| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Alan Salmoni, PhD, School of Kinesiology, The University of Western Ontario, London, ON, N6A 3K7 Canada. E-mail: asalmoni{at}uwo.ca
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Falls • Definition • Risk factors • Reasons for falling • Community • Survey
To understand Homer, learn Greek.To understand old people, learn their language.
(Isaacs, 1992, p. 19)
Everyone intuitively knows what a fall is; but, when asked to define it, people struggle for words. It seems difficult to translate this complex event into a concrete definition. People's concept of a fall seems to rest in the domain of tacit knowledge; some things are understood when experienced but are difficult to describe in words. The question is, if everyone tacitly knows what a fall is, do researchers need to define it at all? In a report of the Kellogg International Work Group on the Prevention of Falls by the Elderly (Kellogg, 1987; Appendix A, p. 14), Bernard Isaacs warns about the possibility for "faulty transmission of information between patient and physician and between researchers, with consequent lack of clarity" unless a fall is appropriately defined. Many researchers of falls do not define a fall and leave it to the senior to decide what is meant by a question like, "How many times have you fallen in the past 12 months?" However, these same researchers use the data they have collected in order to compare "fallers" to "nonfallers" when explaining and predicting falls. It is likely that inconsistencies in what researchers and seniors define as a fall cause, in part, differences in risk factors and in the proportion of explained variance in prediction equations. Additionally, if health care practitioners hope to help seniors avoid falls, then all stakeholders should use the same language.
The Kellogg Report (1987) recognized the need (a) to define a fall in order to clearly identify which events could be included and which could not, and (b) to classify different types of falls in order to allow comparability between research results. The authors recommended that "falls be recorded as a disease entity in Index Medicus and in the International Classification of Diseases Xth Revision" (p. 1). Although the ICD–10 (World Health Organization, 1992) does not define a fall, it offers a list of plausible fall types. Under the heading "External causes of morbidity and mortality—Other external causes of accidental injury" (Codes W00–W19) the ICD–10 includes some fall categories applicable to older people. Since the publication of the Kellogg Report, researchers have focused mainly on understanding risk factors for falling and identifying of successful intervention programs. They have not continued to develop appropriate terminology and classifications of falls. The question remains: Do seniors, their health care providers, and researchers speak the same language when discussing falls? The fact that the terms slips, trips, and falls are often used interchangeably shows the confusion in the current situation. When properly defined, a slip (sliding of the support leg) and a trip (impact of the swinging leg with an external object or a body part) clearly represent different events with different causes for loss of balance. Both can cause a fall.
To our knowledge, no researcher has attempted to define a fall by asking seniors what they think a fall is or what they think causes it. Although some researchers have developed definitions with contributions from health care professionals, there is no evidence that anyone has asked a diverse group of stakeholders (including physicians, nurses, physical therapists, home care providers, or pharmacists) about what they believe constitutes a fall. Finally, it is not known whether these diverse groups define falls in the same way. Potential differences can cause two problems. If seniors participating in a research study do not understand what they are asked to report, they may interpret the meaning of the word fall differently, which could compromise the validity of the results. The other problem is that communication between seniors and their health care providers could be less effective. If seniors and their health care providers define a fall differently, there is a real probability that seniors will report only serious occurrences while ignoring or underreporting valuable warning signs—such as non-injurious falls, near falls, or missteps—that often signal increased need for intervention.
Researchers have sought to define a fall since the 1980s. Although there is still no universally accepted definition or consensus as to what a fall should be, some definitions (Kellogg, 1987) are more popular than others. In the early 1990s, Lach and colleagues (1991) observed that the research literature rarely provides operational definitions of falls, thereby preventing use of those data by other investigators. A decade later, Lord, Sherrington, and Menz (2001) suggested that the Kellogg definition was suitable for studies of sensorimotor impairments and balance, but that research concerning cardiovascular health required expansion of this definition to include syncope and dizziness, even though Nevitt, Cummings, and Hudes (1991) had already added syncopal and unexplained falls to the Kellogg definition. The Cochrane Review (Gillespie et al., 2002) recognized that many studies do not define a fall and that those that do use a variety of definitions. The Cochrane Review found that the most commonly used definition was that used for the Frailty and Injuries: Cooperative Studies of Intervention Techniques trials (Buchner et al., 1993). Not surprisingly, researchers often provide definitions for falls that reflect the needs of their particular studies.
Apart from the diversity of fall definitions present in the research literature, organizations that provide services to seniors also have a need to define a fall. Many tailor the definition to fit the characteristics of their target population (Canadian Institute for Health Information, 2002; Florida Hospital Association, 2001; Nurse Assist, 2005). The importance of the lack of an adequate definition in practice is illustrated by the "Q & A with Dr. Tideiksaar" archives at the Nurse Assist Web site, a feature in which Rein Tideiksaar answers questions from the nursing community. Between 2000 and 2005, the most frequent inquiries concerned the definition of a fall.
In preparation for the present analysis, we examined 30 fall definitions from a cross-section of research reports, books, fall assessment manuals, brochures, and Web sites published between 1987 and 2005 (Table 1). The literature review also revealed that the definition of a fall is frequently missing from the published article (American Geriatrics Society, 2001; Fletcher & Hirdes, 2000; Myers, Young, & Langlois, 1996; Talbot, Musiol, Witham, & Metter, 2005).
|
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Participants
Trained volunteers conducted a telephone survey of 477 community-dwelling adults aged 55 years or older in order to collect information about perceptions of falls, exposure to major risk factors, and awareness about falls prevention programs. In addition, we performed in-person interviews with a convenience sample of 31 health care providers who served the community. This group included 3 physicians, 3 pharmacists, 1 optometrist, 2 community-care coordinators, 4 therapists (occupational, physical, recreational, and rehabilitation), 11 nurses, 5 personal support workers, and 2 health administrators. One participating physician had some training and special interest in geriatrics, but there were no geriatricians in this community. However, because of the large population of seniors, all of the health care providers worked with a significant number of seniors. The providers worked in the local hospital, the community-based long-term care facility, a personal support service, the public health unit, a community-care access center, an optometrist's office, two pharmacies, and a family health center. Nine (29%) provider participants were men.
Procedure
We trained a group of 27 volunteers (22 of them seniors) to conduct telephone interviews and record open-ended answers with as much detail as possible and without interpretation. Because of time constraints, we permitted the interviewers to use point form in order to record the key phrases used by interviewees in response to the open-ended questions. For example, "Trip over rug and go down and injure yourself," "Anytime you land on a floor is a fall, when you are off your feet you are falling ...," "Suddenly change position without wanting to," "Slip, sudden reaction, miss step, bending," and "Panic, don't know what to do" were typical definitions of a fall captured in point form. An onsite coordinator supervised the interviewers. The coordinator selected telephone numbers by using systematic random sampling from the community telephone directory. The survey began with the interviewer determining whether one or more eligible participants lived in the household. In cases where two eligible respondents resided together, the interviewer invited both to participate. There were 28 (6%) multiple respondents in the households surveyed. The interviewer introduced the respondent to the falls prevention project, invited him or her to participate in the survey, and assured that all information provided would be kept strictly confidential. Interviewers considered a respondent's agreement to answer the survey questions as consent to participate. We calculated the response rate according to the method outlined by Aday (1996, p. 165–168) for telephone interviewing (i.e., dividing the number of successfully completed interviews into the number of eligibles). In order to calculate the number of eligibles, the equation uses (a) total number of phone calls, (b) known number of ineligibles (e.g., out of service or discontinued number, respondents who did not meet the age or gender criteria); (c) estimated number of ineligibles (e.g., no answer, answering machine, busy signal); and (d) non-interviews (e.g., refusals, unsuccessful callbacks, assumed eligibles who were not interviewed). The response rate was 36.4%. The information reported here is anonymous secondary data extracted from a database that the community established in preparation for the falls prevention program.
Survey
The actual survey was much broader than what is reported here and included questions about the perception of falls, fall history, personal risk, fear of falling, health, medications, physical activity, footwear, alcohol consumption, home safety, awareness about falls prevention programs, and demographic information. We report here only results from questions about perceptions of falls. There were two open-ended questions:
The interviewers used two prompts for Q1 ("What are the important characteristics of a falling event?" and "Please describe what happens in a fall") and one prompt for Q2 ("Any others?") only if the respondent had difficulty answering the question. After the first question, the interviewer asked the respondent for the remainder of the survey to think of a fall as "an unexpected event in which some part of a person's body comes to rest on the ground or some lower level, like a bed or chair."
Analysis
We tabulated all open-ended answers, transcribed all in-person interviews, and used content analysis in order to capture conceptualized thinking. Two investigators independently read a random sample (30%) of all responses looking for patterns in answering the question "What were seniors thinking when asked to define or describe a fall?" We then discussed independently developed preliminary coding schemes and refined them until we produced the final version. We resolved all differences by consensus. We asked two independent researchers to code 30% of the fall definition answers in order to check clarity and comprehensiveness of codes. We used their comments in order to refine the wording of the codes. The first author then used the final version of the coding scheme to code all answers.
Interviewers asked seniors to give three reasons for falling and health care providers to give five. When a respondent had given fewer than the required number of reasons, we coded the missing answers as N/A. As we had done for the definitions, we coded all reasons into categories and assigned them a descriptor (e.g., Motor Control included slips, trips, stumbles, missed steps, and so on). If the reason for falling appeared fewer than 5 times across all respondents it did not qualify as a useful category and we included it in the category Other. If a respondent mentioned two reasons that belonged in the same category (e.g., we coded mats and pets into the category Indoor Obstacles), we coded the second and third reference into the category Repeated. Interrater reliability for the fall definition, based on 30% of the answers, was more than 90% for all categories except body position change (86%) and motor control (67%). For the reasons for falling, interrater reliability ranged between 82% and 93%. We used SPSS (2001) in order to analyze the data.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Definitions of a Fall
Development of the coding scheme for the definition of a fall revealed two distinct categories of higher-level meaning: antecedents to and consequences of falling. Antecedents occurred prior to the fall event and indicated why, where, and how a fall might happen. Consequences described the change in body position, landing, or injury. Figure 1 provides all of the categories identified in the seniors' definitions and their frequencies. Most commonly, seniors included motor control (slip, trip, stumble), followed by loss of balance, injury, environmental landing point (when the participant explicitly stated the place of landing [e.g., pavement or sidewalk]), weather, body position change (when the participant did not explicitly specify the landing site [e.g., from up to down, not vertical]), and obstacles.
|
Content analysis of the fall definitions selected from the research and prevention literature (Table 1) and those provided by seniors generated 10 distinct constructs, composed of multiple descriptors (Table 3). There were many similarities in the definitions provided by seniors and researchers, but seniors talked more about the antecedents to falling and fall consequences, whereas researchers focused mainly on the description of the event itself and on exclusion criteria.
|
|
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Because both seniors and health care providers tended to focus on the consequences of falls, there is a possibility that non-injurious events will often be disregarded. This omission would negate the opportunity for early intervention and diminish the likelihood of preventing more serious falls in the future. It would be beneficial for the health care system to focus attention on preventing injuries by scanning for warning signs like near falls, mishaps, or missteps.
Some research definitions eliminate particular health conditions known to cause falls, but seniors saw health as important and often cited it when defining falls. In our opinion, elimination of any condition that is age sensitive is not desirable. Improving fall-definition reporting practices in the research literature would improve understanding among researchers and facilitate comparison among studies. We also found that the seniors very often produced incomplete definitions of falls and that there was a great deal of variability in the definitions they used. Therefore, if not provided with an appropriate definition, seniors will likely interpret the word fall in many different ways. An older person's definition and reporting of a fall event will depend on that individual's tacit knowledge, experience, and psychological mindset. Although differences in the language that seniors use to define falls might reduce reliability and validity of self reporting, it is difficult to know if this would influence the accuracy with which researchers can predict future fallers.
An example of the necessity to provide an appropriate fall definition in order to compare between research groups can be found in the common database for the Frailty and Injuries: Cooperative Studies of Intervention Techniques trials (Buchner et al., 1993). This multisite collaborative study recognized the need for common measures that would allow meaningful comparisons across the sites, so they clearly defined a fall, a near fall, and other measures of interest. This practice was recently followed by the Prevention of Falls Network Europe consensus group (Lamb, Jørstad-Stein, Hauer, & Becker, 2005), and our present study confirms the importance of this approach. Of special interest are reviews that do not report the use of a fall definition as a search or inclusion/exclusion criteria. We wonder if reviews on falls really summarize comparable results.
Based on the results of the present study, we suggest that one define a fall using wording understandable to seniors whenever falls are being considered—be it in communication with a patient, community prevention programs, educational material, or journal articles (see Table 3). Although an effort has been made in the past to produce one universal fall definition (Kellogg, 1987), such a definition may not be possible or necessary. We suggest that researchers provide study participants with an operational definition explicitly stating inclusion and exclusion criteria. Reporting this operational definition in subsequent publications will allow for appropriate comparison with other studies and for summaries in reviews and meta-analyses. This will be particularly useful when focusing on specific issues such as modifiable risk factors. The next step toward obtaining further clarity in the area of fall definitions could be the organization of a consensus conference with all stakeholders involved in the process.
Reasons for Falling
Some of the risk factors identified in the literature may not be comparable to the reasons for falls provided by the seniors and health care providers in this study. Therefore, one might expect to see differences between the perceptions of seniors and providers and the research literature. This is because researchers typically identify risk factors for falls by using correlational research techniques. Some risk factors (e.g., history of falls, as reported in Table 4), although related to falls, are not causes but simply factors correlated to the likelihood of a fall. Although the term reasons (defined as a cause or an explanation in the Concise Oxford Dictionary [Pearsall, 2001]) diminishes the gap between causes and correlated factors conceptually, it may not have been completely successful in doing so practically (i.e., as used in this survey). It is also sometimes difficult to identify an obvious biomechanical link between a risk factor identified in the literature and a cause for a fall. Fear of falling is a case in point. If fear of falling causes agonist–antagonist co-contraction resulting in increased muscle stiffness, then it clearly can be a cause for a fall. With these limitations noted, we are satisfied that the reasons identified here and the risk factors from the literature are reasonably comparable.
Considering the large number of risk factors identified in the literature (Fletcher & Hirdes, 2000; Myers et al., 1996; Nevitt et al., 1991; Tinetti, Speechley, & Ginter, 1988), it was not surprising that seniors and health care providers recognized many reasons for falling. The differences between the perceptions of seniors and providers (as shown in Figure 2) and the comparison of these perceptions with a ranking of risk factors from the literature (Table 4) are of extreme interest to the development of falls prevention programs. Because the community represented by these results is quite small and relatively isolated from other major centers, one might expect the views of seniors and their health care providers to be fairly similar, as one of the roles of the provider is to educate the patient. Although there was a 60% overlap in reasons mentioned, there were also some important differences. The focus by health care providers on medical conditions, medications, vision, and assistive devices seems to reflect the professional training they received and the work they do. From a falls prevention perspective, however, the discrepancy between the views of seniors and providers concerning medication use is problematic, because 80% of the seniors surveyed reported daily prescription medication use (see Table 2). More public education around the issue of prescription medications seems warranted. It may be the case that the links between certain medical conditions and falls may not have been effectively communicated.
The focus by seniors on weather and surface hazards outdoors partially reflects the fact that the survey was conducted in the winter months and that the community is in northern Ontario, where snow and ice are abundant at that time of year. Proper snow removal and safe walking surfaces are essential to falls prevention. Indoor obstacles were frequently mentioned by both seniors and health care providers. Inattention was mentioned frequently by seniors and to a lesser degree by health care providers. Developing an educational program to encourage seniors to pay attention at critical times would be useful.
Table 4 provides further insights into falls prevention strategies. Seniors, health care providers, and the research literature all accurately identified balance as an important risk factor for falls. In contrast, muscle weakness is an important modifiable risk factor mentioned most often in the research literature. It is potentially problematic that the health care providers surveyed did not seem to recognize the relative importance of this factor. Lower-body strength is critical to the maintenance of independent living. We could raise similar concerns about gait deficits, although seniors and health care providers in the present sample clearly recognized the Motor Control category as important (see Figure 2). Therefore, an intervention strategy for falls prevention should ensure that seniors and providers are familiar with all risk factors and their relative importance—in particular those that are modifiable either through personal or environmental strategies.
Conclusion
This study showed how seniors conceptualize falls and reasons for falling. Fall definitions from the literature were similar to those provided by seniors. However, seniors associated falls with antecedents and consequences, whereas research studies understandably focus mainly on a description of the event. Operational definitions of a fall would be valuable for both research consistency and effective falls management and prevention. In all cases one should utilize wording familiar to seniors. Holding a consensus conference of seniors, researchers, and health care providers should be the next step toward obtaining an operational definition of a fall that would maximize the likelihood that seniors would report all falling events.
|
Footnotes |
|---|
1 School of Kinesiology, University of Western Ontario, London, Canada.
2 Department of Epidemiology and Biostatistics, University of Western Ontario, London, Canada.
3 School of Physical Therapy, University of Western Ontario, London, Canada.
Decision Editor: Linda S. Noelker, PhD
Received for publication October 5, 2005. Accepted for publication March 7, 2006.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. E. Tinetti, C. Gordon, E. Sogolow, P. Lapin, and E. H. Bradley Fall-Risk Evaluation and Management: Challenges in Adopting Geriatric Care Practices Gerontologist, December 1, 2006; 46(6): 717 - 725. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
