I posted a question on Twitter asking for thoughts/ideas/SoPs with regards to data sharing and incidental findings. There was quite a bit of discussion, but it was suggested that I should post this here also.
In short, I am working on early draft documentation to support Open Access within our imaging centre. I’ve been trying to put together a data deposition procedure; however, it occurred to me, that it is not straightforward when it comes to IFs. I had 2 scenarios in mind (I didn’t discuss the second on Twitter).
A neuro incidental finding
A screening finding (i.e., screening for exclusionary criteria; where a participant scores outside of the desired range in, for example, a clinical measure)
I was interested to know what people’s thoughts were about this regarding this. There is some discussion on Twitter, but further thoughts/ideas very much welcomed. The first priority is always to conduct research in an ethical way with regard to the participant, but this is of course not straightforward; informed consent requires that we, in our capacity as researchers, consider the implications and convey all of those to the participant.
If anyone has any thoughts on the topic, I’d be grateful of them. I will be putting draft documents up on the Open Science Framework once they are further developed, so will share progress in due course.
Anyone interested generally in the topic, I highly recommend Open Brain Consent I’ve found this resource invaluable.
the participant has consented to the study.
2a) The data are available in a public server and the consent had explicit allowance for it
OR
2b) The data are available in a public server in deidentified form as per current HIPAA guidelines.
A neuro incidental finding
This scenario, even without the data deposition requirement, varies from institution to institution. Some institutions mandate radiology review and many don’t.
From a personal standpoint, I believe that:
the choice of whether such information should be provided to the participant should be left to the participant
Depositing the data in non-identifiable form is not a breach of a consent
From a researcher standpoint, the consent should clearly have the above options:
If an incidental finding is observed, would you like to be notified of it - Yes | No
I consent to having my data shared with other qualified researchers - Yes | No
I consent to having my data shared publicly - Yes | No
I believe this also requires clarity in the consent document of these questions as well as an evaluation of participant understanding of these questions.
A screening finding (i.e., screening for exclusionary criteria; where a participant scores outside of the desired range in, for example, a clinical measure)
I am generally in favor of giving the participant the information. It is better to receive such information from an informed person rather than to randomly chance upon it using some online or other assay. Most screening should happen post consent. Therefore, it can be made clear during the consent process whether the participant wants this information or not.
The first priority is always to conduct research in an ethical way with regard to the participant, but this is of course not straightforward; informed consent requires that we, in our capacity as researchers, consider the implications and convey all of those to the participant.
Very much agree with this sentiment. Another view is to consider the research data as belonging to the participant and allowing the participant to share that data with the researcher or with anyone else. And in this context, to ensure that the participant decides what can be done with the data and that the participant understands at which point the data becomes anonymous.
As this is an evolving area involving many different types of data collection, I would recommend reading this repository and associated documentation: https://github.com/Sage-Bionetworks/PCC-Toolkit