In the artificial intelligence age, is Oswestry spinal risk index still useful? an external validation of Oswestry spinal risk index applied to radiotherapy patients and its relevance.

BACKGROUND AND STUDY AIMS: In 2013, Balain et al. proposed a scoring system, the Oswestry Spinal Risk Index (OSRI), to predict survival in patients with spinal metastases (SM) and hence aid treatment decisions. This study aims to externally validate the OSRI for patients treated exclusively with radiotherapy, and assess its relevance in emergency decision making in the age of nomograms and machine learning. MATERIALS AND METHODS: Data from 150 patients with spinal metastases (SM) treated at our Regional Cancer Centre over a four-year period were retrospectively analysed. OSRI scores were calculated for each patient, and actual survival duration from diagnosis was determined to enable comparison with predicted survival. The cohort, selected over a decade ago, ensured complete survival data for all included patients. RESULTS: Spearman's rank correlation demonstrated a strong positive relationship between actual and OSRI-predicted survival (r = 0.706, p < 0.001). The Mantel-Cox log-rank test revealed statistically significant differences in survival curves across OSRI risk groups (p < 0.001), with lower OSRI scores associated with longer survival. Paired sample t-tests further confirmed this correlation (one-sided p = 0.009; two-sided p = 0.017). Kaplan-Meier survival plots were similar to those of the original study. CONCLUSIONS: This study is the first to externally validate the Oswestry Spinal Risk Index (OSRI) specifically in patients treated exclusively with radiotherapy. Unlike previous validations, which included patients managed surgically or with combined modalities, our cohort focused solely on radiotherapy. We found a strong correlation between predicted and actual survival, highlighting the OSRI's simplicity, accuracy and its practical value for rapid decision-making in emergency settings.