Jason M Etchegaray1, Terry Throckmorton. 1. The University of Texas Medical School at Houston, The University of Texas-Houston Memorial Hermann Center for Healthcare Quality and Safety, 6410 Fannin Street, UTPB 11.08, Houston, TX 77030, USA. jason.etchegaray@uth.tmc.edu
Abstract
OBJECTIVES: To demonstrate a statistical analysis for testing the measurement equivalence of a patient safety survey instrument. The survey instrument examined in the present study is the Medication Administration Error Reporting Survey. METHODS: Surveys were posted to a random sample of registered nurses in the State of Texas, with 435 nurses completing the survey. The surveys contained questions about various error reporting issues, including the 16-item, Medication Administration Error Reporting scale. Nurses were divided into one of two samples--calibration and holdout--to ensure replicability of the results. Within each sample, two groups were created based on nurse tenure on the job. RESULTS: Multiple Group Confirmatory Factor Analysis was conducted across nurses with varying levels of experience for the calibration and holdout samples. For each sample, a baseline model was estimated, where model parameters were allowed to vary across the nursing groups, and compared with more restrictive models. The results provided support for the factor structure of the Medication Administration Error Reporting System but yielded mixed results concerning the equivalence of the measure across nursing groups. CONCLUSIONS: The present study provides an explanation of how to examine the measurement equivalence of survey instruments and demonstrated that the Medication Administration Error Reporting scale might not be equivalent across nurses who differ with respect to experience levels.
OBJECTIVES: To demonstrate a statistical analysis for testing the measurement equivalence of a patient safety survey instrument. The survey instrument examined in the present study is the Medication Administration Error Reporting Survey. METHODS: Surveys were posted to a random sample of registered nurses in the State of Texas, with 435 nurses completing the survey. The surveys contained questions about various error reporting issues, including the 16-item, Medication Administration Error Reporting scale. Nurses were divided into one of two samples--calibration and holdout--to ensure replicability of the results. Within each sample, two groups were created based on nurse tenure on the job. RESULTS: Multiple Group Confirmatory Factor Analysis was conducted across nurses with varying levels of experience for the calibration and holdout samples. For each sample, a baseline model was estimated, where model parameters were allowed to vary across the nursing groups, and compared with more restrictive models. The results provided support for the factor structure of the Medication Administration Error Reporting System but yielded mixed results concerning the equivalence of the measure across nursing groups. CONCLUSIONS: The present study provides an explanation of how to examine the measurement equivalence of survey instruments and demonstrated that the Medication Administration Error Reporting scale might not be equivalent across nurses who differ with respect to experience levels.