Instability in Spinal Infection: Evaluating the Predictive Value of the Spinal Instability Spondylodiscitis Score (SISS)

Retrospective observational cohort study. Evaluating the predictive value of the Spinal Instability Spondylodiscitis Score (SISS) for radiologic progression in conservatively treated patients and its association with functional outcomes in both treatment groups. The SISS was developed to assess spinal instability in spondylodiscitis and assist in treatment decisions. However, its ability to predict radiologic and functional outcomes has not been validated. A total of 176 patients diagnosed with primary pyogenic spondylodiscitis at two centers in Germany and the UK were included. Clinical data, imaging studies, and outcomes were retrospectively collected. SISS scores were calculated pretreatment based on MRI/CT imaging. Functional outcomes were measured using the modified Rankin Scale (mRS). Radiologic progression was assessed in conservatively treated patients. Predictive value analyses, correlation, and multivariate regression were conducted to assess the relationship between SISS and both radiologic and clinical outcomes. On the basis of SISS, patients were classified as stable (14.77%), potentially unstable (44.89%), and unstable (40.34%). In conservatively treated patients (n=54), the "SISS_unstable" category showed moderate predictive value for deformity progression (PPV and NPV both 77.8%, sensitivity 41.2%, specificity 94.5%, AUC=0.68). Vertebral destruction >50% was the strongest predictors of radiologic progression. At a mean follow-up of 121 days correlation between initial SISS scores and unfavorable clinical outcomes (mRS 4-6) was weak (r=0.201, P = 0.002). In multivariate analysis, age, neurological deficits at admission, and systemic inflammatory response syndrome (SIRS) were stronger predictors of poor outcomes, while SISS classification was not significant. Patients triaged as "unstable" were very likely to show radiologic progression, indicating a fair predictive value of the score for spinal deformity. However, instability itself did not correlate with functional outcomes. These findings suggest that instability alone may not capture the multifactorial nature of outcome in spinal infections, but must be interpreted with caution due to small subgroup sizes and the use of a global, less granular outcome measure. Future research should focus on integrating broader patient factors to support clinical decision-making in spondylodiscitis.