A Cluster Randomized Controlled Trial Comparing Early and Late Carbapenems De-escalation in the Medicine Units, Maharaj Nakorn Chiang Mai Hospital

BACKGROUND: The rate of carbapenems use in the institution is high despite antimicrobial stewardship is in place. The current hospital policy allows the primary care team prescribed carbapenems for the first 72 hours, the de-escalation will occur after 72 hours once drug susceptibility available or infectious diseases specialist unapproved. We conduct this study aimed to determine rate of de-escalation within the first 24 hours between early and late carbapenems de-escalation. METHODS: A cluster randomized control trial was conducted among patients receiving care at the medicine units of the Maharaj Nakorn Chiang Mai Hospital between June 2016 and February 2017. Patients were randomly assigned into 2 groups; early de-escalation group where carbapenems were de-escalated within 24 hours or no later than 72 hours of prescription by ID specialist and late de-escalation where carbapenems were de-escalated followed the current hospital policy. RESULTS: A total of 104 patients were enrolled; 51 in early de-escalation group and 53 in late de-escalation group. The median age was 62 years (IQR 52–71 years), 49 patients (79%) were male. The most common diagnosis required carbapenems empirically was bloodstream infection (43 patients, 41.3%), followed by urinary tract infection (28 patients, 26.9%), and pneumonia (20 patients, 19.2%). Carbapenems de-escalation within 24 hours was 17 patients (33.3%) in early de-escalation group, and 4 patients (7.6%) in late de-escalation group (P = 0.001). The median duration of carbapenems use was 2 days (IQR 1, 6) and 4 days (IQR 3, 6) in early and late-de-escalation group, respectively (P = 0.017). The median cost of carbapenems per patients were 2,412 and 4,824 THB in early and late de-escalation group, respectively (P = 0.018). The 90-day mortality, readmission within 30 days, and length of hospital stay were not different between groups. Factor associated with de-escalation within 24 hours was ID intervened within 24 hours or no later than 72 hours of prescription (early de-escalation group). CONCLUSION: Early carbapenems de-escalation with careful clinical integration decreased unnecessary use of carbapenems without compromised the clinical outcomes. DISCLOSURES: All authors: No reported disclosures.