Preoperative Predictors of Spinal Infection within the National Surgical Quality Inpatient Database.

Abstract

Surgical-site infections (SSIs) are a major cause of morbidity and mortality, increasing the length and cost of hospitalization. In patients undergoing spine surgery, there are limited large-scale data on patient-specific risk factors for SSIs. The American College of Surgeons National Surgical Quality Improvement Program database was reviewed for all spinal operations between 2006 and 2012. The rates of 30-day SSIs were calculated, and univariate analysis of selected preoperative risk factors was performed. Multivariate analysis was then used to identify independent predictors of SSIs. A total of 1110 of the 60,179 patients (1.84%) had a postoperative wound infection. There were 527 (0.87%) deep and 590 (0.98%) superficial infections. Patients with infections had greater rates of sepsis, longer lengths of stay, and more return visits to the operating room. Independent predictors of infection were female sex, inpatient status, insulin-dependent diabetes, preoperative steroid use greater than 10 days, hematocrit less than 35, body mass index greater than 30, wound class, American Society of Anesthesiologists class, and operative duration. Analysis of a large national patient database revealed many independent risk factors for SSIs after spinal surgery. Some of these risk factors can be modified preoperatively to reduce the risk of postoperative infection.