Thanks Frank! Agree with these points, most is in the paper. The issue we also address in the scoping review is that most papers do not provide sufficient detail to even understand the output (e.g. no information of reference pop or algo used) and tendency to over-interpret results (causal language)
