Clinical Performance and Safety of Closed-Loop Systems: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

BACKGROUND:: Automated systems can improve the stability of controlled variables and reduce the workload in clinical practice without increasing the risks to patients. We conducted this review and meta-analysis to assess the clinical performance of closed-loop systems compared with manual control. Our primary outcome was the accuracy of closed-loop systems in comparison with manual control to maintain a given variable in a desired target range. The occurrence of overshoot and undershoot episodes was the secondary outcome. METHODS:: We retrieved randomized controlled trials on accuracy and safety of closed-loop systems versus manual control. Our primary outcome was the percentage of time during which the system was able to maintain a given variable (e.g., bispectral index or oxygen saturation) in a desired range or the proportion of the target measurements that was within the required range. Our secondary outcome was the percentage of time or the number of episodes that the controlled variable was above or below the target range. The standardized mean difference and 95% confidence interval (CI) were calculated for continuous outcomes, whereas the odds ratio and 95% CI were estimated for dichotomous outcomes. RESULTS:: Thirty-six trials were included. Compared with manual control, automated systems allowed better maintenance of the controlled variable in the anesthesia drug delivery setting (95% CI, 11.7%–23.1%; percentage of time, P < 0.0001, number of studies: n = 15), in patients with diabetes mellitus (95% CI, 11.5%–30.9%; percentage of time, P = 0.001, n = 8), and in patients mechanically ventilated (95% CI, 1.5%–23.1%; percentage of time, P = 0.03, n = 8). Heterogeneity among the studies was high (>75%). We observed a significant reduction of episodes of overshooting and undershooting when closed-loop systems were used. CONCLUSIONS:: The use of automated systems can result in better control of a given target within a selected range. There was a decrease of overshooting or undershooting of a given target with closed-loop systems