PKP-017 Impact of a bayesian pharmacokinetic dosing programme of vancomycin on clinical outcomes

Background The recommended starting dose of vancomycin is 25–30 mg/kg followed by 15–20 mg/kg/12–8 h (adjusted if there is renal impairment). Early plasma concentrations (PC), after 3 doses, should be obtained as soon as possible to determine if therapeutic levels (TL) have been reached (10–20 µg/ml). Purpose To describe patients and indications, and to analyse treatments and a pharmacokinetic monitoring plan. To assess efficacy and its relation with PC and AUC/MIC, and nephrotoxicity (0.5 mg/dL or 50% creatinine increase). Material and methods Retrospective study of vancomycin treatment guided by pharmacokinetic monitoring (Bayesian method) over 5 months. ICU, haemodyalisis, paediatrics, duration Results 87.9% of treatments were monitored (n = 22). Patients were 64 years (IR=22), CrCl=96 mL/min (IR=71.5) and 77.3% showed some nephrotoxicity risk factor. 22.7% were skin/soft tissue (40% E faecium , 20% MRSA, 20% CNS), intra-abdominal 18.2% (66.7% E faecium , 33.3% CNS), bacteraemia 13.6% (100% CNS), catheter 13.6% (100% CNS), pneumonia 9.1% (100% MRSA), urinary tract 9.1% (100% Enterococcus), 9.1% without a clear focus and 4.5% non-pneumonia respiratory infections (100% MRSA). 100% E faecium showed MIC ≤4, 100% MRSA MIC ≥1.5, 50% CNS MIC ≥2. No loading dose was administered. Initial dosage was appropriate in 31.8%; 68.2% was under dosed. The first PC was obtained after 3 days (IR=2.25); 50% were delayed beyond the third dose and 42% were subtherapeutic. TL were obtained after 5 days (IR=4). Pharmacokinetically guided dosing showed 72.7% of patients achieved TL (18.2% above; 9% under range). Clinical cure rate was 77.3%. By indication: 100% bacteraemia, urinary and non-pneumonia respiratory infections were cured; 80% skin/soft tissues; 75% intra-abdominal; 66.6% catheter; 50% pneumonia; and 50% without focus. By microorganism: 87.5% CNS; 66.7% E faecium ; and 66.7% MRSA. There was no statistically significant difference in clinical cure related to PC or AUC/MIC although there was a tendency to higher PC in the cure group (16.7 µg/mL vs 12.13 µg/mL). 9.1% of patients developed nephrotoxicity. Conclusion Although most treatments were pharmacokinetically monitored, the first level was delayed in half of the patients; 68.2% of treatments were initially under dosed. This led to delay in achieving TL. A relationship was not found between clinical cure and PC or AUC/MIC, probably due to the small sample size. No conflict of interest.