Two recent articles have evaluated the measurement properties of the PEDro scale. This second article summarised the clinimetric properties of the PEDro scale, and is summarised in this post.
The PEDro score has demonstrated ‘fair’ to ‘excellent’ inter-rater reliability (Intraclass Correlation Coefficient 0.53-0.91) for randomised controlled trials of physiotherapy interventions, and ‘excellent’ inter-rater reliability (Intraclass Correlation Coefficient 0.80-0.89) for trials of pharmacological interventions. The inter-rater reliability for the individual PEDro scale items ranges from ‘fair’ to ‘almost perfect’ (Kappa 0.36-1.00) for physiotherapy trials, and from ‘moderate’ to ‘almost perfect’ (Prevalence And Bias Adjusted Kappa 0.51-1.00) in pharmacological trials. The inter-rater reliability for both the total PEDro score and individual PEDro scale items has been shown to increase with consensus ratings generated by groups of two or three raters. Evidence to support construct validity has been reported for the PEDro scale and the total PEDro score, with data supporting the total PEDro score to discriminate between high-quality and low-quality trials. Convergent validity is supported for the total PEDro score through correlations with other quality rating tools: the Jadad scale (r = 0.35) and van Tulder 2003 scale (0.71) for trials of physiotherapy interventions; and the Cochrane Back and Neck Group risk of bias tool (0.61) for trials of pharmacological interventions.
While the authors report that total PEDro scores of 0-4 are considered ‘poor’, 4-5 ‘fair’, 6-8 ‘good’, and 9-10 ‘excellent’, it is important to note that these classifications have not been validated. Furthermore, for trials evaluating complex interventions (e.g., exercise) a total PEDro score of 8/10 is optimal.