John,
It is reassuring that we share sensitivity to the issues researchers would face if they had to individualize consents to meet each patient's unique requirements.
We appear to part ways (at least for purposes of this discussion) on the best ways to compose and memorialize medical research consents.
Regarding consent composition, perhaps, it is best addressed by developing a few model consents that afford different levels of control to data subjects. In this way, review boards might then challenge researchers on why they didn't use the model with the most user control and the researcher would be obliged to justify his selection from among the models. I don't know that allowing subjects full ability to tailor their consent won't raise the objections you cite of sample size and cost. And I don't know that giving subjects such power has much impact on the doctor who recommends participation in the wrong study for the wrong reasons.
Regarding memorializing the consents, I could be persuaded that merely giving a subject a copy of a signed consent may be too burdensome on the subject. Storing and tracking the document (in hard copy of digital form) is such a rare event for most subject that they are unlikely to have a good system for doing so that allows for ready retrieval when an issue arises. But I also suspect that adding such a consent to a blockchain may be unnecessarily costly. It could just as easily be stored in a standard database by some independent administrator of such research consents that would operated like a library. The funding for the administrator could come from the same source as funding for a blockchain.
I also suspect that giving individuals additional consent opportunities with regard to re-use of research in which they were a subject could dramatically reduce the opportunity to re-use the data. If a large study has 10,000 subjects and the data from it could be used for another study, the new researchers have merely to show that their reuse is consistent with the original consent. If they have to go to each subject to obtain a new, separate consent, the database could well become unusable. Subjects may have new addresses (physical or digital) and tracking them down too costly.
I had a similar experience when I wrote a musical several years ago. The entire point of the musical was to re-use existing rock songs that told a story. I assumed that a theater could just pay ASCAP or BMI their licensing fees and be free to have the songs performed in the context of the musical. Not so. Bands can play copyrighted songs under ASCAP and BMI. But a musical requires obtaining unique consent from every publisher. No theater would consider my musical until I first obtained consent from each publisher. I began contacting the publishers. After a few replies such as, "Mr. Morrison needs to know how many seats are in the theater and what is the price per seat," I abandoned the effort. I see individualized consents causing the same problem.
I suppose it could be argued that if we can't obtain such individualized consents, perhaps the research is not in the public interest. Perhaps that current practices is tilted too strongly towards the researchers. But I would not want to be the spokesperson representing individualized consents -- even though I am a strong adherent of the rights of the individual.
Jef