Risk factors for the development of early-onset delirium
and the subsequent clinical outcome in mechanically
Shu-Min Lin MD, Chien-Da Huang MD, Chien-Ying Liu MD,
Horng-Chyuan Lin MD, Chun-Hua Wang MD, Pei-Yao Huang MD,
Yueh-Fu Fang MD, Meng-Heng Shieh MD, Han-Pin Kuo MD, PhD*
Department of Thoracic Medicine, Chang Gung Memorial Hospital, School of Medicine,
Chang Gung University, Taipei 333, Taiwan
Objectives: The aim of the study was to investigate the risk factors of developing early-onset delirium
in mechanically ventilated patients and determine the subsequent clinical outcomes.
Methods: Confusion assessment method for the intensive care unit (ICU) was used to assess the
enrolled mechanically ventilated patients for delirium. The risk factors of developing delirium and
clinical outcomes were determined in these subjects.
Results: Delirium was present in 31 (21.7%) of 143 patients in the first 5 days. In multivariable
analysis, hypoalbuminemia (odds ratio, 5.94; 95% confidence interval, 1.23-28.77) and sepsis (odds
ratio, 3.65; 95% confidence interval, 1.03-12.9) increased the risk of developing delirium in
mechanically ventilated patients. The patients with delirium had a higher in-hospital mortality
(67.7% vs 33.9%, respectively; P = .001) and longer duration of mechanical ventilation (19.5 F 15.8 vs
9.3 F 8.8 days, respectively; P = .003) than patients without delirium. The incidence of nosocomial
pneumonia was increased in delirious patients (64.5% vs 38.4%, P = .01) compared with nondelirious
patients, whereas the lengths of ICU or hospital stay were similar between both groups.
Conclusions: Mechanically ventilated patients with sepsis or hypoalbuminemia were more vulnerable
to develop delirium in their early stay in the ICU. Early-onset delirium is associated with
prolonged duration of mechanical ventilation and higher incidence of nosocomial pneumonia, leading
to a higher mortality.
D 2008 Elsevier Inc. All rights reserved.
Delirium is an acute change in mental status with a
fluctuating course of impaired attention and disorganized
thinking. Delirium is especially common in certain groups
of patients including elderly patients, patients with alcohol
0883-9441/$ – see front matter D 2008 Elsevier Inc. All rights reserved.
BThis project is supported by NSC-92-2314-B-182A-158, National
Science Council, Taiwan, ROC.
* Corresponding author. Tel.: +886 3 328 1200x8467; fax: +886 3
E-mail address: email@example.com (H.-P. Kuo).
Journal of Critical Care (2008) 23, 372–379
abuse, and patients in the intensive care unit (ICU) [1-3].
Several risk factors for the development of delirium 
have been identified, including advanced age, preexisting
brain disease or cognitive impairment, multiple medica-
tions, and multiple systemic illnesses . Evidence
indicates that development of delirium is associated with
longer ICU stay , higher medical cost , and increased
mortality rate [7-9].
Delirium, especially the hypoactive type, is usually
overlooked by many clinicians. Detection of delirium in
patients with mechanical ventilation is even more difficult
because of a lack of verbal communication. The introduc-
tion of nonverbal assessment, confusion assessment method
for the ICU (CAM-ICU), has improved the detection rate of
delirium in those patients . Development of delirium in
mechanically ventilated patients has been shown to be
associated with as much as a 3-fold increase in risk for
mortality even after controlling for preexisting comorbid-
ities, illness severity, and use of sedative medication .
Our recent study also demonstrates that development of
delirium in the early course of ICU stay (early-onset
delirium) is as important as the Acute Physical and Chronic
Health Evaluation III (APACHE III) score in predicting
patients’ outcome and is associated with a higher mortality
rate . Thus, the early onset of delirium in patients with
mechanical ventilation may be regarded as a warning sign
for illness deterioration, and interventions should be
implemented for modifiable factors.
This study is designed to identify the risk factors for the
development of early-onset delirium in mechanically
ventilated patients and to determine the impacts on
subsequent clinical outcomes, including the length of
ICU stay, duration of mechanical ventilation, and occur-
rence of complications.
2. Materials and methods
The institutional review board approved this study, and
informed consent was obtained from the patients or
surrogates. The study recruited mechanically ventilated
adult ICU patients admitted to a university-affiliated
medical center medical ICU. From October 2002 to May
2003, there were 206 patients admitted to the ICU, and
175 patients required mechanical ventilation support.
Among the 175 patients, 151 were enrolled and 24 were
excluded; some of these subjects were included in a
previously published investigation of the association
between delirium and survival . The exclusion criteria
of this study were a history of chronic dementia, psychosis,
mental retardation, or other neurologic disease; patients
receiving antipsychotics, high dose of morphine (N50 mg/d),
or midazolam (N0.09 mg/kg per hour); general anesthesia,
recovering from surgery, or heavily sedated combined with
neuromuscular blocking agents that would confound the
diagnosis of delirium (n = 20); and patient or family refusal
to participate (n = 4). The presence of dementia was
defined as in a previous report . Briefly, in addition to a
history of severe dementia, the Blessed Dementia Rating
Scale  was used at enrollment to screen for previously
unrecognized dementia via surrogate interviews. Patients
were defined as having bsuspected dementiaQ if they met
any of the following 2 criteria: a Blessed Dementia Rating
Scale score of 3 or higher, or a rating by the surrogate of
3 or higher out of 5 as bpossibly having dementia.Q The
definition of heavily sedation was determined by the
Richmond Agitation Sedation Scale  scoring ranging
from ?3 to ?5. In addition, 8 of the 151 enrolled patients
remained comatose throughout the investigation and were
excluded from analysis. Therefore, a total of 143 patients
were included in the final analysis. Exclusion criteria are
outlined in the patient flow diagram (Fig. 1).
2.2. Baseline assessment
Before the beginning of the study, the questionnaires
were pilot tested and research assistants were well trained
and standardized in the interview procedure . Each
patient and their family members were interviewed, and the
Flow diagram of patients in the study cohort.
Risk factors for the development of early-onset delirium 373
medical records were also reviewed to obtain the baseline
information including demographic data, illness severity,
and principal diagnosis for admission to ICU. The APACHE
III and sepsis-related organ failure assessment (SOFA) score
were rated by chart review of data obtained within 24 hours
of ICU admission. The underlying medical history was also
collected including diabetes, hypertension, cerebral stroke,
chronic heart disease (coronary disease, heart failure, and
arrhythmia), and chronic airway diseases (asthma, chronic
airway obstructive disease, and bronchiectasis). In addition,
alcohol abuse  and administration of tranquilizers,
narcotic analgesics, sedative agents, and systemic steroids
were also recorded. The assessors daily evaluated the
presence of fever, sepsis, shock, and laboratory values
including albumin, glucose, nitrogen, creatinine, bilirubin,
sodium, calcium, and arterial oxygen and carbon dioxide
levels. Sepsis was defined by the consensus committee of
the American College of Chest Physicians and the Society
of Critical Care Medicine . The definition of shock was
systolic blood pressure of less than 90 mm Hg or mean
arterial pressure of less than 60 mm Hg.
2.3. Delirium assessment
Delirium assessments of the included patients were
conducted by 2 research assistants independently by means
of CAM-ICU in the first 5 days of their ICU stay. The
assistants performed daily assessments between 9:00 am
and 12:00 pm. If the patients were not accessible because
they were comatose, the research assistants would try
another evaluation between 2:00 pm and 5:00 pm on the
same day. The subjects were excluded only when 3 consec-
utive days of assessment had failed. There were 8 patients
excluded from this study because of persistent coma
throughout the investigation. All the tasks and questions
were designed to be nonverbal. The CAM-ICU contains
4 categories of assessments: I, acute onset of mental status
changes or fluctuating course; II, inattention; III, disorga-
nized thinking; and IV, altered level of consciousness.
Delirium was defined as presence of both categories I and II
plus either III or IV. Patients were recognized as bwith
deliriumQ if they ever met the criteria for delirium with the
first 5 days of ICU admission, whereas those who never
matched the criteria for delirium were recognized as b no
delirium.Q Once the patients were recognized as delirious,
the assessment was continued until the patients became
CAM-ICU–negative in 2 consecutive days. The patients
were followed up daily with the Glasgow Coma Scale 
and the Richmond Agitation Sedation Scale for assessment
of acute onset of mental status changes or fluctuating
course. The caregivers in the ICU were blind to the results
of delirium assessment.
2.4. Outcome assessment
The primary outcome of this study was the development
of delirium within the first 5 days of ICU admission and the
secondary outcomes included ICU mortality, length of ICU
stay, length of hospital stay, and complications after
admission to ICU. The study also determined the 10- and
30-day successful extubation rate, reintubation rate, and
other delirium-associated care problems, such as self-
extubation or self-removal of catheter. The group of failed
extubation was designated as those who failed spontaneous
breath trial and needed reinstitution with mechanical
ventilation, or those who were extubated but were reintu-
bated within 48 hours. In contrast, the group of successful
extubation was designated as those who were extubated
successfully after spontaneous breath trial. The need for
transient noninvasive bi-level pressure ventilation after
extubation was not considered to be a weaning failure.
2.5. Statistical analysis
Descriptive statistics were used to examine the demo-
graphic characteristics of the study population. Data were
expressed as median and interquartile range (IQR).
Colinearity of variables in the model was evaluated using
Spearman correlation coefficient before running the anal-
ysis. The baseline characteristics, disease, and laboratory
variables between the delirious and nondelirious patients
were compared using the Wilcoxon rank sum test for
continuous variables and the v2or the Fisher exact test
(if total sample size is b20) for categorical variables.
Univariate analyses (Wilcoxon rank sum test or v2test)
were primarily used for selection of variables, based on a
Patient characteristics of mechanically ventilat-
(n = 31)
(n = 112)
Age, y, median (IQR)
Female sex, n (%)
BMI, median (IQR)
APACHE III score,
Principal admission diagnosis, n (%)
Chronic lung disease
Congestive heart failure
Ischemic heart disease
8 (5-11)6 (4-9)
aInclude diabetic ketoacidosis or hyperosmolar hyperglycemic
S.-M. Lin et al.374
P value of less than .1. The selected variables including
diabetes mellitus, sepsis, and hypoalbuminemia were
entered into a multivariate stepwise regression analysis to
identify the net effects of each individual factor with
control of the others. Odds ratios (ORs) and their 95%
confidence intervals (CIs) were used to assess independent
contribution of significant factors.
Survival times were measured from the date of enroll-
ment, and death from all causes within the period of hospital
stay was taken as outcome. The comparisons of in-hospital
survival between patients with delirium and those without
delirium were performed by Cox proportional hazard
regression with variables including age, sex, mean arterial
pressure, arterial pH value, Pao2/Fio2, white blood cell
count, preexisting medical conditions, sepsis, and shock. A
P value of less than .05 was considered as statistical
significance. Analyses were performed using SPSS software
version 10.0 (SPSS, Chicago, Ill).
3.1. Patient characteristics
The 143 enrolled patients were divided into 2 groups
depending on whether they developed delirium in the first
5 days of ICU stay. The 2 groups had similar baseline
characteristics including demographic variables, SOFA,
APACHE III score, or principal admission diagnoses
(Table 1). All of the 143 patients were eligible for the
nonverbal CAM-ICU test and 31 (21.7%) patients were
found to be delirious during the first 5 days of ICU stay.
Delirium developed mostly on the second day (n = 14,
45.2%) after admission to ICU. The mean duration of
delirium was 3.2 F 1.7 days.
3.2. Univariate analysis
Univariate analysis (Table 2) allowed us to identify the
possible risk factors of developing early-onset delirium
during the ICU stay in mechanically ventilated patients.
We analyzed the underlying comorbidities, laboratory
variables, and administration of medications on the day
of admission to ICU and identified 3 possible risk factors
for development of early-onset delirium among our patient
population. A history of diabetes increased the risk of
delirium (OR, 2.79; 95% CI, 1.23-6.29). Meanwhile,
patients admitted to the ICU with sepsis had a higher
incidence of delirium development in their early stay in the
ICU (OR, 3.13; 95% CI, 1.19-8.22). Hypoalbuminemia
detected on admission to ICU was also a significant
predictor of the development of delirium in these patients
(OR, 6.27; 95% CI, 1.8-21.87).
3.3. Multivariate analysis
Furthermore, we entered those 3 selected factors of
delirium into multivariate stepwise regression analysis to
identify the net effects of individual factor with control of
Factors associated with delirium in mechanically ventilated patients
FactorDelirium (n = 31), n (%) No delirium (n = 112), n (%) OR 95% CI
History of stroke
Chronic airway disease
Chronic heart disease
Elevated serum level of creatinine
Elevated serum level of urea nitrogen
Elevated serum level of total bilirubin
Administration of medications
Risk factors for the development of early-onset delirium375
the others (Table 3). Hypoalbuminemia and the presence of
sepsis on admission to ICU appeared to be independent
contributing factors to the development of early-onset
delirium, increasing the odds by a factor of 5.94 (95% CI,
1.23-28.77) and 3.65 (95% CI, 1.03-12.9), respectively. A
history of diabetes seemed not to independently increase the
risk to be delirious in the early ICU stay.
3.4. Early-onset delirium-associated outcomes
During the hospital stay period, 67.7% (21/31) of the
patients in the delirium group died vs 33.9% (38/112) of the
patients in the nondelirium group (P = .001) (Table 4).
Fig. 2A showed survival curves during hospital stay among
the patients in both groups, with significantly higher
mortality among patients with delirium (Cox regression
analysis, P = .003; hazard ratios, 2.69; 95% CI, 1.41-5.10).
Survival times were measured from the date of enrollment,
and death from all causes within the period of hospital stay
was taken as outcome. However, the causes of death were
not significantly different between 2 groups. Septic shock
and respiratory failure were the first and second popular
causes of death among the both groups, respectively.
Patients with delirium received longer duration of mechan-
ical ventilation than patients without delirium (median,
16 days [IQR, 6-30 days] vs 6 days [IQR 3.5-12.5 days],
respectively; P = .003) (Table 4). Kaplan-Meier plots of the
with delirium in mechanically ventilated patients
Multivariate analysis of major factors associated
FactorOR 95% CI
Clinical outcomes and complications of patients
(n = 31)
(n = 112)
Days of hospital stay,
Days of ICU stay,
Days of mechanical
10-d extubation rate
30-d extubation rate
pneumonia, n (%)
edema, n (%)
Reintubation, n (%)
Self-extubation, n (%)
catheter, n (%)
21 (67.7%)38 (33.9%).001
29.5 (18.5-55) 26.5 (14.5-45) .402
13 (7-29.5)8 (5-18) .115
16 (6-30) 6 (3.5-12.5) .003
3 (9.7%)13 (11.6%).646
20 (64.5%) 43 (38.4%).010
6 (19.4%) 11 (9.8%).211
aP value for v2test or Fisher exact test (if total sample size is b20)
in the case of categorical variables, and for Wilcoxon rank sum test in
the case of quantitative variables.
patients with and without delirium B, The probability of remaining
on mechanical ventilator over time between patients with and
without delirium. Dashed line represents patients with delirium;
continuous line, patients without delirium. Censored cases are
shown by solid circles.
A, The in-hospital survival proportion according to
S.-M. Lin et al. 376
probability of successful extubation over time between
patients with and without delirium are shown in Fig. 2B
(P = .0007; hazard ratio, 0.48; 95% CI, 0.27-0.70). Patients
with delirium had lower 10-day (35.5%) and 30-day (70%)
successful extubation rate when compared with patients
without delirium (59.8% and 94.6%, P = .016 and P = .0001,
respectively). However, there was no difference in length of
ICU or hospital stay between 2 groups. The incidence of
self-extubation, reintubation, and removal of catheter was
similar in 2 groups. The patients with delirium had a higher
incidence of development of nosocomial pneumonia during
the ICU stay compared with patients without delirium
(64.5% and 38.4%, respectively, P = .01) (Table 4), whereas
the incidence of development of adult respiratory distress
syndrome and pulmonary edema during ICU stay were
similar between 2 groups.
This study demonstrated that the presence of sepsis and
hypoalbuminemia on admission to ICU were independent
predictors for development of early-onset delirium in
patients requiring mechanical ventilator support, even after
adjusting for subjects’ age, sex, illness severity, comorbid-
ities, and laboratory variables. The development of early-
onset delirium was associated with grave clinical outcomes,
including in-hospital mortality, duration of mechanical
ventilation, and 10-day as well as 30-day extubation
rate. During the ICU stay, patients with delirium had
higher incidence of development of nosocomial pneumonia
when compared with nondelirious patients. However,
the length of ICU and hospital stay, reintubation, self-
extubation rate, and self-removal of catheter incidence in
patients with delirium did not differ from those patients
There are numerous risk factors associated with the
development of delirium . Traditional factors associated
with delirium in general ward patients may not be
applicable to critically ill patients . Compared with the
common contributing factors of delirium in ICU in
previous reports [2,9], our results indicated that early-onset
delirium in mechanically ventilated patients was not
attributed to use of narcotics or sedatives, abnormal
bilirubinemia, or hypertension. Compatible with previous
reports [3,17], patients with sepsis are more vulnerable to
be delirious in the early stay in the ICU. The relationship
between sepsis and delirium could be attributed to the
adverse effects of bacterial endotoxin or cytokines evoked
by sepsis on central nervous system (CNS). Previous
reports have demonstrated that endotoxins are responsible
for the depressed CNS function in sepsis [18-20]. Further-
more, cytokines including interleukin (IL) 1b, IL-10, and
IL-13 have been reported to be implicated in the CNS
dysfunction of sepsis . Decreased regional cerebral
blood flow has been reported in patients with delirium ;
thus, sepsis-induced alteration of CNS blood flow could be
one of the mechanisms responsible for development of
delirium among the septic patients.
Previous studies have revealed that hypoalbuminemia
increases the risk of development of delirium in hospital-
ized patients [23,24]. Hypoalbuminemia detected in the
first 5 days of ICU stay may indicate patients have been in
a status of long-term poor nutrition or protein loss due to
proteinuria, or decreased protein synthesis such as liver
cirrhosis. The decreased osmotic force due to hypoalbumi-
nemia fails to maintain adequate intravascular volume and
might therefore decrease brain perfusion, vulnerable to
develop a delirious state. In addition, albumin has a role in
binding and transporting many medications, leading to
higher free drug concentrations in hypoalbuminemia and,
therefore, more potential for precipitating delirium. Never-
theless, hypoalbuminemia may be a common feature of
multiple comorbidities that contribute to the development
Although all of our study patients were admitted to ICU
for mechanical ventilator support for their respiratory
insufficiency, only 28.8% of our study patients died of
respiratory failure. Most of the other causes of death were
due to nonpulmonary major organ dysfunctions. There was
no significant difference in the causes of death between the
delirious and nondelirious groups in terms of commonly
mentioned major organ dysfunction. Because the develop-
ment of delirium in the early course of ICU stay was
independently associated with a marked increase in in-
hospital mortality, delirium, representing brain dysfunction,
may therefore be regarded as an important nonpulmonary
major organ dysfunction contributing to mortality.
Our study demonstrated that patients with delirium
during the early course of ICU stay required longer duration
of mechanical ventilator support and a lower extubation rate
within 10 and 30 days than patients without delirium.
Prolonged duration of mechanical ventilation in patients
with delirium developed during ICU stay may be caused by
altered consciousness [10,25] or more complicated comor-
bidities , which delay the weaning process. There are
several clinical factors other than delirium or other
encephalopathy related to the failure to wean from
ventilator. Critical illness neuromyopathy induced by sepsis
or medication may play a more important role than delirium
among ICU patients. Further investigation is needed to
explore whether delirium itself or the underlying comorbid-
ities contribute to prolonged duration of mechanical
ventilation. Patients with delirium were more likely to have
nosocomial pneumonia during their later stay in ICU.
Prolonged mechanical ventilation has been documented to
be a major risk factor for nosocomial pneumonia .
However, nosocomial pneumonia may also contribute to
prolong the duration of mechanical ventilation. Our study
did not intend to explore the cause-and-effect relationship
between prolonged mechanical ventilation and nosocomial
pneumonia, but indicated both clinical outcomes were more
Risk factors for the development of early-onset delirium 377
commonly occurred to patients with early-onset delirium.
Prolonged mechanical support and increased incidence of
pneumonia may further precipitate other major organ
dysfunction leading to mortality. Thus, early detection of
delirium in mechanically ventilated patients with preventive
or therapeutic interventions may alter the outcome, as those
in non-ICU hospitalized patients .
Some limitations of this study need to be noted. First,
assessment of delirium only in the first 5 days of ICU
admission may underestimate the prevalence of delirium in
this cohort. There must be more incidence of delirium in the
later ICU stay, but the underlying mechanisms and the
contributions to patients’ outcome may be different. It is
possible that risk factors for the development of delirium in
ventilated medical ICU patients during their hospitalization
differ from those identified in this study within the first
5 days and may be associated with poor outcomes.
In addition, this study excluded the patients in a state of
coma when they were enrolled. Development of coma may
be associated with delirium beyond the 5-day period studied
and negative clinical outcomes may have been under-
estimated by eliminating those patients with more than
3 days of coma. Furthermore, excluding patients who
received high doses of morphine or midazolam may also
limit the prevalence of delirium detected, as well as mask
the risk of developing delirium due to sedative medications.
Lastly, administration of narcotics and sedatives has
been reported to be associated with development of
delirium in ICU patients [2,3], but our study revealed that
neither sedatives nor narcotics was a predictor for
development of delirium in the first 5 days of ICU stay.
The discrepancy could be raised from the different
protocols for delirium assessments between those studies.
Our study detected delirium only in the first 5 days of ICU
stay, whereas the previous studies accessed the develop-
ment of delirium during the whole course of ICU stay.
Previous report has demonstrated that development of
delirium was associated only with administration of high
dose of sedatives, but not low dose of sedatives . Our
study patients received only minor dose of sedatives and
narcotics in the first 5 days of ICU stay. In the early course
of ICU stay, patients may not receive as much narcotics or
sedatives as the patients with delirium that developed later
in the course of ICU stay. There must be more incidence of
delirium in the later ICU stay, but the underlying
mechanisms and the contributions to patients’ outcome
deserve a further study in the future.
In conclusion, this study demonstrated mechanically
ventilated patients with sepsis or hypoalbuminemia were
more vulnerable to develop delirium. The development of
delirium in those patients was associated with prolonged
duration of mechanical ventilation and higher incidence
of nosocomial pneumonia, and higher mortality. Our study
raised the awareness that delirium should be intensively
monitored in ventilated patients of ICU, and earlier preven-
tion or treatment of delirium might affect the outcomes.
The authors express their appreciation to Dr Ling-Ling
Hsieh for her assistance in statistical analysis.
 Inouye SK, Bogardus Jr ST, Charpentier PA, et al. A multicomponent
intervention to prevent delirium in hospitalized older patients. N Engl
J Med 1999;340:669-76.
 Dubois MJ, Bergeron N, Dumont M, et al. Delirium in an
intensive care unit: a study of risk factors. Intensive Care Med
 Ely EW, Gautam S, Margolin R, et al. The impact of delirium in the
intensive care unit on hospital length of stay. Intensive Care Med 2001;
 Practice guideline for the treatment of patients with delirium.
American Psychiatric Association. Am J Psychiatry 1999;156:1-20.
 Lukan JK, Reed Jr DN, Looney SW, et al. Risk factors for delirium
tremens in trauma patients. J Trauma 2002;53:901-6.
 Milbrandt EB, Deppen S, Harrison PL, et al. Costs associated with
delirium in mechanically ventilated patients. Crit Care Med 2004;32:
 Kakuma R, du Fort GG, Arsenault L, et al. Delirium in older
emergency department patients discharged home: effect on survival.
J Am Geriatr Soc 2003;51:443-50.
 Lawlor PG, Gagnon B, Mancini IL, et al. Delirium as a predictor of
survival in older patients with advanced cancer. Arch Intern Med
 Ely EW, Shintani A, Truman B, et al. Delirium as a predictor of
mortality in mechanically ventilated patients in the intensive care unit.
 Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically
ventilated patients: validity and reliability of the confusion assess-
ment method for the intensive care unit (CAM-ICU). Jama 2001;286:
 Lin SM, Liu CY, C.H. W, et al. The impact of delirium on the survival
of mechanically ventilated patients. Crit Care Med 2004;32:2254-9.
 Blessed G, Tomlinson BE, Roth M. The association between
quantitative measures of dementia and of senile change in the cerebral
grey matter of elderly subjects. Br J Psychiatry 1968;114:797-811.
 Ely EW, Truman B, Shintani A, et al. Monitoring sedation status over
time in ICU patients: reliability and validity of the Richmond
Agitation-Sedation Scale (RASS). JAMA 2003;289:2983-91.
 O’Connor PG, Schottenfeld RS. Patients with alcohol problems.
N Engl J Med 1998;338:592-602.
 Bone RC, Sibbald WJ, Sprung CL. The ACCP-SCCM consensus
conference on sepsis and organ failure. Chest 1992;101:1481-3.
 Teasdale G, Jennett B. Assessment of coma and impaired conscious-
ness. A practical scale. Lancet 1974;2:81-4.
 Aldemir M, Ozen S, Kara IH, et al. Predisposing factors for delirium
in the surgical intensive care unit. Crit Care 2001;5:265-70.
 Eidelman LA, Putterman D, Putterman C, et al. The spectrum of septic
encephalopathy. Definitions, etiologies, and mortalities. JAMA 1996;
 Wyler F, Rutishauser M, Weisser K. Endotoxin induced regional
circulatory reactions in the rabbit with and without halothane
anesthesia. J Surg Res 1972;13:13-9.
 Coelho A, Fioramonti J, Bueno L. Brain interleukin-1beta and tumor
necrosis factor–alpha are involved in lipopolysaccharide-induced
delayed rectal allodynia in awake rats. Brain Res Bull 2000;52:223-8.
 Wong ML, Bongiorno PB, Rettori V, et al. Interleukin (IL) 1beta, IL-1
receptor antagonist, IL-10, and IL-13 gene expression in the central
nervous system and anterior pituitary during systemic inflammation:
S.-M. Lin et al.378
pathophysiological implications. Proc Natl Acad Sci U S A 1997;94: Download full-text
 Yokota H, Ogawa S, Kurokawa A, et al. Regional cerebral blood flow
in delirium patients. Psychiatry Clin Neurosci 2003;57:337-9.
 Trzepacz PT, Francis J. Low serum albumin levels and risk of
delirium. Am J Psychiatry 1990;147:675.
 Ljubisavljevic V, Kelly B. Risk factors for development of delirium
among oncology patients. Gen Hosp Psychiatry 2003;25:345-52.
 McNicoll L, Pisani MA, Zhang Y, et al. Delirium in the intensive care
unit: occurrence and clinical course in older patients. J Am Geriatr Soc
 O’Keeffe S, Lavan J. The prognostic significance of delirium in older
hospital patients. J Am Geriatr Soc 1997;45:174-8.
 Hoffken G, Niederman MS. Nosocomial pneumonia: the importance
of a de-escalating strategy for antibiotic treatment of pneumonia in the
ICU. Chest 2002;122:2183-96.
Risk factors for the development of early-onset delirium 379