Interprofessional Antimicrobial Stewardship Influencing Clostridioides difficile Infection: An 8-Year Study Using Antimicrobial Use Density

Junichi Yoshida, Tetsuya Kikuchi, Takako Ueno, Akiko Mataga, Ikuyo Asano, Kazuhiro Otani, Tetsuro Tamura, Masao Tanaka Division of Infection Control, Shimonoseki City Hospital, Shimonoseki, JapanCorrespondence: Junichi YoshidaDivision of Infection Control, Shimonoseki City Hospital, 1-13-1, Shimonoseki 750-8520, JapanTel +81 83 231 4111Fax +81 83 224 3838Email yoshidaj@uicalumni.orgPurpose: To reduce Clostridioides difficile infection (CDI), we implemented interprofessional antimicrobial, infection control, and diagnostic stewardship (ipAS) conducted by physicians/pharmacists, infection control nurses, and medical technologists, respectively. As a numerical indicator for ipAS, we used antimicrobial use density (AUD) in an 8-year study to validate its efficacy in CDI reduction.Patients and methods: This was an observational study. CDI was defined as stool samples or C. difficile isolates containing toxin A and/or B from a patient with diarrhea occurring three or more times per day. From 2011–2018 at a 10-ward single site the subjects were in-patients with CDI, and the following data were collected: AUDs for 23 antibiotics, and antimicrobial test results. By 2015, we had established ipAS, consisting of culture submission before the administration of broad-spectrum antimicrobials, the promotion of point-of-care testing for diagnosis-based antimicrobials, perioperative prophylactic antibiotics, intervention at positive diagnosis of blood culture, team round for diarrhea, and inspection on contact precautions and disinfection in CDI cases. The study outcomes included annual numbers of CDI patients and blood culture sets. We compared annual AUDs between former (2011–14) and latter (2015–18) periods using Kruskal–Wallis tests and examined the correlation between AUDs and CDI numbers.Results: Of a total 50,970 patients, 1,750 patients underwent C. difficile toxin tests, of whom 171 patients (9.8%) were positive for CDI. Between the former and latter periods, AUDs for flomoxef (11.96 to 2.71 by medians), panipenem/betamipron (0.30 to 0.00), and clindamycin (3.87 to 2.19) significantly decreased (P