Postoperative Delirium:
Who’s at Risk & How to Predict?
Deborah J. Culley
Harvard Medical School
Brigham & Women’s Hospital
Conflicts
R21 AG048522
R56 AG055833
RO1 GM132668
Risk Factors for Delirium
Medical - Fever, electrolyte abnormalities, AF, Frailty, etc.
Physiologic - Low SaO2, Hct, albumin
Pharmacology - Medication history; anticholinergics,
ketamine, propofol, neuroleptics
Patient - Age, pre-existing cognitive impairment,
genotype, depression, sensory deficits
Surgical - Ortho, cardiac, major vascular & thoracic,
emergency
J Gen Intern Med. 2018;33:500-509
Risk Factors for Delirium
Advanced AGE
High Vulnerability Major Insult
Minor InsultLow Vulnerability
Vulnerability
Factors
Precipitating
Factors
Adapted from Sharon Inouye (Harvard)
Postoperative Complications
in Geriatric Patients
• Liu LL, et al., JAGS 48:405, 2000
• Moller JT, et al. Lancet 351: 857, 1998
• Monk TG, et al., Anesthesiology
108:18, 2008
Complication Incidence
Pulmonary Embolism 0.5%
ARDS 0.8%
Stroke 1%
Myocardial infarction 2%
Pneumonia 4%
Death 5%
Heart Failure 6%
Delirium 15-60%
POCD 10-12%
Anesthesiology. 2017;127:765-774
Preoperative cognition predicts delirium
Jones RN, et al. J Geriatr Psychiatry Neurol 2016; 29:3320-27
Frailty
Robinson TN, et al., J Am Coll Surg 2015; 221: 1083-92
Preoperative Frailty Predicts Delirium
Brown CH IV, et al. Anesth Analg 2016; 123: 430-35
N
o
r
m
a
l
C
I
V
e
r
b
a
l
F
l
u
e
n
cy
C
I
M
ini-
C
o
g
F
r
a
i
l
F
r
a
i
l
+
C
I
0
10
20
30
40
50
% Delirium
Brief Preoperative Screening for Frailty and Cognitive
Impairment Predicts Delirium Following Spine Surgery
Susano & Culley, Under Review, JAMA Surgery
79 Days
J Am Coll Surg 2012; 215: 453-66
A&A 2018; published before print
But ... Are there things we can do?
Oh, ES, et al.
JAMA 2017; 121: 318: 1161-74
JAMA Netw Open2018;1(4):e181405.
VA Hospital System, 27,625 patients, 98% Male,
76 ± 9 years of age
Delirium Screening Targets:
preexisting cognitive impairment
Infection
sodium level
≥ 80 years or older
JAMA Netw Open.2018;1(4):e181018.
Training sample: 14,227
Test set 3,996
≥ 18 years of age
Can we use MPOG to identify patients at risk for delirium?
Nationally
State Wide
Regionally
Are there things we can do?
Oh, ES, et al.
JAMA 2017; 121: 318: 1161-74
Br J Anaesth. 2019;123:479-487
N = 102 patients
Br J Anaesth. 2019
Br J Anaesth. 2019
< 95K patients
1% Mortality
Br J Anaesth. 2019
Can we do this for delirium?
• It will likely be more difficult
• Prior examples had end firm end points
• 30-day Postoperative Mortality in a Large Number of Patients
• Sedations Scores in All Patients
• Predictors may be in the electronic record
• Diagnosis of delirium is less likely to be accurate in the medical record
• This would require a commitment to screen a large number of people for
delirium for the Machine Learning and Validation
• It is worth doing …
•
Are there things we can do?
Oh, ES, et al.
JAMA 2017; 121: 318: 1161-74
Informed Consent
Patient and Family Centered Care
Could you imagine a world where you could predict
the number one cause of postoperative morbidity in
older surgical patients with an AUC of 0.8?
We have the opportunity
to reduce the risk by up to 40%!
Thank You!
http://www.asahq.org/brainhealthinitiative#news
