Efy Vayena, Anna Mastroianni, and Jefrey Kahn recently published a beautiful policy piece on informed consent in Science Translational Medicine titled, “Caught in the Web: Informed Consent for Online Health Research.”

The paper addresses the growing imbalance between the restrictions of traditional informed consent approaches and the way the Web enables rapid, cross-discipline data collection. It expresses a hopeful set of ideas about this imbalance as an opportunity to open up a conversation about informed consent, not a “problem” to be “solved”. From the paper:

Respect for individual autonomy, balanced and equitable distribution of risks and benefits, and the right to privacy have not lost normative weight because the social Web challenges the approaches we have created to respect them. On the contrary, the challenge presents an opportunity to explore more nuanced and innovative ways of interpreting these values.

A significant portion of the paper is devoted to Sage Bionetworks’ pioneering work in developing Portable Legal Consent – the idea that, when it comes to personal health data, consent should be in the hands of the individual rather than the institution or clinic running a study. The authors are generally favorable to PLC but note four perceived potentially problematic areas:

(i) PLC participants must be willing to give up control over their personal health data, including a limited right to withdraw and the choice to decline specifc research projects;

(ii) deidentifying of data limits its usefulness for some research projects;

(iii) PLC relies on a self-selected, well-informed population of computer-savvy users who are unlikely to be representative of the population at large; and

(iv) PLC cannot be used for Web data collected for nonresearch purposes.

Let’s look at each of these areas in turn.

First, point (i) is absolutely correct. The entire idea behind PLC is data donation, not a conditional gift like an open source copyright license. In copyright licenses, I can give something away for free but also assign conditions to it – if you don’t give me attribution when you make a copy or a use, your license evaporates, and I can sue you for infringement. As there is no copyright on our personal data, this power doesn’t exist. That makes it extremely difficult to achieve both openness and conditionality. We’ve chosen to create a truly free zone in our consent work because we’re pretty sure companies will charge into the space to offer, under contracts with punitive terms and conditionals, more granular approaches to privacy. And that’s as it should be – a well functioning world should welcome both the commons and the entrepreneur.

Second, the de-identification issue. This may be conflating the arguments made by Paul Ohm in his paper Broken Promises of Privacy: Responding to the Surprising Failure of Anonymization – which lays out the tension between the anonymization of data and its usefulness. The most famous epigram from which is that data may either be anonymous or useful, but not both. But the data in PLC are in fact not anonymized – the identifiers are removed, but we also create a unique identifier for each participant so that data entered over time connects to data entered earlier. This creates an increasing possibility of identification over time, as the scientific value of the data increases as well. We cover this as part of our consent process specifically. We also permit re-contact of consented individuals via these identifiers. The problem with de-identification is that it’s not perfect, not that it limits usefulness, and that must be addressed in consent.

Point (iii) is to date unfortunately correct. Our initial population is clearly coming from a self-selected group who are predisposed to accept the terms, risks, and benefits. We will be looking to expand our pool of participants enormously this year through the Bridge platform, and through other efforts in an attempt to address this.

Point (iv) is factually inaccurate. In our consent form (available for download) there is no restriction to what data a participant may upload, whether gathered for research purposes or not. I’ve uploaded my genotype, purchased from 23andme, and my electronic medical records, gathered from my physician. Data can come from anywhere and be uploaded. Indeed one of the core functions of PLC is to work through people who are motivated to convert their regular health data into research data. The Personal Genome Project achieves a similar function.

This is the real power of portable consent. It lets a person who wants to be part of research to join the conversation through their data. Until now, the only ways to get involved were passive – enroll in a study someone else ran, whose data you never saw, or donate money whose returns were murky. Portable consent – and its progenitor open consent – invert the power structure and give the individual a seat at the research table.

We praise Vayena et al for a lovely policy piece, and our comments are simply clarifications to an article that deserves wide attention.