Crowd-Sourcing as Virtual Medical Research

An interesting article came through the email pertaining to the The Internet of Humans.  The gist is that an increasing amount of information is being generated by the machine versus the human.  While this is an interesting concept in its own right (so interesting I shamelessly reproduced it below), I was more intrigued by the following paragraph:

PatientsLikeMe.com, which, as Weinberger explained, “not only enables patients to share details about their treatments and responses but gathers that data, anonymizes it, and provides it to researchers, including to pharmaceutical companies.  The patients are providing highly pertinent information based on an expertise in a disease in which they have special credentials they have earned against their will.”  The intriguing term Weinberger used to describe the source of this crowd-sourced amateur knowledge was human sensors.

Now compare this to the 2011 announcement of Alberta’s own Personal Health Portal or PHP which will allow for things such as:

  • Secure log-in for individuals who want to track their own personal health data such as blood pressure readings, insulin levels, weight, immunizations, and much more.
  • When complete in 2015, MyHealthAlberta will provide individual access to Alberta’s electronic health record system.

My thoughts when I combine the Information Management article with the PHP is a made in Alberta crowd-sourcing opportunity.  Assuming suitable provisions for privacy and informed consent, there is an opportunity for Albertans to contribute to the body of knowledge of both their specific disease condition and their state of wellness.

At its most voluntary basis, PHP subscribers with a specific disease condition, would be asked to contribute pertinent information (blood pressure readings, glucose levels, thoughts of depression, weight gain/loss, etc.) via a computer or smart phone application.  The results, would then be used by both health planners and researchers for the effectiveness of treatments, disease prevalence and health resource planning.  Key to this contribution would be to honor those patients (and their families) making contributions as being part of the larger benefit to society.  My own view is that a person suffering from a chronic condition would likely provide very credible and valuable information – if they new this information was being used and was valued by their peer-patients; healthcare providers, researchers and future patients.

In a less voluntary model, instead of a health premium, for example, one could agree to participate and provide factual information about their health condition. This would capture not only information about a specific disease condition but also the general health state of the population.  e.g. how many 30 year old males are overweight, smokers, moderate drinkers who have high blood pressure.

Once again, insert strong privacy considerations here along with a myriad of details about data collection, quality, validity and audit controls.  Nevertheless – a very interesting thought exercise!  Feel free to comment with your own thoughts.

DISCLOSURE and DISCLAIMER: the above comments and thoughts are exclusively those of Frank Potter.  Although Frank is an employee of Alberta Health, the above opinions do not necessarily represent the opinions of the Department or Ministry. 


Information Development has posted a new item, ‘The Internet of Humans

The Internet of Things became a more frequently heard phrase over the last decade as more things embedded with radio-frequency identification (RFID) tags, or similar technology, allowed objects to be uniquely identified, inventoried, and tracked by computers.  Early adopters focused on inventory control and supply chain management, but the growing fields of application include smart meters, smart appliances, and, of course, smart phones.

The concept is referred to the Internet of Things to differentiate its machine-generated data from the data generated directly by humans typing, taking pictures, recording videos, scanning bar codes, etc.

The Internet of Things is the source of the category of big data known as sensor data, which is often the new type you come across while defining big data that requires you to start getting to know NoSQL.

In his book Too Big to Know, David Weinberger discussed another growing category of data facilitated by the Internet, namely the crowd-sourced knowledge of amateurs providing scientists with data to aid in their research.  “Science has a long tradition of embracing amateurs,” Weinberger explained.  “After all, truth is truth, no matter who utters it.”

The era of big data could be called the era of big utterance, and the Internet is the ultimate platform for crowd-sourcing the knowledge of amateurs.  Weinberger provided several examples, including websites like GalaxyZoo.orgeBird.org, and PatientsLikeMe.com, which, as Weinberger explained, “not only enables patients to share details about their treatments and responses but gathers that data, anonymizes it, and provides it to researchers, including to pharmaceutical companies.  The patients are providing highly pertinent information based on an expertise in a disease in which they have special credentials they have earned against their will.”  The intriguing term Weinberger used to describe the source of this crowd-sourced amateur knowledge was human sensors.

In our increasingly data-constructed world, where more data might soon be constructed by things than by humans, I couldn’t help but wonder whether the phrase the Internet of Humans needs to be frequently heard in the coming decades to not only differentiate machine-generated data from human-generated data, but, more importantly, to remind us that humans (amateurs and professionals alike) are a vital source of knowledge that no amount of data from any source could ever replace.