A Simple Scoring Algorithm Predicting Vascular Infections in Adults With Nontyphoid Salmonella Bacteremia

Abstract

Nontyphoid Salmonella (NTS) can cause fatal vascular infections. This study aims to establish a predictive scoring algorithm to identify adults aged ≥ 50 years with NTS bacteremia who are at risk for vascular infections. There were 358 adults aged ≥ 50 years with NTS bacteremia at 2 medical centers in southern Taiwan included in this study. Multiple logistic regression was used to identify risk factors for imaging-documented vascular infections. The prediction capability of the proposed scoring algorithm was indicated by a receiver operating characteristic curve and measures of sensitivity and specificity. Sixty patients (16.8%) with vascular infections were noted. The 4 risk factors significantly associated with vascular infections-male sex, hypertension, coronary arterial disease, and serogroup C1 infections-were each assigned +1 point to form the NTS vascular infection (NTSVI) score. In contrast, malignancy and immunosuppressive therapy were each assigned -1 point, owing to their negative associations with vascular infections. Based on the proposed NTSVI scoring, the prevalence of vascular infections in patients with ≤ 0, 1, 2, 3, or 4 points was 2.2% (3 of 138 patients), 10.6% (13 of 123 patients), 39.4% (26 of 66 patients), 55.2% (16 of 29 patients), and 100% (2 of 2 patients), respectively (P< .0001). The scoring algorithm shows an area under the curve of 0.83 (95% confidence interval, .78-.89; P < .0001). A cutoff value of +1 represents a high sensitivity (95.0%) and an acceptable specificity (45.3%). This simple scoring algorithm can be used to identify patients with NTS bacteremia with a high risk of vascular infections. The cost-effectiveness of this algorithm should be further studied.