Some position papers and opinions have restricted its benefit to few patients with very bad prognostic features, only because the benefit didn't show intragroupal statistic signification, ignoring the concept of "interaction". Methodological reviewers criticize this practical as a common mistake in subgroup analysis.
In this case, patients with not so high burden of disease don't have a statistically different outcome with docetaxel, as seen by the non-significant p-value for interaction. So, their overall survival benefit is not expected to be different to the full study results*. It seems inadequate to deprive those patients of an effective treatment.
Read the Abstract
Written By: García de Paredes Esteban JC1, Alegre Del Rey EJ1, Asensi Díez R2.
1. Hospital Universitario de Puerto Real, Cádiz (Spain)
2. Hospital Regional Universitario de Málaga, Málaga (Spain)
Sun X, Ioannidis JPA, Agoritsas T, Alba AC, Guyatt G. How to Use a Subgroup AnalysisUsers’ Guide to the Medical Literature. JAMA. 2014;311(4):405-411