Development and preliminary validation of a Greek-language outpatient satisfaction questionnaire with principal components and multi-trait analyses

Abstract

Abstract Background In the recent years there is a growing interest in Greece concerning the measurement of the satisfaction of patients who are visiting the outpatient clinics of National Health System (NHS) general acute hospitals. The aim of this study is therefore to develop a patient satisfaction questionnaire and provide its preliminary validation. Methods A questionnaire in Greek has been developed by literature review, researchers' on the spot observation and interviews. Pretesting has been followed by telephone surveys in two short-term general NHS hospitals in Macedonia, Greece. A proportional stratified random sample of 285 subjects and a second random sample of 100 outpatients, drawn on March 2004, have been employed for the analysis. These have resulted in scale creation via Principal Components Analysis and psychometric testing for internal consistency, test-retest and interrater reliability as well as construct validity. Results Four summated scales have emerged regarding the pure outpatient component of the patients' visits, namely medical examination, hospital environment, comfort and appointment time. Cronbach's alpha coefficients and Pearson, Spearman and intraclass correlations indicate a high degree of scale reliability and validity. Two other scales -lab appointment time and lab experience- capture the apparently distinct yet complementary visitor experience related to the radiographic and laboratory tests. Psychometric tests are equally promising, however, some discriminant validity differences lack statistical significance. Conclusion The instrument appears to be reliable and valid regarding the pure outpatient experience, whereas more research employing larger samples is required in order to establish the apparent psychometric properties of the complementary radiographic and laboratory-testing process, which is only relevant to about 25% of the subjects analysed here.