Big data code-a-thon creates sleep deprivation resource

The power of collaborative challenge-based competitions to find big data solutions to real world problems was demonstrated at a recent US ‘code-a-thon’, focussed on the health industry.

Health 2.0 – a community based organisation which runs conferences and support services for users of health technology – organised the event on May 12 in Boston, to coincide with its ‘Spring Fling’ conference, in association with AT&T and Massachusetts Institute of Technology.

Given data from a host of different sources, including multiple US government offices and Athena Health which provided real patient data, the 85 contestants were split into teams with the aim of extracting value from the data and presenting it in a new, operational format to be used by patients or care providers. The teams had a 24 hour window in which to come up with a solution.

The winning team chose to focus their efforts on demonstrating how medical conditions caused poor sleeping patterns as well as things like how obesity can cause road accidents. The team created a website – nosleepkills.org – which allows users to enter personal information and details of their sleep patterns to determine whether they are currently achieving a healthy level of sleep. There is also information which allows medical professionals to spot those that might be sleep deprived.

The group hopes that the website, as it develops, will have a lobbying role in convincing public bodies to provide funding to test people for sleep apnoea – a condition where people stop breathing temporarily when they sleep and something that is a leading factor in adverse sleeping patterns.

The use of data from a wide range of sources was the leading factor in the teams win, which landed them the first prize of $4,000. A spokesperson from Health 2.0 said: “Almost two million drivers crash because of fatigue each year. The No Sleep Kills app creates a portal where patients/professionals can review and access data regarding sleep risks by state and compare to national averages.

“The application uses big data to compare sleep-related car crashes with rates of sleep apnoea and related conditions like obesity. The team’s next step is to compare sleep-related accident data with sleep apnoea testing expenditures on a state by state basis. No Sleep Kills showed us an application of big data to a little publicised issue that is actually a serious problem.”

The team that came second mined almost one million user profiles on the effectiveness and side effects of medicinal drugs and was able to draw conclusions as to which drugs would be best suitable for an individual based on their previous experiences.
The participants in the code-a-thon represented a range of organisations, including Harvard, Yale, Philips Medical, MIT and Aetha.

Health 2.0 runs a range of different research based challenges for the medical industry and has awarded over $1m in prize money so far.