Topic Archives: Symposium
Larry Smarr’s major contributions to scientific progress are well known. A physicist and the founding director of the National Center for Supercomputing Applications (NCSA), he helped bring the power of computing to scientific research at a time when computers will still highly specialized instruments. Today he is the Director of the California Institute for Telecommunications and Information Technology (Calit2), one of the most innovative research institutes in the world. He’s also an avid self-tracker, using his own data to correctly self-diagnose the onset of Crohn’s disease. In preparation for our upcoming Quantified Self Public Health Symposium I asked Larry about his idea for a large scale, non-commercial, broadly accessible infrastructure for improving access to self-collected data for both personal and public benefit.
Gary Wolf: What’s the role of the public health sector and of the academic research community in a world where individuals and consumer-oriented tech companies are taking on increasingly complex questions of personal and population health?
Larry Smarr: The fundamental role is bridging the gap between N=1 and N=a lot. Any time in the last 30 years when I’ve seen a technical innovation that mattered, like a software tool, the first approaches aren’t ready for prime time. They are not developed with professional-level software engineering, version controls, documentation and all that. Similarly, scaling up of biomedical observations made by N=1 quantified individuals is going to require the professional methodologies of the public health sector.
GW: Can this be left to industry?
LS: Not entirely, although startups are doing a fabulous job of getting tracking tools into the hands of tens of millions of individuals. The problem is how to do research on the data produced by that broad population. Too often these days I see researchers from the university going to tracking companies and asking for access to the company’s raw data feeds, for instance to heart rate or exercise time series, and the company says no. They will give you the weekly or daily average, but you can’t get to the raw data. If you go to them and say, I’ve got this really great innovation that can be used to understand this data, more often than not they decline. They have an installed base and market share to protect, which naturally tends to make them conservative. I think there is a real opening for companies to make this anonymized broad population data available to academic researchers. That’s when a raft of scientific discoveries will be made from the quantified population.
GW: Those are the consumer fitness companies, but what about the healthcare IT world?
LS: Again there is a disconnect between the consumer fitness cloud-based apps for millions of individuals and the electronic health records in your healthcare provider. If you’re a doctor in a medical office, unlike a data science researcher, you don’t want all this data. What you want to know is: did my patient do 1000 steps or 10,000 steps today, did you get aerobic exercise or not, are they getting enough sleep? So it’s not like you need a vast dumping place inside electronic health records. Again, I think pilot experiments are the way to get started.
GW: You’re arguing that the incentives aren’t there.
LS: These are currently major structural barriers. Who is going to work on the bridging we are discussing? There aren’t incentives for the commercial tracking companies to work on it. Neither are there incentives for the electronic medical record companies to work on it. NIH isn’t going to support bridging between commercial companies. It falls between the stools. You need to have the research community, and health care IT experts, the commercial tracking companies, and the individual self-trackers all come together and collaborate.
GW: You envision some kind of technical system so that individuals and health care providers and researchers could all benefit from access to data. What does your experience tell you about how long this would take to have a working prototype that would be practically useful?
LS: It’s a three-to-five year project. I think if a major funder did a call for proposals requiring a health care provider, university research community and the self-tracking community to come together a prototype a solution, I think they would get some very interesting proposals.
GW: In a talk you gave in 2011, you said “science is not enough.” You pointed out that we’ve known the link between smoking and cancer for over half a century, and yet global cigarette consumption has tripled during this time. So we have all this possibility for new discoveries with self-tracking data, but how is that going to help make people healthier?
LS: Yes, just knowledge of what causes negative impacts on health is not enough. My former UC San Diego colleague Naomi Oreskes documents how economic interests slowed down the logical social reaction to smoking health threats and climate change in her Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming (2010). We are seeing similar delaying and disinformation tactics in the obesity/diabetes epidemic, which has been building for four decades. It is sobering to me to see someone as politically skilled as New York City mayor Bloomberg defeated in his efforts to ban jumbo sugary drinks. My best guess is that we face a multi-decadal battle, just as we have had with tobacco and climate change, to get our society to move to healthy eating and drinking. The bright spots are subcultures of healthy living, often empowered by tracking and social media, that are developing across the country. My hope is that these will spread and scale over the next decade.
GW: It seems you are also pointing toward activism, since that’s been so important with smoking.
LS: Activism is essential given the enormous power of the entrenched economic interests. Activism can lead to regulatory reform, which over time can make huge social changes. For example, when I grew up in the 1950s and early 60s my father didn’t smoke, but he was embarrassed that you had to have ashtrays in your house, because he said you couldn’t tell people not to smoke in your own home. Socially, you just couldn’t. About that time the Surgeon General’s report on smoking was published. Fifty years later, huge chunks of society are smoke-free, such as all the University of California campuses, restaurants, and large social gatherings. Just think of what an enormous shift that has been! We are beginning to see similar activism in getting pension fund investors to boycott carbon fuel companies in order to slow down climate change. So can we imagine a boycott against sweetened beverages and high glycemic prepared foods? I believe that there is a huge role for health-related individual and organized activism in the near future.
GW: At the last Quantified Self Public Health meeting, you suggested that this emerging field needs a new kind of journal where individuals can report their discoveries. In light of the big challenges you’ve been describing, challenges that can’t be solved by academic and research publication alone, what kind of contribution could a new journal make?
LS: Let’s go back to the issue of scaling we discussed. Imagine the journal articles are fairly short, describing how the data was generated, but the back end is a publicly available cloud of data so that you could begin growing a large dataset of N=1 projects. Then the research community could pick up on the ideas coming out of the Quantified Self community, explore the data, and take it further. That’s how things grow.
GW: You want to be on the editorial board?
LS: No, I want to submit a paper!
As part of the Quantified Self Public Health Symposium, we invited a variety of individuals from the research and academic community. These included visionaries and new investigators in public health, human-computer interaction, and medicine. One of these was Jason Bobe, the Executive Director of the Personal Genome Project. When we think of the intersection of self-tracking and health, it’s harder to find something more definitive and personal than one’s own genetic code. The Personal Genome Project has operated since 2005 as a large scale research project that “bring together genomic, environmental and human trait data.”
We asked Jason to talk about his experience leading a remarkably different research agenda than what is commonly observed in health and medical research. From the outset, the design of the Personal Genome Project was intended to fully involve and respect the autonomy, skills, and knowledge of their participants. This is manifested most clearly one of their defining characteristics, that each participant receives a full copy of their genomic data upon participation. It may be surprising to learn that this is an anomaly in most, if not all, health research. As Jason noted at the symposium, we live in an investigator-centered research environment where participants are called on to give up their data for the greater good. In Jason’s talk below, these truths are exposed, as well as a few example and insights related to how the research community can move towards a more participant-centered design as they begin to address large amounts of personal self-tracking data being gathered around the world.
I found myself returning to this talk recently when the NIH released a new Genomic Data Sharing Policy that will be applied to all NIH-funded research proposals that generate genomic data. I spent the day attempting to read through some of the policy documents and was struck by the lack of mention of participant access to research data. After digging a bit I found the only mention was in the “NIH Points to Consider for IRBs and Institutions“:
[...] the return of individual research results to participants from secondary GWAS is expected to be a rare occurrence. Nevertheless, as in all research, the return of individual research results to participants must be carefully considered because the information can have a psychological impact (e.g., stress and anxiety) and implications for the participant’s health and well-being.
It will not be surprise to learn that the Personal Genome Project submitted public comments during the the comment period. Among these comments was a recommendation to require “researchers to give these participants access to their personal data that is shared with other researchers.” Unfortunately, this recommendation appears not to have been implemented. As Jason mentioned, we still have a long way to go.
This week we’re taking a look back at our 2014 Quantified Self Public Health Symposium and highlighting some of the wonderful talks and presentations. We convened this meeting in order to bring together the research and toolmaker communities. Both of these groups have questions about data, research, and how to translate the vast amount of self-tracking data into something useful and understandable for a wider audience.
As part of our pre-conference work we took some time speak with a few attendees who we thought could offer a unique perspective. One of those attendees was Margaret McKenna. Margaret leads the Data & Analytics team at RunKeeper, one of the largest health and fitness data platforms. In our conversation and in her wonderful talk below Margaret spoke about two important issues we, as a community of users, makers, and researchers, need to think about as we explore personal data for the public good.
The first of these is matching research questions with toolmaker needs and questions. We heard from Margaret and others in the toolmaker community that there is a near constant stream of requests for data from researchers exploring a variety of questions related to health and fitness. However, many of these requests do not match the questions and ideas circulating internally. For instance, she mentioned a request to examine if RunKeeper user data matched with the current physical activity guidelines. However, the breadth and depth of data available to Margaret and her team open up the possibility to re-evaulate the guidelines, perhaps making them more appropriate and personalized based on actual activity patterns.
Additionally, Margaret brought up something that we’ve heard many times in the QS community – the need to understand the context of the data and it’s true representativeness. Yes, there is a great deal of personal data being collected and it may hold some hidden truths and new understanding of the realities of human behavior, but it can only reveal what is available to it. That is, there is a risk of depending too much on data derived from QS tools for “answers” and thus leaving out those who either don’t use self-tracking or don’t have access or means to use them.
Enjoy Margaret’s talk below and keep an eye out for more posts this week from our Quantified Self Public Health Symposium.
Personal data, personal meaning. That’s the guiding principle of much of the work we do here at QS Labs. From our show&tell talks and how-to’s, to our worldwide network of meetups and carefully curated unconferences, we strive to help people make sense of their personal data and inspire others to do the same. However, over the last few years we’ve started to see that there is a third actor in the Quantified Self space. Data collected in the ordinary course of life can hold clues about some of our most pressing questions related to human health and wellbeing. Personal data might be a resource for public good.
On April 3, 2014 Quantified Self Labs with support from the Robert Wood Johnson Foundation, the US Department of Health and Human Services, and Calit2 at UCSD hosted the first Quantified Self Public Health Symposium. We gathered over 100 researchers, toolmakers, science leaders, and pioneering users to open up a discussion about what it means to use personal data for the public good. Over the course of the day we hosted a variety of talks, discussions, and toolmaker demonstrations. This week we’ll be highlighting some of the outstanding talks delivered at the symposium and we’re kicking it off with one of our favorites.
Susannah Fox has been a friend and colleague for many years. Her pioneering work at the Pew Internet and Life Project has inspired us many times over and remains the standard for research pertaining to self-tracking. We asked Susannah to help us open up the meeting by discussing some of her research findings as well as her thoughts on self-tracking in the broader landscape of health and healthcare.
(A transcript of Susannah’s talk can be found on her website here.)