Friday, 12 August 11:00am – 12:00pm

This seminar will be delivered in Chemistry Lecture Theatre 4 and Online Zoom Please email chemistry.researchsupport@sydney.edu.au for zoom link and password.

Speaker: Dr Sophie Stocker; Sydney Pharmacy School, The University of Sydney

Host: Prof. Liz New

Title: Implementation of model informed precision dosing into routine clinical practice

Abstract: Introduction. The use of population pharmacokinetic models can facilitate individualisation of therapy to ensure efficacy whilst minimising toxicity. Various factors require consideration when implementing a model informed therapeutic drug monitoring (TDM) approach including model selection, establishment of a model of care and understanding sources of variability which may influence the accuracy of dose predictions. Limited data are available on the effect of these factors on improvement of therapeutic outcomes with model informed precision dosing into clinical practice.

Aims. This research program examines model selection processes and factors contributing to the variability in the accuracy of precision dosing strategies. The impact of a therapeutic drug monitoring TDM service for vancomycin on therapeutic target attainment in routine clinical is evaluated.

Methods. The predictive performance of population pharmacokinetic models used to predict drug exposure were evaluated to identify suitable pharmacokinetic models for vancomycin and tacrolimus in specific patient populations. Observed concentration data and drug exposure (AUC) was compared with that predicted by population pharmacokinetic models. Quantitative and qualitative approaches were used to assess the accuracy of the time of drug administration, blood sample collection and optimal sampling time. Monte Carlo simulation was used to predict the impact of discrepancies in time of drug administration on dose adjustments for vancomycin. Interrupted time series analysis was employed to evaluate monthly measures of the median proportion of vancomycin therapy spent within the target range. Indices of time to target attainment were also assessed before and after implementation of the vancomycin TDM service utilising Bayesian forecasting software.

Results. The median discrepancy between actual and documented administration times of antimicrobial agents was 16 min (range, 2-293 min; Roydhouse et al Br J Clin Pharmacol. 2021). The observed discrepancies in vancomycin administration time resulted in a different dose recommendation in 57.4% of cases (28.9% higher, 28.5% lower). Whilst blood collection times were accurately recorded, phlebotomists’ high workload, insufficient communication between health professionals and workflow practices impede ‘on-time’ blood collections. For vancomycin, a sample obtained between 4-6 hours post infusion is optimal (Shingde et al J Antimicrob Chemother 2020). Post-Service implementation the median proportion of time spent within the therapeutic range increased by 10.4% (95% CI, 1.2-19.6%, P=0.03) and target attainment at 48 hours by 7.8 % (95% CI 1.3-14.3%, P=0.02; Stocker et al Clin Pharmacol Ther 2021).

Conclusions. Selecting an unsuitable pharmacokinetic model and inaccuracies in data sources used to inform optimal dosing strategies can result in inappropriate dose recommendations. A consultative TDM Service utilising Bayesian forecasting software can facilitate attainment of vancomycin therapeutic targets. Understanding the factors which contribute to variability in model-informed dose recommendations can inform the design or optimisation of interventions to improve the accuracy of precision dosing strategies.