Advertisement

Opinion: A Truly Public Lab?

Should the healthcare system support health and wellness data tracking for the purposes of long-term observational studies?

By | March 21, 2012

Framingham State University, Framingham, MassachusettsWIKIMEDIA COMMONS, DADEROT

There should be one day each year where we turn to face the small town of Framingham, Massachusetts, and join in a global standing ovation of thanks. Never heard of Framingham? Almost everything we know about heart disease and health risk factors has come from the very public contributions of the Framingham community.

The Framingham Heart Study (FHS) began in 1948, when more than 5,000 residents of the town were enrolled in a long-term public health study. In 1971, the FHS Offspring study was launched. In 2002, the third generation of Framingham residents and family members was enrolled. The studies are observational, meaning data is periodically collected and aggregated, but there is no intervention by the physicians. All subjects receive standard preventative health and wellness care, a robust history is recorded for each, and a physical is performed every 2 years; all of this healthcare information is collected in a massive database.

This shared data has taught us fundamental lessons about the risks of smoking, of “bad” and “good” cholesterol, of high blood pressure, and of diabetes. More recently, the data has been used to answer questions that the FHS founders never would have thought to ask. By looking at relationships and friendships over time, new research teams have begun to explore how obesity, alcohol consumption, or mood trends can move through a community. From these newer studies, we are beginning to learn about social influences on health.

But more than any of these individual insights, there is a greater lesson to be learned from the FHS: large-scale collection and sharing of healthcare data can go a long way towards advancing our understanding of health and disease, and sometimes it can answer questions we didn’t even know we had.

The big picture and the “big data”

All of the lessons learned from the FHS program are possible because of the structured collection of health information—great things can happen when health information is shared, collected, aggregated, and studied in the right way.

First, any tool that supports the collection and standardization of health information can provide immediate individual benefits—the simple act of measurement provides awareness and motivation for the individual often leading to healthier decisions.

Second, sharing health information and finding ways to measure our data and experiences against others’ can provide perspective—how people are tracking over time, and how they compare to people like themselves. More recently, we have learned a lot about the value of online peer-to-peer health communities—the more engaged a user is, and the more she contributes to the community, the greater the benefit she herself enjoys.

Third, sharing health data can benefit the public health—allowing research organizations to aggregate health data across a growing subset of the population can provide novel insights into broader health correlations and trends. The computational sciences behind social network analysis and the greater availability of “big data” allow health and wellness patterns to be studied in ways never before possible. We can track contagion, we can track mood, and we can find patterns of health and illness that may inform research agendas, accelerate drug development, and alter public health policy. In short, our ability to overcome today’s emerging healthcare challenges would be severely hampered without such openness.

A public laboratory for healthcare research

The benefits of sharing healthcare data are exemplified by the FHS. But the FHS was a controlled environment: data was shared, but it was not truly public, and to this day, permission must be sought from the FHS review committee to access this data for research. More recently, it’s been argued that to maximize the fullest potential of such health research efforts, these and other data should be placed in open-access databases available to scientists all around the world.

In keeping with a more “public laboratory” model, we are now beginning to see small pilot public health research studies launch in conjunction with various health tracking sites and devices. These studies run the gamut from “quantified self” sites allowing you to collect data on weight, diet, sleep cycles, activity levels, or mood; to patient communities like PatientsLikeMe allowing patients to place their disease history, current symptoms, and narrative experience into the cloud. But where these pilots will take us is largely unknown. The data collection itself serves to unify patient communities, and if the data can be structured, there is every reason to believe that we can produce more and more Framingham-like experiences.

Of course, such projects will no doubt come with tremendous risks as well—health and wellness information gets at the heart of the most personal and vulnerable elements of being human. In the abstract, this data reveals our imperfections, casts a light on disease, and may stigmatize the afflicted. In the concrete, this data impacts our employability, insurability, and financial security. The fact is that we have no track record from which to predict whether broader public sharing will end up doing more good than harm, and we are still grappling with ensuring the anonymity and security of these data.

So while some, such as journalist and author Jeff Jarvis, argue that the greater good of humanity depends on pushing “publicness” as far as it can go, at this moment, I am a fair bit more guarded. When it comes to healthcare data and information, there is too much at stake to blindly promote a public laboratory for research. For now, my suggestion is to take ownership of your health, ensure that you have access to your data, and begin to digitize and track your own health story—this can only help when confronting an unexpected injury or illness—but hold off on placing your data in the “cloud” and be cautious when sharing your data with third-party companies to aggregate and publically mine.

At this point, it is impossible to balance known benefits against unknown risks. But if you ask me again in 2 or 3 years, I bet my perspective will have changed.

Brian S. McGowan is an independent consultant and author focused on solving the healthcare quality crises. He has worked as a medical scientist and educator and he is the author of the forthcoming fall 2012 release of #SOCIALQI: Simple Solutions for Improving Your Healthcare. 

Advertisement

Add a Comment

Avatar of: You

You

Processing...
Processing...

Sign In with your LabX Media Group Passport to leave a comment

Not a member? Register Now!

LabX Media Group Passport Logo

Comments

Avatar of: Sky_is_Falling

Sky_is_Falling

Posts: 20

March 21, 2012

As an advocate for open databases, I urge you to read Dr. John Ioannidis' work, such as http://www.plosone.org/article...
where many publications do not make data available for independent analysis, where many studies are biased http://www.plosone.org/article..., and where significance is exaggerated http://www.plosmedicine.org/ar....

There may be many invalid conclusions that waste a lot of taxpayer's money, as well as information that benefits us. Restrictions that limit access to databases delays research, but we cannot measure the cost of lost opportunity, so nothing is done to lift them. Scientifically, we know that epigenetics is more important than genomics, so to me, the risk of revealing my DNA sequence is less than having my financial information exposed online.

Avatar of: JohnMunroJr

JohnMunroJr

Posts: 2

March 21, 2012

Few question what on the surface or what in use in daily life appears to offer great benefits and no immediately identifiable hazards, all the way from use of lead in plumbing for millenia to use of Google to provide access to vast amounts of information very quickly and conveniently.  We now know of the serious health hazards of lead in paint and gasoline and are learning about the potentials for serious problems associated with data mining the vast amounts of data collected by Google, Microsoft (Bing), Yahoo, and the National Security Agency.  We use networked computers to operate most of our critical infrastructure and are still coming to terms with vulnerabilities to attack from an adversary in this area.  It is probably not possible, due to our limited experience, to understand the serious problems of abuse of health data that will be available in any national health service arrangement.  But all this data and more are being collected anyway: by local health clinics, insurance companies and private health providers.  All the health information is "out there" in some form or another, whether a government health program collects it or not.  As long as it is "out there" it is available for mining, either easily or with some difficulty.  The first task is to have some idea what questions to ask.

Avatar of: lonjones

lonjones

Posts: 17

March 21, 2012

Absolutely!!! Look at the value of Denmark, where they have records on everyone, and the ability they have to quickly tell if something is of value or not. The vaccine controversy is the most telling case in point.

From the other side, having local centers where new ideas could be tried out is much more effective than dissecting out reactions for new drugs that are filled with unintended consequences because of all the networked interactions they don't see in the analysis.

How effective, for example, is washing one's nose in preventing the nasal related problems of ear and sinus infections, allergies, and asthma? A local health center could check this out quickly and the whole country/world would benefit.

March 21, 2012

Thanks for the reply! 

This is the marvel of 'big data' in healthcare, there are so many great case studies, and many of these seem quite viable even if scaled up to the broadest national levels. Lessons from nations with socialized healthcare systems are promising. Lessons from clinical registries are promising. And, most recently lessons from groups like ACOR, Womensheart, and PatientsLikeMe are promising. 

But the question to be raised (and eventually answered) is 'what, if any, are the downsides for the individual whose data is supporting the public laboratory?'

March 21, 2012

Thanks for the shares and RT's!

Avatar of:

Posts: 0

March 21, 2012

As an advocate for open databases, I urge you to read Dr. John Ioannidis' work, such as http://www.plosone.org/article...
where many publications do not make data available for independent analysis, where many studies are biased http://www.plosone.org/article..., and where significance is exaggerated http://www.plosmedicine.org/ar....

There may be many invalid conclusions that waste a lot of taxpayer's money, as well as information that benefits us. Restrictions that limit access to databases delays research, but we cannot measure the cost of lost opportunity, so nothing is done to lift them. Scientifically, we know that epigenetics is more important than genomics, so to me, the risk of revealing my DNA sequence is less than having my financial information exposed online.

Avatar of:

Posts: 0

March 21, 2012

Few question what on the surface or what in use in daily life appears to offer great benefits and no immediately identifiable hazards, all the way from use of lead in plumbing for millenia to use of Google to provide access to vast amounts of information very quickly and conveniently.  We now know of the serious health hazards of lead in paint and gasoline and are learning about the potentials for serious problems associated with data mining the vast amounts of data collected by Google, Microsoft (Bing), Yahoo, and the National Security Agency.  We use networked computers to operate most of our critical infrastructure and are still coming to terms with vulnerabilities to attack from an adversary in this area.  It is probably not possible, due to our limited experience, to understand the serious problems of abuse of health data that will be available in any national health service arrangement.  But all this data and more are being collected anyway: by local health clinics, insurance companies and private health providers.  All the health information is "out there" in some form or another, whether a government health program collects it or not.  As long as it is "out there" it is available for mining, either easily or with some difficulty.  The first task is to have some idea what questions to ask.

Avatar of:

Posts: 0

March 21, 2012

Absolutely!!! Look at the value of Denmark, where they have records on everyone, and the ability they have to quickly tell if something is of value or not. The vaccine controversy is the most telling case in point.

From the other side, having local centers where new ideas could be tried out is much more effective than dissecting out reactions for new drugs that are filled with unintended consequences because of all the networked interactions they don't see in the analysis.

How effective, for example, is washing one's nose in preventing the nasal related problems of ear and sinus infections, allergies, and asthma? A local health center could check this out quickly and the whole country/world would benefit.

Avatar of:

Posts: 0

March 21, 2012

Thanks for the reply! 

This is the marvel of 'big data' in healthcare, there are so many great case studies, and many of these seem quite viable even if scaled up to the broadest national levels. Lessons from nations with socialized healthcare systems are promising. Lessons from clinical registries are promising. And, most recently lessons from groups like ACOR, Womensheart, and PatientsLikeMe are promising. 

But the question to be raised (and eventually answered) is 'what, if any, are the downsides for the individual whose data is supporting the public laboratory?'

Avatar of:

Posts: 0

March 21, 2012

Thanks for the shares and RT's!

Avatar of:

Posts: 0

March 23, 2012

John - not sure I could say it any better. Thank you.

But the first question I would ask is should the uncertainties that you and I both see in the evolving healthcare public laboratory lead us to a state of paralysis or should we begin to work through the end user challenges/habits of data tracking and empowerment such that once our questions begin to be answered we are quickly in a position to move forward (or retreat)?

Avatar of:

Posts: 0

March 23, 2012

Brian I fully support efforts to understand the risks and benefits of data mining health records.  Data mining will occur in any case, on a scale determined by ease or difficulty of access to the data: there is too much financial incentive not to do so.  Public discussion and education on what is happening needs to occur with much more urgency than it is getting currently.  In that sense we certainly need to move forward.

Avatar of:

Posts: 0

March 23, 2012

Thanks for the post!

I try to be as specific as possible when I open the door to 'big data' discussions. Whether it is patients self tracking, or healthcare systems quality data, or even basic science and open science in the truest sense - there are different arguments to be had in favor of and against each. I too am an advocate for open data (across all three silos). 

As it relates to biomedical science, I have argued to move beyond the research article as the 'dose of science' and believe that science should be shared at the level of the data, the experiment, or even the research question...the fact that we rely on 17th century publication models seems rather...17th century! But as you suggest we currently have a problem with misaligned incentives supporting a culture status quo.

The validity of science and the value proposition of gov't dollars going to support research seems to benefit from transparency and sharing, especially as it compares to the rigor of our current peer-review process. So we need to direct attention to the folks that are leading the transparency/sharing/big data efforts.

** shameless plug ** in my forthcoming book I discuss these issues at length and try to highlight groups that are innovating in this regard. Keep an eye out for "#socialQI: Simple Solutions for Improving Your Healthcare" to be published this fall!

March 23, 2012

John - not sure I could say it any better. Thank you.

But the first question I would ask is should the uncertainties that you and I both see in the evolving healthcare public laboratory lead us to a state of paralysis or should we begin to work through the end user challenges/habits of data tracking and empowerment such that once our questions begin to be answered we are quickly in a position to move forward (or retreat)?

Avatar of: JohnMunroJr

JohnMunroJr

Posts: 2

March 23, 2012

Brian I fully support efforts to understand the risks and benefits of data mining health records.  Data mining will occur in any case, on a scale determined by ease or difficulty of access to the data: there is too much financial incentive not to do so.  Public discussion and education on what is happening needs to occur with much more urgency than it is getting currently.  In that sense we certainly need to move forward.

March 23, 2012

Thanks for the post!

I try to be as specific as possible when I open the door to 'big data' discussions. Whether it is patients self tracking, or healthcare systems quality data, or even basic science and open science in the truest sense - there are different arguments to be had in favor of and against each. I too am an advocate for open data (across all three silos). 

As it relates to biomedical science, I have argued to move beyond the research article as the 'dose of science' and believe that science should be shared at the level of the data, the experiment, or even the research question...the fact that we rely on 17th century publication models seems rather...17th century! But as you suggest we currently have a problem with misaligned incentives supporting a culture status quo.

The validity of science and the value proposition of gov't dollars going to support research seems to benefit from transparency and sharing, especially as it compares to the rigor of our current peer-review process. So we need to direct attention to the folks that are leading the transparency/sharing/big data efforts.

** shameless plug ** in my forthcoming book I discuss these issues at length and try to highlight groups that are innovating in this regard. Keep an eye out for "#socialQI: Simple Solutions for Improving Your Healthcare" to be published this fall!

Avatar of: alexandru

alexandru

Posts: 1457

March 26, 2012

Mark 7.21 - 22 - *For from the inside, from a person's heart, come the evil ideas which
lead him to do immoral things, to rob, to kill, commit adultery, be greedy, and
do all sort of evil things; deceit, indecency, jealousy, slander, pride, an
folly.*

Actually, in addition with Adam mtDNA
inheritance theory and Paul, Hebrew 4.12, I scientifically explained this
paragraph because paternal mitochondria DNA, existed only in xiphoid process,
controls the heart pulse and the figurative elements of the circulating blood.

Attention please! To be health, do not live
in *informational dirt place*!
 

Avatar of:

Posts: 0

March 26, 2012

Mark 7.21 - 22 - *For from the inside, from a person's heart, come the evil ideas which
lead him to do immoral things, to rob, to kill, commit adultery, be greedy, and
do all sort of evil things; deceit, indecency, jealousy, slander, pride, an
folly.*

Actually, in addition with Adam mtDNA
inheritance theory and Paul, Hebrew 4.12, I scientifically explained this
paragraph because paternal mitochondria DNA, existed only in xiphoid process,
controls the heart pulse and the figurative elements of the circulating blood.

Attention please! To be health, do not live
in *informational dirt place*!
 

Avatar of:

Posts: 0

April 20, 2012

As a data protection officer, my primary concern is the protection of individual privacy. You seem to state that the direct positive value for the participating individual, is derived from the act of participating. As such, there seems no need for references in the data toward the identity of individual participants. Anonymisation is standard practice for social, as well as financial research using sensitive data. Do you see a reason why that should be less effective here?

Avatar of:

Posts: 0

April 20, 2012

Thanks Lewis - I think the answer to your question lies in how (and on what level) anonymity is managed. For 'big data' the more information collected (appropriately) the better for the analysis and more specific the conclusion that can be drawn. For the social engagement around the data, perhaps more anonymity (with some degree of identity) may be acceptable. For instance, engaging with LJ23 vs Lewis Johnson may be valuable, if I can still have some sense of who LJ23 is, what his experiences are...enough that LJ23 has some credibility - LJ23 has identity, but is still anonymous. 

My other on-going concern is around how anonymous can someone (or their data) actually be. Again for the data collection and analysis I think there is one line of innovation, but for the social participation around the data, I think we are learning that overtime the cloak of anonymity is often lost...eventually with enough participation and inspection I have seen anonymous participants be uncloaked.

That is to the point, that both the science of online anonymity and the systems for such participation both need to evolve. For now it seems that we may not know enough to take a hard line perspective on how this will all play out. I think I see the greater good value proposition, but time will tell whether (and when) we can get there.

Hope this helps,

Brian

Avatar of: Lewis Johnson

Lewis Johnson

Posts: 1457

April 20, 2012

As a data protection officer, my primary concern is the protection of individual privacy. You seem to state that the direct positive value for the participating individual, is derived from the act of participating. As such, there seems no need for references in the data toward the identity of individual participants. Anonymisation is standard practice for social, as well as financial research using sensitive data. Do you see a reason why that should be less effective here?

April 20, 2012

Thanks Lewis - I think the answer to your question lies in how (and on what level) anonymity is managed. For 'big data' the more information collected (appropriately) the better for the analysis and more specific the conclusion that can be drawn. For the social engagement around the data, perhaps more anonymity (with some degree of identity) may be acceptable. For instance, engaging with LJ23 vs Lewis Johnson may be valuable, if I can still have some sense of who LJ23 is, what his experiences are...enough that LJ23 has some credibility - LJ23 has identity, but is still anonymous. 

My other on-going concern is around how anonymous can someone (or their data) actually be. Again for the data collection and analysis I think there is one line of innovation, but for the social participation around the data, I think we are learning that overtime the cloak of anonymity is often lost...eventually with enough participation and inspection I have seen anonymous participants be uncloaked.

That is to the point, that both the science of online anonymity and the systems for such participation both need to evolve. For now it seems that we may not know enough to take a hard line perspective on how this will all play out. I think I see the greater good value proposition, but time will tell whether (and when) we can get there.

Hope this helps,

Brian

Follow The Scientist

icon-facebook icon-linkedin icon-twitter icon-vimeo icon-youtube
Advertisement

Stay Connected with The Scientist

  • icon-facebook The Scientist Magazine
  • icon-facebook The Scientist Careers
  • icon-facebook Neuroscience Research Techniques
  • icon-facebook Genetic Research Techniques
  • icon-facebook Cell Culture Techniques
  • icon-facebook Microbiology and Immunology
  • icon-facebook Cancer Research and Technology
  • icon-facebook Stem Cell and Regenerative Science
Advertisement
UVP
UVP
Advertisement
The Scientist
The Scientist