Predictive modeling of slips and falls.
NIOSH 2000 Dec; :1-117
Slip and fall accidents, caused by complex and interacting environmental and human factors, are a major ergonomic concern. In order to address the long-term goal of reducing slip/fall accidents, specific objectives were identified: (1) investigate changes in gait biomechanics when anticipating slippery environments, (2) describe foot dynamics during slip events, (3) compare center of pressure trajectory and joint biomechanics between no-slip and slip/fall events, (4) investigate gait biomechanics related to slips when carrying a load and (5) describe and possibly model the relationship among the probability of slip/fall, frictional requirements of walking and available friction. Ground reaction forces and sagittal plane body/foot motion dynamics of 16 subjects walking under various slippery and non-slippery environmental conditions were recorded at 350 Hz. Subjects were asked to walk as naturally as possible even without knowledge of the floor's contaminant condition. In spite of these instructions, subjects adopted postural and temporal gait changes when uncertain of the contaminant condition. These changes in gait patterns were associated with decreases (16 to 33%) in slip potential compared to recordings on known dry surfaces. During slipping experiments, corrective biomechanical reactions occurred from 25 to 45% of stance phase (about 190 - 350 ms after heel contact). These corrective reactions included increased flexion moment at the knee and extensor activity at the hip. The effects of this active knee/hip moment on body motion were reflected in the kinematics in an attempt to bring the foot back under the body and thus recover from a slip event. Load carrying experiments found decreased slip potential compared to free walking, with decreased required friction and more controlled heel contact dynamics. Logistic regression models proved to be useful in describing the relationship between the probability of falls and the difference between required and available friction (COFdiff). However, logistic regression functions of COFdiff alone did not provide a satisfactory fit of the probability of slip. The inaccurate assessment of the frictional characteristics of the shoe/floor interface is postulated to be the dominant reason for this unsatisfactory fit. This project revealed important biomechanical findings relevant to slips/falls prevention research. In addition, the research indicated that current slip resistance testing devices might be inadequate to predict slips.
Models; Injuries; Injury-prevention; Environmental-factors; Humans; Ergonomics; Biomechanics; Surface-properties; Accident-prevention
Department of Bioengineering, 749 Benedum Hall, University of Pittsburgh, Pittsburgh, PA 15213
Final Grant Report
National Institute for Occupational Safety and Health
University of Pittsburgh, Department of Bioengineering, Pittsburgh, PA