Overcoming Pharmacokinetic and Pharmacodynamic Variability to Improve Antimicrobial Optimization

By the bioMérieux Connection Editors

Limiting antibiotic misuse and overuse begins with aligning prescribing activities using the 5 Ds of antimicrobial therapy: right drug, correct dose, right drug-route, suitable duration, and timely de-escalation to pathogen-directed therapy. Appropriate and precise antimicrobial dosing is vital in fighting antimicrobial resistance, improving clinical outcomes, minimizing toxicity for the patient, and protecting antibiotic efficacy. Understanding pharmacokinetics and pharmacodynamics (PK-PD) and their associated opportunities and challenges could be pivotal in achieving precision prescribing.

Pharmacokinetics and pharmacodynamics (PK-PD modeling) describe the relationship between the body and a drug (or drugs). Pharmacokinetics encompass what the body does to a drug and conversely, pharmacodynamics encompass how a drug may impact the body. As noted in a recent Nature Reviews study, variability in pharmacokinetics as a result of comorbidities or other medications has the potential to impact antimicrobial pharmacodynamics and furthermore, treatment success.

While there are approaches for addressing explained variability (sources of variability that are measurable and predictable) in antimicrobial dosing, advanced methods of monitoring are needed to aid dosing efforts and react to any residual (unexplained) variability.  The authors of the study describe tactics such as the deployment of real-time biosensing, closed-loop control, and artificial intelligence-driven decision support tools, which may have the capacity to further address variability and support efforts in achieving precision antimicrobial dosing.

Closed Loop Control Systems with Continuous Biosensing for Individualized Antimicrobial Therapy

As noted in the study, closed loop control systems have many clinical applications including blood glucose control for patients with diabetes. Closed loop systems work on the principle of feedback, where a target signal is compared with a real-time biosensing measurement. The data disparities are then analyzed to provide guidance for an accurate rate of drug delivery. Adjustments in drug delivery are implemented based on insights from the data and weighed against the optimal measurement for delivery. The authors write that closed-loop control systems and continuous biosensing have many clinical applications, and when used together, have the potential to deliver truly individualized antimicrobial therapies.

Integrating Diagnostics and Other Data Sources to Improve Artificial Intelligence and Decision Support Systems

Support systems like artificial intelligence-driven systems help to provide the necessary data to facilitate effective antimicrobial treatment in real time. However, according to the study authors, “decision support for optimization of antimicrobial regimens has focused predominantly on implementation of Bayesian forecasting and mobile applications containing traditional dose optimization software.” To further promote optimized antimicrobial therapy, greater integration of diagnostics, electronic health records, and surveillance systems may be used to support real-time decision-making. A comprehensive system that supports the seamless integration of data and drives cross-departmental collaboration may provide the support needed to achieve precision dosing.

Advanced Technology Supports Antimicrobial Stewardship Initiatives

Implementation of advanced tools such as real-time biosensing, closed-loop control, and artificial intelligence-driven decision support systems may help address antimicrobial resistance. The authors conclude, “Future work must focus on addressing current barriers to implementation of optimized antimicrobial use and support greater connectivity with other developments in the field of antimicrobial prescribing.”


Opinions expressed in this article are not necessarily those of bioMérieux, Inc.

Leave a Reply

Your email address will not be published. Required fields are marked *