Documented Dental Disease in Patients With Staphylococcus aureus Bacteremia

Abstract

Dental and odontogenic infections cause transient oral bacteremia. Whether antecedent dental disease identifies a distinct bloodstream infection (BSI) phenotype is unknown. In multicenter BSI-PATH-v1 (19,834 pathogen-containing episodes from 17,686 patients), we compared five antecedent dental phenotypes (days 1–90) against oral-organism BSI (2,170 episodes, 10.9%) using case-case logistic regression with prespecified confounders and Holm correction. For oral-organism BSI, adjusted odds ratios were 3.22 (1.88–5.49) for periodontal disease, 3.53 (2.38–5.24) for apical/pulpal disease, 5.11 (2.29–11.42) for odontogenic abscess or cellulitis, and 3.91 (2.62–5.84) for composite dental disease. Associations peaked within 7 days and were null by days 31–90. Dental procedures (0.92, 0.57–1.47) showed no association. Post-BSI enrichment suggests organism-aware ascertainment. On a locked test set (n=2,966), the clinical-only model reached AUROC 0.686 and AUPRC 0.247. Dental features added no incremental value ( AUROC −0.0005; AUPRC −0.017). Antecedent odontogenic disease identifies a distinct BSI phenotype but does not improve pathogen prediction.