Comment by SquidMagnet

How do we ensure a proper association between the samples taken and the person in question, in particular as described in the case from the article, especially without some sort of formal framework and authorization before hand?