The Role of Anesthesia Providers in
Perioperative SSI Prevention & QI
Dustin Long MD
Associate Professor
Dept. Anesthesiology & Pain Medicine
University of Washington School of Medicine
ASPIRE Collaborative Meeting
AGENDA
>Top 5: Why surgical site infection (SSI) remains a key
outcome for patients and health systems in 2026
>Top 5: Why this is relevant to anesthesia practice
>Ideas about how MPOG/ASPIRE can drive improvements in
the understanding and prevention of SSI
>Practical examples of anesthesiology-led QI in SSI at UW
Disclosures
>NIH and foundation funding, no FCOI
>Presentation includes unpublished data and one trade
name
What is Surgical Site Infection (SSI)?
Infections of a surgical wound that develop after a planned procedure
with standard antiseptic protocols in the absence of prior infection
Why Surgical Site Infection (SSI)?
American College of Surgeons NSQIP 2019 Participant Use Data File (PMID: 35666980)
#1 It’s common
SSI is the most frequent
complication of surgery
other than bleeding
Why Surgical Site Infection (SSI)?
#1 It’s common
1 in 30
patients
Why Surgical Site Infection (SSI)?
Most costly form of
healthcare-associated
infection
PMID: 23999949
#2 It’s expensive
Why Surgical Site Infection (SSI)?
Most common reason for
hospital readmission
after surgery
PMIDs: 25647204, 33848832
#2 It’s expensive
Why Surgical Site Infection (SSI)?
Survival
#3
Healthcare
Satisfaction
Function Pain
Major impact on patient-important outcomes
Why Surgical Site Infection (SSI)?
#4 Heavily weighted in hospital quality
& reimbursement programs
Why Surgical Site Infection (SSI)?
Current AHRQ National Scorecard on Hospital-Acquired Conditions
#5 Not getting better
PMID: 42456137
Why Surgical Site Infection (SSI)?
https://doh.wa.gov/public-health-provider-resources/healthcare-professions-and-facilities/healthcare-associated-
infections/nhsn-data-and-reporting/hai-annual-report
#5 Not getting better
Why Surgical Site Infection (SSI)?
#5 Not getting
better
Increasing resistance
to surgical antibiotic
prophylaxis
PMID: 26482597
Why Surgical Site Infection (SSI)?
Recap: Common problem with an out-sized
impact on patient outcomes, hospital capacity,
and healthcare economics that is not getting
better and probably getting worse
Why is this relevant to anesthesiologists?
#1 Surgical antibiotic prophylaxis
• Getting it right (or wrong) has a major impact on risk
• ~1/5
th of all inpatient antibiotic exposures
• Huge piece of the antimicrobial stewardship challenge for yo
health system
Why is this relevant to anesthesiologists?
#2 Need for better preoperative optimization
- Most effective infection prevention measures are things that take
place before surgery
Why is this relevant to anesthesiologists?
#2 Need for better preoperative optimization
PMID: 31990733
ASAHQ.org
Why is this relevant to anesthesiologists?
#2 Need for better preoperative optimization
- Heavy focus on cardiopulmonary risk, pain, frailty, anemia…
Why not also infection prevention?
ASAHQ.org
Why is this relevant to anesthesiologists?
#3 We are part of the problem
- OR hand hygiene
- anesthesia workspace cleanliness
- sterile handling of CVCs and IV setups
- contact precautions
- temperature and glucose management
Why is this relevant to anesthesiologists?
#4 Anesthesiology documentation impacts
hospital risk adjustment for SSI
- Current model: ASA score heavily weighted metric
(no adjustment for emergency ‘E’ designation)
- Impacts entire DENOMINATOR
- Upcoming model
2: ASA score + emergency designation still more
heavily weighted than wound contamination status
2https://www.cdc.gov/nhsn/2022rebaseline/tables/ssi-table-1.html
Why is this relevant to anesthesiologists?
#5 The field has a data problem
- Quality of data available to drive research/QI poor relative to
the magnitude of the issue
- You are the experts in big data when it comes to perioperative
outcomes
Where I am Coming From:
Our Core Research Questions
• Where do the bacteria causing surgical
infections come from?
• How do they evade our current prevention
strategies?
• What can we do better?
Why study SSI in spine surgery ?
>High volume
–US ~1M/year
–More $ expenditure
than any other
elective surgery
>High risk
–1 in 30
>Major consequences
Causes of SSI by Operative Level
(PMID: 32796459)
The Backside Microbiome
(PMID: 38598612)
Anatomic Stratification of Colonizing Skin Bacteria
(16S rRNA PCR amplicon sequencing)
Preoperative Skin Microbiome Enrichment
Relative Frequency in Surgical Site Infection
(PMID: 38598612)
Why Does this Matter?
Moving Beyond Association:
Understanding Individual Cases of SSI
Surgery is a
“planned”
head-on
collision with
infection
Moving Beyond Association:
Understanding Individual Cases of SSI
Can we capture the
patient microbiome
before surgery, and
when infection happens,
use that to reconstruct
“what went wrong”?
80 -90% of SSIs
in spine surgery
arise from the
patient
microbiome
Moving Beyond Association:
Understanding Individual Cases of SSI
(PMID: 38598612)
(not the hospital
environment)
A BC
1
10
100
1000
1 2 5 10 25 50 100 250 500 1,000 2,500 10,000 50,000
Sam e−Species SSI Isolate Genetic Distance (Wh ole−Genome Varia nts)
Count
Species
inter−patient: C. koseri
inter−patient: C. striatum
inter−patient: E. coli
inter−patient: E. faecalis
inter−patient: K. pneumoniae
inter−patient: P. aeruginosa
inter−patient: P. mirabilis
inter−patient: S. aureus
inter−patient: S. capitis
inter−patient: S. epidermid is
inter−patient: S. lugdunensis
intra−patient: E. coli
intra−patient: K. pneumoniae
intra−patient: P. aeruginosa
intra−patient: P. mi rabilis
intra−patient: S. aureus
Moving Beyond Association:
Understanding Individual Cases of SSI
(PMID: 38598612)
50-6 0 % are
resistant to the
surgical
prophylaxis
administered
Moving Beyond Association:
Understanding Individual Cases of SSI
(PMID: 38598612)
resistance genes
also come from the
patient and are
detectable prior
to surgery
Moving Beyond Association:
Understanding Individual Cases of SSI
(PMID: 38598612)
Moving Beyond Association:
Understanding Individual Cases of SSI
microbiome
impacts risk even
after adjustment
for host/procedure
factors
Not S. aureus
colonized
Colonized with low-risk
S. aureus strain
Colonized with high-risk
S. aureus strain
• Most strains causing SSI in spine surgery arise from th
patient microbiome (not the hospital environment)
• Resistance to standard prophylactic antibiotics and
immune defenses is common and often foreseeable
Research Recap
Instead of selecting antibiotics based on
anatomic location of the surgical procedure…
…should they be further personalized based on the
individual preoperative microbiome of the surgical site?
Why Does this Matter?
Why does this matter?
If a patient is colonized with a strain of S. aureus that
puts them at a 1 in 4 risk of a devastating surgical
complication, shouldn’t we do everything possible
to decolonize them prior to proceeding with
elective surgery?
Why does this matter?
The fact that some of these infections are explainable
motivates me to think real improvement is possible
Data Problems: Missing Outcomes
It’s hard to improve outcomes or to research outcomes
when the outcomes data are missing or wrong
Especially challenging for healthcare-associated infection:
-multiple data sources over long surveillance window
-mostly free text
-judgement involved
-hard task even for trained humans
Data Problems: Missing Outcomes
ICD-10 codes are notoriously unreliable for healthcare-
associated infection
-under-coding
-non-specific attribution of infections
NSQIP and NHSN (CDC/CMS) rely on human chart review
-expensive ($$$, time, opportunity costs)
-can only review a small % of charts
-imperfect (bias and coding errors persist)
Data Problems:
Missing Outcomes
How about large language models?
OpenAI/Anthropic models via secure
UW API for spine/crani SSI surveillance
-20 cents per case
-Working on operationalizing
-Future: NSQIP-style outcomes for all?
(unpublished)
-Small subset of cases
-Not available for research
-Not linke d to pe riope ra tive or othe r clinica l da ta
-Have info on SSI occurrence, but no microbiology data
-Raw microbiology results captured as element of lab tables
-Linke d to m e a ningfu l clinica l/ de m ogra ph ic da ta , a n tibiotic proph y la xis
-Wide ge ogra ph ic dis tribution
Data Problems: Missing Labs
Why are microbiology data different?
One order…
-culture type
-specimen source
-sampling comments
…multiple bacteria
-Staphylococcus aureus
-Pseudomonas aeruginosa
…multiple antibiotics
-cefazolin
-clindamycin
-levofloxacin
-vancomycin
-cefepime
-meropenem
-piperacillin-tazobactam
Data Problems: Missing Labs
Standard MPOG
data exchange
Backfill historic data from prior
reporting periods
Update application suite
Test data query across sites
Create micro lab concept in
MPOG data dictionary
Develop data specification
Push data to MPOG Central
Perform proof-of-concept analyses
Future Opportunities: ASPIRE Metrics
Further focus focus on antibiotic choice?
Antibiotic timing:
-We rarely get this wrong (3%)
-Pre-incision adherence nearly universal
-Time within 60-minute window not important
-Wrong antibiotic redosing: ↑12% risk
Antibiotic choice:
-We often get this wrong (13%)
-Larger impact on risk: ↑40-80%
Future Opportunities: ASPIRE Metrics
-Varies by procedure
-Institutional protocol variation
-Affected by individual patient factors
-Allergies
-MRSA colonization status
-Therapeutic antibiotics
Why don’t we already have more metrics for this?
It’s more complicated….
…but this is where
anesthesiologists excel!
Future Opportunities: ASPIRE Metrics
There are common modes of failure that can be captured with rule-based logic
-Appropriate provision of cefazolin with PCN allergies
-Anaerobic coverage (e.g., metronidazole) for qualifying abdominal/GYN cases
-Completion and timeliness of preop MRSA screening
-MRSA agent for MRSA+ cases in qualifying cases
-Avoidance of MRSA agent for MRSA- patients in qualifying cases
Future Opportunities: MPOG Research
Practice variation is evidence of the need for more research in this area
Future Opportunities: MPOG Research
1. Observational studies of perioperative infection that leverage
explanatory laboratory data (microbiology results)
- MRSA screening and selective MRSA prophylaxis—standard of care?
2. Antimicrobial resistance reporting and forecasting
- How is resistance to surgical prophylaxis changing over time/geographically?
- How does this impact different demographic and procedural groups?
3. Pragmatic trials of surgical prophylaxis and preoperative decolonization?
- Vancomycin vs. daptomycin vs. vs. linezolid for MRSA coverage?
- Nozin vs. povidone-iodine vs. mupirocin?
Examples of Positive QI Programs
from UW Experience
• Greater anesthesiology and surgery ownership of OR
infection prevention
• Two dedicated anesthesiology resident didactic sess
on infection prevention
• Anesthesiologist review of all reportable case types
(documentation accuracy and QI opportunities)
• Discussion of infectious complications of surgery in
anesthesiology QA/M&M meetings
Examples of Positive QI Programs
from UW Experience
• Increase hand hygiene audits in OR and PACU spac
• Direct feedback to surgeons/anesthesiologists on
antibiotic prophylaxis choice
• Redesigned our programs for preoperative screening
and decolonization
>K23 Mentorship Team
–Stephen Salipante, MD, PhD
–Monica Vavilala, MD
–John Alverdy, MD (U. Chicago)
>Salipante Lab Members
–Adam Waalkes, BS
–Elizabeth Holmes, MSc
–Kelsi Penewit, BS
–Janessa Lewis, BS
>UW Dept. Anesthesiology & Pain Med.
–Burkhard Mackensen, MD, PhD
–Tonya Palermo, PhD
–Margaret Sedensky, MD
ACKNOWLEDGEMENTS
>HMC Anesthesiology
–Adrienne James
–Nathalie Bentov, MD
>Orthopaedic Collaborators
–Celeste Tavolaro, MD
–Jason Hsu, MD
>PPiQSO
–Vikas O’Reilly-Shah, MD, PhD
–Christine Fong, MS
>Funding
–NIH/NIAMS K23
(K23AR080209)
–Society for Healthcare Epidemiology of America
(2019 Epi Competition Award)
THANK YOU!
drlong@uw.edu
