Gravar-mail: Dealing with clinical heterogeneity in meta‐analysis. Assumptions, methods, interpretation