Study limitation reported by the research below: “The study design label 'Meta-Registry Observational Study' is the closest standard label; this is not a traditional meta-analysis but a pooled analysis of an international registry (PVRI GoDeep). Causal inference is not supported.”