BRIEF REPORTS
A Simple Tool to Predict Development of Delirium After
Elective Surgery
Andy Dworkin, MD,* David S.H. Lee, PharmD PhD,† Amber R. An, DO,‡ and
Sarah J. Goodlin, MD‡
OBJECTIVES: To identify a quick clinical tool to assess
the risk of delirium after elective surgery.
DESIGN: Prospective observational study.
SETTING: Preoperative assessment clinic at the Veterans
Affairs Portland Health Care System.
PARTICIPANTS: Community-living veterans aged 65 and
older scheduled for elective surgery requiring general or
major anesthesia.
MEASUREMENTS: Confusion Assessment Method
(CAM) or Family Confusion Assessment Method (FAM-
CAM). Data on education, medications, substance use,
Patient Health Questionnaire (PHQ-9), Study of Osteo-
porotic Fractures Frailty, Mini-Cog, and Charlson-Deyo
score were collected preoperatively.
RESULTS: Of 114 veterans who agreed to participate, 76
completed the final delirium assessment. Ten of the 76
(13%) developed delirium in the 72 hours after surgery as
assessed using the CAM or FAM-CAM. In bivariate analy-
sis, factors that increased the odds of delirium at least
three times were low education; poor PHQ-9, clock draw,
word recall, Mini-Cog, and poor preoperative orientation
scores; alcohol use; and higher comorbidities as measured
using Charlson-Deyo index. Scoring the Mini-Cog from 0
to 5 had a higher predictive power (area under the receiv-
ing operating characteristic curve = 0.77) than other
approaches to scoring the Mini-Cog. Other models did not
significantly improve prediction of postoperative delirium
risk and would be complicated to use in a clinical setting.
CONCLUSION: In this sample of veterans who had elec-
tive surgery with major anesthesia, Mini-Cog score pre-
dicted likelihood of postoperative delirium. J Am Geriatr
Soc 2016.
From the *Department of Medicine, Legacy Health System; †College of
Pharmacy, Oregon State University & Oregon Health and Science
University; and ‡Geriatrics Section, Veterans Affairs Portland Health
Services Center and Oregon Health & Science University, Portland,
Oregon.
Address correspondence to Sarah J. Goodlin, P3-Med VA Portland HCS,
PO Box 1034, Portland, OR 97207. E-mail: sarah.goodlin@va.gov
DOI: 10.1111/jgs.14428
JAGS 2016
© 2016, Copyright the Authors
Journal compilation © 2016, The American Geriatrics Society
Key words: elderly; postoperative; delirium; prediction
D elirium, an acute encephalopathy marked by fluctuat-
ing cognition and attention, is a common, costly,
life-threatening complication of surgery in older adults.
Estimates of the incidence of postoperative delirium range
from 7% to 10% of older adults after simple elective sur-
gery to at least half of older adults undergoing emergency,
cardiac, or orthopedic surgery.1–5 Individuals who develop
delirium have longer hospital stays and are more likely to
be debilitated, require skilled nursing care, and die in the
year after surgery.6–8 The annual cost of delirium associ-
ated with surgical and medical U.S. hospital admissions is
as much as $150 billion.9
Given this heavy burden, many investigators have
tried to identify risk factors for delirium and tools to
predict its development.10 Practical and reliable screening
would help clinicians and patients assess the risks and
benefits of surgery and could affect decisions about oper-
ative procedures, medications, and targeted postoperative
interventions that have been shown to reduce the devel-
opment of delirium significantly.11,12 Individual studies
have identified dozens of risk factors for developing
delirium.13,14 Factors that have been significantly associ-
ated with delirium risk in multiple studies include preex-
isting dementia, depression, comorbidity, poor baseline
physical function, polypharmacy, sensory impairment,
and the use of alcohol or psychoactive drugs. Several
proposed screening tools have combined multiple factors
to predict delirium risk, but none of these tools has
been widely adopted, and many require significant clini-
cal testing, time, or training to administer, limiting their
clinical utility.
The aim of this prospective cohort study was to
develop a quick and practical clinical tool to stratify older
adults undergoing surgery according to their risk for post-
operative delirium. Baseline factors linked to delirium in
multiple prior studies were evaluated for incorporation
into a final model. So that the tool would be clinically
0002-8614/16/$15.00
2 DWORKIN ET AL.
practical, only factors that can be assessed quickly at low
cost and with minimal training were considered.
METHODS
Community-living adults aged 65 and older presenting for
preoperative evaluation at an urban Department of Veter-
ans Affairs (VA) hospital were screened for study eligibil-
ity. Eligible participants had to be undergoing elective
surgery with major anesthesia (general or epidural anesthe-
sia or spinal block with sedation). Individuals undergoing
intracranial surgery or cardiopulmonary bypass, presenting
directly from a hospital or other care facility, or screening
positive for delirium before surgery as assessed using the
Confusion Assessment Method (CAM),15 were excluded.
To assess delirium in individuals discharged within
48 hours of surgery, participants had to identify a study
partner who would see them after surgery and be available
to answer questions about their condition, cognition,
attention, and orientation. Trained researchers described
study methods, risks, and benefits to eligible candidates;
assessed decision-making capacity; and had interested par-
ticipants sign informed consent forms. The institutional
review board of the VA Portland Health Care System
approved the research protocol.
At a regular pre-op clinic visit, research staff asked
study subjects 11 orientation questions (date, day of week,
month, season, year, city, county, state, name of hospital,
type of building and which floor the clinic was on) and
administered several tests: The PHQ-9 was used to assess
depression.16 The Study of Osteoporotic Fractures Frailty
index was calculated to gauge frailty.17 The Mini-Cog, a
sensitive and specific screening tool for dementia in com-
munity-living adults, was administered.18 Participants self-
reported education level, drug and alcohol history, and
vision or hearing impairments not relieved by assistive
devices. Researchers later reviewed the electronic medical
record to confirm age and sex and to record current medi-
cations and active diagnoses, which were used to calculate
the Charlson-Deyo score, a measure of comorbidity.19
Between 48 and 72 hours after surgery, research staff
conducted a second interview in person with participants
still in the hospital or over the telephone with the partici-
pant and study partner for those who had been discharged.
Delirium was assessed using a postoperative CAM that
research staff performed for those still in the hospital and
the FAM-CAM20 (a validated method of identifying delir-
ium in the outpatient setting) that research staff performed
with study partners for participants who had been dis-
charged. The principal investigator ensured interrater relia-
bility for the five research staff participating in this study
by direct observation of CAM and FAM-CAM perfor-
mance.
Data were analyzed using SAS version 9.3 (SAS Insti-
tute, Inc., Cary, NC). Twenty preoperative variables were
included in bivariate analysis of risk for incident postoper-
ative delirium: age; education; orientation; Charlson-Deyo
score; clock draw, word recall, and total Mini-Cog; PHQ-
9; Study of Osteoporotic Fractures; use of Beers list medi-
cations21 or of any antidepressant; uncorrected vision or
hearing impairment; alcohol use; and history of marijuana,
hallucinogens, cocaine, heroin, amphetamines, or other
2016 JAGS
drugs. Type of procedural anesthesia received was also
included.
Stepwise logistic regression was used to test multivari-
ate models for predictive performance using the area under
the receiver operating characteristic curve (AUC); an AUC
of 0.5 indicates random chance, and an AUC of 1.0 indi-
cates that the model offers perfect prediction. The initial
risk factor included in the model was chosen based on
highest AUC in bivariate analysis. Additional variables
were added based on improvement in the AUC, and no
additional variables were included when the Akaike infor-
mation criterion (AIC) decreased. Using this method, if
several candidate models emerged, the most clinically sim-
ple model would be chosen.
RESULTS
One hundred fourteen people agreed to participate and com-
pleted baseline study questions. Postoperative delirium
assessment was completed on 76 participants (66%). The
average age of participants was 71  6, and all but one
were men. Other characteristics of the subjects are listed in
Table 1. Reasons for dropout include canceled surgeries
(n = 15, 13%), sedation other than major anesthesia (n = 5,
4%), or inability to contact participant or study partner
within the 72-hour postoperative window (n = 18, 16%).
There were no significant differences in baseline characteris-
tics between individuals who dropped out and those who
completed the final assessment (Table 1). No participant
was delirious before surgery as assessed using the CAM. No
participant had a dementia diagnosis. Two participants had
diagnoses of organic brain syndrome not otherwise specified
(NOS) and two of cognitive disorder NOS. None of these
four developed delirium.
Ten of 76 participants (13%) developed delirium in the
72 hours after surgery as assessed using the CAM or FAM-
CAM. In bivariate analysis, factors that increased the odds
of delirium at least three times were low education, poor
PHQ-9 or Mini-Cog scores (as well as poor scores on the
individual components of clock draw and word recall), poor
preoperative orientation, alcohol use, and higher comor-
bidities as measured using Charlson-Deyo index
(Appendix S1). Age, frailty, and sex did not significantly
increase delirium risk in this sample. Type of anesthesia was
not associated with development of delirium (P = .63),
although 89% of subjects (n = 68) had general anesthesia.
The remaining participants had endotracheal intubation
and sedation alone (n = 3, 4%) or with intra-articular
(n = 1, 1%), epidural (n = 2, 3%), or spinal block (n = 2,
3%).The 76 participants in the final data set had elective
orthopedic, urological, or bowel surgeries. No participant
studied had prolonged anesthesia or any intraoperative
complications.
In bivariate analysis, Mini-Cog had the highest predic-
tive power. Multiple scoring systems were explored,
including a binary model, in line with the test’s original
description (0–2 = high risk, 3–5 = low risk). In the end,
scoring the Mini-Cog from 0 to 5 (with 1 point for each
word recalled and 2 points each for correct clock) yielded
the highest predictive power (AUC = 0.77, 95% confi-
dence interval (CI) = 0.61–0.93). Participants with a Mini-
Cog score of 0 to 1 had a 50% or greater probability of
JAGS 2016 PREDICTING POSTOPERATIVE DELIRIUM 3

Table 1. Participant Characteristics According to Table 2. Predicted Risk of Delirium

Whether Assessment Was Complete (N = 114)
Mini-Cog Score Predicted Risk of Delirium, %
Completed
Final No Final 0 64
Assessment, Assessment, 1 46
Characteristic n = 76 n = 38 P-Value 2 29
3 17
Age, mean  standard 72  6.7 71  6.8 .65 4 8.5
deviation 5 4.5
Education, n (%)
<High school 4 (7.6) 1 (3.7) .45
High school or GED 8 (15) 2 (7.4)
≥4-year degree 21 (40) 24 (89)
Alcohol consumption, n (%)
Never 39 (37) 15 (48) .82
≥1/month 12 (16) 7 (23)
3–5/week 13 (17) 4 (13)
Daily 11 (15) 5 (16)
Charlson-Deyo Score, n (%)
0 27 (36) 12 (42) .41
1 19 (25) 8 (29)
2 17 (23) 2 (7.1)
≥3 12 (16) 6 (21)
Patient Health Questionnaire-9, n (%)
0 37 (52) 15 (50) .17
1 24 (34) 13 (43)
≥2 10 (14) 2 (6.7)
Hearing difficulty, n (%) 18 (24) 4 (13) .20
Vision difficulty, n (%) 16 (21) 5 (16) .54
Study of Osteoporotic Fractures Frailty Index, n (%)
Robust 39 (52) 14 (45) .59
Prefrail 27 (36) 11 (35) Figure 1. Predicted probability (lines) of delirium according
Frail 9 (12) 6 (19) to Mini-Cog score. The shaded area represents the 95% confi-
Orientation, number wrong, n (%) dence interval. Observed episodes of delirium according to
0 36 (56) 14 (47) .23 Mini-Cog score are represented by the •. [Color figure can be
1 18 (28) 7 (23)
viewed at wileyonlinelibrary.com]
2 5 (7.8) 6 (20)
≥3 5 (7.8) 3 (10)
Mini-Cog score, n (%) delirium and 90% of those without. A score of 0 or 1 cor-
0 2 (2.7) 0 (0.0) .89 rectly predicted 60% of the observed individuals with
1 3 (4.0) 1 (3.2) delirium and 88% of those without.
2 6 (8.0) 2 (6.5)
The clock draw portion of the Mini-Cog by itself had
3 15 (20) 5 (16)
4 16 (21) 9 (29) moderate predictive ability (AUC = 0.72). Six of 17 partic-
5 33 (44) 14 (45) ipants (35%) with abnormal preoperative clock draw tests
Word recall score, n (%) and four of 58 (7%) with normal clock draw tests devel-
0 2 (2.7) 0 (0.0) .39 oped delirium. One participant’s clock was not scored.
1 12 (16) 2 (6.9) Two other models that incorporated more factors
2 22 (29) 12 (41) emerged from the analysis. The first model added age and
3 39 (52) 15 (52) age2 (AUC = 0.86, 95% CI = 0.75–0.97), the second added
Clock draw score
Charlson-Deyo score (AUC = 0.79, 95% CI = 0.59–0.97),
Normal 57 (77) 24 (77) .97
Abnormal 17 (23) 7 (23) with a maximum of 3 points on Charlson-Deyo.
Although the predictive ability of the model including
P-value assessed using t-test for continuous variables and chi-square test age was marginally but nonsignificantly better, the for-
for categorical variables. mula to calculate risk is complex: risk = 1.3 + (Mini-
Cog 9 1.01) + (age65 9 0.27) + (age652 9 0.019).
Age65 is the number of years greater than 65 (e.g., aged 76:
postoperative delirium, those with a score of 3 had an age65 = 11). Risk can be converted to predicted probability
approximately 20% probability, those with a score of 4 using the formula probability = eRisk/(1 eRisk).
had a 13% probability, and those with a score of 5 had
less than 5% probability of delirium (Figure 1).
DISCUSSION
A Mini-Cog score of 4 or worse correctly predicted
21% of the observed individuals with delirium and 94% In this prospective cohort study of older adults undergoing
of those without (positive predictive value = 21%, nega- elective surgery, the Mini-Cog scored from 0 to 5 is a sim-
tive predictive value = 94%). A score of 2 or worse cor- ple tool for assessing the risk of postoperative delirium. A
rectly predicted 36% of the observed individuals with significant increase in risk of delirium was found with each
4 DWORKIN ET AL.
increase in Mini-Cog score. Adding other factors to the
model (age and age2 or Charlson-Deyo comorbidity score)
marginally but not significantly increased predictability,
but both of these models would be more complicated to
use in a clinical setting. The AUC and predictive ability of
the Mini-Cog are similar to those of previous, more-com-
plex models derived to predict risk of delirium.10 The cur-
rent study provides a simple tool to predict the risk of
delirium in older adults considering elective surgery.
The incidence of delirium in this study was 13%,
which aligns with published studies in similar popula-
tions.3,5,10 It is lower than some studies that included sub-
jects who lived in skilled care facilities or needed urgent,
cardiac, or central nervous system surgery. Similar to
many previous studies, a number of factors were identified
that appeared to increase the risk of developing postopera-
tive delirium, including comorbidity, depression, alcohol
use, disorientation, and low education, although these fac-
tors did not substantially improve prediction over the
Mini-Cog score alone.
Several characteristics of the Mini-Cog make it an
attractive screening tool. The test is free, takes 2 to 3 min-
utes to administer, is simple to score, and is robust in popu-
lations of varying ethnicity and educational levels.22
Preexisting cognitive impairment as measured using the
Mini-Cog has previously been found to be a strong predictor
of postoperative delirium.23 A Canadian prospective cohort
study found that general medicine patients without known
dementia who scored from 0 to 2 on the initial Mini-Cog
were four times as likely to develop delirium as peers with
scores of 3 to 5.24 In these studies, the absolute risk associ-
ated with a given Mini-Cog score was not quantified.
Delirium is thought to stem from complex interactions
between medical stressors, environmental cues, and indi-
vidual physiology, including preexisting dementia, that
predispose people to cognitive decline. The Mini-Cog is a
sensitive and specific dementia screening tool.25 Although
none of the participants in the current study had existing
diagnoses of dementia, clinicians often fail to identify
dementia, especially at early stages.26 The Mini-Cog may
therefore have identified some individuals previously undi-
agnosed with dementia who were at greater risk of delir-
ium. The Mini-Cog tests a variety of cognitive domains,
including attention, comprehension, short-term recall,
executive function, visual-spatial ability, and praxis. A low
Mini-Cog score may indicate subtle but multifocal cogni-
tive impairment short of dementia that still predisposes
subjects to delirium in the face of stress and disorientation.
The finding of greater risk of delirium in participants with
a Mini-Cog score of 3, above the screening cutoff for
dementia, supports this possibility
If further studies validate this finding, Mini-Cog
screening could significantly affect clinical care and
research on postoperative delirium. The ability to risk-stra-
tify would help physicians and individuals make more-
informed decisions about the hazards and benefits of
pursuing elective surgery, particularly in older adults and
individuals with multimorbidity. For example, the Mini-
Cog might be administered before consideration of elective
surgery, avoiding offering surgery to individuals at highest
risk of delirium. Practitioners and health systems
could also focus resources and delirium-prevention
2016 JAGS
programs on at-risk individuals, as professional guidelines
recommend.12 This is especially important because current
interventions appear to be more effective at preventing
delirium, especially in individuals at moderate risk, than at
treating delirium once it develops.11 Screening could also
help guide selection of surgical anesthesia or postoperative
analgesia, although the influence of these factors on delir-
ium is debated.27,28 Finally, the Mini-Cog could be used to
stratify individuals’ baseline risk of delirium in future stud-
ies to better understand the interaction between baseline
risk of delirium and the effect of interventions to decrease
postoperative delirium.
This study has several limitations. The dropout rate
was high, and the size of the sample with complete data
was small because of cancellation of surgery, change in
anesthetic plan after baseline screening, or inability to reach
the participant or study partner within 48 to 72 hours after
surgery. Incident delirium was assessed for within 72 hours
of surgery, which forced some individuals who could not
be interviewed in that time frame to be excluded. Although
this may have lowered the total number of participants
with delirium, assessment within this short time window
was important to ensure that delirium was related to sur-
gery, because signs and symptoms of delirium generally
appear within 2 to 3 days of a new hospital admission or
illness.29 The small number of cases of delirium identified
limited power to detect risk factors that are rare or that
convey a small magnitude of risk. It is reassuring that indi-
viduals who did not complete the study did not differ sig-
nificantly in their baseline characteristics from those who
completed the second assessment. Although the CAM and
FAM-CAM are validated tools and acceptable for screening
or research, this study may have resulted in false negatives,
because the criterion standard to diagnose delirium is an
evaluation by a qualified clinician. In addition, more data
about the intraoperative course of participants and the spe-
cifics of the anesthesia treatments administered would be
important to investigate in future studies.
Finally, study participants were mostly male, reflecting
the population at the study hospital. This and the exclu-
sion of individuals residing in institutional settings limit
generalizability to other populations. Further studies in lar-
ger, more-diverse populations are required to validate the
use of the Mini-Cog as a delirium risk prediction tool.
This study has important strengths in development of
a clinical prediction tool for postoperative delirium. It was
a prospective cohort study with extensive baseline infor-
mation on diagnoses and medications provided by the
VA’s electronic Clinical Patient Record System. Most pre-
viously identified risk factors were included in the analy-
ses, including the Mini-Cog, which is a well-validated tool.
Baseline data were collected during normal preoperative
visits without disrupting workflow, indicating that incor-
porating the Mini-Cog into clinical practice is feasible.
Various healthcare team members can perform and score
the Mini-Cog, including medical assistants and nurses, and
it could be easily incorporated into preoperative assess-
ments. Primary care providers could use the Mini-Cog and
might explore nonsurgical options for individuals at high-
est risk of postoperative delirium.
In conclusion, the Mini-Cog scored from 0 to 5 served
as a simple tool with adequate predictive power to identify
JAGS 2016
risk of postoperative delirium in this population of veter-
ans undergoing elective surgery with major anesthesia.
Additional studies will be important to validate the use of
the Mini-Cog to screen for postoperative delirium and to
determine its generalizability.
ACKNOWLEDGMENTS
This project was conducted with funding from the
Research Division of the VA Portland Health Care System.
The authors wish to thank the clinicians and staff of the
Pre-operative Clinic at VA Portland Health Care System
for their support during participant enrollment and initial
data collection and Martha Gerrity, MD, for her assistance
with the manuscript.
Author Contributions: Dworkin: acquisition of data,
drafting and revising the manuscript. Lee: design, analysis,
interpretation of data; critical revision of manuscript. An:
conception and design. Goodlin: conception and design,
acquisition and interpretation of data, critical revision of
manuscript for important intellectual content. All authors:
final approval of version to be published.
Sponsor’s Role: Portland VA Medical Center Research
Division provided initial funding to Dr. Goodlin for
research staff that made data acquisition possible.
REFERENCES
1. Canet J, Raeder J, Rasmussen LS et al. Cognitive dysfunction after minor
surgery in the elderly. Acta Anaesthesiol Scand 2003;47:1204–1210.
2. Neufeld KJ, Leoutsakos JS, Sieber FE et al. Evaluation of two delirium
screening tools for detecting post-operative delirium in the elderly. Br J
Anaesth 2013;111:612–618.
3. Ansaloni L, Catena F, Chattat R et al. Risk factors and incidence of post-
operative delirium in elderly patients after elective and emergency surgery.
Br J Surg 2010;97:273–280.
4. Williams-Russo P, Urquhart BL, Sharrock NE et al. Post-operative delir-
ium: Predictors and prognosis in elderly orthopedic patients. JAGS
1992;40:759–767.
5. Fong HK, Sands LP, Leung JM. The role of postoperative analgesia in
delirium and cognitive decline in elderly patients: A systematic review.
Anesth Analg 2006;102:1255–1266.
6. Leslie DL, Zhang Y, Holford TR et al. Premature death associated with
delirium at 1-year follow-up. Arch Intern Med 2005;165:1657–1662.
7. McAvay GJ, Van Ness PH, Bogardus ST et al. Older adults discharged
from the hospital with delirium: 1-year outcomes. J Am Geriatr Soc
2006;54:1245–1250.
8. McCusker J, Cole M, Abrahamowicz M et al. Delirium predicts 12-month
mortality. Arch Intern Med 2002;162:457–463.
9. Leslie DL, Marcantonio ER, Zhang Y et al. One-year health care costs
associated with delirium in the elderly population. Arch Intern Med
2008;168:27–32.
10. Van Meenen LC, van Meenen DM, de Rooij SE et al. Risk prediction mod-
els for postoperative delirium: A systematic review and meta-analysis. J Am
Geriatr Soc 2014;62:2383–2390.
11. Inouye SK, Bogardus ST Jr, Charpentier PA et al. A multicomponent inter-
vention to prevent delirium in hospitalized older patients. N Engl J Med
1999;340:669–676.
12. American Geriatrics Society Expert Panel on Postoperative Delirium in
Older Adults. American Geriatrics Society abstracted clinical practice
PREDICTING POSTOPERATIVE DELIRIUM 5
guideline for postoperative delirium in older adults. J Am Geriatr Soc
2015;63:142–150.
13. Pompei P, Foreman M, Rudberg MA et al. Delirium in hospitalized older
persons: Outcomes and predictors. J Am Geriatr Soc 1994;42:809–815.
14. Ahmed S, Leurent B, Sampson EL. Risk factors for incident delirium
among older people in acute hospital medical units: A systematic review
and meta-analysis. Age Ageing 2014;43:326–333.
15. Inouye SK, van Dyck CH, Alessi CA et al. Clarifying confusion: The Con-
fusion Assessment Method. A new method for detection of delirium. Ann
Intern Med 1990;113:941–948.
16. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depres-
sion severity measure. J Gen Intern Med 2001;16:606–613.
17. Ensrud KE, Ewing SK, Taylor BC et al. Comparison of 2 frailty indexes
for prediction of falls, disability, fractures, and death in older women. Arch
Intern Med 2008;168:382–389.
18. Borson S, Scanlan J, Brush M et al. The Mini-Cog: A cognitive ‘vital signs’
measure for dementia screening in multi-lingual elderly. Int J Geriatr Psy-
chiatry 2000;15:1021–1027.
19. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index
for use with ICD-9-CM administrative databases. J Clin Epidemiol
1992;45:613–619.
20. Steis MR, Evans L, Hirschman KB et al. Screening for delirium using fam-
ily caregivers: Convergent validity of the Family Confusion Assessment
Method and interviewer-rated Confusion Assessment Method. J Am Geri-
atr Soc 2012;60:2121–2126.
21. American Geriatrics Society 2012 Beers Criteria Update Expert Panel.
American Geriatrics Society updated Beers Criteria for potentially inappro-
priate medication use in older adults. J Am Geriatr Soc 2012;60:616–631.
22. Borson S, Scanlan JM, Watanabe J et al. Simplifying detection of cognitive
impairment: Comparison of the Mini-Cog and Mini-Mental State Examina-
tion in a multiethnic sample. J Am Geriatr Soc 2005;53:871–874.
23. Robinson TN, Raeburn CD, Tran ZV et al. Postoperative delirium in the
elderly: Risk factors and outcomes. Ann Surg 2009;249:173–178.
24. Alagiakrishnan K, Marrie T, Rolfson D et al. Simple cognitive testing
(Mini-Cog) predicts in-hospital delirium in the elderly. J Am Geriatr Soc
2007;55:314–316.
25. Scanlan J, Borson S. The Mini-Cog: Receiver operating characteristics with
expert and naive raters. Int J Geriatr Psychiatry 2001;16:216–222.
26. Ismail Z, Rajji TK, Shulman KI. Brief cognitive screening instruments: An
update. Int J Geriatr Psychiatry 2010;25:111–120.
27. Rasmussen LS, Johnson T, Kuipers HM et al. Does anaesthesia cause post-
operative cognitive dysfunction? A randomised study of regional versus
general anaesthesia in 438 elderly patients. Acta Anaesthesiol Scand
2003;47:260–266.
28. Maldonado JR, van der Starre PJ, Wysone A. Post-operative sedation and
the incidence of ICU delirium in cardiac surgery patients. Anesthesiology
2003;99:A465.
29. Saravay SM, Kaplowitz M, Kurek J et al. How do delirium and dementia
increase length of stay of elderly general medical inpatients? Psychosomat-
ics 2004;45:235–242.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Appendix S1 Candidate Variables to Predict Risk of
Postoperative Delirium.
Please note: Wiley-Blackwell is not responsible for the
content, accuracy, errors, or functionality of any support-
ing materials supplied by the authors. Any queries (other
than missing material) should be directed to the corre-
sponding author for the article.