Everybody wants the data . . . That’s the phrase I hear most often bandied around in healthcare discussions these days. Employers will get involved with employee healthcare and promote digital apps to collect data. Providers want to collect data and /or sort through the data they have on their populations. Drug discovery companies want access to data from specific patient groups.
But is it the data or what comes out the other end after crunching the data? Sure, there are some companies well-positioned for processing large data sets, designing algorithms to sort, correlate and pull out new never-before-realized associations. But the bulk of interested parties? Not so much. The interest there is most likely to be the end-results of associating data with phenotypes. In other words, if I had a good data set on people relevant to my organization or cause, and a powerful tool box, I could make predictions of behavior, health results or the like with a reasonable chance of success.
So, these powerful tool boxes – what’s the deal? These seem to be the pesky methods that the patent office classifies as “abstract” and fall outside of the eligible area for patenting. Or if the algorithms are specific enough or coupled to a specific piece of hardware, patenting may be possible, but the risk of competitor design-around becomes the issue.
An alternate approach – the trade secret route. But then today I read a comment by one healthcare company that illustrated the frustration on this side as well. The example, sensors and apps used to record data for patients in clinical trials. The problem – the accuracy and variability of the devices and difficulties in standardization. But the pharma companies dealing with these devices couldn’t get at the algorithms to solve this problem because they were closely protected trade secrets of the device company.
What would happen if instead of trade secret, these algorithms could be adequately protected by patents? Now you might be taking one of at least two potential views here. You might argue that algorithms in some forms are protectable, you just need a good and experienced patent attorney in this area. I won’t disagree completely with this view. But even with the expertise, it’s not an easy path.
The second view you might take is the “oh no we can’t allow patents to cover natural and abstract ideas.” I suppose if you patent an entire field, perhaps like patenting E=mc2, that could have some issues. But patenting some slice when the discovery is how to use that formula for something more than an abstract equation, that to me seems reasonable. With the ability to protect these tools from infringement with patents, companies might be more willing to share and integrate their systems.
Back on the data side . . . who does or who will own all of this data that everyone is so interested in? The individual data points, does any of that reside in the ownership or control of the individual? Shortly after JP Morgan, I ran across Nebula Genomics which is using a combination of blockchain and cryptocurrency to give individuals some part in guiding the use of their data and potential reimbursement. It would seem to me that a micropayment system and a system for specific permissions (e.g., opt-ins) for access to stored health data might work. I’ve also heard suggestions of data consortiums, where patient groups with a shared disease could consolidate their data and offer it to interested companies for some sort of compensation, or early involvement in the drug development.
Last week at a Nasdaq healthcare-related event, an informal audience poll suggested many people would not object to organizations collecting and accessing health-related data if they could be assured that the data would not then be “against them,” i.e., for insurance rates or job-related decisions. However, I wonder if a more participatory case-by-case opt-in system could bring along benefits. For example, knowing that a company was collecting fitness app data and having to opt-in might employees’ increase awareness and curiosity about the app and the fitness goals it promotes.
Putting the data and the tools together, what will we do with all this information? This question is not so much on the R&D side, but for the general populace. Knowing correlations and predictions in healthcare doesn’t always drive the outcomes we might aspire to. We all know the link between smoking and cancer – but I still find plenty of cigarette butts lying in the streets of SF every day. We know about the dangers of a diet loaded with refined sugar, but the grocery stores are still full of cheap sugar-loaded foods out-numbering the healthy options. Perhaps we’ll find the magic correlation that will drive awareness to behavior change – I’ll keep my eyes open for that one.