Table 3. Independent Predictors of Post-Operative Adverse Outcomes: Analysis of 150 patients with methicillin-resistant Staphylococcus aureus (MRSA) surgical site infections (SSI) compared with 231 uninfected controls to determine the independent effect of SSI due to MRSA on outcomes of surgical patients.

Independent Predictor	Odds Ratio [95% Confidence Interval]
Readmission within 90 days of surgical procedure a
SSI due to MRSA	35.0 [17.3–70.7]
Need assistance with ≥3 ADLs	4.28 [1.52–12.0]
Death within 90 days of surgical procedure b
SSI due to MRSA	7.27 [2.83–18.7]
Need assistance with ≥3 ADLs	6.73 [2.80–16.2]
Age≥65	4.45 [1.41–14.0]
Orthopedic procedure	0.27 [0.10–0.71]
Increased length of hospitalization during 90 days following surgical procedure c
SSI due to MRSA	4.36 [3.31–5.75]
Procedure at tertiary care hospital	1.41 [1.30–1.54]
Need assistance with ≥3 ADLs	1.35 [1.25–1.46]
Post-operative serum glucose >200 mg/dL	1.18 [1.15–1.22]
Orthopedic procedure	0.68 [0.62–0.75]
Hospital charges during 90 days following surgical procedure d
SSI due to MRSA	4.44 [2.68–7.34]
Procedure at tertiary care hospital	2.97 [2.23–3.95]
Coronary artery bypass graft procedure	1.34 [1.26–1.43]
Surgical duration >75th NNIS percentile	1.27 [1.22–1.32]
Procedure on same day as admission	0.75 [0.72–0.79]

^aPatients who died during the index admission (n = 23) were excluded from this analysis. Final model controlled for confounding effect of ASA score and contained term for interaction between SSI due to MRSA and Need assistance with ≥3 ADLs. Reference model also included the following variables: sex, history of congestive heart failure, history of cerebrovascular accident, McCabe score = 1, and surgery on same day as admission.

^bReference model also included the following variables: admitted from home, Charlson score ≥3, McCabe score = 1, wound class >2, ASA score ≥3, surgery on same day as admission, serum glucose >200 mg/dL, and repeat procedure at same operative site.

^cFinal model controlled for confounding effects of Caucasian race, McCabe score = 1, male sex, coronary artery bypass graft procedures, and surgical duration >75^th NNIS percentile and contained a term for the interaction of MRSA SSI and need assistance with 3 or more ADLs. Reference model also contained the following variables: BMI≥30, age≥65 years, admission from home, Charlson score ≥3, ASA score ≥3, and repeat procedure at same operative site.

^dFinal model controlled for confounding effects of ASA score ≥3 and contained an interaction term for the interaction between MRSA SSI and need assistance with 3 or more ADLs and an interaction term for the interaction between MRSA SSI and procedure at a tertiary care hospital. Reference model also contained the following variables: need assistance with 3 or more ADLs, receipt of immunosuppressive medications, McCabe score = 1, post-operative serum glucose >200 mg/dL, and orthopedic procedure.