SurvivalPOSNs

SurvivalPOSNs predicts individualized probability of synchronous distant metastasis and overall survival for patients with primary osseous spinal neoplasms (POSNs) using statistical prognostic models.

Key Features:

• Data Source: Development is based on the Surveillance, Epidemiology, and End Results (SEER) database (2004–2017) comprising 1,345 POSN patients, of whom 238 cases (17.70%) had synchronous distant metastasis at diagnosis.
• Predictive Models: Includes a diagnostic nomogram to estimate probability of distant metastasis and a prognostic nomogram to predict overall survival for POSN patients with synchronous DM.
• Risk Factor Analysis: Applies univariate and multivariate logistic regression to identify significant risk factors for distant metastasis, including grade, T stage, N stage, and histological type.
• Prognostic Factor Analysis: Uses Cox regression to determine independent prognostic factors for patients with synchronous DM, including age, surgery, and histological type.
• Validation Techniques: Assesses model performance and clinical utility using receiver operating characteristic (ROC) curves, calibration curves, decision curve analysis (DCA), and Kaplan–Meier survival analysis.

Scientific Applications:

• Clinical decision support: Provides quantified probabilities of DM development and individualized survival estimates to inform treatment planning and prognosis for POSN patients.
• Research utility: Supplies a validated statistical framework for analyzing prognostic factors and outcomes in studies of POSNs with synchronous distant metastasis.

Methodology:

Univariate and multivariate logistic regression for DM risk factor identification, Cox regression for prognostic factor identification, construction of diagnostic and prognostic nomograms, and validation using ROC curves, calibration curves, decision curve analysis, and Kaplan–Meier survival analysis.