In a situation where somebody has wound up in a hospital, there are a million things that can go wrong at any given second, either through the patient's body failing somehow, pure bad luck or the hospital staff either making a mistake, or not having an essential piece of knowledge about the patient and situation they're dealing with. Google-owned A.I. startup DeepMind, who you may remember as the creators of the A.I. that beat the current world champion in the ancient Chinese board game of Go, has partnered with U.K. hospitals in the Royal Free NHS Trust to help reduce the occurrence of that last type of mishap by alerting staff to a major and somewhat common risk factor; risk of kidney failure.
Naturally, determining which patients are at risk for kidney failure and making sure staff are informed requires access to a vast amount of patient and hospital data, which is exactly what the Royal Free NHS Trust has granted to DeepMind. To be precise, the A.I. firm now has access to a fair amount of data on things like tests and health records for about 1.6 million patients within the Royal Free NHS Trust's hospital network. This data will allow them to assess which patients may be at an elevated risk of kidney failure, which can be agitated by tons of things such as medicine or certain surgical procedures, and intelligently alert any staff that will be working with that patient via a special smartphone app.
The move is not without its share of controversy, of course; there are those who would rather not have their own data in Google's hands, just as there are those who would prefer that Google not have extensive data on 1.6 million patients. According to the Royal Free NHS Trust, however, this data transmission is not only protected in the same ways as normal transmission of medical data, including via heavy encryption, but patients need only to contact the hospital network's data protection officer. As more data is gathered, however, it's quite possible that DeepMind could expand the alert beyond kidney issues by creating a medical testing record algorithm of sorts, giving staff a laundry list of caveats whenever working with a patient in the data set. It's worth noting that Lord Darzi, a prominent surgeon in the network, has given the app and the organization behind it his full endorsement and a vote of confidence that the data sent back and forth will be fully secure and only used for medical purposes.