Crowdsourcing EMR

by Jeff Mongelli

Recently, a group of online gamers resolved a roadblock that AIDS scientists couldn’t get past for more than a decade. This structured effort utilized a game designed specifically for solving scientific riddles, and the challenge was offered through an open call on the internet. The achievement opens the door to a new generation of antiretroviral drugs and marks a significant accomplishment for the emerging field of crowdsourcing.

Crowdsourcing is defined by Wikipedia as “the act of outsourcing tasks, traditionally performed by an employee or contractor, to an undefined, large group of people or community through an open call.” What makes the idea intriguing is the ability to get input from people with disparate backgrounds, thereby opening the thought process to fresh ideas from unique perspectives. The concept has rapidly been gaining traction. In fact, GE put out an open call named “The Challenge,” soliciting the crowd to submit breakthrough ideas for accelerating early detection and enabling more personalized treatment for breast cancer. Underwriting the effort are heavyweight venture capital firms Kleiner Perkins Caufield & Beyers, Venrock Capital, and others.

Crowdsourcing for Medical Professionals

Crowdsourcing is also at play on patient-oriented websites like Curetogether.com and Patientslikeme.com. On these sites, patients share their experiences with certain conditions and treatments, and they can do so anonymously. This collective experience creates real-world data and has already resulted in many patients reporting improvement in their conditions based upon feedback they’ve received from the sites. Some physicians may shudder at the thought of patients sharing treatment experiences with others—after all, each patient is unique and outside factors, like ancillary health issues, could play a role in a treatment option that may not get conveyed to others.

Those concerns aside, these sites are producing self-directed clinical trials that are being taken seriously by the medical community. A case in point, in June of this year, an Italian study reported that lithium carbonate showed potential for slowing the progression of ALS. Patientslikeme users, by the hundreds, brought this information to their treating physicians. Patientslikeme is now conducting its own patient-driven, clinical trials of the treatment. While the results are presently inconclusive, the power of crowdsourcing is once again proving itself.

Parenthetically, crowdsourcing is a global phenomenon where the only barriers to participation are internet access and language, in some cases. Imagine having the ability to reach beyond your own patient base and empirical experience to discover patterns that you’ve long suspected, simply by participating in a similar website. Imagine putting an end to the debate regarding a link between ultrasounds and autism. Crowdsourcing could pool together outcomes from across the world to evaluate, if indeed correlations exist.

Is Crowdsourcing a Beneficial Option for EMRs?

Will crowdsourcing play a role in EMR? It could surface in a number of areas. First let’s consider its potential impact on EMR design. Crowdsourcing for software development is already being used by NASA in a joint venture with Harvard University to develop software for complex data processing and to solve other complex problems. Why not bring the concept to the EMR field and see what gets created? It’s no secret to physicians that many EMRs appear to have been hatched by programmers developing in a vacuum. The end result has been systems that have been difficult to use, resulting in reduced productivity and in dire need of rapid evolution. Opening screen designs and workflows to crowdsourcing could result in more user-friendly screens and more efficient workflow-oriented designs.

To a certain degree, EMR vendors do this already by soliciting feedback from their users and incorporating their input into future builds. Crowdsourcing takes user input to a new level. In a structured manner, participants could collectively contribute to designs that maximize and enhance productivity. The best vehicle to implement a strategy like crowdsourcing would be with the handful of open source EMR s that currently exist—a list of which can be found here. Speeding up the evolution of EMR s would benefit everyone, most importantly, the physicians that are using them in hopes of achieving Meaningful Use. After all, what’s the point of pursuing additional reimbursement if the road to getting there results in reduced productivity?

Better Design for EMRs

While design would be enhanced by a crowdsourcing approach, a larger benefit would be the use of crowdsourcing in data management. Meaningful Use reporting will create a massive database. Actually making use of all of this data is an incredibly daunting task. The crowd could accomplish in less time what a single development team could take years to achieve. In the interest of privacy, it could be done using sample data since the data structure is more important for development than the actual data. The result would be viable analytics created in a faster time frame and the unlocking of potential correlations that are currently locked away in paper charts and disconnected EMR databases.

The AIDS/ gamer experience taught us that the crowd can use intuition and higher level thought processes than computers are currently capable of achieving. As powerful as computers are, they are trapped in linear thought processes, each successive step following the previous. Intuition and speculation are still inherently human characteristics that computers have yet to emulate. That’s why computers didn’t solve the AIDS problem. Crowdsourcing could ultimately prove to be the greatest social contribution of the internet. Crowdsourcing is in its infancy and will continue to find its way into more and more uses. For additional examples of crowdsourcing in use today, you can visit crowdsourcing. org. To see how Netflix used crowdsourcing to improve its streaming algorithm, take a look at Buzzfeed's article.


comments powered by Disqus