b'The New Model: Leveraging AI to Unlock Value in Hospital OperationsHow can hospitals and health systems be more like airlines in such an environment? How can we both predict changing volumes in real time and match our staffing levels to optimize efficiency while balancing operational priorities such as patient satisfaction and throughput?Over the past year, SCP Health has built proprietary tools to help our hospital and health system partners do precisely this:FORECAST SIMULATION SOLVER VolumePatient Forecast DemandAI Additive Digital Twin Linear ProgrammingRegression Sim Model OptimizerThe team built an AI model with deep data and algorithms, creating the baseline. The AI model in operation predicts volume with 90-95% accuracy and is improving every day.1. Leveraging data to predict volume. We use historical volume data to train sophisticated algorithms designed topredict volume into the future. Our goal is to predict volume 60 to 90 days in advance: specifically, the number ofpatients on an hourly and daily basis, in 15-minute increments. Machine learning gets smarter over time by lookingat the actual volumes and incorporating that into continuously refining its predictive model. A digital twin simulation tool accurately predicts volume and case mixand helps operational leaders respond to volatility in real-time. 2. Using digital twin simulation engines to predict case mix. So-called digital twin simulations work in real-timeto emulate a virtual representation of a real-life system. For example, airlines run a digital simulation of every planeengine currently in the air. Those engines send back continuous data as they fly. Suppose the real-time data thatcomes back is different from the digital copy. In that case, that alerts airlines to potential problems which need to beaddressed in real-time, whether it is a vibration sensor that is picking up a mechanical issue or a heat sensor thatsignals a need for future maintenance. In health care, the cloud-based computing power has now enabled leveraging these powerful simulation engines tomirror hospital operations. The simulation engine will run hundreds of thousands of potential scenarios for who willshow up to the ED, at what times, and at what levels of acuity (or intensity of clinical intervention) until it arrives atthe most likely patternand it learns over time. Together, we healTogether, we heal 4Together, we heal SCP HEALTHIINVESTING IN AI: PROVEN STRATEGIES TO FUEL TRANSFORMATION IN HOSPITAL OPERATIONS'