Questions: Does the PEDro scale measure only one construct ie, the methodological quality of clinical trials? What is the hierarchy of items of the PEDro scale from least to most adhered to? Is there any effect of year of publication of trials on item adherence? Are PEDro scale ordinal scores equivalent to interval data?
Design: Rasch analysis of two independent samples of 100 clinical trials from the PEDro database scored using the PEDro scale.
Results: Both samples of PEDro data showed fit to the Rasch model with no item misfit. The PEDro scale item hierarchy was the same in both samples, ranging from the most adhered to item random allocation, to the least adhered to item therapist blinding. There was no differential item functioning by year of publication. Original PEDro ordinal scores were highly correlated with transformed PEDro interval scores (r = 0.99).
Conclusion: The PEDro scale is a valid measure of the methodological quality of clinical trials. It is valid to sum PEDro scale item scores to obtain a total score that can be treated as interval level measurement and subjected to parametric statistical analysis.