Your data monetization questions answered! By Atul Butte and Jim Stolze
Your data monetization questions answered! By Atul Butte and Jim Stolze27 April 2022 | 6min
In February we had the pleasure of hosting both Atul Butte and Jim Stolze in our latest virtual event – Data monetization in healthcare: How to safely share your data for the sake of science. We invited you to join our two speakers as they attempted to answer some of the difficult questions surrounding data monetization in healthcare. They discussed many topics including how to leverage the billions of dollars worth of data being poured into our healthcare system and how artificial intelligence (AI) systems can help us to do this safely.
In case you missed the live event and prefer to read all about it, take a look at our summary article: Data monetization in healthcare: Sharing data for the sake of science.
As always, we received a huge volume of interesting questions from our audience during our live Q&A session. While Atul and Jim did their best to answer as many as possible, many were still left unanswered.
So, in this special follow-up article, Atul and Jim give their answers to your insightful questions.
The future of blockchain in healthcare
Q1: Is tokenization and blockchain technology the future for secure data sharing?
Jim Stolze: I believe it is. If you look beyond the hype around bitcoin and consider cryptocurrencies as just one application of blockchain technology, you come to the objective conclusion that ‘the blockchain’ has never been hacked. Despite many attempts and the obviously very lucrative results show if one would have been able to do so. So, if you wish for a future in which data sharing is safe, you need to add ‘robust’ to that wishlist and blockchain may grant you that wish.
At the same time, there are also other technologies that are considered safe and robust. You could store your data on an encrypted hard drive, in a vault, in an atomic bunker with no in or out coming internet traffic. But how useful is that? What I think blockchain brings to the table, is that it’s robust and programmable. The data is not stored in one place and you can create so-called smart contracts that adhere to the elements of your data. It’s you that is in charge. You can give and revoke access for specific use and specific users.
The role of the patient in the future of data monetization
Q2: In the spirit of the low code/no code trend, is it possible to simplify AI tools so that even patients and consumers could potentially analyze their data and manage their care in the future?
Jim Stolze: What a great observation. In my opinion, this is what AI should do. It shouldn’t just benefit the medical professionals, but the patients just as much. It shouldn’t create extra layers of complexity, it should make things easier. Just like updating a website in 2000 was a lot of work, you needed to know about FTP (File Transfer Protocol) and HTML (HyperText Markup Language), but today it’s just an app on your phone. AI will do the same for complex statistical analysis, with a lens that is tailored to your personal situation.
Atul Butte: Yes, indeed I think AI-driven tools to help patients understand their own data is the future. We (at least in the United States) are certainly delivering more and more health data to the patient, through standards like FHIR (Fast Healthcare Interoperability Resources). But what’s still missing are the tools to help a patient understand their own data. That’s going to change, as AI trained with health data is used to best inform a patient what could be next for them, and what might be missing in their health care.
Q3: Do you think the best place to store the data is in the hand of the patient themself?
Jim Stolze: This could be a metaphor that I’m not familiar with. But I don’t think it’s the patient’s responsibility to store and guard the data. This should be facilitated so that the patient only needs to worry about who gets access and how he/she is included in the process.
Atul Butte: Certainly patients should have as much data on themselves as possible (and when appropriate, their family members). But we as health systems cannot depend on patients having their data when we need it. We also need to keep the data we have on our patients as well. For example, we often will need to reach out to patients between encounters, to remind them of preventive care measures they are missing. So health systems need to safely keep and maintain health records as well
Tackling bias in data sets
Q4: How do you ensure the accuracy of AI algorithms given that there may be bias in the data sets?
Jim Stolze: Accuracy is a percentage that you try to get as high as possible, without doing too much harm to your model (over-fitting). Bias tends to get in the way of this. So every AI project starts with an assessment of the data. Never assume that your dataset is perfect because it’s not. Just like human beings have biases, algorithms are neutral until they are applied to data.
Atul Butte: We believe that AI trained on larger and larger data sets, with data from more diverse individuals, is the best way to ensure the most generalizable and robust algorithm possible. But algorithms need to be externally validated, by regulatory bodies such as the US Food and Drug Administration. Algorithms coming into health systems should be stewarded as well, the same way we carefully manage new drugs coming into a health system and being listed in the formulary.1
Q5: Is there a risk that Real-World Data (RWD) could be a source of bias considering that the data generated is from patients who have access to relatively better medical care to start with?
Jim Stolze: Good point. The technical term is survivor bias. You only get the results of cases that made it all the way through. This means that you know nothing about why or what other cases didn’t make it into the dataset.
Atul Butte: Yes, but this really depends on what the desired use is for the data, i.e. the questions one wants to ask. Of course, health systems do not have data on people who do not come to that health system, so one always has to recognize what a data set is and is not good for answering.
Globalization and the potential impacts of data monetization
Q6: How do you see the integration and monetization of data vary across different geographies/countries, different systems, billing methods, and technologies? Especially when you are trying to pull in a specific population’s health/medical journey across different years and different countries.
Atul Butte: Of course, medical practice and how medicine is paid for is different across the world. So the regulations around how data is collected and responsibly used are quite different, but the value of having computers search for what’s actually working and not working in medicine, what might be missing in a patient’s care, is eventually going to be a worldwide need.
Q7: How does data monetization influence commercial models in a healthcare setting?
Atul Butte: There are now many hundreds of AI algorithms already approved by the US Food and Drug Administration, and most of these are being developed by commercial entities. So having access to data to train computers in a safe, responsible way, is likely going to be quite valuable for health systems, companies, and patients.
Want to learn more about data monetization and healthcare? Then check out our article with Bryn Roberts, The possibilities and pitfalls of monetizing healthcare data.
Jim Stolze is a tech-entrepreneur and a prominent figure in the European startup scene. In 2009 he was approached by TED.com to become one of their twelve ambassadors worldwide. Between then and 2016 he was the driving force behind TEDxAmsterdam and many other TEDx events in Europe, the Middle East and even the Caribbean. An alumnus from the prestigious Singularity University (California) Jim Stolze is a thoughtleader and changemaker in the field of exponential technologies. Since 2017 Jim focuses on Artificial Intelligence (AI). With his platform Aigency he connects algorithms from PHD’s and startups to data-sets and challenges from big corporates. This initiatieve was labeled by the media as “the world’s first employment agency for artificial intelligence”.
Atul Butte, MD, PhD is the Priscilla Chan and Mark Zuckerberg Distinguished Professor and inaugural Director of the Bakar Computational Health Sciences Institute (bchsi.ucsf.edu) at the University of California, San Francisco (UCSF). Dr. Butte is also the Chief Data Scientist for the entire University of California Health System and has authored over 200 publications, with research repeatedly featured in the New York Times, Wall Street Journal, and Wired Magazine. Dr. Butte was elected into the National Academy of Medicine in 2015, and in 2013, he was recognized by the Obama Administration as a White House Champion of Change in Open Science for promoting science through publicly available data. Dr. Butte is also a founder of three investor-backed data-driven companies: Personalis (IPO, 2019), providing medical genome sequencing services, Carmenta (acquired by Progenity, 2015), discovering diagnostics for pregnancy complications, and NuMedii, finding new uses for drugs through open molecular data. Dr. Butte trained in Computer Science at Brown University, worked as a software engineer at Apple and Microsoft, received his MD at Brown University, trained in Pediatrics and Pediatric Endocrinology at Children's Hospital Boston, then received his PhD from Harvard Medical School and MIT.
- Eaneff et al. (2020). JAMA, 324, 1397–1398. Paper available from https://jamanetwork.com/journals/jama/article-abstract/2770772 [Accessed April 2022]