Why does Kylie Jenner’s tweets matter for healthcare data?
Patient-Generated Health Data (PGHD) such as the one produced by connected devices, social media or mobile apps, can be very useful to better understand the patient’s health in a continuous manner and complement nicely the discrete information collection in EMRs at the point of care. We would have a better understanding of the patient’s activity between two encounters with health professionals, and potentially improve observance in chronic diseases, we would have visibility into the patient’s living conditions, which could be very useful in diseases such as Alzheimer’s.
Combining the real-life PGHD with clinical data produced by healthcare providers in EMRs and genomics databases would offer providers and researchers an incredibly powerful data repository to improve treatment outcomes, preventive care and discover innovative, highly targeted treatment options beyond the healthcare product.
The potential is mind-boggling and it’s critical to overcome foreseeable challenges when it comes to practical implementations of this concept.
First and foremost, securing the trust from the patient is paramount and trust in our era where social communities rule is volatile. Social leader opinion leaders often carry more weight than institutional leaders : as a recent example, the February 21st Kylie Jenner tweet on her doubts about the new Snapchat app tanked the stock by a billion dollars on the same day ! Imagine a social media fitness opinion leader expressing concerns about the handling of her PGHD, his followers will immediatly stop sharing their health data and trigger a similar trend among their own followers. Patients need to be in full control of their PGHD and always be aware of who uses it, trust protocols such as the one brought by blockchain can be a good enabler here.
Second, knowing that a Google probably accumulates 1 or more petabyte of data every data (that’s 1 followed by 15 zeros !), we will need to upgrade our big data repository from petabytes to exabytes (1 followed by 18 zeros) or even zettabytes (21 zeros !). Perhaps more challenging is the lack of standards adoption in the world of healthcare data: as we approach 2020, data interoperability between EMR is still embarrassingly slow and painful, how in the world are we going to take the next steps and interoperate with all the other PGHD and genomics systems? It’s really urgent for regulators and lawmakers to take forceful action to drive standards adoption and get us back on track here.
Finally, there’s a fundamental question of model: centralized or decentralized. Up to now, healthcare big data repositories were often built using a centralized model: there’s a central entity which governs the repository and defines the rules by which participants access and share the data. The French « Dossier Médical Partagé » is an example of centralized model. Decentralized means participants have an equal weight in the governance of the system — including the patient, and participants agree amongst themselves the rules by which information is shared and monetized. Blockchain is an example of decentralized model. Don’t get me wrong, I am not suggesting centralized is all bad, and decentralized is all good. Years back, you didn’t have any other choice than to do a centralized model for healthcare data sharing, there were no other way to deal with data privacy and security safeguards. I am suggesting that adopting a new decentralized model for healthcare data sharing can upgrade the patient’s position in the ecosystem and bring more flexibility and momentum to providers and researchers for the benefit of patients.
