Journal Article
Clinical predictors, microbiological profile, and outcomes of bacteremia in patients with cellulitis: a 6-year retrospective cohort study
Ana Cotos Suárez; Amparo López-Bernus; Montserrat Alonso-Sardón; Virginia Velasco Tirado; Laura Sancha Domínguez; Beatriz Rodriguez Alonso; Josue Pendones Ulerio; Juan Luis Muñoz Bellido; Moncef Belhassen-García
European Journal of Clinical Microbiology & Infectious Diseases · 2026
Abstract
Background Cellulitis is a common cause of hospital admission, yet the clinical relevance and predictors of bacteremia remain incompletely characterized. Identifying patients at risk of bloodstream infection may help optimize diagnostic strategies and clinical management. Methods We conducted a retrospective observational study at a tertiary-care hospital including all consecutive patients hospitalized with cellulitis between January 2019 and December 2024. Patients without blood cultures were excluded. Clinical severity was stratified using the CREST classification. Microbiological data, antibiotic therapy, hospital outcomes, and follow-up at 1 and 3 months were analyzed. Results Among 311 episodes included, bacteremia was confirmed in 46 cases (14.8%). Patients with bacteremia showed significantly higher clinical severity at admission, with CREST grades III–IV present in 56.5% compared with 36.6% in non-bacteremia patients ( p < 0.001). Sepsis occurred in 65.2% of bacteremia cases versus 24.5% of non-bacteremia cases ( p < 0.001). Bacteremia was associated with longer hospital stay (18 ± 18.3 vs 13 ± 9.7 days; p = 0.003) and more frequent antibiotic modifications during admission ( p = 0.023). Streptococcal species and Staphylococcus aureus showed a higher propensity for bloodstream invasion. In-hospital mortality was higher among bacteremia patients (13.0% vs 7.9%), although the difference was not statistically significant. Three-month mortality reached 14.3% in the bacteremia group compared with 9.9% in non-bacteremia patients. Conclusion Bacteremia complicating cellulitis is uncommon but associated with greater clinical severity, longer hospitalization, and increased therapeutic complexity. Severity stratification tools such as CREST may assist in identifying patients at increased risk of bacteremia; however, given their limited discriminatory performance, they should not be used in isolation to guide blood culture collection or other clinical decisions.