UCSF

Background

Patients may be inadvertently harmed while undergoing dental treatments. To improve care, we must first determine the types and frequency of harms that patients experience, but identifying cases of harm is not always straightforward for dental practices. Mining data from electronic health records is a promising means of efficiently detecting possible adverse events (AEs).

Methods

We developed 7 electronic triggers (electronic health record based) to flag patient charts that contain distinct events common to AEs. These electronic charts were then manually reviewed to identify AEs.

Results

Of the 1,885 charts reviewed, 16.2% contained an AE. The positive predictive value of the triggers ranged from a high of 0.23 for the 2 best-performing triggers (failed implants and postsurgical complications) to 0.09 for the lowest-performing triggers. The most common types of AEs found were pain (27.5%), hard tissue (14.8%), soft tissue (14.8%), and nerve injuries (13.3%). Most AEs were classified as temporary harm (89.2%). Permanent harm was present in 9.6% of the AEs, and 1.2% required transfer to an emergency room.

Conclusion

By developing these triggers and a process to identify harm, we can now start measuring AEs, which is the first step to mitigating harm in the future.

Knowledge transfer statement

A retrospective review of patients' health records is a useful approach for systematically identifying and measuring harm. Rather than random chart reviews, electronic health record-based dental trigger tools are an effective approach for practices to identify patient harm. Measurement is one of the first steps in improving the safety and quality of care delivered.