[Critical analysis of the ISUIA study: the methodological point of view]
Fiche publication
Date publication
mai 2008
Lien Pubmed
Résumé
ISUIA is a prospective study that ran from 1991 to 1998 and included 4060 patients from 161 centers. Analysis of the methodology focused on: (1) sampling: inclusion was more intensive towards the end of the study period, which was likely to generate a selection bias; (2) a possible indication bias, as direct comparison between observation and treatment groups showed that patients receiving preventative treatment were more at risk of complications; (3) the statistical analyses, which never compared these groups directly, although that would have been the best approach for answering the research question. The size and site of aneurysm appear to be risk factors that were also never included in the multivariate analysis; (4) the outcome criteria, which were either not comparable (spontaneous rupture rate), not compared (disability rate) or presented in a non-comparable way (mortality rate by the Kaplan-Meier method versus raw estimates); (5) interpretation of the results, which should be made with caution as, with no statistical calculations provided, it is assumed that the risk of rupture is constant over time. In spite of the practical difficulties such a study would raise, only a randomized controlled comparison of strategies would offer a clear basis for improved clinical practices, given the fact that increasingly smaller aneurysms can now be identified.
Référence
J Neuroradiol. 2008 May;35(2):104-8