Switching Vancomycin Monitoring From Trough Concentration to Area Under the Curve Estimation by Bayesian Forecasting: A Short Communication on a Cost-Benefit Study in Resource-Limited Settings

Background: This study was conducted to evaluate the cost-benefit indicators of a vancomycin monitoring protocol based on area under the curve estimation using commercial Bayesian software. Methods: This quasi-experimental study included patients who were aged >18 years with a vancomycin prescription for >24 hours. Patients who were terminally ill or those with acute kidney injury (AKI) ≤24 hours were excluded. During the preintervention period, doses were adjusted based on the trough concentration target of 15–20 mg/L, whereas the postintervention period target was 400–500 mg × h/L for the area under the curve. The medical team was responsible for deciding to stop the antimicrobial prescription without influence from the therapeutic drug monitoring team. The main outcomes were the incidence of AKI and length of stay. Cost-benefit simulation was performed after statistical analysis. Results: There were 96 patients in the preintervention group and 110 in the postintervention group. The AKI rate decreased from 20% (n = 19) to 6% (n = 6; P = 0.003), whereas the number of vancomycin serum samples decreased from 5 (interquartile range: 2–7) to 2 (interquartile range: 1–3) examinations per patient ( P < 0.001). The mean length of hospital stay for patients was 26.19 days after vancomycin prescription, compared with 17.13 days for those without AKI ( P = 0.003). At our institution, the decrease in AKI rate and reduced length of stay boosted yearly savings of up to US$ 369,000 for 300 patients receiving vancomycin therapy. Conclusions: Even in resource-limited settings, a commercial Bayesian forecasting–based protocol for vancomycin is important for determining cost-benefit outcomes.