The Safety Organizing Scale: development and validation of a behavioral measure of safety culture in hospital nursing units

Med Care. 2007 Jan;45(1):46-54. doi: 10.1097/01.mlr.0000244635.61178.7a.

Abstract

Background: Evidence that medical error is a systemic problem requiring systemic solutions continues to expand. Developing a "safety culture" is one potential strategy toward improving patient safety. A reliable and valid self-report measure of safety culture is needed that is both grounded in concrete behaviors and is positively related to patient safety.

Objective: We sought to develop and test a self-report measure of safety organizing that captures the behaviors theorized to underlie a safety culture and demonstrates use for potentially improving patient safety as evidenced by fewer reported medication errors and patient falls.

Subjects: A total of 1685 registered nurses from 125 nursing units in 13 hospitals in California, Indiana, Iowa, Maryland, Michigan, and Ohio completed questionnaires between December 2003 and June 2004.

Research design: The authors conducted a cross-sectional assessment of factor structure, dimensionality, and construct validity.

Results: The Safety Organizing Scale (SOS), a 9-item unidimensional measure of self-reported behaviors enabling a safety culture, was found to have high internal reliability and reflect theoretically derived and empirically observed content domains. The measure was shown to discriminate between related concepts like organizational commitment and trust, vary significantly within hospitals, and was negatively associated with reported medication errors and patient falls in the subsequent 6-month period.

Conclusions: The SOS not only provides meaningful, behavioral insight into the enactment of a safety culture, but because of the association between SOS scores and reported medication errors and patient falls, it also provides information that may be useful to registered nurses, nurse managers, hospital administrators, and governmental agencies.

Publication types

  • Multicenter Study

MeSH terms

  • Humans
  • Medication Errors / prevention & control*
  • Nursing Service, Hospital / organization & administration
  • Nursing Service, Hospital / statistics & numerical data*
  • Regression Analysis
  • Safety*
  • Surveys and Questionnaires
  • United States