Background: The accuracy rates of several effective serological markers of surgical site infection following spine surgery are unclear. We aimed to verify the accuracy of each significant marker and identify the most suitable and effective combination of these markers for the diagnosis of surgical site infection following spine surgery.
Methods: This retrospective study enrolled 329 patients who underwent spine surgery for causes other than infectious spondylitis, including 9 patients with surgical site infection. Complete blood cell count, differential counts, and C-reactive protein levels were measured preoperatively and postoperatively (days 2 and 7). Serological data were compared among non-surgical site infection and surgical site infection cases. Cutoff values for items presenting significant differences were determined using receiver operating characteristic curves. Ratios in each serological factor at each time-point were compared. Combinations of these factors on postoperative day 7 and ratio items were investigated to determine the most suitable combination comprising the least number of items.
Results: Significant differences were noted among four factors on postoperative day 7, except for the lymphocyte count. For the ratio items, significant differences were observed among 6 items. The combination of these ten markers was examined; each factor was assigned 1 point. The most suitable combination comprising 4 items, including neutrophil count, neutrophil-to-lymphocyte ratio, lymphocyte count ratio, and C-reactive protein ratio, presented an AUC of 0.95, with a cutoff value, sensitivity, and specificity of 3 points, 0.89, and 0.92, respectively.
Conclusion: The combination of four markers is the most suitable criterion for the surgical site infection scoring system following spine surgery, where scores of ≥3 points strongly indicate surgical site infection. This criterion may be a strong tool for detecting surgical site infection.
Copyright © 2021. Published by Elsevier B.V.