Background Candidaemia is a serious complication in critically ill patients, and associated with high mortality. The University Hospital of South Manchester, UK (UHSM) candidaemia guidelines for the ICUs recommend early initiation of antifungal therapy based on strong clinical suspicion and the presence of speciﬁc risk factors. Echinocandins are the ﬁrst line therapy, and micafungin is the echinocandin on formulary. Serum b-1-3-D-glucan (BG) should be used to guide discontinuation of therapy in the absence of other microbiological evidence of candidaemia as it has a very high negative predictive value. The aim of this study was to evaluate the compliance of the 3 ICUs (including cardio-thoracic & transplant, burns, acute medical) in UHSM (total of 71 beds) with the current guideline, the impact of the UHSM antifungal stewardship programme on antifungal consumption.
Methods The usage of micafungin at UHSM between April and July 2014 was reviewed using pharmacy databases, and by reviewing patient records. All prescriptions were reviewed by the audit team and a decision made as to whether or not the use of micafungin was appropriate in each individual case. Antifungal consumption data was analysed for the general adult ICU (19 beds) for the past 24 months.
Results A total of 72 patients admitted to ICU were started on micafungin during the 4-month study period. Of these, 45 were treated for suspected or proven candidaemia, and 25 for suspected or proven aspergillosis or scedosporiosis. Of those treated for candidaemia, seven (16%) had Candida spp. isolated from blood cultures. Two of these were ECMO patients. In 14 cases (31%) the local guideline was not fully followed. In eight cases the documented reason for starting micafungin was isolation of Candida spp. in a single superﬁcial sample such as sputum or urine. In four cases microbiological tests had not been taken as per guideline, and in two cases negative BG result was not acted upon. Of the 38 cases without evidence of invasive candidosis when the patients were reviewed by a member of the Infectious Diseases team, micafungin was discontinued in 19 cases (50%). In two further cases, discontinuation of micafungin was recommended, however the primary team (both surgical) chose not to follow the advice given. Two patients had been prescribed higher (150 mg od) dose in the absence of clinical suspicion of aspergillosis or oesophageal candidosis. There were no deaths due to candidaemia during the study period. The monthly micafungin expenditure on the AICU was some 13 000€ per month before the audit and was reduced to £9200€ during the audit and has stayed on that level until now.
Conclusion True candidaemia is a medical emergency and micafungin a well-tolerated drug. Therefore, it would be inappropriate to restrict the initiation of ﬁrst-line therapy in cases of suspected candidaemia. Our study shows that with the active presence of Infectious Diseases Consultants on ICU, and with the appropriate use of diagnostics, it is possible to safely discontinue echinocandin therapy when it is not indicated. This has important implications for antifungal stewardship and cost saving in intensive care.