A research team at the University of Chicago uses analytics to determine how long a patient will stay in the hospital, and, for long stays, to attempt to determine who might need a case manager for long-term support.

Initially, the team determines that the most accurate predictor for assessing hospital stay is a person's zip code. However, this raises a red flag for many, who assert that because zip codes are often race- and class-based due to geographic segregation, using zip codes can end up separating conversations about hospital stays from the larger context of racism and economics.

As a result, the University of Chicago research team decides not to use the zip-code based algorithm because of the implicit racial bias built into it, and due to not having a cohesive strategy for addressing racism and poverty within this framework.