You’ve got harm:

A look at the Automated All Cause Harm trigger system

The prevention of avoidable harms has been a goal of healthcare since day one, but it was given fresh life in 2010 when the Office of Inspector General (OIG) urged that healthcare facilities report all types of harms: medical complications, preventable harms, and system failures and errors. However, harms data has been chronically under-reported for years. Even the adverse events compiled in The Joint Commission’s Sentinel Event database only represent a small fraction of safety events.

Then in April 2017, Christine Sammer, DrPH, RN, director of corporate patient safety at Adventist Health System (AHS), and co-authors published a new study that provided some encouragement.

 For 11 months, two AHS hospitals tried out an automated system called the Automated All-Cause Harm Trigger System (ACHTS). The system’s software uses 41 algorithms to monitor electronic medical records (EMR) for signs that harm has befallen a patient, with flagged charts sent to a reviewer to examine. By the end of the study, the ACHTS caught 2,696 cases of patient harm, compared to the 132 harms caught using the old sampling method. The system also saved reviewers time and gave physicians actionable information for helping patients. AHS published its results in The Joint Commission Journal on Quality and Patient Safety.

“Electronic safety event detection systems hold much promise to remove the bottlenecks of more traditional, more analog, and more manual methods,” wrote Eric Kirkendall, MD, MBI, associate chief medical information officer at Cincinnati Children’s Hospital Medical Center, in the study’s accompanying editorial. “Sammer et al. have demonstrated this nicely with their work. Now it's time to take the next steps of validating tools such as the Risk Trigger Monitoring system and determine how to widely implement them, with the goal of helping all healthcare providers ‘do no harm.’ ”

This is an excerpt from members-only content. Please log in or become a member to access the full content.