Ability to Influence Behavior Clinical Trial
Official title:
A Randomized Effectiveness Trial of Early Hemodynamic Support in Patients Demonstrating a Double and Triple Low Combination of Mean Arterial Pressure, and Either End-tidal Anesthetic Concentration or Bispectral Index.
The objective of this study is to use a decision-support system combined with an electronic anesthesia record to detect double low clinical conditions (Mean Arterial Pressure (MAP) < 75 mmHg, and Bispectral Index (BIS) < 45) and generate alerts suggesting hemodynamic support. Specifically, the project will test the hypothesis that providing Double Low alerts reduces 90-day mortality. Secondary outcomes will be the fraction of alerts that generate early clinician responses, subsequent changes in MAP and hospital length of stay.
Data that is routinely collected as part of standard of care intraoperative monitoring
including blood pressure, amount of anesthetic administered and a form of processes
electroencephalogram called the bispectral index, which is a measure of depth of anesthesia
all enter in an electronic record keeper. This project monitors for a combination of a blood
pressure below mean of 75mmHg and a BIS of 45. When this is detected, half of the
anesthesiologists caring for patients receive a text message on the computer screen and
their pagers indicating that a double low situation exists and that they should consider
intervening. The other half of practitioners are not provided with the additional messages,
although all of the same information is available in the operating room (OR). Most patients
undergoing general non-cardiac anesthesia at Mount Sinai are eligible. Documentation of
consent was waived. All patients are informed of the ongoing project upon admission and are
given a variety of methods of opting out of the project.
The analysis seeks to determine if the messages were effective at influencing clinician
behavior and whether that change would make a difference in either length of stay or 90 day
mortality.
;
Allocation: Randomized, Intervention Model: Parallel Assignment, Masking: Single Blind (Subject), Primary Purpose: Prevention