Background: Previously, we presented a proposed treatment algorithm using peptide nucleic acid fluorescence in-situ hybridization (FISH) testing for C. albicans candidemia with fluconazole susceptibility testing (FST). This algorithm was evaluated for physician compliance, impact on patient outcomes and costs following implementation in 2005. Methods: A candidemia treatment algorithm was implemented in 2005. The algorithm incorporated the FISH test, FST using disk diffusion, prior hospitalization (OSH) within 72 hours of candidemia and fluconazole exposure during hospital stay. Data were collected on physician compliance with the algorithm, patient outcomes and the wholesale pharmacy acquisition costs. The comparable group were patients in whom the algorithm was not followed. Results: There were 158 episodes of candidemia in 133 patients. 49 of 133 (37%) occurred within 72 hours after prior OSH. The major Candida sp involved were C. albicans in 70/133 (53%), and the major non-C. albicans sp (NCA) were C. glabrata with 27/133 (20%) and C. parapsilosis with 20/133 (15%). Fluconazole MICs >64 μg/ml were seen in 1 C. albicans, 1 C. lusitaniae and 5 C. glabrata, all of which were from patients with prior OSH. The treatment algorithm compliance was 58% (77/133), with no difference in mortality between the two treatment groups 22% and 25% (p=0.4) respectively. The non-algorithm group received empiric ineffective therapy 17% of the time and this also resulted in excess pharmacy costs from of $27,350 with $6,300 (23%) coming from the 70 mg caspofungin (15 doses) for C. albicans alone. Conclusions: The treatment algorithm developed for candidemia resulted in appropriate and cost effective therapy when utilized by physicians under normal practice conditions. This validated our previous study using rapid identification, prior risk factors and fluconazole treatment for candidemia.
Full conference title:
46th Interscience Conference on Antimicrobial Agents and Chemotherapy
- ICAAC 46th