Google has made a deal for access to patient records from HCA, which which operates 181 hospitals and more than 2,000 healthcare sites in 21 states, so the tech company can develop healthcare algorithms, The Wall Street Journal reports.
Google will store anonymized data from patient health records and internet-connected medical devices. That data will be used to build programs that could inform medical decisions made by doctors. The deal is described as “multiyear” by the WSJ, without specifying how many years.
This feels uncomfortable on so many levels.
I do not know the specifics of this project. However given its sensitive nature, and the strength of HIPAA laws, it would be reasonable to assume privacy would be taken seriously.
That being said, many hidden patterns of drug, disease, protein and gene interactions can only be inferred by looking at very large samples of data, provided no individual is singled out (https://en.wikipedia.org/wiki/Differential_privacy).
Previously, this was done be individual doctors, reading lots of research papers.
Then there are methods developed to go over existing medical literature to help those researcher discover these patterns (let’s plug out paper here: https://pubmed.ncbi.nlm.nih.gov/22693501/)
That could be then used to provide better treatment plans based on individuals (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5898022/ this was one of my professors).
And now this is being done with machine learning by insurance companies, NHS like countrywide systems, or private companies.
The future will definitely have better understanding of disease. As long as we don’t assume the systems as infallible (they do make mistakes), and not single out individuals Gattaca style.