I was between A and D - selected A first, then D in blind review. The explanation for A says that it is descriptively inaccurate because "many instances" are not being described to justify the conclusion.

The reason I thought the stimulus was describing many instances is because of the "(as happens to roughly half of all test subjects)" and "(as also often occurs)." So I thought that they were saying because there are many instances of people accepting the risk and then receiving placebos or ineffective medicines, that the columnist doubts the medical ethicists' claim. I thought the many instances referred to all of the people.

I think I was overthinking it. The part talking about the many test subjects being given placebos or ineffective medicine just isn't part of the premise.