695235

research-article2017
ACC0010.1177/2048872617695235European Heart Journal: Acute Cardiovascular CareFalsini et al.

EUROPEAN
SOCIETY OF
Original scientific paper CARDIOLOGY

European Heart Journal: Acute Cardiovascular Care

Long-term prognostic value of 110

The European Society of Cardiology 2017
Reprints and permissions:
delirium in elderly patients with sagepub.co.uk/journalsPermissions.nav
https://doi.org/10.1177/2048872617695235
DOI: 10.1177/2048872617695235
acute cardiac diseases admitted journals.sagepub.com/home/acc

to two cardiac intensive care units:

a prospective study (DELIRIUM
CORDIS)

Giovanni Falsini1, Simone Grotti1, Italo Porto2,

Giulio Toccafondi3, Aureliano Fraticelli1, Paolo Angioli1,
Kenneth Ducci1, Francesco Liistro1, Maurizio Pieroni1,
Tamara Taddei4, Serena Romanelli1, Roberto Rossi1
and Leonardo Bolognese1

Abstract
Background: Delirium is a frequent in-hospital complication in elderly patients, and is associated with poor clinical
outcome. Its clinical impact, however, has not yet been fully addressed in the setting of the cardiac intensive care
unit (CICU). The present study is a prospective, two-centre registry aimed at assessing the incidence, prevalence and
significance of delirium in elderly patients with acute cardiac diseases.
Methods: Between January 2014 and March 2015, all consecutive patients aged 65 years or older admitted to the
CICU of our institutions were enrolled and followed for 6 months. Delirium was defined according to the confusion
assessment method.
Results: During the study period, 726 patients were screened for delirium. The mean age was 79.17.8 years. A total
of 111 individuals (15.3%) were diagnosed with delirium; of them, 46 (41.4%) showed prevalent delirium (PD), while
65 (58.6%) developed incident delirium (ID). Patients 85 years or older showed a delirium rate of 52.3%. Hospital
stay was longer in delirious versus non-delirious patients. Patients with delirium showed higher in-hospital, 30-day and
6-month mortality compared to non-delirious patients, irrespective of the onset time (overall, ID or PD). Six-month
re-hospitalisation was significantly higher in overall delirium and the PD group, as compared to non-delirious patients.
KaplanMeier analysis showed a significant reduction of 6-month survival in patients with delirium compared to those
without, irrespective of delirium onset time (i.e. ID or PD). A positive confusion assessment method was an independent
predictor of short and long-term mortality.
Conclusions: Delirium is a common complication in elderly CICU patients, and is associated with a longer and more
complicated hospital stay and increased short and long-term mortality. Our findings suggest the usefulness of a protocol
for the early identification of delirium in the CICU.
Clinicaltrials.gov: NCT02004665

1Cardiovascular Diseases and Neurology Department, San Donato Corresponding author:

Hospital, Italy
2Cardiovascular and Thoracic Diseases Department, Fondazione
Policlinico Gemelli, Italy
3Center for Clinical Risk Management, Italy
4Cardiology Unit, Santa Maria alla Gruccia Hospital, Italy
Giovanni Falsini, Cardiovascular Department, San Donato Hospital, Via
Pietro Nenni 20, 52100 Arezzo, Italy.
Email: giovanni.falsini@uslsudest.toscana.it
2 European Heart Journal: Acute Cardiovascular Care
Keywords
Delirium, cardiac intensive care unit, mortality, elderly, confusion assessment method
Date received: 13 August 2016; accepted: 2 February 2017
When difficult breathing and delirium occur in a fever that
is not of the intermittent type, the case is mortal.
Hippocrates. Aphorisms IV: 51
Delirium is a clinical syndrome characterised by inattention
and acute cognitive dysfunction, which results from the
interaction of vulnerability of the patient (i.e. the presence
of predisposing conditions, such as cognitive impairment,
severe illness, or visual impairment) and hospital-related
insults (e.g. medications and procedures), It is often misdi-
agnosed and mismanaged.1,2
The risk of delirium is increased in selected subsets,
such as elderly people with pre-existing cognitive impair-
ment3 or terminal illnesses,4 and, among hospitalised
patients, those admitted to an intensive care unit (ICU) are
more likely to develop such a condition.5
In non-cardiovascular critical care, increasing aware-
ness of delirium in the last decade has been followed by
the development of methods for early detection and risk
factor assessment, and has led to the creation of targeted
intervention strategies aimed at limiting its consequences.6
With a prevalence as high as 80% among mechanically
ventilated patients, delirium is quite common in the ICU,7,8
and its association with poor clinical outcomes is well
established in this setting.9,10
However, less is known about delirium and its signifi-
cance in the cardiac intensive care unit (CICU). According
to the growing evidence supporting its increased burden
in cardiac care settings,11 delirium has recently been rec-
ognised as a highly prevalent comorbid condition among
CICU patients.8,12,13 In a recent retrospective analysis,
Pauley etal. found a correlation between the prevalence
of delirium and both poor survival rate and greater
resource consumption.14
Notwithstanding this evidence, the current body of data
stems from studies with clear limitations (small sample size,
mixed medicalsurgical populations and retrospective design).
We therefore sought to assess prospectively the inci-
dence, prevalence and significance of delirium in a large,
homogenous cohort of elderly patients with acute cardiac
diseases.
Methods
Study design and patients
This study is a prospective, two-centre, observational
cohort study aimed at evaluating the incidence, prevalence
and significance of delirium in CICU patients. It was
approved by the local ethics committee and was carried out
in accordance with the Helsinki declaration. All patients
provided written informed consent.
Between January 2014 and March 2015, all consecutive
patients aged 65 years or over admitted to the CICU of our
institutions were enrolled. Patients were excluded if the pri-
mary diagnosis was a non-cardiovascular condition. Of note,
patients who need orotracheal intubation are not routinely
cared for in the CICU (as opposed to the general ICU) in our
network, and were excluded, whereas patients treated with
non-invasive ventilation were included. Enrolled individuals
were evaluated once daily during their CICU stay (Figure 1).
Measurement of delirium
Patients were routinely assessed using the Richmond agi-
tation sedation scale (RASS)7 at admission (Supplementary
material Table 1) and then once daily. Delirium was
addressed using the confusion assessment method (CAM)
(Supplementary material Table 2)15 scale at admission and
then if the RASS was 3 or greater to avoid including
comatose/unconscious patients. The CAM diagnostic
algorithm is based on four pivotal features of delirium: (1)
acute onset and fluctuating course; (2) inattention; (3) dis-
organised thinking; (4) altered level of consciousness. The
diagnosis of delirium by CAM requires the presence of
features 1, 2 and either 3 or 4. A patient was considered
delirious if the CAM score was positive or non-deliri-
ous if the score was negative. Delirium was considered
prevalent if CAM was positive within 24 hours from
admission and incident if CAM was positive after 24
hours from admission until discharge.
At hospital admission the nursing staff administered a
patient information sheet to the relatives, produced as part
of this study, to provide information for people who have
experienced delirium and a paper copy of the modified
short form of the informant questionnaire on cognitive
decline in the elderly (Short IQCODE). Short form
IQCODE, a simplification from the original 26-item ques-
tionnaire, is a 16-item informant questionnaire aimed at
retrospectively assessing changes in cognitive and func-
tional performance over a 10-year time period.16 IQCODE
is designed to screen individuals for potential dementia,
and was administered to the relevant proxy. Once IQCODE
was completed, the nursing staff registered the IQCODE
score on an electronic form and performed CAM.
Management of patients with delirium
Every patient meeting the selection criteria for delirium was
reported to the physician and re-evaluated at each nursing
Falsini et al. 3

Figure 1. Study flow chart. Of the original 726 patients enrolled, 111 patients were classified as CAM positive (delirium); patients
with positive CAM were then classified as prevalent delirium cases if CAM was positive within 24 hours or incident delirium cases
if CAM was positive after 24 hours from admission. CICU: coronary intensive care unit; CAM: confusion assessment method.

shift (three times a day) with CAM. The risk factor and etiol-
ogy checklist was updated at the time of delirium diagnosis.
In CAM-positive patients, nursing and medical staff
applied a flowchart for delirium treatment, which was
agreed upon by all investigators at the time of study plan-
ning (Figure 2). First, the presence of treatable medical fac-
tors that can underscore the delirium etiology is evaluated,
then non-pharmacological treatment and, where possible,
environmental strategies of reorientation of the patient are
applied. Only in the case of inefficacy of these methods is a
pharmacological strategy adopted. In the case of doubt, a
neurologist/psychiatrist is consulted, for classification of
mental status and appropriate treatment.
The pharmacological strategy foresees two phases; dis-
continuation of deliriogenic drugs (e.g. benzodiazepines,
antipsychotics and neuroleptics), and, only subsequently,
the active use of drugs based on the RASS scale assess-
ment. According to the current guidelines,17 the treatment
of choice involved the use of intravenous haloperidol bolus
2.55 mg repeated every 1530 minutes until sedation
(with monitoring of the QT interval). Maintenance therapy
involved the use of haloperidol (2.55 mg every 6 hours,
orally) or atypical antipsychotics (e.g. quetiapine,
olanzapine), the use of which took place usually after neu-
rological/psychiatric counselling.
At the 6-month follow-up, patient survival was deter-
mined by telephone contact.
The analysis of clinical data was done using electronic med-
ical records. Data were entered into a custom-made database.
Data management
In-hospital clinical data were prospectively recorded.
Patient demographics, comorbidities, primary cardiac acute
illness, medical treatment, laboratory tests, length of stay
(LOS), clinically significant in-hospital acute adverse
events and mortality data were collected. A risk factor sheet
was specifically designed for this study after review of the
literature relevant to risk factors for delirium in cardiac
patients. Thirty-day and 6-month clinical follow-up data
were obtained by outpatient visit or telephone contact.
Statistical analysis
Continuous data are expressed as mean values SD and
were compared using either the t test or non-parametric
4 European Heart Journal: Acute Cardiovascular Care
Figure 2. Flow chart protocol of delirium management in the CICU. CICU: coronary intensive care unit; CAM: confusion
assessment method; RASS: Richmond agitation sedation scale.
Wilcoxon rank sum test when relevant. Categorical varia-
bles were compared with the use of Fishers exact test;
P<0.05 was considered statistically significant.
KaplanMeier curves (log-rank (MantelCox) test)
were used to assess event-free survivals and cardiac death
at hospital discharge, 30 days and at 6 months. Multivariate
logistic regression was used to assess independent predic-
tors of delirium and the effect of delirium on mortality risk.
All statistical computations were performed using the
SPSS 22.0 statistical package (IBM Corp., Armonk, NY,
USA).
Results
During the study period, 814 patients aged 65 years or older
were admitted to the CICU. Seven hundred and twenty six
of them (89.2%) were screened for delirium and these
patients formed the study sample. The mean age of study
sample was 79.17.8 years (range 65100 years).
A total of 111 individuals (15.3%) were diagnosed with
delirium using CAM; of them, 46 (6.3%) had delirium at
admission (prevalent delirium (PD)). Among the 680
patients who were non-delirious at first assessment, 65
(8.9%) showed positive CAM after 24 hours (incident
delirium (ID)). The prevalent cases accounted for 41.4% of
all delirious patients, whereas incident cases accounted for
58.6%. Patients aged 85 years and older showed a delirium
rate of 52.3%.
The baseline characteristics and known risk factors for
delirium are shown in Table 1. Patients with delirium had a
lower education level, were older, more likely to be non-
smokers, with a history of end-stage renal disease, previous
myocardial infarction or stroke, hearing impairment, cogni-
tive impairment, a history of depression or delirium and
dehydration. A conservative strategy was more commonly
adopted among delirious versus non-delirious patients.
Table 2 shows the admitting diagnosis and predisposing
factors of delirium, according to data present in the
Falsini et al. 5

Table 1. Baseline characteristics of the study population. Pharmacological treatment for delirium was adminis-
tered in 41.3% of cases of ID and in 53.8% of cases of PD.
No delirium Delirium P value

(n=615) (n=111)
Clinical outcome
Age, years 78.3 7.7 83.2 7.5 <0.005
6574 204 (33.17) 15 (13.51) <0.0001 Hospital stay was longer among patients with delirium; in
7584 264 (42.93) 38 (34.23) 0.136 this subgroup, ID individuals showed prolonged CICU and
85 147 (23.90) 58 (52.25) <0.0001 overall hospital stay compared to PD patients.
Instruction A total of 22 in-hospital acute adverse events were
Elementary school 384 (62.44) 79 (71.17) 0.053 recorded: six cases of vascular site access haematoma, one
Middle school 162 (26.34) 20 (18.02) 0.076 pneumothorax, one case of cardiac tamponade, two cases
High school 58 (9.43) 7 (6.31) 0.503 of pressure sores, eight nosocomial infections, one fall, two
University 11 (1.79) 3 (2.70) 0.733 episodes of anaphylaxis and one overdose of benzodiaz-
Male gender 357 (58.05) 56 (50.45) 0.201 epines. The rate of in-hospital acute adverse events was
History of CAD 43 (6.99) 8 (7.21) 0.899 significantly higher in delirious versus non-delirious
Active smoker 67 (10.89) 3 (2.70) 0.008 patients and in ID versus non-delirious patients (Table 4).
Hypertension 400 (65.04) 65 (58.56) 0.433 The duration of delirium was similar between PD and ID
Diabetes 198 (32.20) 35 (31.53) 0.974 patients (respectively, 2.151.69 days and 21.29 days,
Type I diabetes 12 (1.95) 2 (1.80) 0.788 P=0.6).
Hyperlipidaemia 220 (35.77) 34 (30.63) 0.418 Patients with delirium showed higher in-hospital, 30-day
Obesity (BMI >30 kg/m2) 60 (9.76) 4 (3.60) 0.042 and 6-month mortality compared to non-delirious patients,
Renal failure 138 (22.44) 46 (41.44) <0.0001
irrespective of the onset time (overall, ID or PD). Among
Dialysis 12 (1.95) 4 (3.60) 0.241
delirious individuals, 40 died from cardiovascular causes,
Previous MI 97 (15.77) 33 (29.73) <0.0001
four from cerebrovascular causes, three from complications
Previous PCI 93 (15.12) 21 (18.92) 0.222
of neoplasia and four from acute respiratory diseases. No
Previous CABG 37 (6.02) 8 (7.21) 0.550
Previous stroke 76 (12.36) 22 (19.82) 0.005
significant differences were detected between ID and PD
Hypothyroidism 46 (7.48) 7 (6.31) 0.746 patients in terms of mortality (Table 4). In-hospital mortal-
PAD 88 (14.31) 20 (18.02) 0.227 ity was higher in the hypoactive pattern of delirium com-
COPD 76 (12.36) 16 (14.41) 0.194 pared with the hyperactive and mixed pattern (Figure 3).
Cancer 36 (5.85) 9 (8.11) 0.483 Six-month re-hospitalisation was significantly higher in
Visual impairment 48 (7.80) 10 (9.01) 0.615 the overall delirium and PD group compared to non-deliri-
Hearing impairment 20 (3.25) 9 (8.11) 0.032 ous patients, while ID individuals showed a trend towards
Cognitive impairment 23 (3.74) 46 (41.44) <0.0001 significance versus the non-delirious group (Table 4).
Dehydration 7 (1.14) 11 (9.91) <0.0001 KaplanMeier analysis showed a significant reduction
Depression 21 (3.41) 11 (9.91) 0.002 of 6-month survival in patients with delirium compared
Previous delirium 4 (0.65) 12 (10.81) <0.0001 to those without, irrespective of delirium onset time
IQCODE 3.0 0.2 3.1 0.4 0.005 (Figure 4).
At multivariable logistic regression analysis, positive
CAD: coronary artery disease; BMI: body mass index; MI: myocardial
infarction; PCI: percutaneous coronary intervention; CABG: coronary CAM was an independent predictor of in-hospital, 30-day
artery bypass graft; TIA: transient ischaemic attack; PAD: peripheral and 6-month mortality (Table 5).
artery disease; COPD: chronic obstructive pulmonary disease;
IQCODE: Informant Questionnaire on Cognitive Decline in the Elderly.
Age and IQCODE are expressed as meanSD; all others are expressed
as n (%).
Discussion
The present study prospectively evaluated the preva-
literature. The occurrence of delirium was found to be unre- lence, incidence and clinical significance of delirium in
lated to the primary cardiac condition. elderly patients admitted to CICU. The major finding of
At multivariate regression analysis, independent predic- this paper are:
tors of delirium were: age, cognitive impairment, previous
delirium, urinary catheterization, benzodiazepines and Delirium is a frequent condition in the setting of
insulin use, ventricular arrhythmias, fever, hypernatraemia CICU and its occurrence is not directly related to the
and conservative strategy (Table 3). primary cardiac condition
Among the 111 delirious patients, the most common Hospital stay is longer among patients with delirium
subtype of delirium was hyperactive (n=70, 63.1%) versus Patients with delirium show higher mortality com-
the hypoactive form (n=14, 12.6%), whereas 22 patients pared to non-delirious patients, irrespective of the
(19.8%) showed mixed delirium. onset time
6 European Heart Journal: Acute Cardiovascular Care

Table 2. Predisposing factors for delirium in the CICU.

No delirium (n=615) Delirium (n=111) P value

Primary cardiac condition
ST elevation MI 143 (23.25) 22 (19.82) 0.799
NSTE-ACS 172 (27.97) 25 (22.52) 0.356
Tako-Tsubo syndrome 21 (3.41) 1 (0.90) 0.622
CAD 12 (1.95) 1 (0.90) 0.725
Conduction diseases 62 (10.08) 19 (17.12) 0.99
Arrhythmias 43 (6.99) 8 (7.21) 0.946
Acute heart failure 78 (12.68) 15 (13.51) 0.966
Acute valvular disease 17 (2.76) 2 (1.80) 0.821
Pericardial effusion/tamponade 6 (0.98) 3 (2.70) 0.207
Pulmonary embolism 14 (2.28) 1 (0.90) 0.639
Critical limb ischaemia 7 (1.14) 1 (0.90) 0.686
Other 41 (6.67) 13 (11.71) 0.358
Clinical and pharmacological features
Sleep deprivation 27 (4.39) 6 (5.41) 0.595
Urinary catheterisation 83 (13.50) 55 (49.55) <0.0001
Restraints 4 (0.65) 6 (5.41) <0.0001
Inotropes 7 (1.14) 7 (6.31) <0.0001
Antibiotics 133 (21.63) 32 (28.83) 0.94
Corticosteroids 36 (5.85) 9 (8.11) 0.362
Oral anticoagulants 130 (21.14) 23 (20.72) 0.927
Morphine 26 (4.23) 7 (6.31) <0.0001
Benzodiazepines 46 (7.48) 35 (31.53) 0.004
Antipsychotics 9 (1.46) 9 (8.11) <0.0001
Diuretics 281 (45.69) 56 (50.45) 0.595
Oral antidiabetic agents 107 (17.40) 15 (13.51) 0.317
Insulin 38 (6.18) 16 (14.41) 0.005
Digitalis 27 (4.39) 8 (7.21) 0.305
Beta-blockers 401 (65.20) 56 (50.45) 0.005
ACEI/ARBs 265 (43.09) 27 (24.32) <0.0001
Calciumcantagonists 86 (13.98) 13 (11.71) 0.743
Statins 380 (61.79) 47 (42.34) <0.001
Anti-arrhythmic drugs 78 (12.68) 14 (12.61) 0.914
PPIs 418 (67.97) 62 (55.86) 0.019
ASA 431 (70.08) 71 (63.96) 0.754
P2Y12 inhibitors 306 (49.76) 46 (41.44) 0.371
NSAIDs 13 (2.11) 3 (2.70) 0.598
Anti-mineralocorticoids 110 (17.89) 20 (18.02) 0.702
Anti-epileptic drugs 6 (0.98) 1 (0.90) 0.997
Coronary angiography 305 (49.59) 36 (32.43) 0.005
Multivessel CAD 117 (19.02) 18 (16.22) 0.631
Anterior MI 115 (18.70) 22 (19.82) 0.340
CABG indications 11 (1.79) 0 (0.00)
CHF 142 (23.09) 37 (33.33) 0.033
PCI 247 (40.16) 35 (31.53) 0.022
LVEF <30% 110 (17.89) 31 (27.93) 0.021
Shock 27 (4.39) 14 (12.61) 0.001
Atrial fibrillation 117 (19.02) 23 (20.72) 0.607
Ventricular arrhythmias 27 (4.39) 13 (11.71) 0.004
Cardiac arrest 18 (2.93) 9 (8.11) 0.016
Fever 104 (16.91) 30 (27.03) 0.014
Nosocomial infections 11 (1.79) 7 (6.31) 0.012
Hyponatraemia 17 (2.76) 6 (5.41) 0.233
Hypernatraemia 1 (0.16) 4 (3.60) <0.0001
(Continued)
Falsini et al. 7

Table 2.(Continued)

No delirium (n=615) Delirium (n=111) P value

Hypokalaemia 34 (5.53) 19 (17.12) <0.0001
Hyperkalaemia 10 (1.63) 4 (3.60) 0.268
Anaemia 70 (11.38) 18 (16.22) 0.176
Hypoglycaemia 2 (0.33) 2 (1.80) 0.195
Pain 11 (1.79) 19 (17.12) <0.0001
Hypoxia 23 (3.74) 22 (19.82) <0.0001
Stroke 6 (0.98) 9 (8.11) <0.0001
Dehydration (during hospital stay) 3 (0.49) 5 (4.50) <0.0001

CICU: cardiac intensive care unit; NSTE ACS: non-ST segment elevation acute coronary syndrome; CAD: coronary artery disease; CHF: congestive
heart failure; MI: myocardial infarction; ACEI/ARB: angiotensin-converting enzyme inhibitors/angiotensin receptor blockers; PPIs: proton pump
inhibitors; ASA: acetyl-salicylic acid; NSAIDs: non-steroidal anti-inflammatory drugs; CABG: coronary artery by-pass graft; CHF: congestive heart
failure; PCI: percutaneous coronary intervention; LVEF: left ventricular ejection fraction.
Data are expressed as n (%).

Table 3. Predictors of delirium. non-selected population was significantly younger.21 In our

sample, delirious patients were older and, in particular,
Predictors of delirium OR (95% CI) P value
about half of patients aged 85 years and older experienced
Age 1.194 (1.0751.326) 0.001 delirium. No significant association was found between
Cognitive impairment 24.656 (4.437137.032) <0.0001 delirium and the primary cardiac condition; the occurrence
History of delirium 185.604 (5.845589.804) 0.003 of delirium was rather related to the presence of haemody-
Urinary catheterisation 4.634 (1.21817.632) 0.025 namic or electrical instability, infectious complications or
Insulin 44.351 (3.862509.32) 0.002 disorders of fluids and electrolytes. Different from previous
Benzodiazepines 26.849 (5.029143.343) <0.0001 findings,8,13,19 haemodynamic complications seem to be
Conservative strategy 46.82 (1.584138.687) 0.026 strongly related to delirium, although are not an independ-
Ventricular arrhythmias 10.404 (1.55869.467) 0.016 ent predictor of its occurrence. Other series suggest a rela-
Fever 11.638 (2.39356.59) 0.002 tionship between delirium and cardiac status: Uthamalingam
Hypernatraemia 141.47 (56.59422.496) 0.004
etal. reported that left ventricular ejection fraction less than
CI: confidence interval. 40% was a risk factor for delirium,13 while Naksuk etal.
Results from logistic binary forward stepwise regression analysis. found a correlation between delirium and haemodynamic
Variables included in the equation were: age, gender, active smoker, instability (cardiogenic shock and heart failure) in their
obesity, renal failure, previous MI, previous stroke, hearing impairment,
multi-organ failure, cognitive impairment, dehydration, history of large cohort.21
depression, history of delirium, under-nourishment, IQCODE, In our study, LOS was longer among delirious individu-
urinary catheterisation, restraints, inotropes, insulin, morphine, als, and this finding was more evident among ID patients.
benzodiazepines, antipsychotics, anti-Parkinson agents, insulin, beta- McCusker etal. showed that ID but not PD was associated
blockers, converting enzyme/angiotensin antagonist (ACEi/ARBs),
statins, proton pump inhibitors, no coronary angiography (conservative with a significantly longer hospital stay.22 This phenomenon
strategy), multi-vessel coronary artery disease, chronic heart failure, could be explained by several points: first, ID may result
left ventricular ejection fraction <30%, cardiogenic shock, ventricular from intercurrent illnesses or complications that are the
arrhythmias, cardiac arrest, fever, infections, hypernatraemia,
hypokalaemia, in-hospital acute adverse events, pain, hypoxia and stroke.
underlying cause of the longer stay; second, the presence of
several comorbidities (in our sample, delirious patients were
more likely to be older, with a history of coronary artery
Delirium is an independent predictor of short and disease, stroke and renal failure) and deterioration in physi-
long-term mortality cal function that may render discharge premature and inap-
The hypoactive pattern of delirium is associated propriate until the patients improvement; third, a new
with worse in-hospital outcome. diagnosis of delirium may prompt further evaluation and
tests, which require longer stays.
Delirium is known to be a frequent complication of ICU A large body of data demonstrated that delirium is asso-
stay, but evidence in the setting of acute, non-surgical car- ciated with poor prognosis, particularly among the ICU
diac diseases is limited. In the setting of the CICU, we population.5,2325 Such findings have recently been con-
found a prevalence of 15% among elderly, non-intubated firmed in the setting of the CICU. Pauley etal. showed that
patients; similar data were shown in other series, with the presence of delirium was a robust predictor of in-hospi-
prevalence ranging from 16% to 21.5%.14,1820 Naksuk tal morbidity and mortality among critically ill patients
etal. observed a lower rate of delirium; however, their treated in the CICU.14 Sato etal. reported an excess of
8 European Heart Journal: Acute Cardiovascular Care

Table 4. Clinical outcome of CICU patients.

No delirium PD+ID PD ID P value

(n=615) (n=111) (n=46) (n=65)

Hospital stay (days: mSD) 5.6 4.8 7.3 5.8 6.0 4.9 8.2 6.3 <0.0001;a 0.478;b <0.0001;c; 0.007d
CICU stay (days: mSD) 3.0 1.68 2.8 1.7 2.46 + 1.79 3.41 + 1.90 0.250;a 0.037;b 0.065;c 0.009d
In-hospital acute adverse events (n, %) 13 (2.1) 9 (8.1) 3 (6.7) 6 (9.2) 0.002;a 0.143;b 0.013;c 0.959d
In-hospital mortality (n, %) 16 (2.6) 19 (17.1) 11 (23.9) 8 (12.3) <0,0001;a <0.0001;b <0.0001;c 0.187d
30-Day mortality (n, %) 29 (4.7) 33 (29.7) 17 (37.9) 16 (25.6) <0.0001;a <0.0001;b <0.0001;c 0.237d
6-Month mortality (n, %) 55 (8.9) 51 (54.9) 25 (54,3) 26 (40.0) <0.0001;a <0.0001;b <0.0001;c 0.2305d
6-Month re-hospitalisation (n, %) 153 (24.8) 48 (43.2) 20 (43.5) 28 (43.1) <0.0001;a 0.001;b 0.056;c 0.979d

CICU: cardiac intensive care unit; PD: prevalent delirium; ID: incident delirium.
aNo delirium vs. overall delirium.
bNo delirium vs. PD.
cNo delirium vs. ID.
dPD vs ID.

60-day mortality among acute, non-intubated cardiac

Figure 3. In-hospital mortality and clinical patterns of delirium
in CICU. CICU: coronary intensive care unit; D: delirium
patients.
Figure 4. KaplanMeier analysis for survival. At 6 months,
there was a significant difference in cumulative survival between
non-delirium and delirium patients (P<0.001); no significant
differences were detected between PD and ID delirium groups
(P=0.1). PD: prevalent delirium; ID: incident delirium.
patients.19 Naksuk etal. found a strong association between
delirium and both in-hospital and one-year mortality in
their large population of acute cardiac patients.21
In our study, CICU patients with delirium showed worse
short and long-term clinical outcomes. The onset time of
delirium did not affect prognosis, as no significant differ-
ences were observed between ID and PD patients in term of
mortality. Finally, the presence of delirium itself had a
strong clinical impact, being an independent predictor of
in-hospital, 30-day and 6-month mortality.
The adverse outcomes observed among delirious
patients could be due to several factors, including severe
effects on the central nervous system, complications of
prolonged hospitalisation and underlying comorbidities.
Moreover, a long-term cognitive impairment after acute
delirium has been observed;26,27 such a prolonged neuro-
logical disorder may interfere with patients recovery and
rehabilitation.
In agreement with previous data,2831 we found a strong
association between the hypoactive pattern of delirium and
in-hospital death. Hypoactive motor-type delirium is more
commonly missed as a diagnosis than in hyperactive
patients, so a delay in recognition and treatment may be a
cause for this.32 Furthermore, hypoactive patients may pre-
sent with a more severe global pathology, increased com-
plications of inactivity and issues related to immobility
(e.g. dehydration, pressure sores, infections, hypoventila-
tion and venous thrombosis).33
So far, the importance of delirium recognition has not
been well studied. While the occurrence of this syndrome is
clearly related to worse outcomes, the clinical impact of an
early identification is under debate. In particular, the avail-
able literature still does not clarify the nature of this asso-
ciation and, to date, it is unknown whether delirium
represents a potentially modifiable risk factor for adverse
outcomes in critically ill patients in hospital,25 or is a marker
of organ dysfunction or systemic disease and an early
Falsini et al. 9

Table 5. Predictors of mortality. Limitations

OR (95% CI) P value First, our findings may not be generalisable to other hospi-
In-hospital mortality tals with different care models and CICU configurations.
Renal failure 5.30 (1.6517.71) 0.005 Second, we did not assess disease severity scores. Pauley
Cancer 6.29 (1.3928.48) 0.017 etal.,14 however, documented no significant interaction
No beta-blockers 10.19 (2.7637.61) <0.0001 between the APACHE II and CAM scores in this critically
Shock 13.97 (4.0348.42) <0.0001 ill cardiac population, despite the fact that both were pre-
Cardiac arrest 39.10 (8.28184.50) <0.0001 dictive of survival. This suggests that both of them may be
In-hospital stroke 9.41 (1.4461.46) 0.019 useful for risk stratification.
Positive CAM 4.97 (1.7014.52) 0.003 Third, in our study there was not a confirmation of delir-
30-Day mortality ium by a routine psychiatric consultation, which was
Age 1.08 (1.021.14) 0.009 reserved for harder cases.
Renal failure 2.63 (1.195.83) 0.017
Cancer 7.83 (2.5124.42) <0.0001
No beta-blockers 4.01 (1.739.29) 0.001 Conclusions
No statins 4,78 (1.8712.23) 0.001 Delirium is a common complication in patients aged 65
STEMI 5.14 (1.8214.48) 0.002 years and older admitted to the CICU, with a high preva-
LVEF <30% 3,82 (1.539.52) 0.004 lence and incidence. This condition is not directly related
Shock 8.68 (2.8626.33) <0.0001 to the primary cardiac condition, while it seems to be
Cardiac arrest 7.11 (1.5732.20) 0.011
associated with haemodynamic complications at the
Nosocomial infections 10.23 (2.5241.57) 0.001
admission and during hospitalisation of the patient.
Positive CAM 5.57 (2.5112.40) <0.0001
Delirious individuals show worse outcomes, with a more
6-Month mortality
complicated hospital stay and increased short and long-
Age 1.10 (1.051.16) <0.0001
PAD 2.47 (1.185.17) 0.016
term mortality.
COPD 2,38 (1.144.96) 0.021 Our findings suggest the usefulness of a protocol for the
Cancer 4.72 (1.7812.53) 0.002 identification of delirium (both prevalent and incident) in
MOF 3.96 (1.5210.29) 0.005 the CICU, although the clinical impact of an early diagno-
Under-nourishment 7.03 (1.1443.46) 0.036 sis is unknown. Large, prospective, multicentre studies are
Urinary 2.20 (1.174.12) 0.014 needed to investigate the clinical impact of strategies for
catheterisation delirium prevention and treatment.
No statins 3.95 (2.127.37) <0.0001
LVEF <30% 3.08 (1.586.01) 0.001 Acknowledgements
Shock 7.68 (2.7221.59) <0.0001 Thanks to the tireless work of the nursing staff of the Arezzo and
Pain 6.01 (1.9218.83) 0.002 Montevarchi CICU that made possible the realisation of this study.
Positive CAM 2.39 (1.164.90) 0.017

CI: confidence interval. Conflict of interest

Results from logistic binary forward stepwise (Wald) regression analysis.
Variables included in the equation were: age, gender, renal failure,
previous myocardial infarction, peripheral artery disease (PAD), chronic
obstructive pulmonary disease (COPD), cancer, visual impairment,
hearing impairment, multi-organ failure (MOF), cognitive impairment,
dehydration, under-nourishment, urinary catheterisation, restraints,
inotropes, morphine, oral anticoagulation, beta-blockers, converting
enzyme/angiotensin antagonist (ACEi/ARBS), statins, ST-elevation
myocardial infarction (STEMI), congestive heart failure (CHF), left
ventricular ejection fraction (LVEF) <30%, cardiogenic shock, ventricular
arrhythmias, cardiac arrest, nosocomial infections, in-hospital acute
adverse events, pain, hypoxia, stroke and confusion assessment method
(CAM) test positive.
indicator for complications in the setting of intensive care.34
In a very recent prospective, observational cohort study,
higher delirium monitoring adherence was independently
associated with a reduction of in-hospital mortality in
mechanically ventilated patients,34 and one cannot exclude
that similar outcomes could be found in an acute non-ven-
tilated population.
The authors declare that there is no conflict of interest.
Funding
This research received no specific grant from any funding agency
in the public, commercial, or not-for-profit sectors.
References
1. Inouye SK. Delirium in hospitalized older patients: recogni-
tion and risk factors. J Geriatr Psychiatry Neurol 1998; 11:
118125.
2. Inouye SK. Predisposing and precipitating factors for delir-
ium in hospitalized older patients. Dementia Geriatr Cognit
Disord 1999; 10: 393400.
3. Levkoff SE. Delirium. The occurrence and persistence of
symptoms among elderly hospitalized patients. Arch Intern
Med 1992; 152: 334340.
4. Lawlor PG, Gagnon B, Mancini IL, etal. Occurrence, causes,
and outcome of delirium in patients with advanced cancer.
Arch Intern Med 2000; 160: 786.
10 European Heart Journal: Acute Cardiovascular Care
5. Ely EW, Inouye SK, Bernard GR, etal. Delirium in mechani-
cally ventilated patients. JAMA. 2001; 286: 2703.
6. Hsieh SJ, Ely EW and Gong MN. Can intensive care unit
delirium be prevented and reduced? Lessons learned and
future directions. Ann Am Thorac Soc 2013; 10: 648656.
7. Ely EW, Truman B, Shintani A, etal. Monitoring sedation
status over time in ICU patients. JAMA 2003; 289: 2983.
8. McPherson JA, Wagner CE, Boehm LM, etal. Delirium in
the cardiovascular ICU. Crit Care Med 2013; 41: 405413.
9. ackson JC, Gordon SM, Hart RP, etal. The association
between delirium and cognitive decline: a review of the
empirical literature. Neuropsychol Rev 2004; 14: 8798.
10. Pompei P, Foreman M, Rudberg MA, etal. Delirium in
hospitalized older persons: outcomes and predictors. J Am
Geriatrics Soc 1994; 42: 809815.
11. Katz JN, Shah BR, Volz EM, etal. Evolution of the coronary
care unit: clinical characteristics and temporal trends in health-
care delivery and outcomes. Crit Care Med 2010; 38: 375381.
12. Uguz F, Kayrak M, ek E, etal. Delirium following acute
myocardial infarction: incidence, clinical profiles, and pre-
dictors. Perspect Psychiatr Care 2010; 46: 135142.
13. Uthamalingam S, Gurm GS, Daley M, etal. Usefulness of
acute delirium as a predictor of adverse outcomes in patients
>65 years of age with acute decompensated heart failure. Am
J Cardiol 2011; 108: 402408.
14. Pauley E, Lishmanov A, Schumann S, etal. Delirium is a
robust predictor of morbidity and mortality among critically
ill patients treated in the cardiac intensive care unit. Am
Heart J 2015; 170: 7986; e71.
15. Ely EW, Margolin R, Francis J, etal. 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: 13701379.
16. Jorm AF. The Informant Questionnaire on Cognitive Decline
in the Elderly (IQCODE): a review. Int Psychogeriatr 1999;
16: 275293.
17. Young J, Murthy L, Westby M, etal. Diagnosis, prevention,
and management of delirium: summary of NICE guidance.
BMJ 2010; 341: c3704.
18. Bamalwa M, Mahmood SN and Praharaj SK. Delirium in
cardiac ICU patients. Ann Clin Psychiatry 2016; 28: 5155.
19. Sato K, Kubota K, Oda H, etal. The impact of delirium on
outcomes in acute, non-intubated cardiac patients. Eur Heart
J Acute Cardiovasc Care. Epub ahead of print 21 December
2015. DOI: 10.1177/2048872615624239.
20. Noriega FJ, Vidan MT, Sanchez E, etal. Incidence and
impact of delirium on clinical and functional outcomes in
older patients hospitalized for acute cardiac diseases. Am
Heart J 2015; 170: 938944.
21. Naksuk N, Thongprayoon C, Park JY, etal. Clinical impact
of delirium and antipsychotic therapy: 10-year experi-
ence from a referral coronary care unit. Eur Heart J Acute
Cardiovasc Care. Epub ahead of print 29 June 2015. DOI:
10.1177/2048872615592232.
22. McCusker J, Cole MG, Dendukuri N, etal. Does delirium
increase hospital stay? J Am Geriatr Soc 2003; 51: 15391546.
23. Leslie DL, Zhang Y, Holford TR, etal. Premature death asso-
ciated with delirium at 1-year follow-up. Arch Intern Med
2005; 165: 16571662.
24. McCusker J, Cole M, Abrahamowicz M, etal. Delirium predicts
12-month mortality. Arch Intern Med 2002; 162: 457463.
25. Salluh JI, Wang H, Schneider EB, etal. Outcome of delirium
in critically ill patients: systematic review and meta-analysis.
BMJ 2015; 350: h2538.
26. Pandharipande PP, Girard TD, Jackson JC, etal. Long-term
cognitive impairment after critical illness. N Engl J Med
2013; 369: 13061316.
27. van den Boogaard M, Schoonhoven L, Evers AW, etal.
Delirium in critically ill patients: impact on long-term
health-related quality of life and cognitive functioning. Crit
Care Med 2012; 40: 112118.
28. Jackson TA, Wilson D, Richardson S, etal. Predicting out-
come in older hospital patients with delirium: a system-
atic literature review. Int J Geriatr Psychiatry 2016; 31:
392399.
29. Meagher DJ, Leonard M, Donnelly S, etal. A longitudinal
study of motor subtypes in delirium: relationship with other
phenomenology, etiology, medication exposure and progno-
sis. J Psychosom Res 2011; 71: 395403.
30. Yang FM, Marcantonio ER, Inouye SK, etal.
Phenomenological subtypes of delirium in older persons:
patterns, prevalence, and prognosis. Psychosomatics 2009;
50: 248254.
31. Kiely DK, Jones RN, Bergmann MA, etal. Association
between psychomotor activity delirium subtypes and mor-
tality among newly admitted post-acute facility patients.
J Gerontol A Biol Sci Med Sci 2007; 62: 174179.
32. Inouye SK, Foreman MD, Mion LC, etal. Nurses rec-
ognition of delirium and its symptoms: comparison of
nurse and researcher ratings. Arch Intern Med 2001; 161:
24672473.
33. ORegan NA, Fitzgerald J, Timmons S, etal. Delirium: a
key challenge for perioperative care. Int J Surg 2013; 11:
136144.
34. Luetz A, Weiss B, Boettcher S, etal. Routine delirium moni-
toring is independently associated with a reduction of hospi-
tal mortality in critically ill surgical patients: a prospective,
observational cohort study. J Crit Care 2016; 35: 168173.