The effect of patient falls is far-reaching. Falls are the leading cause of injury and account for 6.5 percent of reported sentinel events in hospitals, with a direct medical cost of $34 billion annually, according to the Centers for Disease Control and Prevention. While healthcare stakeholders generally agree that the physical design of a space contributes to patient falls, there’s little understanding of the specific role design plays. After all, how could one definitively identify physical environment attributes without witnessing an actual fall event in context?

We know that there are two kinds of factors that contribute to falls: intrinsic and extrinsic. Intrinsic factors consist of one’s physical, mental, and physiological state, including age, gender, influence of medication, visual impairment, psychological condition, etc. The physical environment is a type of extrinsic factor.

According to a 2011 study by T.P. Haines and N.G. Waldron in the Journal of Safety Research, more than 70 percent of patient falls are unwitnessed, which renders understanding of the role of the physical design largely impossible. The only tools available to the designer are educated guesswork and a few broad-brush correlational studies, including a 2012 study, “Contribution of the Designed Environment to Fall Risk in Hospitals” by Margaret Calkins, Stacey Biddle, and Orion Biesan for The Center for Health Design—a report that was used as a foundation for a more recent look at the issue.

That study examined archival data from 27 inpatient units in 12 hospitals to identify associations between frequency of fall events and a range of physical design factors. The current study focused on some of the factors showing significant differences.

Theory in action
Armed with a grant from the National Patient Study Foundation in 2013, the study team (Dr. Debajyoti Pati of the Texas Tech University department of design and Dr. James Yang of the Texas Tech department of mechanical engineering) decided to view fall phenomena from a biomechanical perspective. Although social, cultural, and psychological entities, humans are also physical entities, which means their physical activities involve forces inside the body.

The human body consists of physical matter, and all physical actions are subject to physical forces (gravity, friction, etc.). The human body, like all physical objects, also has a center of mass that constantly accelerates and decelerates, unless one is in a resting position. Observation of a consistent increase in what’s known as the “jerk trajectory” of the center of mass indicates the beginning of a fall, according to engineering theories. Jerk represents the rate of change in acceleration of the center of mass with respect to time, or the smoothness of a movement.

To study this theory in action, 30 elderly subjects were asked to follow a scripted list of tasks inside a clinician zone and bathroom constructed as part of a patient room mock-up at the Human-Centric Design Research Lab at Texas Tech University in Lubbock, Texas. There were a total of 600 trials, and the script was developed by the falls committee of Covenant Health System in Lubbock. The physical mock-up was based on an extensive review of patient room designs in the archives of HKS Inc. (Dallas). The experiments included variations in bathroom locations, door locations, door swing, and which hand was attached to IV tubing with a mobile pole.

Motion capture technology, in combination with software specializing in identification of core of mass and calculation of jerk trajectory helped identify the trajectory of the center of mass for each individual subject as well as the moment representing the initiation of fall. Actual falls were prevented by building a harness to which all subjects were attached continuously during the trials. The harness works like a car seat belt, lifting a person up in response to a certain level of acceleration in the body.

Analysis identified a total of 730 potential fall events. The exact fall moments were subsequently extracted from digital videos, which were used to capture all activities in the mock-up simultaneously with the motion capture system. The fall video clips were subsequently examined and coded by experts of the Covenant falls committee and experienced healthcare designers. A variety of statistical analyses were conducted on the coded data.

Worst offenders
Among the key findings, the research team determined that when the relationship between number of falls and intrinsic, environmental, and postural factors were examined separately, the following demonstrated statistical significance: age and gender (intrinsic); bathroom location (environmental); and turning, grabbing, pushing, and pulling (postural). However, when all factors were examined together, age, gender, and bathroom location lost significance, while postural factors remained significant—meaning, from a physical design perspective, designing for the appropriate postures and motions constitutes the most important consideration.

The top offending postures in the bathroom were turning, grabbing, pulling, and pushing (in that order), with pushing and pulling demonstrating significance in the clinician zone. These were associated with, among others: configurational factors forcing turns; managing the IV pole when walking or conducting tasks; interacting with the door and moving through the entry; a lack of adequate space; and attributes of specific elements such as toilets (seat height, flush valve location), grab bars (number, location, and orientation), patient chairs (absence of armrest, seat height, and as obstruction in travel path), and over-bed tables (as obstruction in travel path).

Step forward
The following are some testable strategies developed from the findings. It’s important to note that these are not tested ideas yet and should be treated as hypotheses.

  • Design bathrooms to reduce turning as much as possible. The number and degree of turning will be affected by the relative locations of the bathroom door, toilet, and sink.
  • Don’t require motions that involve two or more of the offending postures, such as turning and grabbing or grabbing and pulling, etc.
  • Avoid swinging doors in patient bathrooms and use sliding doors instead.
  • Use wider door openings to accommodate the patient and an IV pole.
  • Explore different heights for the toilet seat (e.g., increasing the height).
  • Provide folding grab bars on both sides of the toilet, along with space to accommodate a caregiver.
  • Provide automatic flushing or place the flush handle/button at a location that enables operation from a standing position.
  • Explore operational or design interventions to keep the toilet seat at a lowered position.
  • Examine adjacencies of/distances between bathroom fixt
    ures to reduce sideways shuffling.
  • Assess patient room configurations that may allow the bathroom door to be left open while maintaining privacy from the hallway.
  • Design patient room configurations and operations that eliminate unnecessary obstructions in the clinician zone and the path from the bed to the bathroom.
  • Use over-bed table designs that allow proximal location but quick, effortless removal by the patient (e.g., integrated with the bed).
  • Find IV pole designs with smaller bases or robotics solutions where the pole can independently track and follow a patient but stay clear of the patient’s movement path.

With baby boomers aging across countries and continents, falls will become an immediate problem to millions—in hospitals, long-term care facilities, and homes. This study shows a few vital areas for design thinking and innovation for design, equipment manufacturing, and care delivery.

The following individuals made substantial contributions to the study: Dr. James Yang, Shabboo Valipoor, and Aimee Cloutier (Texas Tech University), Thomas E. Harvey Jr. (HKS Architects), and Patricia Freier (Covenant Health System).


Debajyoti Pati, PhD, FIIA, IDEC, LEED AP, is professor and Rockwell Endowment Chair in the Department of Design at Texas Tech University (Lubbock, Texas). He can be reached at