Retrospective data collection and analytical techniques for patient safety studies

• Published in: Journal of biomedical informatics Vol. 36; no. 1; pp. 106 - 119
• Main Authors: Weinger, Matthew B., Slagle, Jason, Jain, Sonia, Ordonez, Nelda
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01-02-2003
• Subjects: Anesthesiology - methods; Anesthesiology - statistics & numerical data; Database Management Systems; Databases, Factual; Decision Support Techniques; Documentation; Information Storage and Retrieval - methods; Medical Audit - methods; Medical errors; Medical Errors - methods; Medical Errors - prevention & control; Medical Errors - statistics & numerical data; Medical Records Systems, Computerized; Models, Statistical; Patient Care Management - methods; Patient safety; Retrospective Studies; Risk Assessment - methods; Risk Management - methods; Safety Management - methods; Statistics as Topic - methods; Systems analysis; Patient safety; Systems analysis; Medical errors
• ISSN: 1532-0464; 1532-0480
• DOI: 10.1016/j.jbi.2003.08.002

To enhance patient safety, data about actual clinical events must be collected and scrutinized. This paper has two purposes. First, it provides an overview of some of the methods available to collect and analyze
retrospective data about medical errors, near misses, and other relevant patient safety events. Second, it introduces a methodological approach that focuses on non-routine events (NRE), defined as
all events that deviate from optimal clinical care. In intermittent in-person surveys of anesthesia providers, 75 of 277 (27%) recently completed anesthetic cases contained a non-routine event (98 total NRE). Forty-six of the cases (17%) had patient impact while only 20 (7%) led to patient injury. In contrast, in the same hospitals over a two-year period, we collected event data on 135 cases identified with traditional quality improvement processes (event incidence of 0.7–2.7%). In these quality improvement cases, 120 (89%) had patient impact and 74 (55%) led to patient injury. Preliminary analyses not only illustrate some of the analytical methods applicable to safety data but also provide insight into the potential value of the non-routine event approach for the early detection of risks to patient safety before serious patient harm occurs.