Implementation Of An Early Warning System Improves Patient Safety, But Is It Worth The Costs?

IMPLEMENTATION OF AN EARLY WARNING SYSTEM IMPROVES PATIENT SAFETY, BUT IS IT WORTH THE COSTS?

Justin Glasgow, MD PhD, Stephen Hoover, MS, UAB Health

Meeting: Hospital Medicine 2019, March 24-27, National Harbor, Md.

Abstract Number: 212

Keywords: Early warning system, ICU Transfer, Patient Safety, Rapid Response Team

Background: Numerous early warning systems (EWS) exist as potential tools to improve patient safety. Our system recognized higher than peer rates of rapid response (RRT) utilization as well as higher than desired out-of-ICU code blue rates leading to a desire to implement a EWS system. Over a three-year period we reviewed the literature, developed, and implemented an innovative EWS.

Methods: To assess the impact of the implemented EWS we monitored RRT, ICU transfer, code blue and inpatient mortality rates. These rates were monitored using statistical process control charts with t-tests used for statistical testing. We also monitored time-to-patient stabilization, time-to-ICU transfer, and time-to-RRT. Time-to-RRT and ICU-transfer were defined as the period of time from the first high or critical risk score in the EWS until the corresponding event occurred. Time to stabilization was similarly defined as the time from the first high or critical risk score until the patient had a consistent medium or low risk score for at least 8 hours. Statistical testing was completed using t-tests. We focused on comparing the same 6 calendar months from the pre-intervention year to the 6 calendar months after full implementation of the EWS.

Results: Post-intervention there were statistically significant decreases in RRT (13.66 pre, 12.16 post, p = 0.03) and ICU transfer rates / 1,000 patient days (5.18, 4.14, p = 0.01). Due to the rarity of code blue and inpatient mortality events statistical testing did not reveal any differences and control charts suggested a stable system. None of the time-to-event evaluations showed statistical improvement, but they did suggest some promising trends. Patient stabilization time decreased to a mean 12.79 hours from 13.13 hours, p = 0.07; time to RRT decreased to a mean of 12.84 from 13.9, p = 0.13; and time to ICU transfer decreased to a mean of 10.88 from 12.55, p=0.29.

Conclusions: At the initial six-month analysis the EWS contributed to decreased RRT and ICU transfer rates with at worst stable code blue and inpatient mortality rates. However, components of our EWS system significantly impacted nursing workflow and burden. Overall, the system felt that this burden outweighed the EWS benefits and the decision was made to switch to a scaled down EWS system with less impact on nursing workflow.