Daniel E Hall, MD, MDiv, MHSc

Professor of Surgery, Anesthesia and Perioperative Medicine, University of Pittsburgh

Medical Director, High Risk Populations and Outcomes, Wolff Center at UPMC

Staff Surgeon, VA Pittsburgh Healthcare System

Core Investigator, Center for Health Equity Research and Promotion, VA Pittsburgh Healthcare System

National Diffusion of Excellence Fellow, VA

THE SURGICAL PAUSE:

Preoperative Frailty Screening and Prehabilitation

MSQC/ASPIRE April 8, 2022

Outline

• The Risk Analysis Index & the Surgical Pause

• Origin story

• Conceptual framework

• Data—It works

• Your Questions

•NOT going to tell you

who should/should not

have surgery

•May cause anxiety

•New ways of thinking

•Changing culture is hard

Observed/Expected Mortality at the Omaha VAMC

(Red points are > 90% Confidence Interval)

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

2008 2009 2010 2011 2012 2013 2014

Mortality O/E Ratio

Fiscal Year

Origin Story & Conceptual Framework

Omaha: We’ve got a problem

Silver Tsunami

• 1/3 of patients had surgical

interventions in last year of life

• Majority occurred in month

before death

• Surgery associated with

• More admissions

• Longer LOS

• Greater ICU LOS

Kwok AC. Lancet. 2011;378(9800):1408-1413.

We know some patients don’t do well

Smith T, et al., JAMA Surg. 2016;151(5):417-422.

But surgeons are optimists!

“Where all the surgeons are strong, all

the anesthesiologists are good looking,

and all the patients are above average.”

Now how accurate is your eyeball?

Systematic, multifactorial, risk assessment

• “Foot of the bed” assessments of cardiac risk not reliable due to

disagreement between clinicians.

• Hii TB, et al. Heart Lung Circ. 2015;24(6):551-556.

• Multifactorial tools are superior to single-item assessments.

• Afilalo J, et al. Circulation. 2017;135(21):2025-2027

• Hurria A, et al. J Clin Oncol. 2011;29(25):3457-3465.

• Fried L, et al. The Journals of Gerontology: Series A, 2004; 59(3):M255–M263

• Vascular Surgeons effectively estimate mortality, but underestimate

complications and long-term disability compared to multifactorial tool.

• George EL, et al. J Surg Res. 2020;248:38-44.

• Modified Geriatric Assessment (mGA) effectively identifies frailty among

patients that oncologists considered non frail (e.g. ↑ sensitivity).

• Kirkhus, et al. Br J Cancer 117, 470–477 (2017)

Frailty is the Best Predictor of Postoperative

Outcomes….

Makary MA, et al., J Am Coll Surg. 2010;210(6):901-908

• Mortality

• Complications

• Failure to Rescue

• Length of Stay

• Readmission

• Loss of Independence

Why Frailty?

A clinical syndrome of decreased

physiological reserve

• process whereby small deficits

accumulate in multiple adaptive

systems, any one of which might be

clinically insignificant, but together

they produce significant vulnerability

to stress that can lead to catastrophic

decompensation.

• multiple causes and contributors

• characterized by diminished strength,

endurance, nutrition, and cognitive

capacity

• More than just age or the sum of

comorbidities (not captured by

standard risk stratification tools like

ASA or Eagle criteria).

Robert, C. M., & Sean, M. B. (2014). Physiological Reserve and Frailty in

Critical Illness. Oxford, UK: Oxford University Press.

Patient Presents to

Outpatient Surgical Clinic

Risk Stratify by Assessing Frailty

SURGICAL PAUSE (Make No Promises)

• Further Risk Assessment & Mitigation

• Risk-informed Shared Decision

Making Process

Surgery

Non Operative

Management

Usual

Care

Prehabilitation

The

Surgical

Pause

So what happened in Omaha?

• Modified an existing frailty measure (MMRI) for use in surgery

• RISK ANALYSIS INDEX (RAI)

• Made it mandatory to book OR time

• Conducted weekly review of all surgeries scheduled on frail patients.

• Spoke with surgeon to review operative decision making.

• Spoke with anesthesiologists to optimize anesthetic plan.

• Spoke with intensivists to encourage post-operative rescue from near certain

complications.

• Aggressive referral for preoperative palliative care to clarify goals.

Observed/Expected Mortality at the Omaha VAMC

(Red points are > 90% Confidence Interval)

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

2008 2009 2010 2011 2012 2013 2014

Mortality O/E Ratio

Fiscal Year

Outcomes: Decreased Mortality

Start Frailty Screening Initiative

Omaha Frailty Screening Initiative (FSI)

• 180-day mortality among frail fell from 23.9% to 7.7%

(p<0.001)

• 3-fold survival advantage after FSI implementation

(OR 2.87 [95%CI 1.98-4.16]), controlling for:

• Age

• Frailty

• Predicted mortality based on VA risk-adjustment

Hall, DE. et al. JAMA Surgery 152(3) doi:10.1001/jamasurg.2016.4202 (Nov 23).

FSI Changed Perioperative Palliative Care

• Changed Pattern of Perioperative Palliative Care Care Consult

• Rate increased from 32 to 56 per year.

• More often ordered by a surgeon (56.7% vs 24.4%; p< 0.05).

• More often ordered before surgery (52.0% vs 26.3%; p< 0.05).

• Controlling for age, frailty and whether the patient had surgery, Preoperative

Palliative Care Consult reduced risk of death when:

• ordered by a surgeon (AOR 0.50[95% CI 0.30-0.83], p=0.007).

• ordered before surgery (AOR 0.52[95% CI 0.30-0.90], p=0.02).

• ordered by surgeon before surgery (AOR 0.27[95% CI 0.11-068], p=0.006)

Ernst, K. F., et al(2014). JAMA Surg, 149(11), 1121-1126.

Decreased Mortality at VA Pittsburgh

Frailty Screening Initiative begins 12/15/15

Decreased Mortality at UPMC

• Interrupted Time Sequence Analysis with segmented

Poisson regression.

• 51,385 patients July 2016-November 2020

• 23,153 before BPA Implementation

• 28,232 after BPA Implementation

• Overall 180-day mortality reduction

• aOR 0.76 [95% CI 0.65-0.88 ]

• 2-fold survival advantage among frail.

• aOR for survival 2.14 [95% CI 1.42-3.21]

• Cut raw mortality among frail from 14% to 7%

• Lag-adjusted ITS model

• 0.03 fewer 180-day mortalities/1,000 procedures/month

_cons .3495465 .123177 2.84 0.005 .1081239 .590969

lag_180 .1069392 .0404603 2.64 0.008 .0276385 .1862398

interaction -.0278729 .0082757 -3.37 0.001 -.0440931 -.0116528

intervention -.2864171 .1280697 -2.24 0.025 -.5374291 -.0354051

death_time .0206505 .0037606 5.49 0.000 .0132799 .0280211

rate_180 Coefficient std. err. z P>|z| [95% conf. interval]

Robust

OK, but… I can get anyone through a minor procedure:

RAI, Operative Stress and Mortality

• Delphi consensus methodology to rate

operative physiological stress.

• 566 surgical procedures that account for 90% of

all VA surgery

• Ratings by panel of surgeons and

anesthesiologists

• Consensus reached after 3 rounds of rating.

• 5-point Operative Stress Score:

1-cystoscopy, hydrocele, ganglion cyst

2-inguinal or umbilical hernia, arthroscopy of knee

or shoulder

3-cholecystectomy, CEA, arthroplasty of knee,

shoulder or hip

4-open colectomy, prostatectomy, pulmonary

lobectomy or segmentectomy

5-abdominal aortic aneurysm,

pancreaticoduodenectomy, esophagectomy

Shinall, Myrick C. et al. JAMA Surgery 10.1001/jamasurg.2019.4620 (Nov 13).

180 day Mortality Rate

1 2 3 4 5

Physiological Stress Category

<=20 21-29

30-39 40+

RAI-A Category

180 day Mortality by Physiologic Stress and RAI-A Categories

NO SUCH THING AS LOW-RISK SURGERY FOR THE FRAIL

Practical Implementation at UPMC

Frailty Screening

Online RAI

RAI Survey

Frailty

Risk

Analysis

Index

Functional

Activities of

Daily Living

Physical

CHF

CKD

Dyspnea

Cancer

Cognitive

Mental

Status

Nutritional

Weight Loss

Appetite

Social

Age

Sex

Living

Location

Arya et al. Ann Surgery 2019; Shah, et al, J Am Geriatrics 2020; Varley, et al, Ann Surgery 2020

Risk Analysis Index (RAI)

• 14 Variables; weighted scale

• Grouped into 4 categories

with increasing frailty severity

Robust: 0-29

Average: 30-36

Frail: 37-44

Very Frail: ≥ 45

• Most thoroughly validated

measure of surgical frailty,

and only shown feasible for

point-of-care testing

RAI Validation in Veterans and Private Sector

VASQIP & ACS-NSQIP

VASQIP (c=0.842, n=480,731) ACS-NSQIP (c=0.870, N=1,391,785)

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

RAI-rev

Predicted

Observed

C-statistic= 0.842 (0.839-0.845)

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

RAI-rev

Predicted

Observed

C-statistic= 0.870 (0.867-0.873)

Arya, S. et al. Annals of Surgery doi 10.1097/SLA.0000000000003276 (2019, March 23).

RAI Implementation at UPMC: Feasible

Time to assess: 30 seconds (IQR 23-53)

More than 450,000 Assessments to Date

Varley PR, et al Ann Surg. 2020 10.1097/SLA.0000000000003808.

Including me!

5/20/2021

A torn achilles made

me less mobile than

the day before…

but not frail yet.

Now available in Epic

as a Clinical Program

Implementation Map

Active Sites (28)

Exploratory/Interested

Sites (14)

Puerto Rico

Map Legend

Pending Sites (6)

• Total of 50 engaged sites across ALL 18 VISNs

• 28 Sites are active

• 16 sites are considered fully

implemented

• 12 sites are nearing full

implementation within the next few

months.

• FY22 Goal:

• 35 fully implemented sites across 18 VISNs

Threshold for Action: Revised RAI ≥ 37

• Riskiest 10% of population;

• At least twice the average 6-month mortality

• 12% vs 6%

• Twice the rate of 30- and 90-day readmission

• 22% vs 12%

• Twice the rate of long term ICU stay ≥ 5 days

• 6% vs. 3%

• Modest positive predictive value: 19%

• Strong negative predictive value: 96%

• Safe to operate on patients with Revised RAI<37 (e.g. most patients)

Step 1:

Assess Frailty

Step 2:

Do Something About It

Review

↓Mortality

25% to 8% (6-month Mortality)

•3-Fold Survival Advantage

Changes Care Plans

21% declined surgery

↑Functional Performance

↑Endurance

↑Gait Speed

↑Respiratory Pressures

Changed Care Plans

UPMC

Feb 1- Sep 20, 2018

49.654 Patients Screened

2149 (4.3%) RAI ≥ 37

549 (27%) Surgery Planned

472 (86%)

Shared Decision

132 (24%)

CPC

147 (27%)

PCP

28 (21%)

Non-Op

Mngmt

73 (56%)

Delay for

Prehab

5 (4%)

Less

Complex

26 (20%)

Delay

~ 1 Day Reduction in LOS compared to historical controls

Baseline to Day of Surgery

Significant Changes in Physical Performance

Measure

Baseline

Mean (SD)

Day of Surgery

Mean (SD)

Mean Difference

(Standard Error)

P value

Minimum

Clinically

Important

Difference

Extended TUG

(seconds)

N=42

21.9 (12.5)

N=33

17.8 (4.6)

-2.3 (0.5)

<0.001 2.4s

Gait Speed

(meters/second)

N=42

1.11 (0.32)

N=33

1.24 (0.30)

+0.1 (0.03)

0.002 0.1m/s

5 Chair Rise

(seconds)

N=38

13.3 (5.7)

N=33

11.8 (4.6)

-1.6 (0.6) 0.007 2.3s

Six Minute Walk

Test

(meters)

N=40

348.6 (109.1)

N=30

380.6 (102.2)

+29.3 (15.6) 0.060

30m

SPPB Score

N=41

10.2 (1.9)

N=33

10.8 (1.1)

+0.6 (0.3) 0.068 1 unit

Implementation Nuts and Bolts

Two Step Process

• Step 1: Measure Frailty

• Don ‘t Triage the Triage Tool (Measure on Everyone)

• Must measure frailty before booking surgery date

Two Step Process

• Step 2: Do Something About It

• Surgeon champion review

• Interdisciplinary Review Panel

• Surgery, Anesthesia, Palliative Care, Geriatrics, IMPACT Clinic

• Real time or Time Asynchronous

• Goal Clarification & Shared Decision Making

• ”Not a candidate” is NOT shared decision making

• Avoid mental model of “fixing it”

• I’m worried that no matter what we do life will never be the same for you

• Best, Worst, and Most Likely Scenarios of at least 2 options

• Who has this conversation?

• Palliative care has skill but not necessarily the knowledge

• Surgeons have the knowledge, but not necessarily the skill

• Training options available

Lessons Learned

• It’s not a math problem

• Maximizing c-statistics is a distraction

• No algorithm can determine what we should/should not do

• RAI signals need to shift from fast to slow thinking

• It’s about insight not technique

• Shared decision making is really challenging, but it is the next frontier

• Focusing on all-cause mortality creates opportunity

• The RAI works because it is simple, fast, and guides intervention

• Phenotypical frailty may be more “pure” but not feasible for wide screening

• Don’t try to triage the triage tool

• Light, flexible touch—not too much structure

• With a gentle nudge, surgeons step up

• So adapt to your site’s requirements

• 1-2 hours/week of surgical champion

Many thanks to growing Research network.

• Health Systems with RAI

• Atlanta-Emory/VA

• Nashville-Vanderbilt

• Phoenix-VA

• Pittsburgh-UPMC/VA

• Palo Alto-Stanford/VA

• Omaha-UNMC/VA

• Richmond-VA

• Houston-Baylor/VA

• Salt Lake-Utah/VA

• San Antonio-UTH/VA

• Indiana-University

• University of New Mexico

• RAI Workgroup

• Jason, Dan, Shipra

• Ricky Shinall

• Nader Massarweh

• Rupen Shah

• VQI workgroup

• Philip Goodney

• Matthew Mell

• Benjamin Brooke

• Larry Kraiss

• Team Hall/UPMC/VAPHS

• Ada Youk

• Andrew Bilderback

• Jacob Hodges

• Jeff Borrebach

• Mary K Wisniewski

• Tami Minnier

• Steve Shapiro

• Mark Wilson

• Joel Nelson

• Bob Arnold

• Johanna Bellon

• Dan Forman

• Kelly Allsup

• Jonas Johnson

• Stephen Esper

• Jenn Holder-Murray

• Team Arya/Stanford/ VA

Palo Alto/ VA Atlanta/

Emory

• Sebastian Perez

• Amber Trickey

• Rui Chen

• Kelly Blum

• Elizabeth George

• Kara Rothenberg

• Jordan Stern

• Arden Morris

• Mary Hawn

• Ronald Dalman

• Paula Tucker

• Luke Brewster

• Theodore Johnson

• Jason Hockenberry

• Team Johanning/

UNMC/NWICHS/

VISN 23

• Tom Lynch

• Kendra

Schmid

• Kaeli Samson

• Georgia Lyles

• Krishna

Chaitanya

• Karen Taylor

• Tom Edes

• Richard

Allman

• Scott Shreve

• Jahnigen

Scholars

• Health and

Aging Policy

Fellowship

Questions?

hallde@upmc.edu

….Including Cost (Reality Check)

2,530,000

802,872

Robinson TN, et al., Am J Surg. 2011;202(5):511-514.

RAI & Cost: Direct and Net Hospital Costs

Total cost odds ratio (with 95% CI, as demonstrated

by box plot) on logarithmic scale. Calculated as

odds of significantly frail patients costing greater

than the median cost for inpatient elective

operations stratified by service and Risk Analysis

Index with unfrail patients (not depicted) as the

reference value (*p < 0.05).

Univariate Analysis:

↑length of stay (0.8 v. 2.1 days)

↑ total cost ($6,934 v. $13,319)

↓ net hospital income ($5,447 v. $3,129)

Multivariate analysis:

↑ direct cost (OR 2.2)

↑ indirect cost (OR 1.9)

↑ total cost (OR 2.2)

↓ net income (OR 0.8)

(all p<0.001)

Wilkes JG, et al. J Am Coll Surg. 2019;228(6):861-870.

UPMC Charges normalized to 'Normal 30-36' total charges

Category

Robust

≤29

Normal 30

to 36

Frail 37 to

Very frail

≥45

Inpatient

Inpatient Surgical DRG

0.34 0.34 0.35 0.34

ER to Inpatient Surgical DRG

0.03 0.05 0.07 0.08

Inpatient Medical DRG, General, Specialist and Observation

0.06 0.11 0.14 0.19

Inpatient Rehabilitation

0.00 0.00 0.01 0.01

Inpatient Behavioral Health

0.00 0.00 0.00 0.00

Ambulance from Facility to Facility

0.00 0.00 0.00 0.00

Subtotal Inpatient Charges

0.43 0.51 0.57 0.62

Outpatient

Outpatient Surgery

0.04 0.07 0.05 0.06

Outpatient Hospital and Specialized Facility

0.09 0.28 0.21 0.24

Outpatient Office, PCP and Other

0.01 0.02 0.02 0.03

Therapy Service (Is this like Outpatient Rehab/PT?)

0.01 0.00 0.00 0.00

Outpatient Behavioral Health

0.00 0.00 0.00 0.00

ER Discharged to Home

0.01 0.01 0.01 0.01

Observation, from ER or Office

0.01 0.01 0.01 0.01

Subtotal Outpatient Charges

0.17 0.39 0.32 0.36

Post

Acute

Nursing, Skilled and General

0.01 0.02 0.05 0.07

Home Care

0.03 0.05 0.06 0.07

Subtotal Post Acute Charges

0.04 0.08 0.11 0.14

Other

Other (e.g., Lab, OB/GYN, Maternity, Urgent Care)

0.00 0.01 0.01 0.00

Shock Claims

0.04 0.01 0.05 0.04

Subtotal Other Charges

0.04 0.02 0.05 0.04

Total Charge 0.69 1.00 1.05 1.17

OK, but…it’s only for those (other) surgeons:

RAI, Operative Stress, Mortality and Specialty

George EL, et al., JAMA Surg. 2020:e205152. 10.1001/jamasurg.2020.5152

“But the RAI is too subjective….”

Do ”objective” biomarkers help?

0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88

RAI-A

RAI-Ar

RAI-Ar + BMI

RAI-Ar + Creatnine

RAI-Ar + Hematocrit

RAI-Ar + Albumin

RAI-Ar + Biomarkers

RAI-Ar + Blood Biomarkers

C-Statistic

Pandalai, et al, ACS Clinical Congress, 2020

Maybe, but is the juice worth the squeeze?

RAI RAI+Hematocrit

100

0 20 40 60 80

RAI-A-r Recalibrated Score

Observed Mortality Predicted Mortality

95% Confidence Interval

100

0 20 40 60 80

RAIA-r Recalibrated Score w/ Hct

Observed Mortality Predicted Mortality

95% CI

Pandalai, et al, ACS Clinical Congress, 2020

RAI Survey Implementation at UPMC

Discrimination

C= 0.815 (95% CI 0.788-0.842)

Calibration

95.6% of predicted deaths within

95% CI of observed deaths

Varley PR, et al Ann Surg. 2020 10.1097/SLA.0000000000003808.

Multi-Mode Frailty Assessment

• 2-step process

• Screen with RAI (30-seconds)

• Physical function measures for the potentially frail (RAI≥37)

• Grip Strength

• Gait Speed

• TUG

• MiniCog

• Additional History (medication, admission, etc)

Clinic Runway

Goal Clarification

Best Case Worst Case Scenario Planning

• Developed by and for surgeons for

preoperative conversations

• Presents a choice between two options.

• Uses story telling to describe what is likely

under the best, worst and most likely

scenarios.

• Sparks a conversation about patient goals,

values, fears and aspirations.

• Memorialized in a graphic aid. (Check out

the white board video)

• Requires substantial communication skills.

6 Steps for BC/WC

• Recognize that bad/serious news needs to be broken

• Create a visual aid

• Surgery vs. Non Op Management

• Treatment A vs. Treatment B

• Gets you clear in your own head

• Simplifies language

• Physical deposit for family

• Break bad news

• Tell stories about best, worst and most likely scenarios

• Why stories: Scenario Planning

• Elicit preferences: What is important to you now?

• Make a recommendation

UPMC Adaptation

• Gathering information on operative and non-operative options and

outcomes (e.g., “cat herding”

• Making a recommendation to the surgeon rather than the patient

Surgery

Worst Case:

Best Case:

Worst Case:

Best Case:

Most Likely:

• ICU 2-3 days

• Hospital 1-2

weeks

• Eat again

• Nursing home

• ICU with complications

• Much weaker than

before

• Nursing home, never go

home again

• Long surgery

• Die in ICU,

unable to

talk to family

Supportive

care

• Obstruction

resolves

• Back and forth to

hospital

• Nursing home

• Obstruction doesn’t

resolve

• Unable to eat

• Pain controlled

• Die in the hospital, time

to say goodbye

• Obstruction

doesn’t resolve

• Die in the

hospital

Older Woman with Cervical Spine Fracture

• Mrs. Goldstein is an 83 y/o woman with CAD with prior CABG, hypertension, COPD, CKD, and prior

CVA who tripped and fell down a flight of stairs at home. She was brought to the ED by ambulance

and found to have no feeling or movement in her legs and arms. Imaging reveals an unstable cervical

spine fracture at C5. Prior to her fall she had been in usual health. She denies angina, palpitations,

or syncope, and has some mild dyspnea on exertion. She lives with her husband, and was otherwise

independent in her ADL’s and IADL’s. Husband is 7 years her junior and spry.

• Exam: Awake, alert, fully oriented, no acute distress. No elevated JVP. Lungs clear without use of

accessory muscles. Cardiac rhythm regular, no murmurs. No leg edema. Neurologic exam consistent

with C5 quadriplegia.

• VS: T 36.4 BP 128/76 P 92 RR 20 O2 sat 98% on 2 lpm

• Labs: CBC normal, BMP with baseline creatinine of 2.5

• EKG: Sinus tachycardia, no ischemic changes

• CXR: clear, no acute infiltrates

With Surgery Without Surgery

Best Case

•

Long procedure in OR to stabilize cervical spine

• Post-op stay in ICU

• Evaluation by PM&R with transfer to the

inpatient spinal cord rehab program

• Eventual return home with adaptations to live

with quadriplegia

• Lives another 12-24 months with constant

assistance from husband and visiting nurses.

• Avoidance of surgical risk

• Focus on comfort managed by

hospice

• Family can remain near

• C-collar removed after 6 weeks,

except during transfers.

• Likely pulmonary complications

• Lives 4-12 months before

terminal pneumonia

Worst Case

• Surgical complications requiring one or more

additional surgeries

• Prolonged ICU stay

• Failure to wean from ventilator requiring

tracheostomy

• Post-op pneumonia

• Complications of quadriplegia including skin

breakdown, DVT

• Death in the ICU in 2-4 weeks

•

Pain requiring narcotics, possibly

sedating

• Phantom pain; spasms.

• Never goes home because

require inpatient hospice

• Early pneumonia or mucous

plugging

• Death in 1-3 weeks

Most Likely Case

• Prolonged but technically successful

stabilization surgery

• Multiple days in ICU

• Likely respiratory complications

• Prolonged hospital stay

• Extended rehab in a skilled setting

• Permanent placement in SNF

• Lives another 6-18 months

• Successful ability to control

symptoms of pain and shortness

of breath

• Fracture remains unstable

requiring C-collar most of the

time.

• Home hospice

• Survives 2-6 months.

Risk Analysis Index (RAI)—Initial Validation

• Administrative RAI (RAI-A)

• Computed from VASQIP/NSQIP variables

• Predicts 180 day mortality (C= 0.823)

• Clinical RAI (RAI-C)

• 14 Item survey instrument

• Administered by RN, APP or MD

• Linear scale from 0-81

• < 2 minutes to complete

• > 10,000 measurements from 2011-2014.

• Predicts 180 day mortality (C= 0.772)

• Correlation RAI-A:RAI-C=0.547

Hall, DE. et al. JAMA Surgery 152(2) doi:10.1001/jamasurg.2016.4202 (Nov 23).

RAI seems to work, but…..

• Scoring system never calibrated in surgical population

• Validation limited to a single VA hospital

• What about other VA hospitals

• What about non-VA hospitals

• What about women?

• Questions remain:

• Would the “objectivity” of biomarkers help?

• What procedures matter most (should everyone be screened)?

• Are some specialties exempt?

RAI-A Validation in Veteran Patients (VASQIP)

(N=480,731)

Original RAI-A (c=0.813) Revised RAI-A (c=0.842)

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

RAI-A

Predicted

Observed

C-statistic= 0.813 (0.810-0.817)

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

RAI-rev

Predicted

Observed

C-statistic= 0.842 (0.839-0.845)

Arya, S. et al. Annals of Surgery doi 10.1097/SLA.0000000000003276 (2019, March 23).

RAI-A Validation in Men and Women

ACS-NSQIP (N=1,391,785)

Men (c=0.845, N=584,698 ) Women (c=0.885, N=807,087)

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

RAI-rev

Predicted

Observed

20%

40%

60%

80%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

RAI-rev

Predicted

Observed

C-statistic= 0.885 (0.881-0.889)C-statistic= 0.845 (0.841- 0.85)

Arya, S. et al. Annals of Surgery doi 10.1097/SLA.0000000000003276 (2019, March 23).

The impact of frailty on mortality in non-surgical head

and neck cancer treatment: Shifting the paradigm.

• Objective: Compare survival

treated with surgical and

non-surgical management,

stratified by frailty using RAI.

• Cohort: 165 patients with

malignant disease & RAI

• 59 Major Surgery

• 62 Major Surgery + Adjuvant

• 54 Non-surgical therapy

Mady LJ, et al., Oral Oncol. Mar 2022;126:105766.

10.1016/j.oraloncology.2022.105766

• Methods: Multivariable cox

proportional hazard models

• RAI, stage, tumor site, tumor type

• Results: ↓Survival Non-Surgical

• Overall (N=165)

• Among the Frail (69)

• HR 2.5 (1.19,5.23) surgery

• HR 3.91 (1.94,7.89) multimodal

• ↓Survival with ↑Frailty

• Conclusions:

• Non-surgical management is

worse than surgical management

across all levels of frailty

• Challenge assumption of “too frail

for surgery”

• RCTs needed to clarify treatment

of frail patients

Mady LJ, et al., Oral Oncol. Mar 2022;126:105766.

10.1016/j.oraloncology.2022.105766

The impact of frailty on mortality in non-surgical head

and neck cancer treatment: Shifting the paradigm.

• Objective: Compare survival

treated with surgical and

non-surgical management,

stratified by frailty using RAI.

• Cohort: 165 patients with

malignant disease & RAI

• 59 Major Surgery

• 62 Major Surgery + Adjuvant

• 54 Non-surgical therapy

Mady LJ, et al., Oral Oncol. Mar 2022;126:105766.

10.1016/j.oraloncology.2022.105766

• Methods: Multivariable cox

proportional hazard models

• RAI, stage, tumor site, tumor type

• Results: ↓Survival Non-Surgical

• Overall (N=165)

• Among the Frail (69)

• HR 2.5 (1.19,5.23) surgery

• HR 3.91 (1.94,7.89) multimodal

• ↓Survival with ↑Frailty

• Conclusions:

• Non-surgical management is

worse than surgical management

across all levels of frailty

• Challenge assumption of “too frail

for surgery”

• RCTs needed to clarify treatment

of frail patients

Mady LJ, et al., Oral Oncol. Mar 2022;126:105766.

10.1016/j.oraloncology.2022.105766

• Methods: Univariate and multivariable linear and logistic

regression

• RAI, stage, tumor site, tumor type

• Results:

• Univariate ↑Flourishing with age, normal diet, employment, & income

• ↓Flourishing associated with

• ↑Depression

• ↑Anxiety

• ↑Swallowing Dysfunction

• ↑Neck Disability

• ↑Insomnia

• Results (continued)

• Interesting and

meaningful patterns in

sub-domains of flourishing

• Conclusions:

• Common late-term side

effects of HNC treatment

associated with

↓Flourishing

• Further data of this kind

may inform treatment

decisions consistent with

patients’ goals

Harris A, Cancer Med. Mar 11 2022;10.1002/cam4.4636

That’s our story, and then there is peer review:

George EL, Hall DE, Youk A, Chen R, Kashikar A, Trickey AW, Varley PR, Shireman PK, Shinall MC,

Massarweh NN, Johanning JM, Arya S. Patient frailty and postoperative mortality after

noncardiac surgery—does specialty matter? JAMA Surgery (in press).

Agarwal N, Goldschmidt E, Taylor T, Souvik R, Altieri-Dunn S, Bilderback AL, Friedlander RM,

Kanter AS, Okonkwo DO, Gerszten PC, Hamilton DK, Hall DE. Impact of Frailty on

Outcomes following Spine Surgery: A Prospective Cohort Analysis of 668 Patients.

Neurosurgery (in press).

Shinall, M. C., Jr, Youk, A., Massarweh, N. N., Shireman, P. K., Arya, S., George, E. L., & Hall, DE.

(2020). Association of Preoperative Frailty and Operative Stress With Mortality After

Elective vs Emergency Surgery. JAMA Network Open, 3(7), e2010358-e2010358.

doi:10.1001/jamanetworkopen.2020.10358

Shah R, Borrebach JD, Hodges JC, Varley PR, Wisniewski MK, Shinall MC, Jr., Arya S, Johnson J,

Nelson JB, Youk A, Massarweh NN, Johanning JM, Hall DE. Validation of the Risk Analysis

Index for Evaluating Frailty in Ambulatory Patients. J Am Geriatr Soc. 2020;68(8):1818-

1824.

Varley PR, Borrebach JD, Arya S, Massarweh NN, Bilderback AL, Wisniewski MK, Nelson JB,

Johnson JT, Johanning JM, Hall DE. Clinical Utility of the Risk Analysis Index as a

Prospective Frailty Screening Tool within a Multi-practice, Multi-hospital Integrated

Healthcare System. Ann Surg. 10.1097/SLA.0000000000003808. 2020. PMID: 32118596.

Rothenberg KA, George EL, Trickey AW, Barreto NB, Johnson TM, 2nd, Hall DE, Johanning JM,

Arya S. Assessment of the Risk Analysis Index for Prediction of Mortality, Major

Complications, and Length of Stay in Patients who Underwent Vascular Surgery. Ann Vasc

Surg. 10.1016/j.avsg.2020.01.015. 2020. PMID: 31935435.

Li Z, Habbous S, Thain J, Hall DE, Nagpal D, Bagur R, Kiaii B, John-Baptiste A. Cost-effectiveness

analysis of frailty assessment in older patients undergoing coronary artery bypass grafting

(CABG) surgery. Canadian Journal of Cardiology.

https://doi.org/10.1016/j.cjca.2019.09.025. 2020;36(4):490-499. PMID: 32220386.

Shinall MC, Jr, Arya S, Youk A, Varley P, Shah R, Massarweh NN, Shireman PK, Johanning JM,

Brown AJ, Christie NA, Crist L, Curtin CM, Drolet BC, Dhupar R, Griffin J, Ibinson JW,

Johnson JT, Kinney S, LaGrange C, Langerman A, Loyd GE, Mady LJ, Mott MP, Patri M,

Siebler JC, Stimson CJ, Thorell WE, Vincent SA, Hall DE. Association of Preoperative Patient

Frailty and Operative Stress With Postoperative Mortality. JAMA Surgery. 2019: e194620-

e194620. 10.1001/jamasurg.2019.4620. PMID: 31721994 PMCID: PMC6865246

Rothenberg KA, Stern JR, George EL, Trickey AW, Morris AM, Hall DE, Johanning JM, Hawn MT,

Arya S (2019). Association of Frailty and Postoperative Complications on Unplanned

Readmissions Following Elective Outpatient Surgery. JAMA Network Open, 2(5)

doi:10.1001/jamanetworkopen.2019.4330. PMID: 31125103. PMCID: PMC6632151

Arya, S., Varley, P., Youk, A., Borrebach, J. A., Perez, S., Massarweh, N. N., . . . Hall, DE. (2019).

Recalibration and External Validation of the Risk Analysis Index: A Surgical Frailty

Assessment Tool. Ann Surg. doi:10.1097/SLA.0000000000003276. PMID: 30907757.

Shah, R., Attwood, K., Arya, S., Hall, DE, Johanning, J. M., Gabriel, E., . . . Massarweh, NN. (2018).

Association of Frailty With Failure to Rescue After Low-Risk and High-Risk Inpatient

Surgery. JAMA Surg, 153(5), e180214. doi:10.1001/jamasurg.2018.0214. PMID: 29562073

Hall, DE., Arya, S., Schmid, K. K., Carlson, M. A., Lavedan, P., Bailey, T. L., Purviance, G.,

Bockman, T, Lynch, T.G., Johanning, J. M. Association of a Frailty Screening Initiative with

post-operative survival at 30, 180 and 365 days. JAMA Surgery. doi:

10.1001/jamasurg.2016.4219 Epub Nov 30, 2016. PMID: 27902826, PMCID: PMC7180387.

Hall, DE, Arya, S., Schmid, K. K., Blaser, C., Carlson, M. A., Bailey, T. L. Purviance, G., Bockman, T.,

Lynch, T.G., Johanning, J. M. Development and initial validation of the Risk Analysis Index

(RAI) for measuring frailty in surgical populations. JAMA Surg 2017 152(2):175-182 doi:

10.1001/jamasurg.2016.4202 Epub Nov 23, 2016. PMID: 27893030. PMCID: PMC7140150.

Ernst K, Hall DE, Lynch TJ, Schmid, KK, Seevers, G, Lavedan, P, Lynch, T, Johanning J. Surgical

palliative care consultations over time in relation to system wide frailty-screening. JAMA

Surgery. 2014: 149(11):1121-1126 DOI: 10.1001/jamasurg.2014.1393. PMID: 25207603.

PMCID: PMC4603652.

Codman, Hawthorne, and End Results of a Watched System.Johanning JM, Arya S.JAMA Surg.

2016 Dec 1;151(12):1165. doi: 10.1001/jamasurg.2016.2901.PMID: 27653007

Preoperative Frailty Increases Risk of Nonhome Discharge after Elective Vascular Surgery in

Home-Dwelling Patients.Arya S, Long CA, Brahmbhatt R, Shafii S, Brewster LP,

Veeraswamy R, Johnson TM 2nd, Johanning JM.Ann Vasc Surg. 2016 Aug;35:19-29. doi:

10.1016/j.avsg.2016.01.052. Epub 2016 Jun 2.PMID: 27263810

Frailty and Mortality After Noncardiac Surgery in Elderly Individuals: Metrics, Systems, and the

Elephant.Johanning JM, Hall D, Arya S.JAMA Surg. 2016 Jun 1;151(6):545-6. doi:

10.1001/jamasurg.2015.5235.PMID: 26790639

A frailty index identifies patients at high risk of mortality after tracheostomy.Johnson MS, Bailey

TL, Schmid KK, Lydiatt WM, Johanning JM.Otolaryngol Head Neck Surg. 2014

Apr;150(4):568-73. doi: 10.1177/0194599813519749. Epub 2014 Jan 16.PMID: 24436464

Preoperative frailty predicts postoperative complications and mortality in urology

patients.Isharwal S, Johanning JM, Dwyer JG, Schimid KK, LaGrange CA.World J Urol. 2017

Jan;35(1):21-26. doi: 10.1007/s00345-016-1845-z. Epub 2016 May 12.PMID: 27172940

Preoperative frailty Risk Analysis Index to stratify patients undergoing carotid

endarterectomy.Melin AA, Schmid KK, Lynch TG, Pipinos II, Kappes S, Longo GM, Gupta PK,

Johanning JM.J Vasc Surg. 2015 Mar;61(3):683-9. doi: 10.1016/j.jvs.2014.10.009. Epub

2014 Dec 9.PMID: 25499711

Risk Prediction Tools to Improve Patient Selection for Carotid Endarterectomy Among Patients

With Asymptomatic Carotid Stenosis.Keyhani S, Madden E, Cheng EM, Bravata DM, Halm

E, Austin PC, Ghasemiesfe M, Abraham AS, Zhang AJ, Johanning JM.JAMA Surg. 2019 Apr

1;154(4):336-344. doi: 10.1001/jamasurg.2018.5119.

Summary of the panel session at the 38th Annual Surgical Symposium of the Association of VA

Surgeons: what is the big deal about frailty?Anaya DA, Johanning J, Spector SA, Katlic MR,

Perrino AC, Feinleib J, Rosenthal RA.JAMA Surg. 2014 Nov;149(11):1191-7. doi:

10.1001/jamasurg.2014.2064.PMID: 25230137

Assessing preoperative frailty utilizing validated geriatric mortality calculators and their

association with postoperative hip fracture mortality risk.Dwyer JG, Reynoso JF, Seevers

GA, Schmid KK, Muralidhar P, Konigsberg B, Lynch TG, Johanning JM.Geriatr Orthop Surg

Rehabil. 2014 Sep;5(3):109-15. doi: 10.1177/2151458514537272.PMID: 25360340

The CPRS RAI

Reminder

Dialogue

Template

National

Release

9/13/21

Cerner Next

Complimentary Initiatives

SAGE QUERI

Safer Ageing through

Geriatric-informed

Evidence-based practices

•Pittsburgh, PA

•Philadelphia, PA

•Lebanon, PA

•Wilkes-Barre, PA

•Wilmington, DE

PAUSE Trial

HSR&D IIR RCT

Frailty Screening followed

by Multidisciplinary Clinic

•Palo Alto, CA

•Houston, TX

•Nashville, TN

HSR&D IIR

Improving Surgical

Decision-Making by

Measuring and Predicting

Long-Term Loss of

Independence after

Surgery

GECDAC Partner

Residential History File