Brain Health Screening in Older
Surgical Patients
A quick review and lessons learned
Alexander T. Abess, MD | Geisel School of Medicine at Dartmouth | Dartmouth Health
MPOG RETREAT 2026
GEISELMED.DARTMOUTH.EDU
Disclosures
Co-Founder of medical device startups
Advisor to Simbex
Use of AI
Claude (Anthropic) was used to help prepare these slides
GEISELMED.DARTMOUTH.EDU
Acknowledgements
Co-authors
Nirav J. Shah, MD
University of Michigan
Elizabeth L. Whitlock, MD, MSc
UC San Francisco
Hedwig Schroeck, MD
Dartmouth Health
Donna Ron, MD
Tel Aviv University
(formerly Dartmouth Health)
Sandra Becker Rozek, MA
University of Michigan
Pablo Martinez-Camblor, PhD
Dartmouth Health
Anne L. Donovan, MD
UC San Francisco
Katie J. Schenning, MD, MPH
Oregon Health & Science University
Stacie G. Deiner, MD, MS
Dartmouth Health
MPOG Coordinating Center
and all MPOG collaborators and participating sites
GEISELMED.DARTMOUTH.EDU
Published in Anesthesia & Analgesia
Abess AT, et al. Anesth Analg. 2025;141:1097–1106
GEISELMED.DARTMOUTH.EDU
Why do this study?
Common and high impact
Perioperative neurocognitive disorders
are associated with a range of adverse
outcomes
Strong recommendations for
screening
ASA, ACS, ESA all recommend cognitive,
frailty and delirium screening
Unknown state of
implementation
Who is screening, and how?
Hard to study at scale
Research is mostly small, single-site
studies; no ability for large multicenter
pragmatic studies
Born from challenges at our own institution → MPOG working group formed
GEISELMED.DARTMOUTH.EDU
Key Question
What is the current state of implementation of recommended perioperative
geriatric screening for cognitive impairment, frailty, and delirium, and what is
the availability of this data for multi-center research?
Hypotheses
1. Few geriatric screenings would be
retrievable in MPOG
2. Common barriers block both doing
screenings locally and capturing
them in MPOG
Retrospective MPOG analysis
908,158
records, age ≥ 65
57
institutions
2018–22
nonemergent inpatient
surgery
Survey of MPOG sites
49
sites invited
43
responses (88%)
Oct 2022
launched at the retreat
GEISELMED.DARTMOUTH.EDU
What we observed: screenings were rarely present
0.89%
had a preop cognitive screen
8,054 records · mostly MoCA
13.6%
had a postop delirium
screen
123,114 records · 99% CAM
0
frailty screens found
in 908,158 records
Increase in screenings over time:
10.7% → 17.7% of records had any
screen (2018-19 vs 2020-22)
Screens happen but aren’t
captured: Mini-Cog was most used
per the survey, but MoCA was most
common in MPOG
GEISELMED.DARTMOUTH.EDU
Table 1. Patient characteristics
Similar patients
Characteristics were similar among
patients with and without screening
tests; those with tests were slightly
sicker
Institution mattered most
The most significant factor associated
with presence or absence of a test was
the institution where the surgery
occurred
GEISELMED.DARTMOUTH.EDU
Who screens, and with what tools
Figure 1. Proportion of institutions that reported performing each screening
10 sites do all three 13 sites do none
GEISELMED.DARTMOUTH.EDU
Barriers to screening
Figure 2. Barriers to cognitive, frailty, and delirium screening (33 responding sites)
No standard EHR
tools
Top significant barrier (13
sites)
Time and workflow
Most common moderate
or significant barriers (26
sites each)
Patient resistance
Rated the least important
barrier
System barriers dominate: embedded EHR tools could help
GEISELMED.DARTMOUTH.EDU
Sites want these data in MPOG
Broad agreement
Contributing these data to MPOG matters to most
survey respondents.
Figure 3. Importance of contributing cognitive, frailty, and delirium screening data
GEISELMED.DARTMOUTH.EDU
Key Takeaways
Low level of retrievable screenings
Only 14% of 908,158 records had any screen; under 1% had a preoperative cognitive screen and none had a
frailty screen.
Common barriers impede implementation of screening
About half of MPOG sites do not routinely screen. Predominantly System Barriers: lack of EHR tools, time, and
workflow.
Easy-to-use, embedded EHR tools: a high-yield next step
Standardized, structured screening tools could support both clinical care and multicenter research.
GEISELMED.DARTMOUTH.EDU
Lessons Learned and Limitations
Looking backwards
to find the absence
of something
Variability in
screening
Survey bias and
fatigue (dropout)
GEISELMED.DARTMOUTH.EDU
Future directions
Co-Develop Screening tools with Electronic Health Record Vendors
Standard, embedded screening tools with structured fields (31 of 33 sites use Epic)
ASA/APSF Working Group
Leverage and cooperate with other organizations
Renewed effort to connect variables
We need you to help us connect your site!
GEISELMED.DARTMOUTH.EDU
Quick poll: what does your site do?
QR code linking to the audience poll
Scan to vote
forms.cloud.microsoft/r/mUTnmifQVE
1
Does your institution routinely perform preoperative cognitive
screening for older patients?
Yes · No · Unsure
2
Does your institution routinely perform preoperative frailty
screening for older patients?
Yes · No · Unsure
3
Does your institution routinely perform postoperative delirium
screening for older/at-risk patients?
Yes · No · Unsure
4
If your institution does not routinely perform one or more of
these screenings, what is the primary barrier?
EHR tools · Time · Workflow · Clinical interest · Other
Thank you
Questions?
alexander.t.abess@dartmouth.edu
