Cognitive Computing Strategy to Reduce Inpatient Falls

Specific Aim 1 – Optimize Caregiver Notification of Fall Risk. During the first year of this project, we made enhancements in the EMR that optimize caregiver notification that the patient is high risk of fall. Utilizing highly visible banner notifications, caregivers are alerted that the patient is at high risk of fall. These notifications are driven by the JHFRAT score currently. We will be modifying these notifications so that the Epic Model score can drive them instead. Specific Aim 2 – Prospective Evaluation of an Epic Model Based Falls Workflow on Pilot Units Output from the Epic Model can be used by nurses to accurately characterize a patient’s fall risk and develop a customized and dynamic Safe Activity Plan. To prospectively evaluate fall rate and nursing engagement with the falls workflow, the Epic Model data will be incorporated into a standardized falls workflow in 6 inpatient units at NMH. We will use data from the Epic Model to inform the falls workflow only on these intervention units. 6 similar NMH inpatient units which use the current JHFRAT workflow will be the control group. We will evaluate the implementation of the Epic Model and determine if there is a statistically significant impact of the new workflow on fall rate and nursing engagement. 12 months of data will be required to achieve this specific aim. Specific Aim 3 – Optimize Communication About Fall Risk Improving communication among health care workers, patients and family members about individualized fall risk will be necessary to achieve the greatest fall risk reduction. We will modify the handoff tool to automatically include individual fall risk data that can be used during nursing shift change or unit transfer. Fall risk level will also appear on documentation for patient transport technicians. Because we anticipate that the new nursing workflow will take less time, we will also explore opportunities to better engage the patient and family in the Safe Activity