DEMONSTRATION PROJECT TO REFINE, AUTOMATE AND TEST A NOVEL EMERGENCY DEPARTMENT TRIGGER TOOL

Washington University, Saint Louis MO

Investigators

Linked publications, trials & patents

Abstract

Abstract The overarching aim of this 3-year project is to develop a trigger tool (EDTT) specifically tailored for detecting and characterizing adverse events in the emergency department (ED). Trigger tools such as those pioneered by the Institute for Healthcare Improvement (IHI) have outperformed traditional methods for detecting and characterizing patient harm in other clinical settings, to the point that they have been featured by the Agency for Healthcare Research and Quality as ?the premier measurement strategy for patient safety.? The need for an EDTT adapted for use in ED is especially salient given: (i) that current ED surveillance methods are outdated and often inefficient; (ii) that EDs now account for 1/3 of all acute care visits, nearly 50% of hospital admissions in the US and are a major source of care for minority and vulnerable populations; and (iii), that numerous factors such as increasing acuity, limited data for decision making, time pressures and hospital crowding and boarding create an environment with a high potential for adverse events (AEs). In developing the EDTT, we will build on the results of a prior multicenter preliminary study in which we identified candidate triggers. In selecting these triggers, we balanced expected content with ease of application to electronic health records. We will first test and optimize the EDTT (refining trigger selection, eliminating poorly performing ones) on 2,400 randomly selected ED visit records, using dual independent review. This ?gold standard? sample will be used, following a slight delay to allow for data accrual, to develop a computerized version of the EDTT. Then we will validate the optimized, automated EDTT on an independent random sample of up to 20,000 charts, in order to identify 6,000 records with triggers. We will review all of these records for the presence of AEs. Analyses will focus on yield (roughly, ability to detect AEs), reliability, and associations with patient characteristics. This will provide a new standard for detection of harm in the ED setting, establish a baseline against which to measure improvement efforts and help direct resources for improving patient safety and quality. Our consultants come from several institutions nationwide, so that our work will always maintain a view toward multisite implementation and testing.