From IBM’s press release:

IBM today announced its Health Cloud and Watson cognitive computing capabilities will support health data entered by customers in iOS apps using Apple’s ResearchKit and HealthKit frameworks. The move, which complements IBM’s new Watson Health business unit, will arm medical researchers with a secure, open data storage solution, as well as access to IBM’s most sophisticated data analytics capabilities.

And:

Apple’s HealthKit enables developers to create consumer health apps designed to give users a more comprehensive way to manage their health and fitness, and ResearchKit is an open source software framework that gives medical researchers the tools to accelerate medical studies. HealthKit can be used to create apps that monitor health behaviors and help encourage users to adhere to their treatment plan. Participants who opt-in to apps using ResearchKit can also contribute their health information to medical research/science/medicine.

IBM will de-identify and store health data in a secure, scalable cloud system that enables researchers to access and share data in an open ecosystem environment, as well as have access to IBM’s data-mining and predictive analytics capabilities. Health and fitness app developers and medical researchers will be able to draw on data at a scale that until now has never been available. For apps using HealthKit and ResearchKit, IBM will provide a delivery platform through Health Cloud to easily store, aggregate and model data, combining it with other data sources and types to enrich research findings and identify the next frontiers of medical discovery.

This is a perfect partnership. Apple puts products in people’s hands to help gather the health data, then IBM steps in with analytics via Watson.

Interesting use of the word de-identify. I’ve not heard that word before, but from context I assume that means IBM will strip out all identifying information from the health data. That’s a tough thing to do. There are the obvious things like name and ID, but sometimes the health data itself can be used to identify someone. The combination of traits can act like biometric markers, in a manner similar to fingerprints. Data miners long ago figured out how to reverse engineer a specific set of browser habits into a specific identity, all without an IP address.

Not a complaint or worry, just an observation.