Study design: This is a multivariate analysis of a prospectively collected database.
Objective: To determine preoperative, intraoperative, and patient characteristics that contribute to an increased risk of postoperative wound infection in patients undergoing spinal surgery.
Summary of background data: Current literature sites a postoperative infection rate of approximately 4%; however, few have completed multivariate analysis to determine factors which contribute to risk of infection.
Methods: Our study identified patients who underwent a spinal decompression and fusion between 1997 and 2006 from the Veterans Affairs' National Surgical Quality Improvement Program database. Multivariate logistic regression analysis was used to determine the effect of various preoperative variables on postoperative infection.
Results: Data on 24,774 patients were analyzed. Wound infection was present in 752 (3.04%) patients, 287 (1.16%) deep, and 468 (1.89%) superficial. Postoperative infection was associated with longer hospital stay (7.12 vs. 4.20 days), higher 30-day mortality (1.06% vs. 0.5%), higher complication rates (1.24% vs. 0.05%), and higher return to the operating room rates (37% vs. 2.45%). Multivariate logistic regression identified insulin dependent diabetes (odds ratios [OR] = 1.50), current smoking (OR = 1.19) ASA class of 3 (OR = 1.45) or 4 to 5 (OR = 1.66), weight loss (OR = 2.14), dependent functional status (1.36) preoperative HCT <36 (1.37), disseminated cancer (1.83), fusion (OR = 1.24) and an operative duration of 3 to 6 hours (OR = 1.33) or >6 hours (OR = 1.40) as statistically significant predictors of postoperative infection.
Conclusion: Using multivariate analysis of a large prospectively collected data from the National Surgical Quality Improvement Program database, we identified the most important risk factors for increased postoperative spinal wound infection. We have demonstrated the high mortality, morbidity, and hospitalization costs associated with postoperative spinal wound infections. The information provided should help alert clinicians to presence of these risks factors and the likelihood of higher postoperative infections and morbidity in spinal surgery patients.