A team of Michigan engineers, pharmacists, and clinicians has successfully deployed an AI model at Michigan Medicine to help reduce the spread of C. difficile, a highly contagious and dangerous infection typically contracted within hospitals. Their system helped prescribe antibiotics without increasing patient mortality or length of stay. The interdisciplinary effort took nearly a decade to bring from predictive modeling to real-life implementation. While infection rates showed only a slight decline, the study proves it is possible to use AI to inform daily clinical decision-making.
