Evaluation of an active decision support system for hemodynamic optimization during elective major vascular surgery.

BACKGROUND: Active decision support systems implementing goal directed therapy may be an approach to reduce disparities in outcome between different health care providers. We assessed feasibility of and adherence to an active decision support system (ADSS) comprising fluids, vasopressors, and dobutamine to optimize hemodynamics during high-risk vascular surgery. METHODS: In this prospective observational trial a closed-loop goal-directed therapy protocol, employing the minimally-invasive LiDCOrapid device, was used to actively provide advice to the anesthesiologist during surgery. All given suggestions and all interventions were recorded. Every intervention without or against the given advice had to be justified. The primary outcome parameters were the number of interventions done according to the ADSS and its duration of use. Reasons for non-compliance served to describe its limitations. RESULTS : The active decision support system was employed in 32 patients for 137 hours. Median (IQR) use of the ADSS as percentage of surgery time was 100% (94-100%) with 743 interventions being executed. 634 interventions were done according to ADSS proposals. Reasons to act against or without advice were: hemodynamic instability (6%), foreseeing a surgical event (2%), medical reasons (2%), awaiting hemodynamic improvement (1%) and orders by senior physician or surgeons (1%). In five patients the anesthesiologist decided to modify intervention thresholds of the underlying protocol. CONCLUSIONS: High rates of compliance underline clinical acceptability and feasibility of this ADSS during vascular surgery. It may therefore facilitate the work of anesthesiologists and reduce disparities in patient outcomes due to different healthcare providers. Particularly, rapidly developing hemodynamic perturbances as well as co-factors the ADSS as of now does not anticipate are current limitations. These findings may serve to further improve this stand-alone real-time ADSS.