LAUSR

OBJECTIVES To determine factors predictive of postoperative surgical site infection (SSI) after fracture fixation and create a prediction score for risk of infection at time of initial treatment. DESIGN Retrospective cohort study. SETTING Level I trauma center. PATIENTS/PARTICIPANTS Study group, 311 patients with deep SSI; control group, 608 patients. INTERVENTION We evaluated 27 factors theorized to be associated with postoperative infection. Bivariate and multiple logistic regression analyses were used to build a prediction model. A composite score reflecting risk of SSI was then created. MAIN OUTCOME MEASURES Risk of postoperative infection score. RESULTS The final model consisted of 8 independent predictors: 1) male sex, 2) obesity (body mass index ≥30) 3) diabetes, 4) alcohol abuse, 5) fracture region, 6) Gustilo-Anderson type III open fracture, 7) methicillin-resistant Staphylococcus aureus nasal swab testing (not tested or positive result), and 8) American Society of Anesthesiologists classification. Risk strata were well correlated with observed proportion of SSI and resulted in a percent risk of infection of 1% for ≤3 points, 5% for 4 points, 6% for 5 points, 10% for 6 to 8 points, 17% for 9 points, and 41% for ≥10 points. CONCLUSION The proposed postoperative infection prediction model seems to be able to determine which patients have fractures at higher risk of infection and provides an estimate of the percent risk of infection before fixation. LEVEL OF EVIDENCE Diagnostic level III.