Professional Documents
Culture Documents
PII: S0883-9441(17)31007-9
DOI: doi:10.1016/j.jcrc.2017.10.013
Reference: YJCRC 52725
To appear in:
Please cite this article as: Angkita Barman, Debasis Pradhan, Prithwis Bhattacharyya,
Samarjit Dey, Anirban Bhattacharjee, Sonali Shinde Tesia, Jayanta Kumar Mitra ,
Diagnostic accuracy of delirium assessment methods in critical care patients. The address
for the corresponding author was captured as affiliation for all authors. Please check if
appropriate. Yjcrc(2017), doi:10.1016/j.jcrc.2017.10.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
ACCEPTED MANUSCRIPT
PT
RI
TITLE: Diagnostic accuracy of delirium assessment methods in
SC
critical care patients
NU
Name of authors with affiliations:
MA
India, dibemy@gmail.com
CE
India, dmkcg85@gmail.com
India, drsamar0002@gmail.com
PT
India, onirbhatt@gmail.com
RI
6. Sixth author: Sonali Shinde Tesia, North Eastern Indira Gandhi
SC
Institute of Health and Medical Sciences, Shillong, Meghalaya,
India
NU
7. Seventh author: Jayanta Kumar Mitra, All India Institute of
MA
PT
shift attention, change in cognition and or perception. The main objective
RI
for the ICU (CAM-ICU) and Intensive Care Delirium Screening
SC
Checklist (ICDSC) among the nursing and medical staff in a
NU
multidisciplinary ICU.
MA
predictive value (94.5%, 92%) equal to that of CAM-ICU. For both the
sensitive than ICDSC. Though sensitivity and DOR were higher for
medical staff, other diagnostic parameters were similar for both medical
PT
specificity
RI
Introduction
SC
Delirium is a common manifestation of acute brain dysfunction in
delirium screening [3]. It is highly disturbing for the nurses and family
widely varying incidence rates ranging from 11% to 87% [4, 5].
PT
cognitive function, behaviour, and mood [6, 7]. Currently, central
RI
cholinergic deficiency is the leading hypothesized mechanism for
SC
delirium supported by several lines of evidence [8]. Delirium may occur
delirium [10].
ED
care delirium without the need for consulting a psychiatrist. There are six
CE
PT
Glasgow coma scale < 9, Richmond Agitation-Sedation Scale (RASS)
RI
score of ≤ -3 and patients with confounding co-morbid conditions (severe
SC
dementia, psychosis or progressive neurologic disease) were excluded.
Data collection
the ICU (CAM-ICU). Each patient was allowed to participate once and
the commencement of the study, all the ICU nurses were educated with
psychiatrist served as a reference rater and evaluated all the patients and
ACCEPTED MANUSCRIPT
of four hours. The psychiatrist, nurse and intensivist did not have access
to patients’ records and other rater’s finding. All assessments were done
PT
four hours for assessment was to balance the intent of having the
RI
evaluations as close together as possible to enhance the reliability and to
SC
avoid the difficulties introduced by patient care or procedures as well as
clinicians’ schedule.
NU
None of the raters had access to the other’s ratings of the assessment and
MA
the patient records. All patients were assessed by RASS during the first
defined when the RASS was persistently positive (+1 to +4 during all
CE
Statistical analysis
computed using the Mann Whitney U test. The diagnostic accuracy of the
PT
odds ratio (DOR) using simple 2 × 2 tables. The statistical analysis was
RI
done using MedCalc for Windows, version 12.5 (MedCalc Software,
SC
Ostend, Belgium) and Microsoft Excel 2007 for Windows (Redmond,
Washington, USA).
NU
Results
MA
months for possible inclusion in this study. From these, we excluded 152
ED
Therefore, 338 patients were assessed for the purpose of the study. In 28
AC
patients, all the three assessors could not complete the evaluations. After
all exclusion, the final study population constituted 310 patients whose
observations were available for final analysis (Figure 1). The study
(140 out of 310). Delirious patients were significantly more ill with
female patients were less prone to be delirious, yet it was not statistically
PT
postoperative patients compared to other groups. Out of 310 patients 162
RI
(52.2 %) were nonventilated and verbally communicating.
SC
Based on RASS assessment, 22 (15.7%), 91 (65%) and 27 (19.3%)
and intensivist was 0.86 (95% CI 0.8-0.9) and 0.89 (95% CI 0.83-0.96)
(95% CI: 84.9%- 98.9%) for CAM-ICU, 77.8% (95% CI: 62.9%-88.8%)
and 94.6 % (95% CI: 84.9%-98.9%) for ICDSC. Overall DORs was
CE
(67.0, 27.0) for CAM-ICU and ICDSC respectively (Table 3). Subgroup
good sensitivity (84.5 %, 95% CI: 75%-91%) when used by nurse and
use, need for mechanical ventilation, and APACHE II >16 were found to
PT
have significantly high relative risk (RR) (Table 5) of being delirious.
RI
Postoperative patients were found to have low relative risk of delirium
SC
(RR - 0.57, P - 0.03).
Discussion
NU
Compared to general hospital population, delirium in critically ill
MA
patients is more severe and of longer duration with long standing bad
impact on the post discharge quality of life. This is because of the lower
ED
sedatives and analgesics in the critically ill adults and had recommended
ICUs (Van Eijk et al., 2008) [11] and 9% of ICUs in Australia and New
PT
Zealand (Shehabi et al., 2008) [12] used a validated screening tool, in
RI
England 25% of intensivists routinely screen for delirium and 14% used a
SC
validated tool in mechanically ventilated patients (Mac Sweeney et al.,
2010) [13]. Two surveys from North America had reported routine
NU
screening of delirium by 40–59% of the intensivists and only 16-33% of
MA
them used specific tools for delirium assessment (Ely et al., 2004 [14],
Patel et al., 2009 [15]). Increased recognition of delirium is the first step
ED
in the ICU.
The following four key delirium criteria and an algorithm are used to
determine the presence of delirium: (a) acute mental status change; (b)
(b) along with either criteria (c) or (d) are present. In 2001, Bergeron et
al. [17] created the ICDSC, which includes eight items based on the DSM
PT
symptom fluctuation), which has a total score ranging from 0 to 8 points.
RI
A patient with more than 4 points is diagnosed as delirium-positive.
SC
In the face of low rates of utilization of delirium assessment tools in
routine ICU care, a variable rate of delirium prevalence has been found
NU
among intensive care patients. In the present study, prevalence of
MA
into regular ICU practice, the day and time of assessment, presence or
CE
papers had identified twenty five risk factors for delirium development
opioid usage and metabolic acidosis [18]. In the present study, age did
PT
risk factor for development of delirium [18, 21].
RI
Increased severity of illness as indicated by higher APACHE II score has
SC
a positive assosiation with delirium. In a study involving hospitalized
represent “severe illness”. Patients with severe illness had a relative risk
delirium [20]. In the present study, 136 (44%) patients with delirium and
PT
and the relative risk was 2.3 (95% CI: 1.4-3.6). Similar to earlier studies,
CE
agreement) when the two methods were used by nurses and intensivists
intensivist and the nurse and 0.95 (95% CI 0.84 - 1.00) between the two
PT
In the present study CAM-ICU had a higher overall sensitivity while
RI
specificity, predictive value and likelihood ratio were similar for both the
SC
scales (Table 3). Similar higher sensitivity of CAM-ICU in diagnosing
and 96% (95% CI: 95%-97%) for CAM-ICU when compared to ICDSC
ED
with pooled sensitivity 74% (95% CI: 65%-81%) and specificity 82%
PT
(95% CI: 76%-86%) [27]. In the present study, we found overall equal
specificities between the two methods and ICDSC had higher specificity
CE
situation may explain this variable finding. When used by nurses, CAM-
[28].
PT
communicating patients. Very few studies have compared the diagnostic
RI
accuracy of sedation scales between different users. Similar to the
SC
findings of Chuang et al [29], higher sensitivity of 82.1 % (95% CI: 79%-
the highest DOR of 120.5 (95% CI- 53.5-271.5) for CAM-ICU when it
ED
was used by intensivist. Overall, the DORs were higher for CAM-ICU
PT
and for the physician observers similar to the findings of earlier meta-
analysis which reported DORs for CAM-ICU and ICDSC as 103.2 (95%
CE
CI: 39.6 to 268.8) and 21.5 (95% CI: 8.51 to 54.4) respectively [27].
AC
a fast application (2-5 minutes) and does not depend exclusively on the
hours, with easy administration and high sensitivity [19]. The patients are
ACCEPTED MANUSCRIPT
ICDSC checklist. An ideal delirium screening tool for clinical use must
PT
is important to ensure that nursing staff are routinely evaluating all
RI
patients with a sedation scale (e.g., RASS) before a delirium assessment
SC
is completed.
However, this study has some limitations. First, delirium being an acute
NU
fluctuating condition the time of entry into the study affects the outcome.
MA
Second, we did not assess the patient during the rest of ICU stay for the
this was not always possible due to logistic difficulties. But this has been
CE
intorduce patient memory based bias and on the other hand when we try
to keep them apart, we may pick up a new onset delirium by one of the
education session. So, the interobserver bias could not be ruled out
Sixth, this result should not be generalised to all types of health care
settings and ICUs. Lastly, We could not determine the impact of our
ICU.
Conclusions
PT
CAM-ICU was found to be a better delirium assessment tool than
RI
the ICDSC. Frequency of delirium was found to be higher in patients
SC
receiving sedatives, mechanical ventilation, and those who were more
or the nurses were comparable and similar to those of the reference rater.
The hospital settings and necessary training of nursing and medical staff
ED
practice in the intensive care unit using a validated tool may improve
CE
patient care.
AC
ACCEPTED MANUSCRIPT
Conflicts of interest
None.
Funding
This research did not receive any specific grant from funding
PT
Acknowledgements
RI
This work would not have been possible without the collaboration
SC
of the nurses of the Intensive care unit. We deeply thank Dr Elantamilan
References
2. Ely EW, Pun BT. The Confusion Assessment Method for the ICU
PT
Impairment Study Group. Vanderbilt University Medical Center,
RI
Nashville. 2002.
SC
3. Jacobi J, Fraser GL, Coursin DB, Riker RR, Fontaine D, Wittbrod
Tet ,et al, Clinical practice guidelines for the sustained use of
NU
sedatives and analgesics in the critically ill adult. Crit Care Med
MA
2002; 30:119-141.
4. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP,
ED
1990;113(12):941-8.
CE
2000;20:28-37.
ACCEPTED MANUSCRIPT
PT
9. Han L, McCusker J, Cole M, Abrahamowicz M, Primeau F, Elie
RI
M. Use of medications with anti cholinergic effect predicts clinical
SC
severity of delirium symptoms in older medical inpatients. Arch
11. Van Eijk MMJ, Kesecioglu J, Slooter AJ. Intensive care delirium
PT
12. Shehabi Y, Botha JA, Boyle MS, Ernest D, Freebairn RC, Jenkins
AC
IR, Roberts BL, Seppelt IM. Sedation and delirium in the intensive
of Medicine 2010;103(4):243–51.
PT
Care Med 2004;32:106–12.
RI
15.Patel RP, Gambrell M, Speroff T, Scott TA, Pun BT, Okahashi J,
SC
Strength C, Pandharipande P, Girard TD, Burgess H, Dittus RS,
Bernard GR, Ely EW. Delirium and sedation in the intensive care
NU
unit: survey of behaviors and attitudes of 1384 healthcare
MA
2007;33(1):66–73.
20. Inouye SK, Viscoli CM, Horwitz RI, Hurst LD, Tinetti ME. A
PT
patients based on admission characteristics. Ann Int Med
RI
1993;119(6):474-81.
SC
21.Lee HB, Mears SC, Rosenberg PB, Leoutsakos JM, Gottschalk A,
Anesthesiology 2006;104(1):21–6.
CE
23. Lin SM, Huang CD, Liu CY, Lin HC, Wang CH, Huang PY, Fang
AC
YF, Shieh MH, Kuo HP. Risk factors for the development of early-
ZV, Moore EE. Risk factors for delirium after major trauma. Am J
Surg 2008;196(6):864-70.
ACCEPTED MANUSCRIPT
method for the intensive care unit (CAM-ICU) with the Intensive
care patients gives high agreement rate (s). Intensive Care Med
PT
2008;34(3):431-6.
RI
26.van Eijk MM, van den Boogaard M, van Marum RJ, Benner P,
SC
Eikelenboom P, Honing ML, van der Hoven B, Horn J, Izaks GJ,
28.Toro AC, Escobar LM, Franco JG, Diaz-Gomez JL, Munoz JF,
29.Chuang WL, Lin CH, Hsu WC, Ting YJ, Lin KC, Ma SC.
Za Zhi 2007;54:45-52.
PT
RI
Cognitive Test For Delirium7
Cognitive Test For Delirium – abbreviated version8
Confusion Assessment for Intensive Care Unit(CAM-ICU)9,10
SC
Intensive Care Delirium Screening Checklist(ICDSC)11
The Neelon And Champagne Confusion Scale(NEECHAM) 12,13
Delirium Detection Score14
NU
Table 1: Delirium assessment methods commonly used in ICU
MA
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
PT
Others 44 35 (24.4 %) 9 (5.45 %) 0.0325
RI
Table 2: Demographic parameters and diagnosis at admission
SC
NU
MA
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
CAM-ICU ICDSC
Sensitivity 84.4% 70.5%-93.5% 77.8% 62.9%-88.8%
Specificity 94.6% 84.9%- 98.9% 94.6% 84.9%-98.9%
Positive Likelihood Ratio 15.48 5.12 - 46.86 14.26 4.7-43.3
Negative Likelihood Ratio 0.16 0.08 - 0.33 0.24 0.14-0.41
Disease Prevalence 45.0% 35.0%- 55.3% 45.0% 35.0%-55.3%
Positive Predictive Value 92.7% 80.1%- 98.5% 92.1% 78.6%-98.3%
Negative Predictive Value 88.1% 77.1%- 95.1% 83.9% 72.3%- 92.0%
Diagnostic odds ratio
Overall 86.1 49.7-149.2 36.9 22.9-59.3
PT
Intensivist 120.5 53.5-271.5 53.0 25.2-110.2
Nurse 67.0 31.2-143.8 27.0 14.4-50.7
RI
Table 3: Diagnostic value of CAM-ICU and ICDS with 95% CI
SC
NU
MA
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
PT
Nurse 84.52(75-91) 90.6(81-96) 92.2(83-97) 81.7(71-90) 9(4.2-19.4) 0.17(0.1-0.28)
Intensivist 94.1(87-98) 90.6(81-96) 92.9(85-97) 92.1(82-97) 10.03(4.7- 0.07(0.03-
21.5) 0.15)
RI
ICDSC (verbally non- communicating)
Nurse 76.19(65.7- 90.6(81-96) 91.43(82- 74.36(63- 8.13(3.8-17.6) 0.26(0.2-0.4)
84.8) 97) 83)
SC
Intensivist 79.8(70-88) 95.3(87-99) 95.7(88-99) 78.2(67-87) 17(5.6-51.6) 0.21(0.1-0.3)
Relative 95 % CI Significance
Risk Level
Sepsis 1.8333 1.2142 to 2.7681 P = 0.0039
Benzodiazepine use 1.7143 1.1177 to 2.6293 P = 0.0135
Mechanical ventilation 1.6064 1.0326 to 2.4991 P = 0.0355
Dyselectrolytemia 1.7051 1.1331 to 2.5659 P = 0.0105
BUN/ S.creatinine > 18 1.6731 1.0772 to 2.5987 P = 0.0220
APACHE II >16 2.2745 1.4455 to 3.5789 P = 0.0004
Postoperative cases 0.5688 0.3366 to 0.961 P = 0.0350
PT
Table 5: Relative risk of various factors along with its significance
RI
SC
NU
MA
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
PT
RI
SC
NU
MA
ED
PT
CE
HIGHLIGHTS
the ICDSC.
PT
intensivists or the nurses were comparable.
RI
3. Sepsis, benzodiazepine use, mechanical ventilation, and APACHE
SC
II >16 were found to have significantly high relative risk of being
delirious.
NU
MA
ED
PT
CE
AC