What kind of evidence would help a design team confirm an important decision, such as the choice of a life support system for an intensive care unit? How can a research-informed or evidence-based design process assist in making such a decision? (Hamilton & Shepley, 2010)The design of a hospital intensive care unit (ICU) patient room, where critical care is provided to seriously ill patients by specialist physicians and nurses, is significantly influenced by the type of life support system chosen for the design, such as a headwall, power column, or overhead boom configuration (Hamilton, 1999; Hamilton & Ulrich, 2008). Why do designers and their clients choose one system over another? The decision often seems to be made based on familiarity with one system or another, or purely on cost. What evidence could be used in making this decision as part of a design project?The headwall model, in which the utilities and physiological monitors are locat- ed on the wall behind the head of the bed (see Figure 1), is the most common choice. Familiarity with this model may influence its use; it has been the prin- cipal system in ICUs since their beginning, when the ICU was an adaptation of the surgical recovery room model. Clinicians who have experienced only this model may understandably favor it. Another factor is that a headwall configu- ration can be the most economical choice. In the case of a crisis, or code blue situation, the bed must be pulled off the headwall to allow someone to stand behind the head to ensure that the patient's airway remains open. Difficulties for staff associated with stepping over lines and cords, and potentially disconnect- ing something important, is an inconvenience at best and a problem with serious health consequences at worst.The power column system is the second most common choice (see Figure 2). The power column is a vertical system, standing floor to ceiling, typically located diagonally off the head of the bed. It has connections for medical gasses, out- lets for electricity, and uninterrupted power. A power column has a bracket on which to mount the physiologic monitor, and usually includes baskets or shelves for necessary items. It often has an attached sphygmomanometer.More recently, ceiling-mounted boom systems have been introduced, with over- head swiveling pendants to support the monitor, gasses, electrical connections, and other utilities (see Figure 3). The overhead boom life support model is similar to a power column, except that it is not attached to the floor, and can be moved to different positions. The ceiling-mounted boom system can include a pair of booms, and thus a pair of pendants with utilities and other features.Because these different systems to deliver life support technologies perform in different ways, they influence staff activity and behaviors (Gallant, 2001), although comparisons are possible (Pati et al., 2008). Flexibility and adaptabil- ity increases as one moves from headwall to power column to overhead booms. The power column obviates the need to move the bed and the risk of umbilical disruption. The overhead boom in turn allows greater movement of the bed and multiple possible positioning options, as well as the ability to swing it out of the way when desired.If I were still in practice as a hospital architect, or in a consulting role for an organization involved in the design of a new critical care unit, I would consider the choice of a life support system to be a key design decision requiring serious attention. Sometimes the program of space requirements will include a specific square footage that may restrict the choice, but design of the patient room can- not be completed without this decision.I have suggested that identifying the key design issue and turning it into a researchable question is an early step in an evidence-based process. To do this we would need definitions of life support systems, identification of the various choices available, and an understanding of the uses to which such systems may be put. …