delirium according to previous
studies [4,20]. Outcome variables included 28-day mortality, ICU
length of stay (LOS), and duration of mechanical ventilation. Patients
were followed up for a maximum of 28 days or until death.
For every included patient, mental status were assessed immediately on admission to the ICU and at least 3 times per day thereafter.
Critical care nursing staff performed all cognitive assessments on
admission and at least once per shift (every 8 hours) using CAM-ICU

A total of 223 consecutive patients were eligible during the 16month study period. The overall prevalence of delirium in our cohort
was 24.2% (54/223). Characteristics of included patients are shown in
Table 1. The mean age of the patients was 57.2 years, and 141 (63.2%)
of them were men. The median APACHE II score was 13 with an
interquartile range of 9 to 17. There were 93 medical patients (41.7%),
74 surgical patients (33.2%), 30 cardiac surgery patients (13.5%), and
26 neurosurgical patients (11.7%). There were 120 patients (54.1%)
with tracheal intubation. Thirty-ve patients (15.7%) lived alone at
home before this ICU admission. One hundred six patients (47.8%) had
physical restraint before the development of delirium. Fifty-eight
patients (26.1%) had a history of heavy alcohol drinking, and 58
(26.1%) reported a history of cigarette smoking. C-reactive protein
levels were not normally distributed (P b .01 for skewness; P b .001 for
Kurtosis), and they were expressed as median and interquartile range
(67.7; 21.5-128 mg/L).

90

Z. Zhang et al. / Journal of Critical Care 29 (2014) 8892

Table 1
Characteristics of included patients
Parameters

Age (y)
Male, n (%)
APACHE II
Type of medical condition,
n (%)
Medical
Surgical
CPB surgery
Neurological
Tracheal intubation
Living alone
Physical restraint
Drinking
Smoking
Hospital LOS before
ICU admission (d)
CRP (mg/L)
Outcome measures
LOS in ICU (d)
Duration of mechanical
ventilation (d)
28-d mortality

Table 2
The main logistic regression model displaying independent risk factors for delirium

Total
(n = 223)

Patients with
delirium
(n = 54)

Patients
without
delirium
(n = 169)

57.2 17.3
141 (63.2)
13 (9-17)

64.5 18.1
37 (68.5)
17 (13-27)

54.9 16.3
104 (61.5)
11 (8-15)

b.001
.354
b.001

93 (41.7)
74 (33.2)
30 (13.5)
26 (11.7)
120 (54.1)
35 (15.7)
106 (47.8)
58 (26.1)
65 (29.4)
6 (0-15)

26 (48.2)
14 (25.9)
1 (1.9)
13 (24.1)
39 (73.6)
16 (29.6)
40 (75.5)
20 (37.0)
18 (34.0)
11 (0-16)

67 (39.6)
60 (35.5)
29 (17.2)
13 (7.7)
81 (47.9)
19 (11.2)
66 (39.1)
38 (22.6)
47 (28.0)
5 (0-13)

.270
.193
.004
.001
.001
.001
b.001
.036
.404
.029

67.7 (21.5128)

120.5 (45.9156)

57.5 (16.6100.2)

5 (3-11)
1 (1-6)

13 (6-21)
6 (1-14)

5 (3-8)
1 (1-2)

34 (15.5)

11 (21.2)

23 (13.7)

.0001

b.001
b.001
.193

Data were expressed as mean SE, median and interquartile range, or number and
percentage as appropriate.

Patients with delirium were signicantly older (64.5 18.1 vs

54.6 16.3 years; P b .001) and had higher median APACHE II score
(17 vs 11; P b .001) than those without delirium. Cardiopulmonary
bypass patients were less likely to develop delirium (1.9% vs 17.2%;
P = .004), whereas neurologic patients were more likely to develop
delirium (24.1% vs 7.7%; P = .001). Delirium was more common in
patients with tracheal intubation (73.6% vs 47.9%; P = .001). Creactive protein levels were signicantly higher in patients with
delirium than those without (median, 120.5 vs 57.5 mg/L; P =
.0001). With respect to the clinical outcomes, patients with delirium
showed signicantly longer LOS in ICU (median, 13 vs 5 days; P b
.001) and longer duration of mechanical ventilation (6 vs 1 days; P b
.001). The 28-day mortality rate was higher in patients with
delirium, but the difference did not reach a statistical signicance
(21.2% vs 13.7%; P = .193).
All potential confounding factors were included into logistic
regression model. Hosmer-Lemeshow test showed a good model t
(P N .05), and the test for multicolinearity was statistically insignificant (Table 2). There was no interaction between CRP and age (data
not shown). C-reactive protein was independently associated with the
development of delirium. With each 10 mg/L increase in CRP, the risk
of delirium was increased by 7% (OR, 1.07; 95% condence interval
[CI], 1.01-1.05). Acute Physiology and Chronic Health Evaluation II
score was also independently associated with the development of
delirium, with an OR of 1.13 (95% CI, 1.06-1.21). Other variables such
as age, tracheal intubation, and living alone were not independently
associated with the development of delirium. Furthermore, CRP was
incorporated in to logistic regression model as a changeable risk factor
(Table 3). Change in CRP greater than 8.1 mg/L showed a signicantly
increased risk of the development of delirium in both unadjusted (OR,
3.18; 95% CI, 1.39-7.26) and adjusted model (OR, 4.47; 95% CI, 1.2815.60) as compared with those with CRP less than 9.8 mg/L.
Diagnostic performance of CRP in predicting delirium development
was less than satisfactory, with an area under curve of 0.68 (95% CI,
0.60-0.76; Fig. 1). At the cutoff value of 154.8 mg/L, the positive
predictive value of CRP in predicting delirium was 0.48. Fig. 2 shows
CRP levels in patients with hypoactive, hyperactive, mixed delirium

Parameters

OR

95% CI

CRP (with each 10 mg/L increase)

Age (with each 1 y increase)
Sex
APACHE II (with each 1-point increase)
Intubated vs nonintubated
Living alone
Physical restraint
Alcohol drinking
Smoking
Type of medical condition
Hospital LOS before ICU admission

1.07
1.002
0.67
1.13
1.50
1.75
2.80
2.23
0.94
1.54
1.01

1.01-1.15
0.97-1.04
0.27-1.62
1.06-1.21
0.56-4.04
0.51-5.94
0.99-7.90
0.84-5.98
0.32-2.79
0.99-2.37
0.97-1.05

.029
.894
.370
b.001
.425
.371
.053
.109
.917
.052
.651

All variables that may interact with CRP and delirium were included in the main model.
Logistic regression diagnostics: variance ination factor for all variables was less than 10.
The model was well tted with P = .392 in Hosmer-Lemeshow goodness-of-t test.

and those without delirium. Patients with mixed and hypoactive types
of delirium had signicantly higher CRP levels than those without
delirium (129.1 vs 57.5 mg/L, P b .001; 120.5 vs 57.5 mg/L, P = .031,
respectively). Kaplan-Meier survival curves were depicted to compare
the difference of the time to development of delirium in patients with
CRP greater than 67.7 mg/L vs patients with CRP less than or equal to
67.7 mg/L (Fig. 3). Patients who died or were transferred to general
ward without the development of delirium were considered as
censored. The result showed that patients with higher CRP values on
ICU entry were more likely to develop delirium (log-rank test, P b .05).
Furthermore, subgroup analysis by APACHE II score quartiles
indicated that the association between CRP and delirium was more
prominent in patients with higher APACHE II scores (Fig. 4).
4. Discussion
Our study showed that CRP on ICU entry was associated with
subsequent development of delirium during ICU stay. With each 10
mg/L increase in CRP, the risk of delirium increased by 7% (OR, 1.07;
95% CI, 1.01-1.05). The predictive value of CRP is more prominent in
patients with higher APACHE II scores. Because CRP is also
independently associated with subsequent development of delirium,
serial measurement of CRP may provide useful information for the
prevention and treatment of delirium.
The result of the present study was in line with the study by
Macdonald et al [14]. However, that study showed a better predictive
value for CRP. By examining the patient characteristics, we found that
the study population were much older in the study of Macdonald et al
(82.7 vs 55.8 years), suggesting that CRP performed well in older
population. However, further well-designed studies are needed to test
this hypothesis. Another study conducted by McGrane et al [15] used
delirium-/coma-free days as the study end point. They found that
higher CRP levels showed trends toward fewer delirium-/coma-free
days, but the difference was not statistically signicant. This study is
Table 3
Unadjusted and adjusted OR for development of delirium by CRP category
CRP quartiles

Less than 9.8

9.8 to 1.4
1.4 to 8.1
N8.1

Adjusted for all variablesa

Unadjusted
OR (95% CI)

OR (95% CI)

1
0.53 (0.19-1.48)
0.70 (0.27-1.83)
3.18 (1.39-7.26)

.228
.469
.006

1
0.27 (0.03-2.10)
1.15 (0.26 to 5.05)
4.47 (1.28-15.60)

.210
.857
.019

Change in CRP was the difference between the CRP values measured on ICU entry and
24 hours later.
a
Variables including age, sex, APACHE II, intubation status, living alone, physical
restraint, alcohol drinking, smoking, type of medical condition, and hospital LOS before
ICU admission.

Z. Zhang et al. / Journal of Critical Care 29 (2014) 8892

91

Fig. 1. Receiver operating characteristic curve showing the diagnostic performance of

CRP in predicting the development of delirium. The area under curve was 0.68 (95% CI,
0.59-0.72).

Fig. 3. Kaplan-Meier plot of the probability of delirium from ICU admission to day 15.
The result showed that patients with higher CRP values on ICU entry were more likely
to develop delirium (log-rank test, P b .05).

limited by its small sample size and low statistical power.

Furthermore, patients in the study of McGrane et al were more
severely ill than that in our study (median APACHE II score, 28 vs
12). Given good evidence that CRP is associated with the severity of
illness in ICU patients [26-28], the severity of illness may confound
the association between CRP and delirium. In our study, the
subgroup analysis by APACHE II score quartiles indicates that the
association between CRP and delirium is more prominent in patients
with higher APACHE II scores.
Neuroinammation is among the most popular hypothesized
mechanism underlying the pathophysiology of delirium. In this
paradigm, the brain is one of the affected organs by systemic
inammation response in critical illness. Thus, overexpression of
proinammatory cytokines is directly associated with delirium. This
notion is supported by the observation that proinammatory
cytokines such as tumor necrosis factor, interleukin 8, and interleukin
6 are increased in patients with delirium [8,29,30]. In animal model,
the inammation response will activate endothelium and result in
enhanced cytokine transport across the blood-brain barrier. The
inltration of these cytokines into the brain will nally lead to
ischemia and apoptosis of neurons [31,32]. However, most cytokines
except CRP are not routinely available in clinical practice, and some of
them are only used for research purpose. Thus, CRP is the best
candidate for clinical utility for the evaluation of delirium. As CRP is

independently associated with delirium, it is of great interest to

examine whether a reduction in CRP is predictive of the recovery of
delirium in a prospective trial.
Our results are clinically relevant in that CRP is a readily available
biomarker and it is a risk indictor of delirium. By measuring CRP,
clinicians are able to identify patients with high risk of delirium, and
preventive strategies can be initiated. In addition, our study
demonstrates that reduction in serum CRP is associated with the
development of delirium, which mandates further investigations to
test the hypothesis that treatment aiming to ameliorate inammatory
response may help to reduce the risk of delirium. When delirium was
classied into subtypes of hypoactive, hyperactive, and mixed
delirium, the result showed that hypoactive and mixed delirium
were the most common types. That is because sedatives are routinely
given to patients with hyperactive delirium, and hyperactive delirium
is usually transformed to the mixed type.
Limitations of the study include the following aspects: (1)
Delirium is a clinical syndrome with complex pathophysiological
mechanisms, and a single biomarker may be insufcient to predict the
development of delirium. C-reactive protein may be only one part of a
pathogenic cascade, and it has to be in association with a vulnerability
trait to cause a delirious state. By incorporating multiple risk factors
into logistic regression model, we found that other clinical variables
were also independently associated with delirium. Thus, physicians
should assess delirium by incorporating all aspects of risk factors in
clinical practice. Many patients with high levels of CRP may not

Fig. 2. Box plot showing CRP concentrations in patients with and without delirium.
Note: , CRP values were signicantly higher in patients with mixed and hypoactive
delirium than those without (129.1 vs 57.5 mg/L, P b .001; 120.5 vs 57.5 mg/L, P = .031,
respectively).

Fig. 4. Subgroup analysis by APACHE II score quartiles showing the OR for delirium
development with each 10 mg/L increase in CRP.

92

Z. Zhang et al. / Journal of Critical Care 29 (2014) 8892

necessarily develop delirium because other factors such as younger

age and low comorbidity burden may favorably protect them from
delirium. (2) Cardiac surgery and associated cardiopulmonary bypass
are important predisposing factors of delirium [33]; however, this
subgroup of patients cannot be explored in greater detail due to
limited sample size (30 patients in total) in current study. Our hospital
is not a cardiothoracic center, and such patients, especially those with
high comorbidity burden and old age, were mostly referred to another
hospital. That is why cardiac surgery patients in our cohort are less
likely to develop delirium. (3) The study population is heterogeneous,
including surgical, medical, and neurological patients. However, this
allows our result to be generalized to overall critically ill patients. (4)
A large number of etiological factors associated with delirium will
produce inammation and elevate CRP. C-reactive protein may act as
a mediator for the development of delirium. Studies have demonstrated that infection/sepsis, surgery, use of opiates, and propofol will
signicantly inuence CRP levels, and it is favorable to include as
much factors as possible in a regression model. However, because of
the limited number of events (patients with delirium), the number of
factors should be restricted to avoid model overtting.
In conclusion, the study showed that CRP measured on ICU entry
was associated with the development of delirium, and this association
was independent from other conventional risk factors of delirium.
Furthermore, because an increase in CRP is found to be associated
with delirium, serial measurements of CRP may provide useful
information for the management of delirium. We propose that
patients with high CRP on ICU admission are at particularly high
risk for delirium, and thus, preventive measures should be taken to
reduce the risk of delirium. However, this is a hypothesis at present,
and it is of great interest to investigate the cost-effectiveness of this
CRP-directed delirium management strategy as compared with
conventional strategies.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.jcrc.2013.09.002.
References
[1] Sharma A, Malhotra S, Grover S, Jindal SK. Incidence, prevalence, risk factor and
outcome of delirium in intensive care unit: a study from India. Gen Hosp
Psychiatry 2012;34(6):63946.
[2] Veiga D, Luis C, Parente D, Fernandes V, Botelho M, Santos P, et al. Postoperative
delirium in intensive care patients: risk factors and outcome. Rev Bras Anestesiol
2012;62(4):46983.
[3] van den Boogaard M, Pickkers P, Slooter AJ, Kuiper MA, Spronk PE, van der Voort
PH, et al. Development and validation of PRE-DELIRIC (PREdiction of DELIRium in
ICu patients) delirium prediction model for intensive care patients: observational
multicentre study. BMJ 2012;344:e420.
[4] Ouimet S, Kavanagh BP, Gottfried SB, Skrobik Y. Incidence, risk factors and
consequences of ICU delirium. Intensive Care Med 2007;33(1):6673.
[5] Girard TD, Pandharipande PP, Ely EW. Delirium in the intensive care unit. Crit Care
2008;12(Suppl 3):S3.
[6] Salluh JI, Soares M, Teles JM, Ceraso D, Raimondi N, Nava VS, et al. Delirium
Epidemiology in Critical Care Study Group. Delirium epidemiology in critical care
(DECCA): an international study. Crit Care 2010;14(6):R210.
[7] Shehabi Y, Riker RR, Bokesch PM, Wisemandle W, Shintani A, Ely EW. Delirium
duration and mortality in lightly sedated, mechanically ventilated intensive care
patients. Crit Care Med 2010;38(12):23118.
[8] de Rooij SE, van Munster BC, Korevaar JC, Levi M. Cytokines and acute phase
response in delirium. J Psychosom Res 2007;62(5):5215.
[9] van den Boogaard M, Kox M, Quinn KL, van Achterberg T, van der Hoeven JG,
Schoonhoven L, et al. Biomarkers associated with delirium in critically ill patients
and their relation with long-term subjective cognitive dysfunction; indications for

[10]
[11]

[12]

[13]

[14]
[15]

[16]

[17]

[18]

[19]
[20]

[21]

[22]

[23]

[24]

[25]
[26]

[27]

[28]

[29]

[30]

[31]
[32]

[33]

different pathways governing delirium in inamed and noninamed patients. Crit

Care 2011;15(6):R297.
Gabay C, Kushner I. Acute-phase proteins and other systemic responses to
inammation. N Engl J Med 1999;340(6):44854.
Ranzani OT, Zampieri FG, Forte DN, Azevedo LC, Park M. C-reactive protein/albumin ratio predicts 90-day mortality of septic patients. PLoS One 2013;8(3):
e59321.
Litton E, Ho KM, Chamberlain J, Dobb GJ, Webb SA. C-reactive protein
concentration as a predictor of in-hospital mortality after ICU discharge: a nested
case-control study. Crit Care Resusc 2007;9(1):1925.
Zhang Z, Ni H. C-reactive protein as a predictor of mortality in critically ill patients:
a meta-analysis and systematic review. Anaesth Intensive Care 2011;39(5):
85461.
Macdonald A, Adamis D, Treloar A, Martin F. C-reactive protein levels predict the
incidence of delirium and recovery from it. Age Ageing 2007;36(2):2225.
McGrane S, Girard TD, Thompson JL, Shintani AK, Woodworth A, Ely EW, et al.
Procalcitonin and C-reactive protein levels at admission as predictors of
duration of acute brain dysfunction in critically ill patients. Crit Care 2011;
15(2):R78.
Ely EW, Inouye SK, Bernard GR, Gordon S, Francis J, May L, et al. Delirium in
mechanically ventilated patients: validity and reliability of the confusion
assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286(21):
270310.
Morandi A, McCurley J, Vasilevskis EE, Fick DM, Bellelli G, Lee P, et al. Tools to
detect delirium superimposed on dementia: a systematic review. J Am Geriatr Soc
2012;60(11):200513.
Oh SW, Moon JD, Park SY, Jang HJ, Kim JH, Nahm KB, et al. Evaluation of
uorescence hs-CRP immunoassay for point-of-care testing. Clin Chim Acta
2005;356(12):1727.
Gastmans C, Milisen K. Use of physical restraint in nursing homes: clinical-ethical
considerations. J Med Ethics 2006;32(3):14852.
Van Rompaey B, Elseviers MM, Schuurmans MJ, Shortridge-Baggett LM, Truijen S,
Bossaert L. Risk factors for delirium in intensive care patients: a prospective cohort
study. Crit Care 2009;13(3):R77.
Ely EW, Margolin R, Francis J, May L, Truman B, Dittus R, et al. Evaluation of
delirium in critically ill patients: validation of the Confusion Assessment Method
for the Intensive Care Unit (CAM-ICU). Crit Care Med 2001;29(7):13709.
Ely EW, Truman B, Shintani A, Thomason JW, Wheeler AP, Gordon S, et al.
Monitoring sedation status over time in ICU patients: reliability and validity of the
Richmond Agitation-Sedation Scale (RASS). JAMA 2003;289(22):298391.
Peterson JF, Pun BT, Dittus RS, Thomason JW, Jackson JC, Shintani AK, et al.
Delirium and its motoric subtypes: a study of 614 critically ill patients. J Am
Geriatr Soc 2006;54(3):47984.
Bagley SC, White H, Golomb BA. Logistic regression in the medical literature:
standards for use and reporting, with particular attention to one medical domain.
J Clin Epidemiol 2001;54(10):97985.
Hosmer DW, Taber S, Lemeshow S. The importance of assessing the t of logistic
regression models: a case study. Am J Public Health 1991;81(12):16305.
Cervoni JP, Thvenot T, Weil D, Muel E, Barbot O, Sheppard F, et al. C-reactive
protein predicts short-term mortality in patients with cirrhosis. J Hepatol
2012;56(6):1299304.
Kim KI, Oh SW, Ahn S, Heo NJ, Kim S, Chin HJ, et al. CRP level and HDL cholesterol
concentration jointly predict mortality in a Korean population. Am J Med
2012;125(8):787795.e4.
Takahashi R, Ito Y, Takahashi H, Ishii H, Kasuga H, Mizuno M, et al. Combined
values of serum albumin, C-reactive protein and body mass index at dialysis
initiation accurately predicts long-term mortality. Am J Nephrol 2012;36(2):
13643.
Alexander JJ, Jacob A, Cunningham P, Hensley L, Quigg RJ. TNF is a key mediator of
septic encephalopathy acting through its receptor, TNF receptor-1. Neurochem Int
2008;52(3):44756.
van Munster BC, Bisschop PH, Zwinderman AH, Korevaar JC, Endert E, Wiersinga
WJ, et al. Cortisol, interleukins and S100B in delirium in the elderly. Brain Cogn
2010;74(1):1823.
Pan W, Yu C, Hsuchou H, Zhang Y, Kastin AJ. Neuroinammation facilitates LIF
entry into brain: role of TNF. Am J Physiol Cell Physiol 2008;294(6):C143642.
Spiegel DR, Chen V. A case of postoperative cognitive decline, with a highly
elevated C- reactive protein, status post left ventricular assist device insertion: a
review of the neuroinammatory hypothesis of delirium. Innov Clin Neurosci
2012;9(1):3541.
Guenther U, Theuerkauf N, Frommann I, Brimmers K, Malik R, Stori S, et al.
Predisposing and precipitating factors of delirium after cardiac surgery: a
prospective observational cohort study. Ann Surg 2013;257(6):11607.