As more technology companies move into biomedical research, Sage Bionetworks' John Wilbanks and the Scripps Research Institute's Eric Topol write at Nature that there's a danger that any data that companies gather will be inaccessible to patients and researchers.
Though the entrance of private tech companies into biomedical research would likely enable large-scale studies and spur the development of new algorithms to analyze such data, Wilbanks and Topol worry that the data and tools will become proprietary "black boxes." They note that pharmaceutical companies keep data on the effects of certain chemicals on blood measurement to themselves and that algorithms developed by Google and Amazon can intensify societal biases: job postings served by Google to female users are less well-paying than those served to men and the algorithm Amazon used to select neighborhoods in which to roll out its one-day delivery service in the US was more likely to choose neighborhoods in which more white people lived.
Wilbanks and Topol argue that the creation of open-resource competitors would help regulate closed markets, and to that end, networks of open projects and people willing to donate their health data are needed. At Sage, the pair writes, patients enrolled in clinical trials are able to download their own data and are given the option to donate their data for reuse by other researchers, and about 75 percent do.
"[I]f the new era of digital health is accompanied by citizen-led pushes for more openness, it may not just be healthcare that is transformed," the duo says.