Abstract Introduction Stewardship programs have been developed to optimize the use of antibiotics, but programs focusing on antifungal agents are less frequent. Objective To evaluate the quality of antifungal prescriptions in a tertiary care hospital, and to test if a simple educational activity could improve the quality of prescriptions. Methods The study comprised three phases: 1) Retrospective audit of all antifungal prescriptions in a 6-month period, applying a score based on six parameters: indication, drug, dosage, route of administration, microbiologic adequacy after results of cultures, switching to an oral agent, and duration of treatment; 2) Creation of text boxes in the electronic medical records with information about antifungal agents, shown during prescription; 3) Retrospective audit of all antifungal prescriptions in a 6-month period, applying the same 6-parameters score, and comparison between the two periods. Results Among 333 prescriptions, fluconazole was the most frequently (80.5%) prescribed agent. Hematology (26.7%), Infectious Diseases Department (22.8%), Internal Medicine (15.9%) and Intensive Care Unit (14.4%) were the units with most antifungal prescriptions. The median score for the 333 prescriptions was 8.0 (range 0 – 10), and 72.7% of prescriptions were considered inappropriate. The median and mean scores in the first and second audit were 8.0 and 6.9, and 8.0 and 7.9, respectively (p<0.001). All items that comprised the score improved from the first to the second audit. Likewise, there was a reduction of inappropriate prescriptions (80.2% in the first audit vs. 64.6% in the second audit, p=0.001). Conclusions A large proportion of inappropriate prescriptions was observed, which improved with the implementation of simple educational activities.

ABSTRACT Background: The impact of central venous catheter (CVC) removal on the outcome of patients with candidemia is controversial, with studies reporting discrepant results depending on the time of CVC removal (early or any time during the course of candidemia). Objective: Evaluate the effect of time to CVC removal, early (within 48 h from the diagnosis of candidemia) vs. removal at any time during the course of candidemia, on the 30-day mortality. Methods: Retrospective cohort study of 285 patients with candidemia analyzing CVC removal within 48 h (first analysis) or at any time (second analysis). Results: A CVC was in place in 212 patients and was removed in 148 (69.8%), either early (88 patients, 41.5%) or late (60 patients, 28.3%). Overall, the median time to CVC removal was one day (range 1-28) but was six days (range 3-28) for those removed later. In the first analysis, APACHE II score (odds ratio [OR] 1.111, 95% confidence interval [95% CI] 1.066-1.158), removal at any time (OR 0.079, 95% CI 0.021-0.298) and Candida parapsilosis infection (OR 0.291, 95% CI 0.133-0.638) were predictors of 30-day mortality. Early removal was not significant. In the second analysis APACHE II score (OR 1.122, 95% CI 1.071-1.175) and C. parapsilosis infection (OR 0.247, 95% CI 0.103-0.590) retained significance. Conclusions: The impact of CVC removal is dependent on whether the optimal analysis strategy is deployed and should be taken into consideration in future analyses.

ABSTRACT Background Candidemia is the most frequent invasive fungal disease in hospitalized patients, and is associated with high mortality rates. The main objective of this study was to evaluate changes in the epidemiology of candidemia at a tertiary care hospital in a 21-year period. Methods We evaluated all episodes of candidemia diagnosed between 1996 and 2016 at a University-affiliated tertiary care hospital in Brazil. We arbitrarily divided the study period in 3: 1996-2002 (period 1), 2003-2009 (period 2) and 2010-2016 (period 3). Incidence rates were calculated using hospital admissions as denominator. Results We observed 331 episodes of candidemia. The incidence was 1.30 episodes per 1000 admissions, with no significant change over time. Candida albicans (37.5%), C. tropicalis (28.1%), C. parapsilosis (18.4%) and C. glabrata (6.9%) were the most frequent species. The proportion of patients receiving treatment increased (65.5%, 79.4% and 74.7% in periods 1, 2 and 3, respectively, p= 0.04), and the median time from candidemia to treatment initiation decreased from 4 days in period 1 (range 0-32 days) to 2 days in period 2 (range 0-33 days) and 2 days in period 3 (range 0-14 days, p< 0.001). We observed a significant decrease in the use of deoxycholate amphotericin B (47.4%, 14.8% and 11.9%), and an increase in the use of echinocandins (0%, 2.8% and 49.1%; p< 0.001). The APACHE II score increased over time (median 16, 17.5, and 22, p< 0.001). The overall 30-day mortality was 58.9%, and did not change significantly over the study period. Conclusions There was an improvement in patient care, with an increase in the proportion of patients receiving treatment and a decrease in the time to treatment initiation, but no improvement in the outcome, possibly because the proportion of sicker patients increased over time.