Intervening to Prevent Falls and Fractures in Nursing Homes: Are We Putting the Cart Before the Horse?

Nursing home (NH) residents tend to fall at a rate that is approximately three times as high as that of older people living in the community. This translates into a fall rate of 1.5 falls per bed per year in long-term care facilities. Serious injuries occur in approximately 12% of these falls and fracture in 4%. Once injured, these individuals are more likely than community residents to die as a result of their fall. NH residents have a disproportionate rate of hip fracture resulting from a fall and higher mortality rates after hip fracture than community-dwelling older people. Of all persons aged 85 and older, 20% of fatal falls occurs in the NH. The four articles in this issue of the Journal of the American Geriatric Society on prevention of falls, fractures, and other debilitating injuries that face NH residents highlight the need to continue looking for answers. Spector et al. add to the body of observational risk factor literature with an analysis exploring the associations of fractures. The authors analyzed data on 2,711 NH residents from the 1996 Medical Expenditure Panel Survey in an effort to determine individual and facility-level characteristics associated with the risk of fractures to help identify individuals at highest risk and to inform the design of interventions to prevent fractures. Unfortunately, because these are observational data, the interpretation of study findings cannot be used as the sole basis for interventions. Interventions to prevent falls have been tried in NH settings and these interventions tend to be multifactorial and resource intensive. Unfortunately, the results of previous NH-based falls interventions have been disappointing. Rask et al. present the results of a quality improvement intervention undertaken by a chain of NHs in an effort to improve their care process documentation while reducing their rate of falls and restraint use. They describe a multilayered approach with intensive support to incorporate best practices into the daily routine of the nursing facilities. Over the duration of their study, care plan documentation pertaining to falls improved, and restraint use declined, but fall rates remained stable. Capezuti et al. present the result of an intervention in 251 residents of four NHs. The intervention required the efforts of an advanced practice nurse (APN) who worked closely with the entire interdisciplinary team. The goal was to develop realistic plans of care at the individual and facility level and specifically to address alternatives to side rails such as bed alarms, floor mats, transfer poles, and low beds. The authors found that side rail use decreased with these interventions at some, but not all, of the study sites. Capezuti et al. noted that the reduction in bed rail use did not result in increased falls. Not all falls are preventable. Consequently, efforts have been directed toward the reduction of injurious falls. In a widely cited study by Kannus et al., hip protectors were viewed as an important intervention to reduce fall-related injuries, although subsequent studies failed to replicate the strongly positive results of that study. Despite these equivocal clinical trial results, hip protectors use has already been widely accepted clinically. Perhaps this is due to the intuitive belief that hip protectors should be effective or the hope that hip protectors will be a relatively easy solution to a challenging and potentially devastating clinical problem. Given the importance of adherence to hip protectors as a key factor in randomized, controlled trials, the study by Honkanen et al. in the current issue of the Journal studied resident and staff factors for their prediction of hip protector use and adherence in a 409 bed facility. Authors of these four articles are to be commended for their efforts in studying important problems in a setting where there are many special challenges to conducting definitive research. In addition to the shortage of funds for these studies, there also is a reluctance to test the efficacy of approaches such as use of hip pads or nonuse of bed rails when clinical practice has already taken hold. Still, one needs to take a step back from the details of the individual studies and ask the more basic question of whether we are ‘‘putting the cart before the horse.’’ That is, do the strategies whose use we are interested in increasing actually have the desired effect on falls, injuries, or fractures. This question is not meant to imply that we must wait for the definitive answer before using any new management strategy thought to be beneficial for a NH population, but it does suggest that we need better data on evaluating the outcome of strategies we use. Without these data on efficacy, we may be misdirecting our limited resources to practices and policies that have not been shown to help and that could even be detrimental. For example, if we put a lot of emphasis on hip protector use without first knowing that they prevent injuries and fractures, we risk expending limited human and financial resources on ensuring the use of an unproven ‘‘protective undergarment.’’ Randomized or otherwise definitive studies are needed before the industry embraces and adopts costly and untested approaches to care management. Conversely, although we may be putting the ‘‘cart before the horse’’ by examining ways of increasing DOI: 10.1111/j.1532-5415.2007.01088.x