Paper
Friday, 21 July 2006
This presentation is part of : Methodological Issues in Translational Research
Factorial Validity of the Chinese Version of Barriers to Medication Administration
Hui Ying Chiang, PhD, Nursing Department, Chi-Mei Foundation Medical Center, Tainan, Taiwan
Learning Objective #1: understand the factorial structure of nurses' perceived barriers to medication error reporting.
Learning Objective #2: understand the construct validation by using factor analyses.

In order to explore nurses’ perceived barriers to medication administration error (MAE) reporting, validation of the Chinese version of Barriers to Mediation Administration Error Reporting (C-BMAER) is imperative. This can provide an instrument involving cultural domains and having evidence of psychometric testing for further research on medication safety in Taiwan.

Objectives: This study was designed to confirm the factor structure of the C-BMAER composing of the Reasons Why Medication Administration Errors Are Not Reported questionnaire and cultural subscales related to collectivistic characteristics: hierarchy of authority and face-saving.

Methods: Forward and backward translations were applied to the 25-item C-BMAER with 6-point Likert-type scale. Data were collected from 597 nurses in a Taiwanese hospital. The data set was randomly split into two parts for factor analysis testing. Item analysis, internal consistency, and expert validity were first examined. A principal component analysis with Varimax rotation was applied to identify a proposed factor structure. Further, the construct validity of C-BMAER was confirmed by confirmatory factor analysis.

Results: A model of C-BMAER included 5 factors: fear, reporting process, power distance, administrative barrier, and coworker-face needs. Cronbach’s alphas of the five subscales ranged from .83 to .66. The total scale coefficient alpha was .89. This model explained 55.10% of the total variance in nurses’ perceived barriers to MAE reporting. The overall measurement of model fix indices showed that the hypothesized model fit the data closely after 4 modifications (χ2 = 514.84 df = 262, root mean square error of approximation = .056, goodness-of-fit index = .88). Discussion: The C-BMAER is a valid instrument with acceptable internal consistency. Three factors of C-BMAER are similar to original scale and two cultural factors are supported. This instrument can be served as a measure of reporting systems for medication safety improvement. 

See more of Methodological Issues in Translational Research
See more of The 17th International Nursing Research Congress Focusing on Evidence-Based Practice (19-22 July 2006)