Lancet meta-analysis have too much credit to theoretical