In collaboration with some other University of Guelph students (Chris Cameron, Ben Ferriman, Eddie Ma), I helped build a tool for an emergency room doctor that allows users to predict how long they will have to wait once they arrive at the ER and are checked in to a particular zone of the hospital. The idea is to allow the patients to feel better served by letting them know more what to expect rather than just sitting around wondering how long it will be. The tool draws on records from an unnamed hospital in Ontario, and is based on over 50,000 patient records. Right now the tool takes into account only a limited number of factors such as the arrival time, zone of the hospital etc. But it could be expanded in the future to include the number of patients queuing, the type of injury etc. Real-time data could also be fed into the system to allow for very accurate wait time predictions. The current version does not have access to this information, so it must make a prediction based solely on past data. Lastly, the tool could be improved if data from other hospitals was provided as well. It would be possible to make extremely accurate predictions on where to take patients based on a combination of wait time, distance and traffic data so that patients may be treated as quickly as possible.
The tool will be available for a limited time to demo at: http://m.socs.uoguelph.ca/
Here are some screenshots: