OUP user menu

Methodology matters. Identifying drug safety issues: from research to practice

TK Gandhi, DL Seder, DW Bates
DOI: http://dx.doi.org/10.1093/intqhc/12.1.69 69-76 First published online: 1 February 2000


Purpose.Adverse drug events (ADEs), or injuries due to drugs, are common and often preventable. However, identifying ADEs, potential ADEs, and medication errors can be a major challenge. In this review, we describe methodologies that have been used to identify these events and give strategies for identification in non-study settings.

Results.Methods such as voluntary reporting, chart review, and computerized monitoring for events have been most commonly used in studies of ADEs in inpatients. However, voluntary reporting, the method most hospitals currently use, has a very low yield of events. Chart review is much more sensitive but the costs are prohibitive. Computerized monitoring for ADEs (using rules or triggers) is a high yield and relatively inexpensive strategy that should be adopted by organizations. A limitation of this strategy, however, is that it identifies few medication errors and potential ADEs, which are also important. These can be captured through pharmacy logs, chart review, and direct observation. Once events have been identified, they can be classified by type of event, severity, and preventability. In non-study settings, the most practical method for identifying ADEs is computerized monitoring, and for identifying prescribing errors it is pharmacy logs of interventions. Once problems are found, a structure (either individual or committee) must be in place to classify them, identify opportunities for improvement, and carry out the necessary changes.

Conclusion.Health care organizations have the technology to significantly improve their detection of ADEs, medication errors, and potential ADEs. Identification and subsequent classification of events is crucial for quality efforts to improve patient safety.

  • Adverse drug events
  • drug safety
  • medication errors
  • methodology