eJohn D. Halamka, MD, MS, is Chief Information Officer of Beth Israel Deaconess Medical Center, Chairman of the New England Healthcare Exchange Network (NEHEN), Co-Chair of the HIT Standards Committee, a full Professor at Harvard Medical School, and a practicing Emergency Physician. He specializes in the adoption of electronic health records and the secure sharing of healthcare data for care coordination, population health, and quality improvement.
The following has been paraphrased from an interview with Dr. Halamka on December 15th, 2017.
(Click here for the full audio version)
What prompted you to write the white paper, A Case Study For Blockchain In Healthcare?
According to the Gartner hype curve, blockchain, machine learning, and AI, are at the peak of hype right now. I think it is important that people understand that blockchain is great, but it is not a panacea. There are certain cases where it can be useful. That paper was to help people decide where to use blockchain and where not to use blockchain, as well as assess its potential to help solve some healthcare problems.
What developments have you seen in applying blockchain technologies to healthcare since you wrote that?
One example comes from malpractice suits brought against clinicians. What often happens is a plaintiff attorney asks for the medical record and then claims that is has been faked. Doctors then present the audit trail only for the lawyer to claim that it has been falsified as well. Well imagine that we took the note a doctor wrote and the second they signed it we hash it and put it in a blockchain, which is tamper proof. Even if 20 years go by we can guarantee that note has not been altered. It adds some built in trustworthiness to the way we collect and use data.
Another example comes from South Africa where I was last weekend. The Gates foundation would like to unify the HIV data there but the infrastructure is very challenging. There are real question about whether the data they have has integrity. A blockchain would allow us to ensure that the data we see has not been corrupted.
How do you see machine learning playing a role in improving interoperability(the ability of computer systems or software to exchange and make use of information)?
Machine learning is an incredibly useful technique, we just have to be careful when stating the use cases. Is IBM Watson going to replace your doctor? No. Will machine learning help your doctor refine your diagnosis? Absolutely. Especially with rare diseases where a doctor will be able to turn to a computer which has scanned through 10,000 similar patients and produce a list of the few things he or she might have. I’ve used these kind of machine learning techniques on over a dozen projects to date and really what they do is help us filter the signal from the noise. It helps us turn data into wisdom.
The challenge with interoperability is that doctors are overwhelmed. If I give them a thousand blood pressure readings from your medical records, what are they going to do with that? Whereas machine learning can read that information and say that it looks like the blood pressure of this patient has trended up.
Remember that a doctor only has 12 minutes to see you, enter 140 data elements, be measured in 40 quality domains, and never commit malpractice. It’s not possible. With machine learning a doctor can actually spend time with you because it will do a lot of those non-cognitive tasks behind the scenes.
How far away are we from, as the title of your recent book states, Realizing the Promise of Precision Medicine? What are the biggest barriers that stand in the way?
We are already using precision medicine today and the example I give is with my own wife’s cancer care. In her case we were able to ask, ‘for 50 year old Korean women who have this kind of cancer, what have their treatments been and what were their side effects and outcomes?’. This was just using an open source tool that is now available across 60 different academic healthcare centers.
As for barriers, take genomic sequencing. We don’t really have a good way of storing sequences and exchanging biomarkers to enable us to better prescribe treatments. There is work to do on standards and on codifying the data beyond the standard medical record. There are a few other things we are going to have to work on as a society to get precision medicine like how to document patients’ preferences and take data that they generate and put that into their medical record, this would also include their mood, tolerance for medication, amount of exercise or even Fitbit data. But that future is already here, it is just not equally distributed. Some places have adopted a lot of these precision medicine techniques but it is by no means ubiquitous. That is still about 5 years away.
Where do you see the most innovation in precision medicine coming from?
It’s going to be from you. What I mean is that it’s not going to come from some giant corporation. It is going to be individuals who create in their garages unique applications or knowledge sources that connect to the existing electronic health records.
I call it the deconstruction of the electronic health record. Just this week I listened to 60 pitches from 26 year olds working out of their garages. The innovations they are suggesting aren’t even on the radar screen of the big corporations. They are generally small improvements done through apps and cloud supported services that make the electronic health records more useful.
Which fields of medicine do you think will have the most difficulty moving away from the ‘one size fits all’ approach to healthcare that we have had up till now?
I am 55, I trained in emergency medicine 30 years ago where I was taught by my chief resident the very best way to do something. Doctors in general are artists, they are apprentices, they are trained through their life experience. I’m not sure I can narrow down one specialty that will be resistant to change as all specialties will be resistant to change. As a 55 year old my brain is relatively set in its ways. But the people being trained today are not memorizing data like I had to, they are figuring out how to be knowledge navigators. It is up to this next generation of doctors to replace the incumbent doctors and embrace precision medicine.
What role do you believe patients should play in advancing medical science and healthcare?
What we really need is a shared medical record where the patients, family members and caregivers are all using the exact same data set, with each equally contributing to that data. Think of it more like a Facebook for medical care than the electronic health records we have today. I think that is the best way to get to a system that has wellness and outcomes as our goals, not just more episodes of care.
You note your frustrations with the system in America in your blog series Dispatches from a Broken Healthcare System. If you could redesign the system from scratch, what foundational principles would you instill?
I travel the world and I see that every country has a slightly different healthcare system. My favorite is Denmark. They have one medical records system, no barriers to patient-doctor data exchange, as well as an understanding that sharing medical data records benefits society. My worry is that what is wrong here in America is not our healthcare system as much as it is our culture. We have a thousand disconnected insurance companies with their own incentives, 500,000 disconnected doctors with their own incentives, 5,000 different acute care hospitals that hate each other, and on top of that a sense that our data is private because ‘if somebody finds out I have high blood pressure it will diminish my standing in the community.’ A good healthcare system should have data liquidity across all of its constituents while still respecting privacy.
For more from Dr. John Halamka visit his blog, Life As A Healthcare CIO.
Great interview! I think that in the near future block will find an ever wider application. For example, in tourism, in the banking sector, in the music industry