Improve Patient Outcomes within Budget
In one organization, contextual analytics were used to decrease falls from 4 falls per 1,000 patient days to 0.6 falls per 1,000 patient days, saving approximately 1.6 million dollars. In our study, the initial model predicted 35% of falls. Presentation of these initial results to quality control employees helped them refine the falls model to increase explanation of falls to 61%. Once the group was able to pinpoint 61% of the circumstances and behaviors contributing to falls, practice changes were implemented and the incidence of falls plummeted.
Math, given the proper formula and accurate variables, can be used to study and improve every healthcare outcome. We have improved outcomes such as readmissions for heart failure, central line associated infections, length of stay, evidence of caring, and many more. The figure below shows a selection of the outcomes we’ve measured and improved using the proper math, correct variables, and real-time feedback from the people closest to the work, to keep each algorithm relevant to the changing dynamics of the environment and the state of the evidence.