D139. Development of an AI-Based Predictive Model for Septic Wrist and a Risk Assessment Tool

Abstract

PURPOSE: This study aimed to create a septic wrist AI prediction model and develop a score-based risk assessment tool. METHODS: An IRB-approved retrospective review was conducted on patients with a presumed septic wrist diagnosis (2003-2022). Kruskal Wallis algorithm was employed to identify predictors of septic wrist based on comorbidities, penetrating trauma, fever, multi-joint involvement, inflammatory markers (ESR/CRP/WBC), serum uric acid, blood cultures, imaging, and synovial fluid analysis. Subsequently, Naïve Bayes classifier was utilized to populate a prediction model. An independent score-based risk assessment tool was developed using multivariate analyses with each predictor receiving a risk score of 1. RESULTS: 205 (70 females, 135 males) patients were included with a median length of hospitalization of 6[8] days and follow-up 1[3] months. 95 (46.3%) patients had septic wrist confirmed with Gram stain/culture, 79 (38.5%) patients received alternative diagnoses, and 31 (15.1%) patients had undetermined diagnoses. The optimized AI prediction model included 7 predictors (no synovial crystals, positive blood culture, multi-joint involvement, age, prior septic arthritis, IVDU, and penetrating trauma), demonstrating 89.5% sensitivity, 80% specificity, and 0.89 AUC. The scoring tool included 4 predictors (no synovial crystals, positive blood culture, no prior crystalline arthropathy, and multi-joint involvement) with a risk score ranging from 0-4. Classifying septic wrist at a score ≥2 yielded a sensitivity of 64%, specificity of 84%, and AUC 0.78. CONCLUSION: Both the AI model and scoring tool offer potential value in diagnosing septic wrist, with key predictors including multi-joint involvement, absence of synovial crystals, and positive blood culture.

Ähnliche Arbeiten