Previous Abstract | Next Abstract
Printable Version
A737
October 16, 2011
1:00:00 PM - 4:00:00 PM
Room Hall B2 Area M
Effect of Automated Decision Support Alerts on the Occurrence of Prolonged Intervals between Intraoperative Blood Pressure Recordings
Wolf H. Stapelfeldt, M.D., Marc R. Reynolds, M.S., Bhaswati Ghosh, M.S., Jacek B. Cywinski, M.D., Piyush Mathur, M.D., George F. Takla, Ph.D.
Cleveland Clinic, Cleveland, Ohio, United States
Existing monitoring and documentation standards suggest a maximum interval of 5 minutes between subsequent blood pressure (BP) recordings in patients undergoing anesthesia. The present study was designed to determine the incidence of BP recordings intervals exceeding this limit and the effectiveness of automated decision support alerts in notifying providers to the absence of BP recordings in the AIMS database within this time window.

Methods: With IRB approval the timing of BP recordings were obtained from the PHDS registry for all patients undergoing anesthesia during the previous three years (since May 2, 2008). For each patient, the maximal time interval was calculated between subsequent BP recordings occurring after induction and before emergence from anesthesia. The analysis further included real-time alert records generated by a proprietary Clinical Decision Support System (DSS) in response to absence of intraoperative BP recordings within 7 minutes of the most recent, immediately preceding recording, a system which was deployed into production on January 8, 2010. Detection of missing BP recordings caused a button to flash on the AIMS screen as well as a descriptive message to appear on a DSS web page which was accessible from within the AIMS screen and required to be acknowledged by the anesthesia provider for the DSS button to stop flashing. Supervising faculty had the option of subscribing to additional alphanumeric display pages in order to be automatically alerted to this condition whenever they were not present in the OR at the time of the alert. Data analysis included the distribution and duration of prolonged BP recording intervals before and after institution of DSS alerts.

Results: The analysis included 25,202 cases completed prior to and 27,210 cases performed after the institution of DSS alerts. There was no significant difference in the distribution of maximal BP recording intervals exceeding the 5 minute target threshold (Fig. 1). Lack of effect was attributable to a rate of only 16% at which BP interval-related DSS alerts were acknowledged by the anesthesia provider(s) within 5 minutes of these being issued. In contrast, alerts which were successfully acknowledged within 5 minutes of being issued were associated with significantly reduced BP recording intervals by up to an average 3.2 minutes (Fig. 2, middle panel). Alphanumeric display paging (voluntarily subscribed to by approximately 1/3 of faculty) caused a further shortening of BP recording intervals for those alerts which were acknowledged within 3 minutes (Fig. 2, bottom panel).

Conclusion: Maximal intervals between successive BP recordings which exceeded 5 minutes were relatively common (33%). Automated DSS alerts to missing BP recordings were capable of reducing prolonged intervals between BP recordings provided these alerts were noticed by anesthesia care team members and acted upon in a timely fashion. A timely response to DSS alerts (blinking DSS button and web alert message) occurred in only one of six cases when judged by the record of alert acknowledgements within 5 minutes of issuing an alert. Additional alphanumeric display paging may facilitate a more timely response to DSS alerts.











Figure 1
Figure 2

Copyright © 2011 American Society of Anesthesiologists