Accuracy of a precision dosing software program for predicting antibiotic concentrations in critically ill patients

Area of Science:

• Pharmacokinetics and Pharmacodynamics
• Critical Care Medicine
• Infectious Diseases

Background:

• Critically ill patients with sepsis experience physiological changes impacting antibiotic effectiveness.
• Achieving target antibiotic exposures is crucial for sepsis treatment success.
• Precision dosing software offers a potential solution to optimize antibiotic therapy.

Purpose of the Study:

• To evaluate the accuracy of a precision dosing software (ID-ODS) in predicting antibiotic concentrations.
• To assess the impact of software-guided dosing adjustments on actual patient management.
• To determine the influence of therapeutic drug monitoring (TDM) data on prediction accuracy.

Main Methods:

• The ID-ODS software predicted concentrations for piperacillin, meropenem, and vancomycin.
• Patient covariate data and TDM data were utilized for predictions.
• Dosing adjustments based on predicted versus measured concentrations were analyzed.

Main Results:

• Software predictions for piperacillin and meropenem showed significant bias, which improved with TDM data.
• Dosing changes aligned with measured concentrations in 89.2% for piperacillin and 71% for meropenem when TDM data was included.
• Vancomycin predictions were accurate, though limited by a small sample size.

Conclusions:

• Precision dosing software shows promise for predicting antibiotic concentrations in sepsis patients.
• Incorporating TDM data enhances the software's predictive performance for key antibiotics.
• This approach aids in anticipating appropriate antibiotic dose adjustments for improved patient outcomes.