Computer control of drug delivery by continuous intravenous infusion: bridging the gap between intended and actual drug delivery.

Abstract

Intravenous drug infusion driven by syringe pumps may lead to substantial temporal lags in achieving steady-state delivery at target levels when using very low flow rates ("microinfusion"). This study evaluated computer algorithms for reducing temporal lags via coordinated control of drug and carrier flows. Novel computer control algorithms were developed based on mathematical models of fluid flow. Algorithm 1 controlled initiation of drug infusion and algorithm 2 controlled changes to ongoing steady-state infusions. These algorithms were tested in vitro and in vivo using typical high and low dead volume infusion system architectures. One syringe pump infused a carrier fluid and a second infused drug. Drug and carrier flowed together via a manifold through standard central venous catheters. Samples were collected in vitro for quantitative delivery analysis. Parameters including left ventricular max dP/dt were recorded in vivo. Regulation by algorithm 1 reduced delivery delay in vitro during infusion initiation by 69% (low dead volume) and 78% (high dead volume). Algorithmic control in vivo measuring % change in max dP/dt showed similar results (55% for low dead volume and 64% for high dead volume). Algorithm 2 yielded greater precision in matching the magnitude and timing of intended changes in vivo and in vitro. Compared with conventional methods, algorithm-based computer control of carrier and drug flows can improve drug delivery by pump-driven intravenous infusion to better match intent. For norepinephrine infusions, the amount of drug reaching the bloodstream per time appears to be a dominant factor in the hemodynamic response to infusion.