One of the greatest missed opportunities in health care is all of the “dark data” out there — this includes all the critically useful data sets detailing specific patient treatments and health outcomes that are hidden to doctors and researchers let alone to the entire medical field. In an age of data and modern machine learning, it shouldn’t be so hard for this data to see the light of day.
But even though such data is key to making medical advances, it remains “dark” because of the structure of the healthcare system: Providers don’t have any incentives to share data with patients or other physicians outside their network; even if they did, legacy systems and antiquated tech makes it difficult. Similarly, there is no natural incentive for Payers to share data either. And then on top of all this, HIPAA is often (incorrectly) perceived as an unsurpassable barrier to the act of data sharing itself. Put together, these factors have built up a lot of friction against sharing data in the healthcare system over the years.
The key to making advances here is to align incentives — obviously beginning with the patient — but it’s considerably easier said than done. Most patients don’t even know they can get their data. And when they do, they face numerous hurdles: First, the data is typically scattered across multiple payers and providers from which they have received care over the years; second, legacy systems — often driven by one-off patient portal systems that vary from payer to payer — lead to data exchanges in practically incomprehensible formats; third, there’s no interoperability among the data formats, which makes any practical use of the data unlikely.
Yet at the same time, technology is already empowering patients in other ways, through trends such as home technology, wearables, telemedicine, and social media. Patients are also already empowered by law and policy to own their own data — they, unlike the many entities they interact with in the healthcare system, are not bound by HIPAA. In fact, patients not only have the right to ask for their entire medical records, they can choose to share or provide them to third parties too.
But how would patients taking command of their data change healthcare? Let’s consider a patient who has just gotten the life-changing diagnosis that they have cancer. Having access to all of their data (blood tests, imaging, genomics, etc.) is critical for coordinating their care, giving them the options, choices, and the chance to make the best possible — fully informed — decisions for their care. The patient could also get immediate and frictionless access to lifesaving therapies via clinical trial matching and personalized treatments. And since many patients need critical help from family members during this difficult time, data sharing would enable others to help them far more easily and fully.
And there’s more: Even though the immediate interest and incentives lie with patients, providers would naturally benefit from that data as well. Doctors could have all the data they need to help their patients right at their fingertips — without friction, without dealing with data aggregation issues — giving them a complete, longitudinal record into their patients’ full medical histories, which are poorly related and interpreted by patients themselves anyway. The combination of patients bringing their data to providers would thus revolutionize healthcare.
Payers, too, would also see numerous benefits — especially given the ongoing push towards value-based healthcare — not only from providers fundamentally increasing their efficiency (i.e., no repeated tests or unnecessary procedures) but also from other areas such as detecting fraud and preventing over-prescription. For instance, there are many social factors that contribute to the opioid crisis, but on the medical side, some of the issues there arise from the lack of longitudinal records. With more data, payers would get a clearer and more complete look into the treatment patients are getting, providing them much-desired transparency into what they’re covering.
But there’s still more that changes once dark data sees the light of day, going beyond how healthcare is performed. Perhaps the greatest impact will be drug development: The drug design process is often limited by the lack of data, especially human data. So the ability to learn from large populations with longitudinal datasets would revolutionize the development of novel drugs and therapies, helping push cures faster. Moreover, pharma would have a means to proactively gather real world evidence, connect patients to clinical trials directly and efficiently, and directly engage with patients (which is practically unheard of today).
Unfortunately, none of the above is possible without breaking down the structural silos, misaligned incentives, and legacy tech issues — and frictions — of the existing healthcare system. But that’s where modern technology comes in: Like a river flowing around hard rocks, it can bend around traditional systems through software that lets the data talk to each other, making the data exchange happen without friction.
Huge sets of unstructured PDFs don’t teach machines to recognize Glioblastoma on a CT scan, because having a bunch of clunky documents and scans that someone sent you on your phone or hard drive is very different than having machine-readable, actionable data. This is where a solution that doesn’t just provide data sets, but rather creates computable data sets, comes in. And without individualized patient records, there will be no much-promised precision medicine in our futures.
Which is why I’m thrilled to announce that the a16z bio fund is leading an investment in Ciitizen, a platform that helps patients securely collect and share their personal health data. By putting the patient in the middle, suddenly everything in the system falls into place. Only it’s not so sudden in implementation: It requires deep understanding of industry know-how, tech, and product to navigate the system and create a truly patient-centered solution like this. Ciitizen has built a world-class team with these very chops, beginning with co-founders Anil Sethi, Premal Shah, and Brian Carlsen. A longtime serial entrepreneur and product visionary in healthcare, CEO Anil’s last company, Gliimpse, was acquired by Apple. It began with the simple idea that anyone should be able to manage their own medical records, but was motivated by the harder reality of helping his own sister fight her battle with cancer. COO Premal has deep experience in working from inside the healthcare system, and especially in managing complex medical data from the molecular to system scale. Brian, who was formerly special projects lead at Apple, will be the Director of Clinical Informatics at Ciitizen. Joining them are other leaders in their respective fields, including Deven McGraw, who has deep experience in HIPAA and medical data sharing, and was formerly the head of privacy at the U.S. Department of Health and Human Services.
This is the future of health care: patient-controlled, portable data, programmable health records, and improved personal and precision medicine. A system that aligns incentives among all the players by putting the patient at the center, beginning with bringing the dark data into the light.
Vijay Pande is a general partner at Andreessen Horowitz where he leads the firm’s investments in companies at the cross section of biology and computer science including areas such as the application of computation, Machine Learning, and Artificial Intelligence broadly into Biology and Healthcare as well as the application of novel transformative scientific advances. This article first appeared here on the a16z website.