Complications Associated With Surgical Antibiotic Prophylaxis
Practice Patterns
Amit Bardia MD,MPH
Program Director : ACTA Fellowship
DACCPM
Massachusetts General Hospital
Investigational team: Hung- Mo Lin, Xiwen Zhao, George Michel, Clark Fisher,
Douglas Colquhoun, Michael Mathis, Robert Schonberger
Disclosures:
The presenter served as a consultant with Takeda Pharmaceuticals.
The activity has no relation to the present work.
This work is directly funded by the Agency of Healthcare Research and Quality
(AHRQ) R01 funding 5R01HS029172-03.
Background
SSI are the leading cause of health care infections in
surgical patients
1
.
Estimated incidence
3
: 2-5%
Accounts for about 1.6 billion in health care related costs
2
.
1. Anderson DJ et al. Infect Control Hosp Epidemiol. 2014;35(6):605-27.
3. de Lissovoy G et al. Am J Infect Control 2009;37:387-97.
2. Bratzler DW et al. Clin Infect Dis 2004;38:1706-15
Surgical Site Infections (SSI)
Study Design
Exclusion Criteria
• Preexisting infection
• Transplant Surgeries, Cardiac surgeries
• ASA 6
• Missing / Unknown antibiotic administration metrics
• Missing CPT
• Missing Weight
Inclusion Criteria
Patients ≥ 18 years of age who underwent non-cardiac procedures
involving a skin incision from 01/01/2014 to 08/31/2022.
Exposure Variable
Non-adherence to the following antibiotic administration
metrics:
• Timing
• Dosing
• Choice (based on CPTs)
• Re-dosing (if the case qualifies)
Primary Outcome:
NSQIP adjudicated surgical site infection:
Any superficial, deep tissue, or organ space infections as
recorded in the NSQIP registry within 30-days
postoperatively.
Substudy
Vancomycin as the sole antibiotic
- Dosing is mg/kg based
- Need for prolonged infusion
- Associated with higher incidence of SSIs
Primary Analyses
A hierarchical generalized linear mixed model using Poisson regression with
surgical patients nested within MPOG institutions was performed, aiming to
investigate possible associations between the non-adherence metrics and SSI.
Models involved adjusting for patient, provider, institutional, and surgical
characteristics.
Results
A total of 119,236 unique cases met the inclusion criteria.
The incidence of SSI was 4.4% in the cohort.
Non-adherence to guideline-based antibiotic was
common with non-adherence to at least one metric in
26.1% cases.
Non adherent choice, dosing, timing and redosing was
13.3%, 9.0%, 3.0% and 13.9% respectively.
Guideline Non-adherence
was significantly
associated
with a higher relative risk
of SSIs.
Multivariate Regression: Overall Non-adherence
Multivariate Regression: Individual Metrics
Non adherent
choice and redosing
were significantly
associated with a
higher relative risk
of SSIs.
Vancomycin Sub-study
A total of 5542 unique cases met the inclusion criteria.
Results:
The incidence of SSI was 5.6% in the cohort.
Non-adherence to guideline-based antibiotic
choice, dosing and timing was 705 (12.7%), 59 (1.1%)
and 197 (3.6%) respectively.
Vancomycin Sub-study
Non adherent
choice and timing
were associated
with a significantly
higher relative risk
of SSIs.
Association of Vancomycin timing and Surgical Site Infections.
Vancomycin only
Vancomycin paired
with another
antibiotic
Limitations:
Observational study design
MRSA status not known
Nutritional status not known
No data on local decontamination practices
Thank you
