Chicago is Using Big Data to Find Health Violations Faster Than Ever
There's one in every neighborhood: that "authentic-looking" hole in the wall beckoning you with the prospect of a those-in-the-know dish. Then you get food poisoning. And when your neighbor questions why they heard a bunch of retching coming from your apartment at 3 a.m. they corroborate that the joint is notorious for the giving anyone who walks a case of the trots. How does that place stay open? Well, likely because no one is shutting it down. City health departments, like any other government organization, are often wrought with inefficiency and slow-moving bureaucracy.
But Chicago, it seems, has found a better way. Typically, the city's Department of Health would run a list of the 16,000 establishments that were due for inspection, and then randomly assign their staff of about three dozen to visit each of them. It was a necessarily tedious and unfortunately faulty process that let many issues linger too long or slip trough the cracks altogether.
That changed with an experiment lunched by the city in 2014. By analyzing variables such as previous violations, nearby sanitation complaints and how long it's been since anyone has checked on a restaurant, the Chicago Department of Innovation and Technology created an algorithm to prioritize potentially troublesome spots. That algorithm found violations 7.5 days faster, on average, than their usual methods. In food poisoning terms, that's a lot of prevented diarrhea and that is definitely a good thing.
What's not such a good thing is that more cities haven't taken Chicago's lead in utilizing data to predict problems sooner than later. Chicago's chief data officer even went so far as to publicly publish the code so other could see just how it worked. Only Montgomery County, Maryland has instituted a similar system and it immediately improved their department’s effectiveness. That county is also conducting experiments that will include everything from construction permits to Yelp reviews in the analyzed data. Hopefully more cities will follow suit, because this simple application of technology could save us quite a few unnecessary trips to the bathroom.