Sharing information is one of the most intensely debated topics in genetic medicine today. What should we share? How should we share it? With whom should we share it?
There are many ongoing discussions and evaluations. But are we being as considered and exhaustive about this complex area as we need to be? Are we asking the right questions? Are we finding the right solutions?
Should we share genetic information?
The first question we need to answer is whether we should share genetic information. The answer is, unequivocally, YES. If we do not accept and embrace the philosophy of sharing genetic information we simply cannot deliver genetic medicine. To serve the individual we need information from the masses.
This is different from many branches of medicine. The use of new drugs and new tests typically relies on validation studies of many hundreds of people. But once we know a drug works, we do not need information from other people to decide how we use the drug to treat the patient in front of us.
In genetic medicine day in, day out, we need and use global genetic information to treat the patient in front of us. When we do a genetic test we find hundreds or thousands of differences from the reference genome. We need information from other people to work out what these differences mean. To decide which are innocuous changes and which cause medical problems.
Balancing benefits and risks
So we must share genetic data to practice genetic medicine. And our first task is to get better at explaining this need, so people can review the benefits and risks appropriately. If we get this right society will decide that, overall, the benefits of sharing far outweigh the risks. This will allow us all to move the discussion on from whether we should share information, to how we should share information.
Mitigating the risks
The benefits of sharing genetic information are incontrovertible, but this doesn’t mean we can be cavalier or dismissive about people’s concerns about sharing their data. We need to be diligently attentive to these concerns, and transparent about how we are mitigating risks. The risks are not so different from the risks of sharing other types of personal information. So we can, and should, be able to learn from other activities. How they are securing data. And, equally importantly, how they are securing confidence.
How should we share information?
Many different individuals, groups and global initiatives are working hard to build the practical and political infrastructures required for genetic information sharing. Inter-operability is perhaps the most important requirement, because we cannot afford to limit our access to useful information.
All health systems need to provide genetic medicine to citizens of diverse ancestry, and they therefore need information about all populations. Moreover, many genetic conditions are very rare and we need the combined knowledge, data and expertise of the world to provide appropriate care for such patients.
It is unfeasible and undesirable to have a single place in the world where all relevant data is held. Rather, we will continue to have a system of different types of data being housed in many centres, in many countries. Secure, simple inter-connectivity between data warehouses is therefore essential.
The Clinical Annotation Reference System (CARS) that the TGMI is developing, is our contribution to this inter-operability. The CARS will help make exchange of genetic testing data easy, consistent and accurate.
Sharing meaning not just dataWe should not forget that sharing information in genetic medicine is not about sending data between computers. This is simply a means to an end. Our goal is sharing information between people. And for this to be successful it is not sufficient for the information to be delivered, the meaning of that information must also be correctly transferred. How often does this happen? Do we measure it? Do we even know how to measure it?
Sharing meaning is hard
Trying to get people to take the same meaning from information can be very hard. Not least because the meaning we attach to information can be influenced by many different factors, even if the information itself is concrete.
For example, in genetic medicine we often tell couples they have a 25% chance of having a child with a genetic syndrome. Sometimes you see relief in their eyes because this is much lower than they had anticipated. Equally often you see heightened anxiety because it is much higher than they had hoped. What we don’t know is whether we have said 25%, but somehow they have equated this with 90%, or if they have heard 25% but had thought it would only be 5%, or some other permutation. Of course I may be wrong in the meaning I am inferring from this facial expression data!
What we do know. What every one of us knows from our personal and professional lives, is that communication between humans is hazardously unpredictable. All too often there is a disconnect between what we think our words mean and what others think they mean.
Using simple, concise, consistent language helps
So sharing meaning is hard, and we all know this. And yet even though we know this we constantly make it even harder by using complex, inconsistent language to transfer information. We use intellectual disability, mental retardation, developmental delay, intellectual impairment and learning disability interchangeably to describe the same problem, often within the same patient record. A variant can be described as ‘pathogenic’, ‘damaging’, ‘deleterious’, ‘functional’, ‘clinically relevant’ or ‘clinically actionable’ to mean the same thing, or sometimes to mean different things!
We need a holistic approach to information sharing
To make information sharing deliver for genetic medicine we need a holistic approach that integrates the diverse practical, philosophical, political and personal considerations.
And we need to keep in our sights, at all times, why we are doing this. We are trying to share the meaning of genetic information to help people make vital decisions about their lives.