Hospital survey on patient safety culture: psychometric analysis on a Scottish sample

Çakıl Sarac, Rhona Flin, Kathryn Mearns, Jeanette Jackson

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

Objective To investigate the psychometric properties of the Hospital Survey on Patient Safety Culture on a Scottish NHS data set. Methods The data were collected from 1969 clinical staff (estimated 22% response rate) from one acute hospital from each of seven Scottish Health boards. Using a split-half validation technique, the data were randomly split; an exploratory factor analysis was conducted on the calibration data set, and confirmatory factor analyses were conducted on the validation data set to investigate and check the original US model fit in a Scottish sample. Results Following the split-half validation technique, exploratory factor analysis results showed a 10-factor optimal measurement model. The confirmatory factor analyses were then performed to compare the model fit of two competing models (10-factor alternative model vs 12-factor original model). An S-B scaled χ(2) square difference test demonstrated that the original 12-factor model performed significantly better in a Scottish sample. Furthermore, reliability analyses of each component yielded satisfactory results. The mean scores on the climate dimensions in the Scottish sample were comparable with those found in other European countries. Conclusions This study provided evidence that the original 12-factor structure of the Hospital Survey on Patient SafetyCulture scale has been replicated in this Scottish sample. Therefore, no modifications are required to the original 12-factor model, which is suggested for use, since it would allow researchers the possibility of cross-national comparisons.
Original languageEnglish
Pages (from-to)842-848
JournalBMJ Quality and Safety
Volume20
Issue number10
DOIs
Publication statusPublished - 2011

Fingerprint Dive into the research topics of 'Hospital survey on patient safety culture: psychometric analysis on a Scottish sample'. Together they form a unique fingerprint.

Cite this