Persistent candidemia caused by different Candida species: Data from a multicenter contemporary cohort

Objectives: To explore persistent candidemia by different Candida spp. Methods: Observational, retrospective, multicenter study including patients with candidemia (Jan 2018–Dec 2022) from 3 hospitals in Italy and Spain. The primary outcome was persistent candidemia, defined as positive blood culture (BC) yielding the same Candida spp≥5 days from the start of active antifungals. Patients with no available follow-up BCs were excluded. A competing risk analysis (competing risk of death) was performed using Fine and Gray regression models. Results: Among 1188 patients, 298 (25.1%) had persistent candidemia. Cancer (sHR 1.335, 95% CI 1.037–1.633, p=0.011), total parenteral nutrition (sHR 1.440, 95%CI 1.062–1.818, p=0.006), Candida parapsilosis (sHR 1.312, 95% CI 1.075–1.633, p=0.03) and Candida auris (sHR 1.549, 95% CI 1.155–2.159, p=0.029) compared to Candida albicans, were associated with increased risk of persistent candidemia, whereas primary candidemia (sHR 0.573, 95% CI 0.321–0.825, p<0.001) and early source control (sHR 0.557, 95% CI 0.401–0.713, p<0.001) were protective. Persistent candidemia was associated with higher 30-day mortality (aHR 1.605, 95% CI 1.176–2.191, p=0.003). Conclusions: Persistent candidemia affects one in four patients with Candida BSI. Infections caused by Candida parapsilosis or Candida auris require individualized management, with early source control being essential to reduce the risk of persistence.