Reliability of the Swedish version of the Evidence-Based Practice Attitude Scale assessing physiotherapist's attitudes to implementation of evidence-based practice

Purpose To translate and apply a cross-cultural adaptation of the Evidence-Based Practice Attitude Scale (EBPAS) in Swedish and investigate its absolute and relative reliability. Methods The original EBPAS (a questionnaire assessing health professionals' attitudes to implementation of evidence-based practice) was translated into Swedish using a forward and backward procedure, including a group discussion and expert committee. To assess reliability, 55 physiotherapists (48 women) aged 23-64 years from different clinical settings in the Stockholm region answered the EBPAS by postal survey twice within an interval of 2 weeks. Results The Cronbach's alpha values for EBPAS were > 0.721. The intraclass correlation (ICC) between test and retest (relative reliability) was moderate to good for the four subscales, with ICC(A.1) and ICC(C.1) values approximately equal and in the range 0.56-0.89. Values for the absolute reliability of the mean score were a standard error of measurement of about 7% and a smallest real difference of about 19%. Conclusion The Swedish version of the EBPAS shows mainly good reliability.