Search the KHIT Blog

Sunday, December 7, 2014

#Data4Health San Francisco listening session

Much to reflect upon. A natural follow-on from my prior post. Props to RWJF for sponsoring this listening roadshow.

I responded immediately to the post-conference Survey Monkey questionnaire. My free-form responses:
9. What did you think of the breakout topics?

They were fine, as far as they went. Severe time constraint, given the breadth and complexity of the topic. A lot of what I heard discussed really just falls under the category of "Upstream" issues. We will also need to more forcefully address going forward problems of "data quality," -- "accuracy," and the potentially damaging implications of its insufficiency.

10. Did you learn something you can use after the meeting?

Beyond the problems with raising public awareness, a principal takeaway for me is just how far we in the space have yet to go in clarifying fundamental topical definitions. And, the reinforcement of the fact that I seem to really have lost the pedantic fight to preserve the lexical reality that "Data ARE"! Writing "data IS" in a refereed scientific publication is the faux pas equivalent of farting out loud during a plenary session (beyond the fact that it would never get past the editors).

13. Do you have anything else to share you didn't get the chance to?

I will be summarizing my reflections on my blog. Haven't had time yet. Though, I wanted to raise the example of the work now being done at Google Research with their "Google Fusion" initiative, which seeks to surmount the heterogeneity of myriad "big data" formats in the service of structured data merging / synthesis for improved analytics.

Finally, I commend you for this important effort. Thanks to everyone involved.
On the "Upstream" thing. See my prior posts When it comes to health, your zip code matters more than your genetic codeand "The ultimate population health "Upstream" issue?"

And, below, from my August 13th post:

On the Google Fusion thing:
The overarching goal of structured-data research at Google is to build tools that enable a rich ecosystem of structured data on the Web. To fuel such an ecosystem we need to build tools for discovery, extraction, annotation, sharing, querying, integration, visualization and publishing of structured data sets.

Google Fusion Tables lies at the center of these efforts, offering a tool for data owners to easily upload, query and share data sets. Once the data is uploaded to Fusion Tables, the user can query the data, find related data sets that other users have made public and combine multiple data sets. Users can also create compelling visualizations easily and publish them on the Web. Fusion Tables has been used by a wide variety of users (see our gallery), most notably by journalists and in crisis response.

In our efforts to enable discovery of structured data sets on the Web we developed the WebTables System that extracts the HTML tables that contain valuable data. We extracted a corpus of over 150 million high-quality tables and have developed methods for searching tables in response to queries. In a previous line of work we developed techniques for tapping the content stored in databases behind forms, known as the deep web (or the invisible web). Our research has produced methods for automatically analyzing web forms and submitting meaningful queries to them, to obtain HTML pages that can be inserted into the Google index.
Worth your time to ponder the possible utility in the healthcare "Big Data" space.

Below, some shots I took:

Conference moderator Ivor Braden Horn, MD, MPH
Nice turnout.
RWJF's awesome Mike Painter, MD, JD
HHS's Karen DeSalvo, MD, MPH, MSc had to remote in.
Andrew Rosenthal of Jawbone touted their products' utility for personal health data.
Above, Jawbone Cloud data showing a huge spike in sleep disruption during the recent Napa earthquake.
I'm not a big Jawbone fan. I bought three a few of years ago, one for me, one for my wife, and one for my daughter. Mine and my wife's both went bad in short order. More hassle than it was worth to try to get support service. I hope they've improved their products. I now have a FitBit. It's not a whole lot better. Moreover, there's some troubling news on the quantified self wearables front. See my post "Big Data" and "Surveillant Anxiety."

apropos, I've just been asked to read and review Vik Khanna's irascibly iconoclastic new book,

My favorite fitness app story is that of an acquaintance who told me how the FitBit had improved his life, prompting him to walk more at work and even have his employer provide him with a standing desk. This would all be a very positive step for the many Americans who move too little and sit too much, but this guy is one of the fittest people I know, a physician and an avid cyclist. His habits have already lowered his manageable health risks as low as they can go; using a FitBit isn’t going to change them further. For him, as for many people, tidbits like the FitBit are toys that entertain. I don’t know anyone who needed a FitBit less. On the flipside, there is not much evidence that apps, gadgets, and websites are really having much of an impact on the health of anyone who really needs to find a way to change. Why is that? Because if you are not internally motivated to succeed, a digital toy will not take you there.

The toys, apps, and websites that we want to do the work for us are the electronic equivalent of unused treadmills and weight machines currently functioning as expensive clothes hangers in basements across the country. Health success doesn’t start with trivia such as “Which running app do you use?” It starts with fundamentals, such as “Do you run [substitute the exercise of your choice] at all?” And, if you don’t, why don’t you? Because somewhere along your road in life, people stopped telling you it was important to do so and that is was okay not to, that it was your choice. Well, see, what happens when you make a lot of bad choices it creates an opportunity for government and institutions that want your money and obedience to come in and start to control your choices. When you start acting like a sheep, someone is going to eventually shear you.

Maybe someday all the hype about wearables will translate into something beyond making money for the people who make the devices and their related apps. But, until then, the wearables you need to invest time and money in are your attitude, your strategy, and how you wear them for the world to see. Funny thing about self-image and self-respect… the people I know who carry themselves with the most confidence and poise don’t use any of this stuff. They stick to the basics, the indispensables. Everything else is just gravy.

Vik Khanna. Your Personal Affordable Care Act: How To Avoid Obamacare (Kindle Locations 317-335).
Vik recently threw some serious cold water into the comments at the THCB post "Health Data Outside the Doctor's Office."
"While there are an enormous number of uses for the data that we can imagine and many more we cannot yet anticipate…” 

This is a fancy way of saying we have not decided how we will use these data to manipulate people, scare them, control them, and make them come to believe that, contrary to the opening of this post, the government and its private sector enforcers, really are your health salvation, because, to paraphrase Deep Throat, that’s where the money is. There is no profit in leaving people alone. 

You need no data about me and how I live my life. You are entitled to nothing that I do not want to share with you voluntarily. The federal government used to gather data the very old-fashioned way, by doing surveys. Is there any evidence that we did not get useful information that way? And, even if it was less than ideal, was it so deficient that sacrificing privacy, consent, and voluntary participation are worth the price of admission to this brave new world of “we need all the data we can get?” Al Lewis and Granpappy Yokum above have it nailed exactly right. 

Now, the data geeks have all taken over and want to bamboozle the rest of us into believing that their technology-and-data-are-magic-fairy-dust mythology. The shroud of benevolence obscures a more subversive agenda. Bureaucrats and technocrats never do things that do not serve their own interests first: bigger budgets, longer titles, tenure (in the case of academics), etc. 

How I eat, exercise, work, enjoy myself, is none of your g.d. business. And neither is how a create a culture of health in the unit that matters the most…my family. 

It is long past time for Americans to tell government bureuacrats [sic] to take a hike...

I am as disdainful and distrustful of big philanthropy as I am of big government. In my view, big philanthropy is one of the government’s private sector enforcement tools because of the facade of benevolence that it projects. In reality, there is a revolving door between big government and big philanthropy leaders, and, even more important a long history of the two groups playing footsie with each other while they all accrue more power, more money, and more authority over the lives of individuals.

I would tell people to openly and vigorously reject cooperating with your initiative.
Ouch. Tell us how you really feel, Vik.

Toward the end of his book, Mr. Khanna addresses the Upstream:
Acknowledge explicitly that the healthcare industry cannot fix social and community dysfunction. Medical care cannot make up for lousy schools (also an area in which there is too little competition and efficiency and too many bloviating bureaucrats), broken families, churches with pastors and priests who break every commandment they preach from the pulpit, and communities with no economic prospects. Medical care will not fix someone who does not respect himself. 

Fund and advise communities on how to make their environments safer for cyclists, runners, and walkers. And, I don’t mean just create lanes, which are a good start. Fund creation of entire networks of parks, greenways, and alternative routes for getting around 24/7, and then pay for the police to patrol those areas and keep them safe. 

Fund health and physical education in every school in America from pre-k through graduate and professional school, and make schools compete for grant money. How is it possible college students arrive at that stage of life not knowing how to take care of themselves? 
This morass is our fault. It is the natural cascade of adversity that results when individuals largely fail to exercise their personal will in a positive manner, and we invest too fully and too naively in professions and industries that have motivations and incentives antithetical to our own. Our transformation into an information culture actually worsened the malady. We are so conditioned to success at the speed of a search engine that, like that person who aspires to retire early, but refuses to save or invest, we’ve forgotten to manage the fundamentals. First, that every healthy lifestyle decision you make today, from diet and exercise to outlook and mood requires thought and an exertion of will. Even in the age of Google, your choices matter, and choosing, not just wisely, but strategically, is an option available to most people. Healthy lifestyle is a lifelong pursuit that requires modification and flexibility as you age and your circumstances change. [op cit, Kindle Locations 3019-3037]
More on his book later.


From a 3rd JASON Report (pdf).
Data for Individual Health

Executive Summary

The promise of improving health care through the ready access and integration of data continues to draw significant national attention and federal investment. Information technology is rapidly expanding this data universe beyond traditional information associated with health care providers to embrace information in the larger spheres of health and wellness. This includes not only electronic health records (EHRs), but also personal health records (PHRs) and sources such as environmental data and social media data, some of which may be related only indirectly to the delivery of health care. To date, federal investments in health data infrastructure development, through mechanisms such as the Centers for Medicare and Medicaid Services (CMS) Medicare and Medicaid Electronic Health Record Incentive Program, have focused on the medical care of individuals. This report discusses how to expand this vision, with a focus on the health of individuals and the development of a Learning Health System...
  1. 1.3 Summary Today, the delivery of health care moves in a linear fashion, proceeding from preventive medicine, to diagnosis, to treatment, and ultimately to outcomes. This process is informed by clinical research, but there is an inadequate feedback loop between health care outcomes and clinical research, reducing opportunities for further learning in this system. Additionally, population health research and community engagement are not adequately connected. A “Learning Health System” would connect the medical system with broader societal inputs, creating important links between health and wellness and health care. This concept highlights natural roles for EHRs and PHRs, but also points to a level of data access, integration, and scalability that goes well beyond the interoperability of EHR systems...
  3. 1.5 Data Associated with Health
    In extending the ideas of the JASON-proposed architecture to the broader realm of health, it will become necessary to expand greatly the types of data that can be ingested and analyzed. In addition to the traditional data associ­ated with health care (e.g., EHR data), it will also be necessary to assimilate data from PHRs. These include, for example, data from personal health devices, patient collaborative networks, social media, environmental and de­mographic data, and the burgeoning data streams that will soon become available through progress in genomics and other “omics.” Despite the pro­ fusion and complexity of new data sources associated with personal health, the architecture for a learning health system would look essentially the same as that proposed by JASON for EHRs except that the data layer must also encompass these highly diverse forms of personal health information. The requirement for interoperability through the adoption of open APIs becomes even more critical here; without this interoperability it will be extremely difficult to scale up today’s health information technology (IT) systems to assimilate and analyze these new data sources...
I study and reflect on all of the JASON stuff. I'll hold my fire on this one until I've fully digested it.

Another of my tweets.

Obtuse analogy? Where do "data" ("do," not "does") fit into concepts such as the (controversial phrase) "art of medicine"? Are we subtly yet significantly more than the sum of our "structured data"?
Philosophia sana in ars medica sana?


Improving the secure availability and use of health information allows individuals to take ownership over their health, partner with their health care providers, and improve their quality of life and health. It strengthens the delivery of health care and long-term services and supports, and allows public health agencies to detect, track, manage, and prevent disease outbreaks. Information also fuels research and innovation, spurring advancements in scientific discovery. 

Health information technology (health IT) allows individuals and health care entities and providers, home- and community-based supports, and public health entities to electronically collect, share, and use health information. The term “health IT” includes a wide range of products, technologies, and services, such as electronic health records (EHRs), mobile and telehealth technology, cloud-based services, medical devices, and remote monitoring devices, assistive technologies, and sensors.
Federal agencies provide direct care and health insurance, protect public health, fund health and human services for certain populations, invest in infrastructure, develop and implement policies and regulations, and advance groundbreaking research. Given this range of activities, the federal government is also positioned to improve health, health care, and reduce costs through the secure use of information and technology. 

The Federal Health IT Strategic Plan 2015-2020 (Plan) identifies the federal government’s health IT priorities. While this Plan focuses on federal strategies, achieving the vision and goals requires collaboration from state, local, and tribal governments. Efforts by health care entities and providers, public health entities, payers, technology developers, community-based nonprofit organizations, home- based supports, and academic institutions are also essential.
Link to the full pdf report here.

More to come...

No comments:

Post a Comment