Patients at risk of developing invasive candida infection can be diagnosed early, using new rapid technologies. Accelerated diagnose allows earlier shift in antifungal therapy for patients insufficiently treated by first choice of antifungal medicine. OBJECTIVES: To analyze the economic consequences for the Intensive Care Units (ICU) of a faster diagnose and when indicated earlier shift in therapy using PCR compared with a traditional blood culture. METHODS: A decision-tree based costing model was designed to compare the two approaches to diagnosing (blood culture alone versus PCR plus blood culture). The model compares the costs of the diagnostic methods, costs of medicines as well as costs of LOS at the ICU for the two alternatives. The data for the costing model was based on data for a half year of patients at Copenhagen University Hospital in 2008 having an invasive candida infection (n=10), and all costs were from a Danish perspective. RESULTS: All patients in risk of developing invasive candida infection are treated with fluconazol until diagnosed positive. Based on a national surveillance of patterns of resistance up to 35% of the patients will be shifted to an echinocandin after diagnose. A faster diagnose and early shift in therapy with echinocandin will imply in average an extra medicine costs between € 254-267. However, at the same time LOS at ICU will in average be reduced by 1.4 days, which will result in savings as high as €4859. Diagnose with PCR and use of echinocandin for the 35% of the relevant patients could therefore result in savings up to €4605 per patient. CONCLUSIONS: Rapid diagnose with PCR and early shift in therapy to echinocandin is cost-effective. However, different budgets (including the DRG payment) at the hospital are affected, why silo thinking might be a barrier, when deciding upon the cost-effective strategy.