What videos can tell us about falling

In 2004, a videotape of a fall by Fidel Castro, then Cuban President, gained extensive press coverage and elicited a range of reactions [1]. The sequence captured a key shot for researchers who study falls. In the film, Castro, after one of his exhausting speeches and probably dazzled by lights, misses a step, starts falling after an unsuccessful stepping attempt, and turns in the air to reduce the impact of his head on the ground at the expense of an upper limb; the fall resulted in a broken shoulder and patella. Falls and fall-related injuries are a major health burden. Despite many epidemiological studies of predisposing risk factors [2], many assumptions and decisions about falls are still based on subjective and often biased information [3]. Fewer than 10% of falls are witnessed and, even when reports are available, they often do not provide detailed and objective information about the context and circumstances of the fall, or what happened during the fall. This absence of understanding is one of the reasons why efforts to prevent falls have had little success, although some progress has been achieved. [4], [5] and [6] In The Lancet, Stephen Robinovitch and colleagues [7] present results of an observational study of videotaped falls. They extensively studied falls in two long-term care institutions in British Columbia, Canada, between 2007 and 2010, using more than 200 public video cameras that were preinstalled for safety purposes. With a well-defined protocol, they were able to match staff incident reports of falls to video footage, making this a unique study. The researchers recorded 227 falls by 130 individuals whose mean age was 78 years (SD 10). Studies of this type are important because falls by elderly people are much more frequent in long-term care facilities than in the community; more than 90% of all hip fractures are caused by falls, and 20% of all hip fractures occur among residents of long-term care [8]. Robinovitch and colleagues' report provides some important findings. Among these is the high occurrence of falls caused by incorrect weight shifting (the most frequent cause of falls, 93 [41%] of 227 falls) and external perturbations, such as hit or bump events (which accounted for 25 [11%] falls). A further notable aspect is the improved understanding of the role of poor ergonomic design and environmental factors—eg, of the 48 falls caused by trip or stumble, 14 were attributable to a foot catching on equipment and 12 to a foot catching on furniture. This understanding should lead to revised housing norms and improved design of furniture and assistive devices. However, the study has some major limitations. Robinovitch and colleagues present data from publicly accessible spaces and not from private areas, such as bedrooms and toilets. More than 50% of falls in long-term care facilities occur in private areas that cannot be supervised by video footage [9]. Other objective approaches are needed to study falls in these rooms, such as sensors worn on the body. Although some findings might apply to people who depend on care but live at home, independent seniors probably have different risk factors and environmental cofactors that contribute to falls. Thus, the findings might not be applicable to community-dwelling seniors. Where could this study take fall prevention research? Robinovitch and colleagues build a strong case for classification of falls and a taxonomy of causes leading to falls. Currently, falls are most often presented as composite endpoints. Video footage, including that captured by members of the public with smartphones, will be one valuable source of information to generate new research hypotheses. High-speed video footage can also be used to study balance recovery reactions and landing responses in other groups, such as children and athletes. To study falls in the community, we will need a technological shift. Evidence provided by Robinovitch and colleagues of the movement patt...