Evaluation of amikacin use and comparison of the models implemented in two Bayesian forecasting software packages to guide dosing

Abstract

Aims Bayesian forecasting software can assist in guiding therapeutic drug monitoring (TDM)‐based dose adjustments for amikacin to achieve therapeutic targets. This study aimed to evaluate amikacin prescribing and TDM practices, and to determine the suitability of the amikacin model incorporated into the DoseMeRx® software as a replacement for the previously available software (Abbottbase®). Methods Patient demographics, pathology, amikacin dosing history, amikacin concentrations and Abbottbase® predicted TDM targets (area under the curve up to 24 hours, maximum concentration and trough concentration) were collected for adults receiving intravenous amikacin (2012–2017). Concordance with the Australian Therapeutic Guidelines was assessed. Observed and predicted amikacin concentrations were compared to determine the predictive performance (bias and precision) of DoseMeRx®. Amikacin TDM targets were predicted by DoseMeRx® and compared to those predicted by Abbottbase®. Results Overall, guideline compliance for 63 courses of amikacin in 47 patients was suboptimal. Doses were often lower than recommended. For therapy >48 h, TDM sample collection timing was commonly discordant with recommendations, therapeutic target attainment low and 34% of dose adjustments inappropriate. DoseMeRx® under‐predicted amikacin concentrations by 0.9 mg/L (95% confidence interval [CI] –1.4 to −0.5) compared with observed concentrations. However, maximum concentration values (n = 19) were unbiased (−1.7 mg/L 95%CI –5.8 to 0.8) and precise (8.6% 95%CI 5.4–18.1). Predicted trough concentration values (n = 7) were, at most, 1 mg/L higher than observed. Amikacin area under the curve values estimated using Abbottbase® (181 mg h/L 95%CI 161–202) and DoseMeRx® (176 mg h/L 95%CI 152–199) were similar (P = .59). Conclusion Amikacin dosing and TDM practice was suboptimal compared with guidelines. The model implemented by DoseMeRx® is satisfactory to guide amikacin dosing.

Are vancomycin dosing guidelines followed? A mixed methods study of vancomycin prescribing practices

“…A total of 1043 plasma vancomycin concentrations were collected; 939 (90%) during intermittent infusions and 104 (10%) during continuous infusions. A median (range) of 5 1–36 vancomycin concentrations were obtained in each course of therapy, collections occurring every 1.6 days (0.6–8 days).…”

“…Underdosing has been observed for other antibiotics (e.g., amikacin). 29 This practice has been attributed to a lack of understanding of guidelines, habitual dosing, as well as concerns about toxicity. 9,30 Given that underdosing of antibiotics can result in poor clinical outcomes, 8,31 such as unresolved infections and the emergence of resistant bacterial strains, interventions that encourage prescribers to prescribe higher, guideline concordant doses are needed.…”

“…Further, the majority of maintenance dose recommendations made for prescribing scenarios were also lower than guideline recommendations, despite interview participants reporting the importance of, and being observed to acknowledge, guidelines and appropriate patient factors. Underdosing has been observed for other antibiotics (e.g., amikacin) 29 . This practice has been attributed to a lack of understanding of guidelines, habitual dosing, as well as concerns about toxicity 9,30 .…”

Model-Informed Precision Dosing Software Tools for Dosage Regimen Individualization: A Scoping Review

“…Software tools for modelling or research. References [14,[17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36] are cited in the supplementary materials.…”

Blood, dose recommendation reports and phone calls: Experiences of a therapeutic drug monitoring advisory service for vancomycin

“…Pharmacists are commonly embedded in treating teams and could facilitate dose recommendation acceptance and improve team knowledge. Although this study only assessed the use of dose-prediction software for vancomycin, expansion of the Service to other drugs [16][17][18] could also improve efficiencies. Formal economic analysis, considering staffing, as well as software-related costs, is required to confirm the cost-benefit of integration and staffing.…”