Aim: In response to emerging vancomycin resistance among gram-positive cocci, it is recommended that hospitals develop guidelines for the appropriate use of glycopeptides and identify situations where glycopeptide use should be discouraged. The aim of this study was to audit the use of vancomycin in Auckland Healthcare hospitals.
Method: Patients prescribed vancomycin were recorded by pharmacy staff at Auckland, Starship, Green Lane and National Women's Hospitals. Clinical and laboratory information was collected for each course of vancomycin treatment. Standard definitions were used to classify prophylactic, empirical or specific directed therapy as appropriate or inappropriate. Continuing vancomycin when cultures were negative for beta-lactam-resistant, gram-positive organisms and/or initial choice of vancomycin when it was not necessary for the presumed source of infection were reasons for inappropriate empirical use. Reasons for inappropriate specific directed therapy included vancomycin prescribed for methicillin susceptible S. aureus and coagulase-negative staphylococci, or penicillin susceptible viridans streptococci when there was no history of beta-lactam allergy.
Results: One hundred and sixty-eight courses of vancomycin were prescribed for 146 patients; 42 in children (<16 years) and 126 in adults. Thirty-two per cent of all vancomycin courses were in renal patients, 26% in surgical specialities, 17% in haematology/oncology patients, 14% in medical specialities and 10% in intensive care unit patients. Eighty-six (51%) courses of vancomycin were considered inappropriate. The majority, 54/86 (63%) of inappropriate use, was for empirical therapy. It was an inappropriate initial choice in 25 instances, the duration of treatment was inappropriate, given no beta-lactam-resistant organisms were isolated in nine instances and both its initial choice and duration were inappropriate in 20 instances. Switching to other antimicrobial agents sooner when culture results and susceptibilities became available would have shortened the duration of 58/86 (67%) of the inappropriate courses. Of the inappropriate courses, 44/86 (51%) were prescribed for renal patients, 22 for empirical use, e.g. for peritoneal dialysis-related peritonitis, wound infections and presumed line infections and 22 for specific therapy of beta-lactam susceptible isolates because of dosing convenience in patients with renal failure.
Conclusion: Half of the vancomycin use in Auckland Healthcare hospitals could potentially be modified. The majority of inappropriate use (63%) was for empirical therapy. The microbiology laboratory's ability to promptly and accurately report culture and susceptibility results and convey these to the prescribing clinician is important in reducing unnecessary doses. This study identified areas where interventions will be focused to reduce vancomycin use.