De-escalation from Echinocandins to Azole Treatment in Critically Ill Patients with Candidemia.

OBJECTIVES:This study aims to further explore the duration of echinocandins and analyze the de-escalation (DE) strategy in patients with candidemia. METHODS:Multivariable logistic regression was used to evaluate the association between the duration of echinocandins (≤ 5-day group vs > 5-day group) and in-hospital mortality. RESULTS:Of the 357 cases of candidemia, 200 patients (56.02%) were identified in the ≤5-day group. The ≤5-day group did not have a higher in-hospital mortality than the >5-day group in the multivariable model (odds ratio [OR] 1.536, 95% confidence interval (CI) 0.837-2.819, P-value = 0.166), and the finding was validated by the propensity score matching and inverse probability of treatment weighting models. Subgroup analyses showed that patients in the ≤5-day group without DE may have a poor prognosis (OR 4.223, 95% CI 1.773-10.055, P-value = 0.001). The patients in the ≤5-day group, with a sequential organ failure assessment (SOFA) score of ≥3 evaluated at the time of stopping echinocandins, may have a poor prognosis (OR 2.164, 95% CI 1.009-4.641, P-value = 0.047). CONCLUSION:In critically ill adult patients with candidemia, the ≤5-day group with DE was feasible. However, the SOFA score was recommended when stopping echinocandins to ensure the safety of DE therapy.