Thanks for your answers! I have also spoken to Adrian Mann at the MDA about this and asked for his opinion.
He said that recording the initial values, recording future tests as a Pass/Fail only and investigating any appreciable change seems a good compromise.
