An Independent Validation of the Brighton Spondylodiscitis Score and a Proposal to Modify the Score

INTRODUCTION: The Brighton Spondylodiscitis Score (BSS) aims to identify patients with pyogenic spinal infections (PSIs) requiring surgery; an independent assessment of the BSS is required. METHODS: We evaluated 60 patients with PSIs. Using the BSS, we determined whether patients with low, moderate, and high risk (LMHR) had different rates of surgery. We proposed a modified score (MS) using a logistic regression (LOGR). Applying the MS, we determined whether patients with LMHR exhibited different rates of surgery. Another LOGR determined the association of the BSS and the MS with surgery. A C-statistic using the BSS and the MS was generated. RESULTS: We studied 60 patients (mean age = 63 years); 37 (62%) were men; 30 (50%) required surgery. Using the BSS, patients with LMHR had similar rates of surgery (P = 0.53). LOGR showed that cervical PSIs had a larger chance of surgery (odds ratio [OR] = 7.3 [1.1 to 51.3]) than other locations. Using the MS, patients with moderate- and high-risk were operated more frequently than low-risk patients (P = 0.04). The BSS did not predict surgery (OR = 1.07; P = 0.31), but the MS did (OR = 1.16; P = 0.02). The C-statistic using the BSS (0.59) improved using the MS (0.69), P = 0.03. DISCUSSION: The discriminatory capacity to predict surgery of the BSS augmented using the MS. LEVEL OF EVIDENCE: II (Diagnostic study: Transverse study).