Improved computer-assisted digoxin therapy. A method using feedback of measured serum digoxin concentrations.

Area of Science:

• Medical Informatics
• Pharmacokinetics
• Control Systems Engineering

Background:

• Digoxin dosage regulation is complex due to individual patient variability.
• Accurate prediction of serum digoxin concentrations is crucial for effective therapeutic management.
• Physician performance in digoxin dosage adjustment can be variable.

Purpose of the Study:

• To develop and validate a computer system using automated feedback control for digoxin dosage management.
• To compare the predictive and achievement accuracy of the computer system against physician performance.
• To assess the system's ability to simulate physician-like learning of patient responses.

Main Methods:

• A prospective, randomized study involving 51 patients was conducted.
• A computer program utilized measured serum digoxin concentrations for feedback control.
• The computer's predictions and dosage achievements were compared with those of physicians.

Main Results:

• The computer system predicted future digoxin concentrations as accurately as physicians across varying feedback levels.
• Both computer and physicians significantly reduced prediction errors with increased feedback (known concentrations).
• The computer system achieved desired digoxin concentrations more accurately than physicians, particularly with multiple data points.

Conclusions:

• The computer system effectively simulates physician-like learning of individual patient responses to digoxin.
• Automated feedback control offers a viable method for improving digoxin dosage accuracy.
• Computer-assisted management has the potential to enhance therapeutic drug monitoring outcomes.