This is a fully automated randomized trial with two randomization branch-points. The first is
inclusion of disease-specific orders in the admission orders based on a predictive model
using real-time data. The second is the use of dynamic orders that are end-user tested rather
than static orders designed by a committee. The primary hypothesis is that automatic
inclusion of disease specific orders with admission orders will improve adherence to
guidelines for patients with COPD. The secondary hypothesis is that clinical and operational
outcomes will improve, thereby improving value.