This study investigates aspects of the reliability of the Maastricht criteria list for quality assessment in systematic reviews, and whether blinded reviewing is necessary to prevent review bias.
We used the data set of 12 articles from a systematic review concerning the efficacy of balneotherapy, in patients with arthritis.
Twenty reviewers participated of which two reviewers, who have been involved in developing the Maastricht criteria list, acted as reference standard.
Half of all assessments were performed blindly.
A high level of agreement was found between the reviewers and a high level of correlation with the reference standard.
The quality scores between the blinded and unblinded assessment did not differ much.
Based on the results we conclude that the Maastricht criteria list is a reliable instrument in quality assessment of clinical trials.
Within the limits of this study we found no evidence that blinding is necessary to prevent review bias.
Mots-clés Pascal : Balnéothérapie, Arthrite, Essai clinique, Evaluation, Qualité, Article synthèse, Critère, Méthodologie, Fidélité test, Accord interjuge
Mots-clés Pascal anglais : Balneotherapy, Arthritis, Clinical trial, Evaluation, Quality, Review, Criterion, Methodology, Test reliability, Interrater agreement
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 98-0221727
Code Inist : 002B30A01A1. Création : 11/09/1998.