Definition of a Fall

For epidemiological and intervention studies measuring the magnitude of falls, the adoption of a standardized definition is necessary for consistency and comparison of outcome measures. Although the meaning of a “fall” seems intuitive, older people, health professionals, and researchers can interpret a fall in different ways (Zecevic, Salmoni, Speechley, & Vandervoort, Reference Zecevic, Salmoni, Speechley and Vandervoort2006; Haines, Massey, Varghese, Fleming, & Gray, Reference Haines, Massey, Varghese, Fleming and Gray2009; Hauer, Lamb, Jorstad, Todd, & Becker, Reference Hauer, Lamb, Jorstad, Todd and Becker2006; Lamb, Jorstad-Stein, Hauer, & Becker, Reference Lamb, Jorstad-Stein, Hauer and Becker2005). Moreover, although there have been diverse definitions in the literature (Zecevic et al., Reference Zecevic, Salmoni, Speechley and Vandervoort2006), the most widely accepted definition is that of inadvertently coming to rest on the ground, floor, or lower level. The definition excludes intentional change in position to rest on furniture, or against a wall or other structure (World Health Organization, 2007). Ascertainment should consider lay perspectives so that participants are asked about falls, including slips or trips, causing loss of balance to land on the ground, floor, or lower level (Lamb et al., Reference Lamb, Jorstad-Stein, Hauer and Becker2005). For coding purposes in injury surveillance systems, accidental falls may be classified as falls on the same level (e.g., from slipping or tripping) or from different levels (e.g., from stairs, from ladder, bed, or other structure).

Measuring Falls Outcomes

Falls data is collected by three main methods: (a) retrospective reporting systems using telephone interview, face-to-face interview, or postal questionnaire; (b) prospective reporting systems using postcards, calendars, or diaries; and (c) routine surveillance systems or abstraction from health care records (Hauer et al., Reference Hauer, Lamb, Jorstad, Todd and Becker2006). More recently, fall detection devices have been developed (Noury et al., Reference Noury, Fleury, Rumeau, Bourke, Laighin and Rialle2007), and wearable devices such as accelerometers have been used to monitor falls in clinical settings (Culhane, O’Connor, Lyons, & Lyons, Reference Culhane, O’Connor, Lyons and Lyons2005). Such methods, however, are unsuitable for large-scale population monitoring.

Several issues confront researchers when collecting falls data, including accuracy of self-reporting and retrospective recall of previous falls (Ganz, Higashi, & Rubenstein, Reference Ganz, Higashi and Rubenstein2005; Mackenzie, Byles, & D’Este, Reference Mackenzie, Byles and D’ Este2006; Peel, Reference Peel2000), as well as differing perceptions of what constitutes a “fall” (Zecevic et al., Reference Zecevic, Salmoni, Speechley and Vandervoort2006; Hauer et al., Reference Hauer, Lamb, Jorstad, Todd and Becker2006). Depending on the length of the recall period, falls may be forgotten (Wijlhuizen, Hopman-Rock, Knook, & van der Cammen, Reference Wijlhuizen, Hopman-Rock, Knook and van der Cammen2006; Cummings, Nevitt, & Kidd, Reference Cummings, Nevitt and Kidd1988). Falls are often unwitnessed and unreported, and there may even be disincentives on the part of the older person for reporting falls, including embarrassment, perceptions that a fall will be seen as a marker of aging, and fear of consequences, such as loss of independence and risk of institutionalization (Allen, Reference Allen2004; Peel, Bell, & Smith, Reference Peel, Bell and Smith2008). Data on falls in institutional settings may also be of questionable quality as staff time pressures, perception of blame, and other factors are thought to contribute to under-reporting (Haines, Cornwell, Fleming, Varghese, & Gray, Reference Haines, Cornwell, Fleming, Varghese and Gray2008).

Strategies are required to increase reporting of falls by older people by increasing their awareness of the value of reporting (e.g., for the early detection of risk and implementing strategies to reduce future risk). Also needed are health professional initiatives to encourage asking all older people at least once each year whether they have had a fall (American Geriatrics Society, British Geriatrics Society, American Academy of Orthopaedic Surgeons Panel on Falls Prevention, 2001). Providing training to institutional staff regarding the purpose of incident reporting and the usefulness of such reports for preventing future falls may improve incident reporting practices (Haines et al., Reference Haines, Cornwell, Fleming, Varghese and Gray2008), and is seen as a necessary step in changing a culture of blame.

Many falls go undetected until after injury and disabilities have occurred. Falls resulting in injury are more likely to be recorded in administrative data sets, such as hospital records and institutional incident reports. The injury experience of a population has been described as a pyramid. The apex represents the relatively small number of fatal cases, and the broader, lower parts of the pyramid represent the more numerous injuries of lesser severity. Injury data availability is in direct proportion to case severity, and in inverse proportion to case frequency. Quite a lot is known about the relatively small number of injury deaths, less about hospital in-patient cases, and still less about cases resulting in neither death nor hospital admission. Fewer than half of fall events are reported to a health practitioner (Hauer et al., Reference Hauer, Lamb, Jorstad, Todd and Becker2006; Shumway-Cook et al., Reference Shumway-Cook, Ciol, Hoffman, Dudgeon, Yorkston and Chan2009), and the percentage is likely to be much lower in cases where a fall occurs and does not require treatment (Allen, Reference Allen2004).

In epidemiological studies, the most-frequent methods of summarizing fall outcomes include (a) reporting data on number of participants sustaining a fall or multiple falls (usually two or more); (b) number of falls; (c) fall rates; and/or (d) time to first fall (Hauer et al., Reference Hauer, Lamb, Jorstad, Todd and Becker2006). The focus may be on falls causing injury or requiring medical care. In calculating rates, the number of falls, fallers, or fall-related injuries is the numerator, while the denominator may be number of people at risk or number of occupied bed (patient) days per unit of time. As an example, annual national reporting on indicators for safety and quality in the Australian health care system requires measurement of falls resulting in patient harm in hospitals and residential aged-care facilities (RACF). The numerator is the number of in-hospital fall separations (or number of RACF falls) in which the fall resulted in harm, and the denominator is the total number of hospital separations (or RACF residents) (Australian Institute of Health and Welfare, 2009).

Incidence of Falls and Fall-Related Injuries

The incidence of falls and fall-related injuries varies among settings and populations

Community Settings

The variation in study designs, study population characteristics, and methods of falls data ascertainment and reporting make comparison of studies difficult. Table 1 outlines a number of studies in older people in community settings selected to represent a range of recent research across different countries. The majority of studies showed that between 20 per cent and 33 per cent fall each year. The relative rates increased approximately 5 per cent per year with cohort aging (Anstey, Burns, von Sanden, & Luszcz, Reference Anstey, Burns, von Sanden and Luszcz2008) and were highest in the old-old (Speechley et al., Reference Speechley, Belfry, Borrie, Bray Jenkyn, Crilly and Gill2005) and oldest old (Fleming, Matthews, Brayne, & CC75C Study Collaboration, Reference Fleming, Matthews and Brayne2008; Iinattiniemi, Jokelainen, & Luukinen, Reference Iinattiniemi, Jokelainen and Luukinen2009; Formiga, Ferrer, Duaso, Olmedo, & Pujol, Reference Formiga, Ferrer, Duaso, Olmedo and Pujol2008) wherein up to 60 per cent fell over a 12-month follow-up period (Fleming et al., Reference Fleming, Matthews and Brayne2008). The lowest rates were recorded in Hong Kong Chinese (Chu, Chi, & Chiu, Reference Chu, Chi and Chiu2005). These studies also showed that a quarter to half of fallers experienced multiple falls (2 or more), with higher rates of recurrent falling (75%) in the oldest-old (Fleming et al., Reference Fleming, Matthews and Brayne2008). Fall rates were higher in older community-dwelling women than men (not tabulated). Up to a third of fallers reported an injury requiring medical attention, with less than 10 per cent having serious injury such as fractures or head injury. Broadly, the findings of the recent studies reflect those of previously published epidemiological studies in the 1980s and 1990s (Masud & Morris, Reference Masud and Morris2001; Lord, Sherrington, Menz, & Close, Reference Lord, Sherrington, Menz and Close2007), despite a generational change.

Table 1: Community studies of fall incidence

PY = person-years

SD = standard deviation

Care Settings

Because of a high prevalence of clinical diseases, functional disability, and age-related physiological changes, older people in care settings are particularly vulnerable, resulting in an increased incidence of falls and fall-related injury (Rubenstein, Reference Rubenstein2006). High rates of falls (27% to 35.9% over a 90-day assessment period) have been shown in studies of older people in receipt of home care services (Fletcher & Hirdes, Reference Fletcher and Hirdes2002; Cesari et al., Reference Cesari, Landi, Torre, Onder, Lattanzio and Bernabei2002). Hip fracture incidence in home care clients of 24.4 per 1,000 (24.4/1,000) person-years of follow-up (Stolee, Poss, Cook, Byrne, & Hirdes, Reference Stolee, Poss, Cook, Byrne and Hirdes2009) was much higher than that reported in other community-dwelling populations (5.7/1,000 person-years) (Sugarman et al., Reference Sugarman, Connell, Hansen, Helgerson, Jessup and Lee2002).

For older people in residential care facilities, the rates of falls and fall-related injuries are even higher (Rubenstein, Reference Rubenstein2006), with residents experiencing an incidence of falls nearly five times more than community-dwelling people of the same age (Australian Commission on Safety and Quality in Health Care, 2009c). Rates between one and five falls per resident per year have been reported (Australian Commission on Safety and Quality in Health Care, 2009c), although fall and injury rates vary according to the case-mix of the institution and are likely to be different for mobile people with dementia compared with dependent people with high care needs (Australian Commission on Safety and Quality in Health Care, 2009c). An Australian study (Chen et al., Reference Chen, Sambrook, Simpson, Cameron, Cumming and Seibel2009) showed a higher crude hip fracture rate (4.6% vs. 3.0% per person-year) for residents living in intermediate-care hostels compared to those living in high-care facilities (nursing homes), suggesting that exposure to risk was a factor. Similarly, in a study in Wales, age-standardized hip fracture rates per 1,000 per year in residents in the community, sheltered accommodation (group housing), residential homes (low care), and nursing homes (high care) were 5.0, 11.3, 36.1, and 17.2 respectively (Brennan et al., Reference Brennan, Johansen, Butler, Stone, Richmond and Jones2003).

In acute care settings, accidental falls in hospitalized patients accounted for 30 to 40 per cent of reported safety incidents, and occurred at a frequency of 4 to 14 per 1,000 bed-days (Oliver & Healy, Reference Oliver and Healy2009). Incidence rates varied between wards and departments in hospitals (Australian Commission on Safety and Quality in Health Care, 2009b). In a geriatric patient cohort, prior to implementing falls prevention strategies, a rate of 10 falls per 1,000 hospital-days was recorded. Of the falls, 49.1 per cent were recorded as a single fall and the remainder as two or more falls (von Renteln-Kruse & Krause, Reference von Renteln-Kruse and Krause2007). In the sub-acute or rehabilitation hospital setting, more than 40 per cent of patients with specific clinical problems, such as stroke, experience one or more falls during their admission (National Ageing Research Institute, 2004). Falls in hospitals lead to injury in about 30 per cent of cases with one to five per cent leading to serious injury (Healey et al., Reference Healey, Scobie, Oliver, Pryce, Thomson and Glampson2008).

As they occur predominantly to older people with frailty and multiple health problems (Healey et al., Reference Healey, Scobie, Oliver, Pryce, Thomson and Glampson2008), even minor injuries may lead to impaired rehabilitation, deconditioning, loss of confidence, fear of falling, and a longer stay, and are a major factor in patients moving to care homes (Oliver & Healy, Reference Oliver and Healy2009). Falls in care settings also lead to caregiver stress and fear of litigation for staff and administration (National Ageing Research Institute, 2004), when falls are seen as a failure of duty of care (Oliver & Healy, Reference Oliver and Healy2009). Since not all falls are preventable, there has to be a compromise so as not to delay rehabilitation efforts to establish or maintain independence (Oliver & Healy, Reference Oliver and Healy2009).

Causes and Circumstances of Falls

There are many causes of falls in older people stemming from the interaction between environmental hazards (or extrinsic factors) and increased susceptibility to hazards from the accumulated effects of age and co-morbidities (intrinsic factors) (Rubenstein, Reference Rubenstein2006), in association with risk-taking behaviors (Lord et al., Reference Lord, Sherrington, Menz and Close2007). Falls among people younger than 75 years are more likely to be associated with extrinsic factors, while intrinsic factors are more important among people aged 80 and older (Lach et al., Reference Lach, Reed, Arfken, Miller, Paige and Birge1991; Todd & Skelton, Reference Todd and Skelton2004).

The occurrence of falls is strongly related to exposure, in that they occur where older people are undertaking their usual daily activities (Lord et al., Reference Lord, Sherrington, Menz and Close2007). For older community-dwelling people, about 50 per cent of falls occur within their homes and immediate surroundings, with most falls occurring in commonly used rooms such as the bedroom, living area, or kitchen (Lord et al., Reference Lord, Sherrington, Menz and Close2007). The remaining falls occur in public places where commonly reported environmental hazards are cracked pavements, curbs, uneven ground, and slippery surfaces (Li et al., Reference Li, Keegan, Sternfeld, Sidney, Quesenberry and Kelsey2006), particularly in places where ice and snow may contribute to slipping (Bulajic-Kopjar, Reference Bulajic-Kopjar2000). In acute and residential aged-care settings, ambulatory patients are more at risk than non-ambulatory patients of fall-related injury (Bradley, Karani, McGinn, & Wisnivesky, Reference Bradley, Karani, McGinn and Wisnivesky2010; Thapa, Brockman, Gideon, Fought, & Ray, Reference Thapa, Brockman, Gideon, Fought and Ray1996). The majority of falls occur in the patient’s room (Thapa et al., Reference Thapa, Brockman, Gideon, Fought and Ray1996; Butler, Kerse, & Todd, Reference Butler, Kerse and Todd2004; Corsinovi et al., Reference Corsinovi, Bo, Ricauda Aimonino, Marinello, Gariglio and Marchetto2009; Krauss et al., Reference Krauss, Nguyen, Dunagan, Birge, Costantinou and Johnson2007), often from the bed (Corsinovi et al., Reference Corsinovi, Bo, Ricauda Aimonino, Marinello, Gariglio and Marchetto2009), while sitting, or during transfers (Thapa et al., Reference Thapa, Brockman, Gideon, Fought and Ray1996). Environmental hazards are implicated in approximately 15 per cent of falls in residential aged-care settings (Thapa et al., Reference Thapa, Brockman, Gideon, Fought and Ray1996; Butler et al., Reference Butler, Kerse and Todd2004).

International Trends

There is a lack of epidemiological data on falls in many parts of the developing world (World Health Organization, 2007), so that aggregated data at the national level for time trends in falls rates are more readily available for developed nations. Data from the United States showed that during the 1993–2003 time frame, the overall rate of fatal falls among persons aged 65 years or older increased (Centers for Disease Control and Prevention, 2006). Despite a decrease in the rate of hospitalizations for hip fractures, during the years 2001– 2005, the change in the overall rate of nonfatal injuries from falls was not statistically significant (Centers for Disease Control and Prevention, 2006). Canada also reported a significant increase in deaths due to falls between 1997 and 2002 (Public Health Agency Canada, 2005). Although fall-related hospitalizations for seniors remained stable over five years from 1998–1999 to 2002–2003 (Scott et al., Reference Scott, Pearce and Pengelly2005; Public Health Agency Canada, 2005), Canadian hospitalization data for the period 1985–2005 showed age-standardized rates of hip fracture declined significantly by 31.8 per cent in females and by 25.0 per cent in males (Leslie et al., Reference Leslie, O’Donnell, Jean, Lagace, Walsh and Bancej2009).

In Australia, death rates for persons aged 65 years and older fell by 13 per cent from 1997–1998 to 2004–2005 (Henley & Harrison, Reference Henley and Harrison2009), while age-standardized relative rates of hospitalized fall injury cases for both males and females increased during the period 1999–2006 in the order of 2.5 per cent (males) and 1.1 per cent (females) per year (Bradley & Pointer, Reference Bradley and Pointer2009). However, as observed in North America, the rates of hospitalized cases of hip fracture due to falls in people aged 65 years and older have been decreasing over the same period (Figure 1) (Bradley & Pointer, Reference Bradley and Pointer2009).

Figure 1: Age-standardized rates for fall injury hospitalizations, Australia 1999–2006.

Source: Bradley and Pointer (Reference Bradley and Pointer2009)

In Europe, fall-related hospital admissions are reported to be increasing in the Netherlands (Hartholt et al., Reference Hartholt, van der Velde, Looman, van Lieshout, Panneman and van Beeck2010), while a recent decline in hip fracture rates has been observed in Finland (Kannus, Niemi et al., Reference Kannus, Niemi, Parkkari, Palvanen, Vuori and Jarvinen2006). The reduction in hip fractures has been attributed to a cohort effect toward a healthier aging population and increased average body weight and improved functional ability (Kannus, Niemi, et al., Reference Kannus, Niemi, Parkkari, Palvanen, Vuori and Jarvinen2006), as well as to public health measures including wider osteoporosis screening and effective treatments (Yoshida, Reference Yoshida2007).

Costs of Falls

Fall-incurred costs are categorized into two aspects. Direct costs encompass health care costs such as medications and health-care-provider consultations in treatment and rehabilitation. Indirect costs are societal productivity losses of activities for individuals and family caregivers (World Health Organization, 2007). A number of studies (Stevens et al., Reference Stevens, Corso, Finkelstein and Miller2006; Scuffham, Chaplin, & Legood, Reference Scuffham, Chaplin and Legood2003; Hendrie, Hall, Legge, & Arena, Reference Hendrie, Hall, Legge and Arena2003) have underscored the substantial economic burden caused by fall-related injuries, regardless of the medical care system. A systematic review of international studies (Heinrich, Rapp, Rissmann, Becker, & Konig, Reference Heinrich, Rapp, Rissmann, Becker and Konig2009) showed that fall-related costs ranged between 0.85 per cent to 1.5 per cent of the total health care expenditure of countries such as North America, Australia, Europe, and the United Kingdom, equating to 0.07 per cent to 0.20 per cent of the Gross Domestic Product. For comparison purposes, costs were expressed in terms of U.S. dollar (USD) purchasing power parities (PPPs). PPPs are currency conversion rates that both convert to a common currency and equalize the purchasing power of different currencies. The mean costs per fall victim, per fall, and per fall-related hospitalization ranged from $2,044 to $25,955; $1,059 to $10,913; and $5,654 to $42,840 USD PPP respectively (Heinrich et al., Reference Heinrich, Rapp, Rissmann, Becker and Konig2009).

Policy Implications

A number of approaches across the continuum of care can be adopted for preventing falls by older people. Figure 2 shows a summary of the guiding principles (Peel et al., Reference Peel, Bell and Smith2008).

Figure 2: Guiding principles for falls prevention

A public health model targets the population as a whole, whereas the personal health model targets the individual. The aim in the personal health or medical model is to identify those at high risk and tailor individual care plans to reduce risk. The public health model of falls prevention, in contrast, aims to shift the whole population’s distribution of risk, since small shifts in some risks in the population can translate into major public health benefits (Peel et al., Reference Peel, Bell and Smith2008). Both models have a place in falls prevention, depending on setting and target group.

Population health strategies approach falls in older age within the framework of the determinants of healthy, active aging to develop effective interventions and policies (World Health Organization, 2007). The aim of promoting healthy aging is to ensure the best possible health outcomes for the current generation of older people and also for the cohort of baby boomers about to enter older age, by optimizing opportunities for improving and preserving physical, social, and mental wellness, independence, and quality of life (Peel, Bartlett, & McClure, Reference Peel, Bartlett and McClure2007).

At the individual level, strategies involve risk factor screening, assessment, and management to ensure timely diagnostics and integrated care by addressing existing health problems and ameliorating the risk condition or retarding its progression. A number of guidelines based on reviews of the evidence have been developed that outline interventions for falls prevention across the continuum of care (Australian Commission on Safety and Quality in Health Care, 2009a, 2009b, 2009c; American Geriatrics Society, British Geriatrics Society, American Academy of Orthopaedic Surgeons Panel on Falls Prevention, 2001; National Collaborating Centre for Nursing and Supportive Care, 2004).

For sustainability in preventing falls, evidence must be integrated into practice. This needs to take into account the views, preferences, and experiences of older people and their caregivers (McInnes & Askie, Reference McInnes and Askie2004). To improve the likelihood of being adopted strategies should emphasize positive messages that focus on promoting health and maintaining independence (Hughes et al., Reference Hughes, van Beurden, Eakin, Barnett, Patterson and Backhouse2008). Despite the availability of policy and practice guidelines, there are still considerable challenges on many levels for integrating best-practice falls prevention strategies into health service delivery to reduce the public health impact of falls in older age.