Introduction/Background In the United States there were 820 major disasters declared since 2000, which included earthquakes, severe storms, terrorist attacks and other events requiring some hospital evacuations.1–5 Limited data are available for hospital evacuations, in particular for complex scenarios such as patients that are: 1) undergoing procedures, 2) immobile, 3) mentally incapacitated or 4) on multiple life support measures. While the Joint Commission requires hospitals to have an emergency operations plan with exercises twice a year, the extent of these exercises is not specified.6 The goal of this study is to evaluate evacuation protocols at the unit level. Specific aims are: 1) benchmark current performance in existing evacuation plans, 2) develop guidelines and best practices, and 3) produce checklists for each team member to standardize evacuation performance. Methods Evacuation simulations were observed at a single hospital using a convenience sample of interprofessional subjects. Single patient vertical evacuations out of the building were performed from operating and recovery rooms, intensive care, long term care and telemetry units using high fidelity manikin or standardized patient, commercial evacuation-sleds, video cameras, smoke simulator, megaphone, and stopwatches. Qualitative data was elicited from participants in the debriefing identifying improvement areas. In this exploratory descriptive project, the following findings were identified as threats to safety and ongoing patient care: 1) Time Performance - Average time to evacuate 29.75 minutes (range 16–42 minutes); 2) Patient Factors - injuries from oxygen tanks, oxygen tank falling out of evacuation sled, morbid obesity, claustrophobia while in evacuation sled and accidental extubation and IV dislodgement; 3) Staff Factors - musculoskeletal injuries - pull straps too short, staff to patient size disproportion, inadequate staffing - less available on nights and weekends, unfamiliar with procedure and rendezvous point, inexperience with evacuation sled; 4) Equipment Issues - evacuation sled use not intuitive, insufficient monitoring equipment for every patient, some beds inoperable without electricity; 5) Supply Issues - no teams evacuated with supplies for ongoing care, teams identified items to bring in subsequent exercises: medications, dressings, IV fluids, gloves and blood products, limited provisions in storage unit for continued patient care; 6) Process - no standardized equipment, supply, or patient packaging procedure, did not ensure clear stairwells prior to entry with patients, bottleneck at stairwells, unit leadership unprepared for triage or patient flow, poorly prepared for inclement weather, nonclinical personnel without designated roles, no mechanism to account for visitor evacuation, posted evacuation maps differ from the planned evacuation route. Results: Conclusion This work represents an initial description of the current performance of vertical evacuation for complex patients in one facility without outside community support. The complexity of the patient population, lack of continuing care, knowledge deficit and incomplete evacuation protocol development contributed to suboptimal performance. The strengths of the simulation were demonstrated in successful evacuation of all patients and accessible evacuation sleds. Disaster plans and equipment are available; however, in a disaster the deficiencies with planning, training and equipment put patients and staff at significant risk of injury not only from the evacuation process but also from an inability to evacuate. This data will be used to establish checklists for each team member. It is our hypothesis that checklists will standardize and expedite performance. In-situ and virtual reality simulation will be used as the training vehicle to check the validity and refine checklists in future studies with increased patient complexity, multiple patients and simultaneous unit evacuations.