An evidence-based toolkit for the development of effective and sustainable root cause analysis system safety solutions

A. Zachary Hettinger*, Rollin J. Fairbanks, Sudeep Hegde, Alexandra S. Rackoff, John Wreathall, Vicki L. Lewis, Ann M. Bisantz, Robert L. Wears

*Corresponding author for this work

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Root cause analysis (RCA) after adverse events in healthcare is a standard practice at many institutions. However, healthcare has failed to see a dramatic improvement in patient safety over the last decade. In order to improve the RCA process, this study used systems safety science, which is based partly on human factors engineering principles and has been applied with success in other high-risk industries like aviation. A multi-institutional dataset of 334 RCA cases and 782 solutions was analyzed using qualitative methods. A team of safety science experts developed a model of 13 RCA solutions categories through an iterative process, using semi-structured interview data from 44 frontline staff members from 7 different hospital-based unit types. These categories were placed in a model and toolkit to help guide RCA teams in developing sustainable and effective solutions to prevent future adverse events. This study was limited by its retrospective review of cases and use of interviews rather than clinical observations. In conclusion, systems safety principles were used to develop guidelines for RCA teams to promote systems-level sustainable and effective solutions for adverse events.

Original languageEnglish
Pages (from-to)11-20
Number of pages10
JournalJournal of healthcare risk management : the journal of the American Society for Healthcare Risk Management
Volume33
Issue number2
DOIs
Publication statusPublished - 2013
Externally publishedYes

Fingerprint Dive into the research topics of 'An evidence-based toolkit for the development of effective and sustainable root cause analysis system safety solutions'. Together they form a unique fingerprint.

Cite this