Tuesday, November 27, 2012

Open Patient Consent and Clinical Research Data



Overcoming Lingering Concerns

In my last several posts, I wrote about how crucial it is to increase the liquidity of clinical research data  particularly clinical trial data  and how we can achieve improved data liquidity with patient-centric systems and software.

But there’s a remaining  and not insignificant  challenge:  patient consent.

Before sharing their health information, people want to know it’s going to be secure and beneficial to do so.  Until you’ve got people who are willing to share their data, it’s tough to justify the investment in building secure systems. A classic chicken-and-egg problem.

This is a very similar dynamic to e-commerce back in the 1990s.  People were afraid to enter their credit card numbers into a Web site. Many people were even saying that no one would ever trust personal financial data to the Web. Today, you can take a picture of a check with your iPhone, deposit the check electronically and throw the physical check away.  The convenience of electronic financial transactions via the Web far outweighed the security risks (both real and perceived).  

I believe that we’re going through a similar transition with electronic health data that we went through with personal financial data back in the 1980s and 1990s.  It may take a decade or two for people to be comfortable sharing their anonymized and aggregated medical information to benefit research...but maybe not.  In my humble opinion, the benefits of portability of our medical information now far outweigh the security risks/concerns.  

Some of the benefits are very pragmatic:
  • Portability: When you switch doctors, you can bring your medical history with you electronically
  • Accessibility:  When you have an emergency and the ER team needs your history immediately and will want to search for all your allergies quickly
  • Reference: When your doctor asks you when you had your last immunizations
You might recall that some high-profile early efforts at personal patient record systems failed ― specifically, Google Health and Microsoft HealthVault.  I think of these not as failures, but as invaluable experiments that helped us all learn what works and doesn’t work in managing and sharing health data securely and efficiently.  And perhaps most importantly, these experiments began to socialize the ideas of medical information being represented electronically and of patients owning their medical records. 

Meet Project Green Button

Project Green Button is an important experiment in creating data liquidity and sharing medical information.
  
You may have heard of Project Blue Button.  This was a fantastic project launched by the US Department of Veterans Affairs, enabling VA patients to download a copy of their own health data from the VA’s systems  by clicking on a Blue Button.  This is a really great example of empowering patients to own their own medical data and improving the portability of their medical information.  

Now let’s think about it the other way around.

Research has suggested that if people were presented with a Green Button (see picture above) that provided a single-click way to share their data with researchers, more than 80% of people would press it.  My close friend and trusted colleague John Wilbanks gave a great TED Talk about this a few months ago and has been doing ground-breaking work on open consent, the philosophy behind the Green Button.

I’m very hopeful that John will be successful in his mission to empower patients to share their own data to benefit research – after all, it is their data  and that most people, when asked, will be willing to share their anonymized information to benefit research.  
The team at the LIVESTRONG Foundation, led by Director of Evaluation and Research Ruth Rechlis-Oekler, Ph.D,  did a great study a few years ago about cancer patients’ and survivors’ willingness to share their information in the interest of improving research.  The results were compelling: the majority of patients and survivors WERE willing to share their de-identified and aggregated health information with researchers in the interest of improving health care for others. 

So, the willingness is there.  All we are missing are the systems to enable this. And the time to create those systems (similar to what I described in my last blog post) is NOW.   One of the most interesting companies working in this area is Avado; founder Dave Chase is one of the thought leaders in this space and has fantastic vision for where the industry needs to go over the next 10 years.  

Wanted: A Trusted “Zone”

In order to prime the pump of online personal health data, we need patient-controlled solutions and a trusted zone where we can connect patients securely with their data. This trusted zone would be a place where:
  • Each patient has a “dashboard” through which he can get access to all data about his own health, regardless of where the referential data is located.
  • Each patient can determine whether and with whom to share pieces of his health data – with his family doctor, relatives or loved ones.
  • Each patient can track his own health, and record and track his own experiences.
The result would be something that feels like a simple dashboard: a single environment, available to the patient, where he can access, share, and manage all his relevant health information.

Once we can “free the data” in a trusted environment like this, it’s possible to bolt on a whole battery of cool apps that you can’t even anticipate today  and that you don’t even have to develop yourself.

With today’s rapid app development technologies, we can build apps for physicians and for many other use cases, and simultaneously make the apps accessible via browser-based applications. 

A great example is the LIVESTRONG Cancer Guide and Tracker for iPad, which collects and combines patient-reported data for patients living with cancer.  We currently have a pilot project under way to connect the Tracker app with traditional clinical systems and a secure Web-based application, enabling the patient and his doctor to collaborate more effectively during office visits.  We're using the fantastic SMART Platform developed at Children's Hospital in Boston under the leadership of Zak Kohane and Ken Mandl. 
  
A Call to Action

So, here are a few action items:

First, we in the biopharmaceutical industry should ask ourselves:  “What data can we ‘free up’ about our products and our studies to better support patients and physicians?"  Let's start with the basic inclusion/exclusion criteria for our existing studies that we already share with our clinical partners.  As I mentioned in my previous post, if we would just increase the liquidity of this data in the biopharmaceutical industry, we could have a HUGE impact on the efficiency of research through the ability for the right patients to find the right studies at the right time.  

Second, we should support the idea of open consent by supporting the Consent to Research project (http://weconsent.us/).  Let's all give the flexibility to patients to support research with their own medical information and, in the process, radically improve the efficiency of the healthcare system by simplifying the complex spider web of consents that our dysfunctional healthcare system has created.  The current system of consents does not protect patients, but rather confuses them.

This blog post came out of a presentation that I recently delivered at Rev Forum, a conference sponsored by Lance Armstrong’s LIVESTRONG Foundation and Genentech. I have worked with LIVESTRONG and various biopharmaceutical companies on new health care information products and apps that take advantage of data liquidity to help patients combat cancer and other difficult diseases.


2 comments:

  1. Respectfully, your chicken and egg may be backwards :)

    The problem isn't lack of secure personal health records. I spent 5 years working on Indivo at Children's Hospital. You mention Google Health and Microsoft HealthVault (which is still at it, but without much fanfare). And there are probably 1000s of mom-and-pop PHR outfits as well.

    Poor PHR adoption has not been due to lack of access to PHRs, or fear about security. In fact, most patients I've interviewed care little about security because they assume their health data is already zooming around the medical system. The problem is lack of ubiquitous access to data. American EHR vendors have utterly failed to make their data available to patients in easy, machine-readable, and automatic ways.

    Remember, when you signed up for that Google Health account, only a tiny handful of hospitals offered data subscriptions... the vast majority of us had to type everything in by hand. Those hospital integrations had to be implemented on a case by case basis... not scalable at all, from business, technology, or healthcare perspectives.

    We need a national health data API specification that every EHR must implement as a requirement of certification -- not manual one-time exports -- so any patient, with any PHR, can aggregate her data from any point of care automatically and at any time.

    This infrastructure would be the healthcare equivalent of the interstate highway system -- bold, transformative, beneficial to all: individuals, government, academia, and industry. It's also boring like the highway -- it's less about the driving and more about the destination, i.e. what can really smart, creative teams build for patients once we've put the data access problem to rest.

    SMART has great potential to deliver really useful healthcare apps -- it could be hugely successful with data flowing freely into secure PHRs.

    ReplyDelete
  2. Jonathan - totally agree - need APIs!

    ReplyDelete