http://informahealthcare.com/dre ISSN 0963-8288 print/ISSN 1464-5165 online Disabil Rehabil, 2014; 36(21): 1798–1803 ! 2014 Informa UK Ltd. DOI: 10.3109/09638288.2013.871073
ASSESSMENT PROCEDURES
Development and validation of a French Canadian version of the falls Behavioral (FaB) Scale Johanne Filiatrault1,2, Louise Demers1,2, Manon Parisien3, Agathe Lorthios-Guilledroit2, Christine Kaegi4, Isabelle Me´nard4, Mary-Grace Paniconi4, and Caroline St-Laurent4
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
1
School of Rehabilitation, Faculty of Medicine, Universite´ de Montre´al, Montreal, Quebec, Canada, 2Research Centre, Institut universitaire de ge´riatrie de Montre´al, Montreal, Quebec, Canada, 3Cavendish Health and Social Services Centre, Montreal, Quebec, Canada, and 4Institut universitaire de ge´riatrie de Montre´al, Montreal, Quebec, Canada Abstract
Keywords
Purpose: To develop a French Canadian version of the Falls Behavioral (FaB) Scale and examine its psychometric properties. Methods: The FaB was adapted in French Canadian (FaB-FC) and validated according to standard guidelines for cross-cultural adaptation of questionnaires. The internal consistency and construct validity of the FaB-FC were studied among 64 community-dwelling adults aged 60 and over. The concurrent validity and test–retest reliability of the FaB-FC were respectively examined among subsamples including 31 bilingual and 33 unilingual participants. Results: The FaB-FC showed good concurrent validity with the original FaB (ICC2 ¼ 0.94; 0.87–0.97), as well as good test–retest reliability (ICC2 ¼ 0.94; 0.88–0.97). The FaB-FC also demonstrated high internal consistency ( ¼ 0.91). Moreover, analyses showed significant associations of the FaB-FC scores with fear of falling and balance confidence scores, attesting to its construct validity. Conclusion: This study provides evidence that the FaB-FC has sound psychometric properties. Since falls are associated with multiple risk factors, including behavioral factors, the FaB-FC is undoubtedly a relevant assessment tool for clinicians and researchers working toward fall prevention among French-speaking community-dwelling seniors.
Aging, falls, fall prevention, fall-related behaviors, fall risk assessment, psychometric properties History Received 26 April 2013 Revised 26 November 2013 Accepted 27 November 2013 Published online 9 January 2014
ä Implications for Rehabilitation
Fall-related behaviors should be addressed in the assessment of community-dwelling seniors’ fall risks. Like the original FaB, the French Canadian version of the tool (FaB-FC) is valid and reliable for assessing fall-related behaviors. The FaB-FC is a relevant complementary assessment tool for identifying seniors at risk for falls. The FaB-FC could also be useful in guiding fall prevention interventions and measuring the impact of these interventions on seniors’ behaviors.
Introduction Falls among older adults is considered a public health issue in most industrialized countries given their incidence and consequences on older adults’ health and quality of life [1–3]. They represent the main cause of fatal and nonfatal injuries among seniors both in the United States and in Canada [4–6]. Falls resulting in injuries such as hip fractures may cause important functional loss, jeopardize independence and impose an important burden on families [7]. The economic burden associated with falls is also substantial. By 2031, health care
Address for correspondence: Johanne Filiatrault, Associate Professor, School of Rehabilitation, Faculty of Medicine, Universite´ de Montre´al, C.P. 6128, Succursale Centre-ville, Montreal, Quebec H3C 3J7, Canada. Tel: +514 343-6111 (ext. 0836). E-mail: johanne.filiatrault@ umontreal.ca
costs for fall-related injuries among seniors will be of $4.4 billion in Canada, according to projections [8]. These considerations, and the aging of the population, have led to the development of national plans for falls prevention among older people and the inclusion of fall prevention interventions among the continuum of care offered to seniors [5,9,10]. A widely recognized strategy to prevent falls among the elderly is to implement multifactorial fall prevention strategies [11]. Consistent with such strategies, the assessment of fall risks should consider the multiple factors involved in falls. While many instruments have been developed to assess intrinsic risk factors (e.g. alteration of balance and vision) and environmental risk factors (e.g. home hazards), behavioral fall risk factors have received relatively less attention in the fall prevention literature [5,12]. To the best of our knowledge, the Falls Behavioral (FaB) Scale is the only available valid instrument reported in the literature to assess behavioral fall risk factors among communitydwelling seniors.
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
DOI: 10.3109/09638288.2013.871073
Adaptation and validation of a scale to assess falls-related behaviors
The FaB was originally developed in Australia by Clemson, Cumming and Heard [13,14]. This questionnaire was designed to assess seniors’ awareness and practice of behaviors that could potentially protect against falling. It includes 30 items related to fall prevention behaviors in 10 dimensions, including cognitive adaptation, protective mobility, avoidance and awareness among others. Respondents are asked to indicate how much each item describes the actions they are in fact doing in their everyday life (e.g. I use a light if I get up during the night). Each item of the FaB is rated on a four-level response scale ranging from 1 (never) to 4 (always). The FaB was designed to be self-administered or administered by an interviewer. It usually takes between 5–10 min to complete. Six items of the FaB (items #7–10, #19 and #23) need to be recoded so that higher item scores reflect safer behaviors [14]. An average item score can be computed for the respondent after completion of the FaB. The FaB has been shown to be reliable and valid for community-dwelling seniors [13,15]. The initial psychometric study revealed that the FaB has good internal consistency (Cronbach a ¼ 0.84), excellent test–retest reliability (ICC ¼ 0.94; p50.01) and a high content validity index (0.93) [13]. In addition, the FaB scores were positively associated with increasing age and negatively associated with greater physical mobility, implying high construct validity for evaluating protective behaviors related to falls. The FaB also demonstrated its potential to discriminate fall-related behaviors between fallers and non-fallers (seniors who reported a fall in the previous year were more inclined to adopt safer behaviors than those who did not) [13]. A second study conducted by Clemson and colleagues further established the validity and robustness of the tool using Rasch analysis [15]. Thus, the FaB seems a relevant tool to use with community-dwelling seniors. However, until now, it was not available in French. This interfered with its use among a large proportion of Canadians. Thus, the goal of this study was to develop a French Canadian version of the FaB (FaB-FC) and to examine the psychometric properties of this adapted version.
Methods The development and validation of a French Canadian version of the FaB was done with the approval of Dr. Lindy Clemson, first author of the original scale. This study was submitted to the ethics research committee of the Institut universitaire de ge´riatrie de Montre´al and received its approval in November 2010. Development of the French Canadian version of the FaB The design used in this study is based on the guidelines proposed by Vallerand [16] for cross-cultural adaptation of questionnaires. In keeping with these guidelines, the original version of the FaB was translated in French by one professional translator and two bilingual health professionals (parallel translations) working in the field of geriatrics and whose mother tongue was French. As shown in Figure 1, the two translated versions were subsequently translated in English (parallel backward translations) by a second professional translator and another team of bilingual health professionals whose mother tongue was English. This backward translation step increases the quality of the adaptation process by ensuring the intercultural equivalence between the original and translated versions of a questionnaire [16,17]. All versions of the FaB (two French Canadian versions, two backward translated English versions and the original version) were examined by an expert committee. This committee was composed of three researchers (JF, LD and MP), one research professional, one professional translator and two health professionals involved in the translation process. After comparing the
1799
Original version of the FaB Translation 1
Translation 2
2 occupational therapists
1 professional translator
Backward translation 2
Backward translation 1
1 professional translator
2 physical therapists
Expert committee 3 researchers 1 research professional 1 occupational therapist 1 physical therapist 1 professional translator
Pilot test (n = 6)
Psychometric study of the FaB-FC (n = 64)
Figure 1. Diagram showing the steps followed for the cross-cultural adaptation of the FaB.
five versions of the FaB, the committee agreed on an optimal French version for each item of the FaB. This led to the elaboration of a pre-experimental version of the FaB-FC, which was pilot tested among six bilingual community-dwelling seniors, recruited among the research participants’ database of the Institut universitaire de ge´riatrie de Montre´al and among acquaintances. Participants were asked to compare each item of the French and original versions of the FaB and indicated if they thought there were discrepancies between versions. They were also asked to report on any ambiguous or incomprehensible terms or expressions in the French version of the FaB. Minor changes were brought to the questionnaire and involved making some items more precise (#1, #22, #26), more concise (#3, #28), substituting a word (#4, #24) and removing a word (#15). This pilot test led to the elaboration of a final version of the FaB-FC. Psychometric study of the FaB-FC Design and participants The study sample included 64 community-dwelling seniors who were primarily recruited from the research participants’ database of the Institut universitaire de ge´riatrie de Montre´al, from a retirement home in a suburb of Montreal and from communitybased organizations. For inclusion in the psychometric study, participants had to be aged 60 or older and be independent for walking (with or without an assistive device). Seniors with cognitive problems, a neurological condition such as Parkinson’s disease, and seniors with uncorrected sensory deficits were not eligible for the study. Half of the participants had to be bilingual (able to understand and speak English and French). The bilingualism level was determined with the questionnaire developed by Vallerand & Halliwell [18]. This tool is composed
1800
J. Filiatrault et al.
of four items measuring a person’s ability to read, write, understand and speak a given language. For each item of the questionnaire, persons are asked to indicate their language mastery level on a response scale ranging from 1 (low) to 4 (excellent). As suggested by Vallerand & Halliwell, a minimal score of 12 was required for both languages (English and French) to consider potential participants bilingual.
Disabil Rehabil, 2014; 36(21): 1798–1803
scores obtained on the FaB-FC and fear of falling, balance confidence scores, and fall status allowed an appreciation of the construct validity of the FaB-FC. This was respectively done using Kruskal–Wallis H test, Pearson correlation and Mann–Whitney U test. Finally, the internal consistency of the FaB-FC was determined by computing a Cronbach a value using the FaB-FC scores at T1. All analyses were performed with the SPSS software version 20 (Armonk, NY).
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Data collection procedures Data were collected at the research center or in the participating retirement home or community organizations. Each participant was interviewed twice in a room free from distraction. A 1-week interval was set between interviews to limit memory bias. During the first interview (T1), the research coordinator invited each person agreeing to participate in the study to sign a consent form. Next, data on various sociodemographic and health variables (age, sex, level of education, marital status, perceived health status, fear of falling, balance confidence and fall status) were collected. In order to examine the concurrent validity of the FaBFC, 31 bilingual participants were administered both the original and French Canadian versions of the FaB. Sixteen of them were administered the original version of the FaB at T1 while 15 were administered the FaB-FC first. The other version of the FaB was administered 1 week later (T2). Thirty-three unilingual participants were administered the FaB-FC at both instances (T1 and T2). This allowed an appreciation of the test–retest reliability of the FaB-FC. Measures Fear of falling was assessed with a single question (‘‘Are you afraid of falling?’’) and a four-level response scale (1 ¼ never; 4 ¼ very often). Balance confidence was evaluated using the French and simplified version of the Activities-specific Balance Confidence (ABC) scale, originally developed by Powell & Myers[19]. The simplified version includes 15 items evaluating the person’s confidence in his/her capacity to maintain balance when performing daily activities of various levels of difficulty [20]. A psychometric study of the simplified scale, or ABC-S, indicated that this scale has good internal consistency (reliability coefficient of 0.86) and good convergent validity demonstrated by significant associations between ABC-S scores and balance performance scores, fear of falling and fall status [20]. Fall status was assessed with the following question: ‘‘How many times have you fallen during the last year?’’. Participants were later categorized as fallers (one fall or more) or non-fallers (no fall). Data analysis Descriptive statistics for each variable were computed to provide a profile of study participants. The concurrent validity of the FaB-FC was examined by comparing its scores with those obtained on the original FaB. The test–retest reliability of the FaB-FC was examined by comparing scores obtained on the French questionnaire at T1 and T2. The test–retest reliability and concurrent validity of the FaB-FC were estimated using weighted kappa (kw) for individual items [21] and intraclass correlation coefficients (ICC2) for mean scores [22]. Our sample size was appropriate to calculate kw values for items with four-level response options [23] and ICC2 values [24]. For both statistics, values closer to 1 reflect higher level of stability in measures. A kw of 0.40 and an ICC of 0.70 were set as the minimal acceptable values. Agreement percentages (Pa) were also calculated for the tool’s items, as kappa calculations can lead to paradoxical results when some levels of the scale are used by few respondents [25]. An examination of associations between
Results Participants profile Table 1 provides information on the sociodemographic characteristics and health status for the entire sample (n ¼ 64), as well as for the two subsamples (bilingual and unilingual participants). The age range of participants was 61–92 years with a mean age of 77.3 years. The sample consisted of a majority of women (78%) and most of the participants lived alone (72%). In terms of mobility, 70% of participants were independent for walking without an assistive device, while nearly a third of participants used one or two types of assistive devices to walk. Almost 75% of the participants declared being afraid to fall, at least occasionally. Finally, about a third of the participants had fallen within the last 12 months. Participants in the bilingual subsample had a higher education level and higher socioeconomic conditions than participants in the unilingual subsample (p50.05). Psychometric results The mean average FaB-FC and FaB scores obtained across the bilingual subsample were quite similar (FaB-FC: 1.79, sd ¼ 0.51; FaB: 1.74, sd ¼ 0.49). The ICC2 value computed in this subsample was 0.94 (95%IC: 0.87–0.97). In the unilingual subsample, the mean FaB-FC scores obtained at T1 and T2 were also comparable (FaB-FCT1: 1.69, sd ¼ 0.52; FaB-FCT2: 1.73, sd ¼ 0.54). The ICC2 value computed in this subsample was 0.94 (95%IC: 0.88–0.97). Table 2 reports the weighted kappas (kw) and agreement percentages (Pa) computed for each item for the analyses of the concurrent validity and the test–retest reliability of the FaB-FC. For the concurrent validity, the kw values ranged from 0.16 to 0.72, but most of the items (63%) had a kw value superior to 0.40. The Pa values between each item of the FaB-FC and the FaB ranged from 33% to 90%, and a majority of items (70%) had a Pa value higher than 50%. With regards to the test–retest reliability, calculated kw values ranged from 0.38 to 0.73. In this case, 85% of the kw were superior to 0.40. The Pa values for each item ranged from 45% to 100%, and all items except one had a Pa higher than 50%. Table 3 reports associations between the FaB-FC scores and fear of falling, balance confidence scores and fall status reported at T1 for the 64 study participants. It shows significant associations between FaB-FC scores and balance confidence scores, as well as fear of falling, but no significant association with fall status. Finally, a Cronbach’s a value of 0.91 was computed for the FaB-FC.
Discussion The purpose of this study was to develop a French Canadian version of the Falls Behavioral Scale and to examine its psychometric properties. Concurrent validity of the FaB-FC The ICC2 value computed for the analysis of consistency between the average mean scores obtained for the FaB-FC and FaB scales is indicative of good concurrent validity of the FaB-FC according
Adaptation and validation of a scale to assess falls-related behaviors
DOI: 10.3109/09638288.2013.871073
1801
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Table 1. Participants profile. Entire sample (n ¼ 64)
Unilingual subsample (n ¼ 33)
Bilingual subsample (n ¼ 31)
Mean (s.d.) or %
Mean (s.d.) or %
Mean (s.d.) or %
77.3 (8.5) 78.1
77.5 (8.9) 81.8
77.2 (8.2) 64.2
7.9 32.9 59.4
15.1 36.4 48.5
0.0 29.0 70.9
39.0 48.4 12.5 71.9
27.3 63.6 9.1 63.6
51.7 32.3 16.1 80.6
70.3 17.2 6.3 6.3
75.8 15.2 3.0 6.1
64.5 19.4 9.7 6.5
26.6 51.6 12.5 9.4 34.7 (7.0) 32.8
27.3 57.6 6.1 9.1 34.9 (7.3) 39.4
25.8 45.2 19.4 9.7 34.3 (6.8) 25.8
A. Sociodemographic variables Age Sex (women) Education level* Elementary school High school College/university Socio-economic conditions* Very wealthy/wealthy Average Poor Living condition (live alone) B. Health variables Mobility No assistive device Cane Walker Cane and walker Fear of falling Never Sometimes Often Very often Balance confidence (ABC-S) Fall status (fallers)
*Significant group difference (p50.05).
Table 2. Weighted-kappa and agreement percentage for each item of the FaB-FC. Concurrent validity (n ¼ 31) Items
n
Kw
IC 95%
1. When I stand up, I pause to get my balance. 2. I do things at a slower pace. 3. I talk with someone I know about things I do that might help prevent a fall. 4. I bend over to reach something only if I have a firm handhold. 5. I use a walking stick or walking aid when I need it. 6. When I am feeling unwell, I take particular care doing everyday things. 7. I hurry when I do things. 8. I turn around quickly. 9. To reach something up high, I use the nearest chair, or whatever furniture is handy to climb on. 10. I hurry to answer the phone. 11. I get help when I need to change a light bulb. 12. I get help when I need to reach something very high. 13. When I am feeling ill, I take special care of how I get up from a chair and move around. 14. When I am getting down from a ladder or step stool, I think about the bottom rung/step. 15. I notice spills on the floor. 16. I use a light if I get up during the night. 17. I have made changes at home to make the lighting better. 18. I clean my spectacles. 19. When wearing bifocals or trifocals, I misjudge a step or do not see a change in floor level. 20. When I buy shoes, I check the soles to see if they are slippery. 21. When I walk outdoors, I look ahead for potential hazards. 22. I avoid ramps and other slopes. 23. I avoid going out on windy, icy or wet days. 24. When I go outdoors, I think about how to move around carefully. 25. I cross at traffic lights or pedestrian crossings whenever possible. 26. I hold onto a handrail when I climb stairs. 27. I avoid walking about in crowded places. 28. I keep shrubbery and plants trimmed back on the pathways to my front/back doors. 29. I carry groceries up the stairs only in small amounts. 30. I ask my pharmacist or doctor questions about side effects of my medications.
31 31 31
0.31 0.56 0.35
0.06; 0.57 0.30; 0.81 0.10; 0.61
31 31 30 31 31 31
0.41 * 0.23 0.54 0.49 0.16
0.16; * 0.04; 0.29; 0.24; 0.09;
31 31 31 27
0.53 0.72 0.72 0.43
0.28; 0.47; 0.46; 0.18;
18
*
31 31 31 28 18
* * 0.38 * *
31 31 31 31 31 31 30 31 3 14 28
Test–retest reliability (n ¼ 33)
Pa (%)
n
Kw
IC 95%
Pa(%)
41.9 58.1 48.4
33 33 33
0.52 0.65 *
0.45; 0.96 0.39; 0.90 *
66.7 66.7 66.7
0.55 0.80 0.75 0.41
48.4 90.3 43.3 58.1 54.8 54.8
33 32 33 33 33 33
0.73 0.78 0.41 0.38 0.49 0.64
0.47; 0.52; 0.15; 0.13; 0.23; 0.39;
0.98 1.03 0.66 0.64 0.74 0.89
69.7 87.5 54.5 54.5 54.5 69.7
0.79 0.97 0.97 0.68
64.5 67.7 74.2 48.1
33 33 33 33
0.72 0.73 0.64 0.68
0.47; 0.47; 0.38; 0.42;
0.98 0.98 0.89 0.93
69.7 72.7 63.6 72.7
66.7
27
*
*
66.7
* * 0.12; 0.63 * *
54.8 64.5 41.9 71.4 77.8
33 33 33 31 20
0.49 0.73 0.40 * *
0.23; 0.74 0.47; 0.98 0.15; 0.66 * *
60.6 72.7 54.5 67.7 45.0
0.53 * * 0.51 0.30 * 0.64 0.47 *
0.28; * * 0.25; 0.04; * 0.39; 0.22; *
67.7 83.9 61.3 58.1 64.5 77.4 76.7 54.8 33.3
33 33 33 33 33 33 33 33 4
* * * 0.55 0.56 * 0.46 0.57 *
* * * 0.30; 0.80 0.31; 0.82 * 0.20; 0.71 0.31; 0.82 *
87.9 75.8 66.7 54.5 75.8 78.8 63.6 54.5 100.0
0.55 *
0.30; 0.81 *
57.1 71.4
19 32
0.53 *
0.27; 0.78 *
57.9 65.6
0.67
*
0.79 0.76 0.55 0.89 0.73
*SPSS could not compute the Kw statistics because one or more response options were not chosen by any participant for this item.
1802
J. Filiatrault et al.
Disabil Rehabil, 2014; 36(21): 1798–1803
Table 3. Associations between FaB-FC scores and fall-related variables (n ¼ 64). Variables
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Fear of falling Balance confidence Fall status
Statistics KW ¼ 8.9; p ¼ 0.03 r ¼ 0.56; p50.001 MW ¼ 441.5; p ¼ 0.89
to the expected criteria [26]. Indeed, for individual items, we found that the majority of kw values were between 0.41–0.72. This indicates that the level of agreement between the French Canadian and the original items was moderate to substantial for most of the items [27]. This confirms that most items have equivalent meaning in the original and French Canadian versions of the FaB. Surprisingly, the item 9 (To reach something up high, I use the nearest chair, or whatever furniture is handy to climb on) had a particularly low kw value (0.16) but a quite good Pa (54.8%). This apparently paradoxical result may be explained by the fact that some levels of the scale were used by very few respondents [25]. Test–retest reliability The ICC2 value computed for the mean average FaB-FC scores at T1 and T2 in the unilingual subsample also indicate that the FaB-FC has good test–retest reliability. The calculated kw values for each item of the FaB-FC were superior to 0.40 for 85% of the items, thus indicating that participants responded in a consistent manner from T1 and T2 for a majority of items. This result was further supported by moderate to high Pa values. Construct validity The FaB-FC was found to have good construct validity with balance confidence and fear of falling, as reflected by significant associations between the mean FaB-FC scores and the ABC-S scores and fear of falling measures. The directions of the measured relationships were in the expected direction, with more protective behaviors associated with increasing fear of falling and low balance confidence scores. In their original study, Clemson and colleagues also found a significant association between fall history and FaB scores (seniors who reported a fall in the previous year were more inclined to adopt safer behaviors than those who did not) [13]. In the current study, the association between FaB-FC scores and fall status was not significant, possibly because of our relatively small sample size (n ¼ 64) compared to the sample (n ¼ 418) in the study of Clemson and colleagues [13]. We nonetheless observed a positive trend; indeed, participants with a fall history in the previous 12 months generally had higher FaB-FC scores (data not shown). Finally, analyses also indicate that the FaB-FC has high internal consistency, reflected by a Cronbach alpha value of 0.91 [26]. Strengths and limits of the study Besides the rigorous methods used to develop the French Canadian version of the FaB [16], the reliance on a bilingual subsample (n ¼ 33) to examine the concurrent validity of the FaBFC is of particular value [28]. However, limitations of this study include its relatively small sample size (n ¼ 64) and the fact that a large proportion of study participants (25%) were recruited from a retirement home. Future studies should be conducted with a larger and more representative sample of community-dwelling seniors. Future research should also examine the sensitivity of the scale to change.
Conclusion Overall, findings of this psychometric study support that the FaB-FC is a valid and reliable tool for assessing fall-related behaviors among community-dwelling seniors, hence providing a complementary tool for clinicians and researchers in the detection of seniors who are at risk for falls. This tool could also be helpful in guiding fall prevention interventions (e.g. engaging discussion with seniors about their risky behaviors and raising awareness about fall prevention strategies) and measuring the impact of these interventions on seniors’ behaviors. Further studies are warranted to examine the tool’s psychometric properties and, more specifically, its sensitivity to change. As falls among community-dwelling seniors are recognized as being multifactorial, integrating a questionnaire on behavioral fall risk factors in the assessment process, such as the FaB or its French Canadian version, is undoubtedly relevant for fall prevention.
Acknowledgements The authors would like to thank seniors who participated in the study and the experts involved in the cross-validation of the FaB-FC. They would also like to acknowledge the assistance of Myriam Beaulieu who coordinated the first phases of the project, Anne-Marie Belley who collected data for the psychometric study, and Francine Giroux who guided the statistical analyses. The authors are also grateful to the Comite´ aviseur pour la recherche clinique (CAREC) of the Institut universitaire de ge´riatrie de Montre´al who financially supported this project.
Declaration of interest This project was financially supported by the Comite´ aviseur pour la recherche clinique (CAREC) of the Institut universitaire de ge´riatrie de Montre´al. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
References 1. Kannus P, Parkkari J, Niemi S, Palvanen M. Fall-induced deaths among elderly people. Am J Public Health 2005;95:422–4. 2. World Health Organization. Global report on falls prevention in older age. Geneva: World Health Organization; 2007. 3. Zecevic AA, Chesworth BM, Zaric GS, et al. Estimating the cost of serious injurious falls in a Canadian acute care hospital. Can J Aging 2012;31:139–47. 4. Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS) [online] 2010. Available from: http://www.cdc.gov/injury/wisqars/pdf/ 10LCID_All_Deaths_By_Age_Group_2010-a.pdf [last accessed 14 Jan 2013]. 5. Public Health Agency of Canada. Report on seniors’ falls in Canada. Ottawa, Canada: Public Health Agency of Canada – Division of Aging and Seniors; 2005. 6. Canadian Institute for Health Information. Health Care in Canada: a Focus on Seniors and Aging [online]. 2011. Available from: https://secure.cihi.ca/free_products/HCIC_2011_seniors_report_en. pdf [last accessed 14 Jan 2013]. 7. Lin PC, Lu CM. Hip fracture: family caregivers’ burden and related factors for older people in Taiwan. J Clin Nurs 2005;14:719–26. 8. Scott V, Wagar L, Elliott S. Falls & related injuries among older Canadians: Fall-related hospitalizations & intervention initiatives. Victoria, Canada: Victoria Scott Consulting [online] 2010. Available from: http://www.hiphealth.ca/media/research_cemfia_phac_epi_ and_inventor_20100610.pdf [last accessed 14 Jan 2013]. 9. Peel NM, Travers C, Bell RAR, Smith K. Evaluation of a health service delivery intervention to promote falls prevention in older people across the care continuum. J Eval Clin Pract 2010;16: 1254–61. 10. Ministe`re de la Sante´ et des Services sociaux. La pre´vention des chutes dans un continuum de services. Cadre de re´fe´rence. Quebec,
DOI: 10.3109/09638288.2013.871073
11. 12. 13. 14.
15. 16.
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
17. 18.
Adaptation and validation of a scale to assess falls-related behaviors
Canada: Direction ge´ne´rale de la sante´ publique, ministe`re de la Sante´ et des Services sociaux; 2004. Chang JT, Morton SC, Rubenstein LZ, et al. Interventions for the prevention of falls in older adults: systematic review and metaanalysis of randomised clinical trials. BMJ 2004;328:680. Perell KL, Nelson A, Goldman RL, et al. Fall risk assessment measures: an analytic review. J Gerontol A Biol Sci Med Sci 2001; 56:M761–6. Clemson L, Cumming RG, Heard R. The development of an assessment to evaluate behavioral factors associated with falling. Am J Occup Ther 2003;57:380–8. Clemson L, Cumming RG, Heard R. The Falls Behavioural (FaB) Scale for the older person. Instruction manual. [online] 2003. Available from: http://sydney.edu.au/health-sciences/staff/docs/ lindy_clemson/FaB_manual_2003.pdf [last accessed 14 Jan 2013]. Clemson L, Bundy AC, Cumming RG, et al. Validating the Falls Behavioural (FaB) scale for older people: a Rasch analysis. Disabil Rehabil 2008;30:498–506. Vallerand R. Vers une me´thodologie de validation transculturelle de questionnaires psychologiques: implications pour la recherche en langue franc¸aise. Can Psychol 1989;30:662–80. Vallerand R, Hess U. Me´thode de recherche en psychologie. Montreal: Gae¨tan Morin e´diteur; 2000. Vallerand RJ, Halliwell WR. One method of trans-cultural validation of psychological questionnaires: implications for sports psychology. Can J Appl Sport Sci 1983;8:9–18.
1803
19. Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. J Gerontol A Biol Sci Med Sci 1995;50A:M28–34. 20. Filiatrault J, Gauvin L, Fournier M, et al. Evidence of the psychometric qualities of a simplified version of the activitiesspecific balance confidence scale for community-dwelling seniors. Arch Phys Med Rehabil 2007;88:664–72. 21. Cohen J. Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 1968;70: 213–20. 22. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. 23. Cicchetti DV. Assessing inter-rater reliability for rating scales: resolving some basic issues. Br J Psychiatry 1976;129:452–6. 24. Streiner D, Norman GR. Health measurement scales – a practical guide to their development and use. 3rd ed. New York: Oxford University Press; 2003. 25. Feinstein AR, Cichetti DV. High agreement but low kappa: I. The problems of two paradoxes. J Clin Epidemiol 1990;43:543–9. 26. Nunally JC, Bernstein IH. Psychometric theory. 3rd ed. New York: McGraw-Hill; 1994. 27. Landis J, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74. 28. Deutscher I. Asking questions cross-culturally: some problems of linguistic comparability. In: Warwick D, Osherson S, eds. Comparative research methods. New Jersey: Prentice Hall; 1973:163–88.
ASSESSMENT PROCEDURES
Development and validation of a French Canadian version of the falls Behavioral (FaB) Scale Johanne Filiatrault1,2, Louise Demers1,2, Manon Parisien3, Agathe Lorthios-Guilledroit2, Christine Kaegi4, Isabelle Me´nard4, Mary-Grace Paniconi4, and Caroline St-Laurent4
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
1
School of Rehabilitation, Faculty of Medicine, Universite´ de Montre´al, Montreal, Quebec, Canada, 2Research Centre, Institut universitaire de ge´riatrie de Montre´al, Montreal, Quebec, Canada, 3Cavendish Health and Social Services Centre, Montreal, Quebec, Canada, and 4Institut universitaire de ge´riatrie de Montre´al, Montreal, Quebec, Canada Abstract
Keywords
Purpose: To develop a French Canadian version of the Falls Behavioral (FaB) Scale and examine its psychometric properties. Methods: The FaB was adapted in French Canadian (FaB-FC) and validated according to standard guidelines for cross-cultural adaptation of questionnaires. The internal consistency and construct validity of the FaB-FC were studied among 64 community-dwelling adults aged 60 and over. The concurrent validity and test–retest reliability of the FaB-FC were respectively examined among subsamples including 31 bilingual and 33 unilingual participants. Results: The FaB-FC showed good concurrent validity with the original FaB (ICC2 ¼ 0.94; 0.87–0.97), as well as good test–retest reliability (ICC2 ¼ 0.94; 0.88–0.97). The FaB-FC also demonstrated high internal consistency ( ¼ 0.91). Moreover, analyses showed significant associations of the FaB-FC scores with fear of falling and balance confidence scores, attesting to its construct validity. Conclusion: This study provides evidence that the FaB-FC has sound psychometric properties. Since falls are associated with multiple risk factors, including behavioral factors, the FaB-FC is undoubtedly a relevant assessment tool for clinicians and researchers working toward fall prevention among French-speaking community-dwelling seniors.
Aging, falls, fall prevention, fall-related behaviors, fall risk assessment, psychometric properties History Received 26 April 2013 Revised 26 November 2013 Accepted 27 November 2013 Published online 9 January 2014
ä Implications for Rehabilitation
Fall-related behaviors should be addressed in the assessment of community-dwelling seniors’ fall risks. Like the original FaB, the French Canadian version of the tool (FaB-FC) is valid and reliable for assessing fall-related behaviors. The FaB-FC is a relevant complementary assessment tool for identifying seniors at risk for falls. The FaB-FC could also be useful in guiding fall prevention interventions and measuring the impact of these interventions on seniors’ behaviors.
Introduction Falls among older adults is considered a public health issue in most industrialized countries given their incidence and consequences on older adults’ health and quality of life [1–3]. They represent the main cause of fatal and nonfatal injuries among seniors both in the United States and in Canada [4–6]. Falls resulting in injuries such as hip fractures may cause important functional loss, jeopardize independence and impose an important burden on families [7]. The economic burden associated with falls is also substantial. By 2031, health care
Address for correspondence: Johanne Filiatrault, Associate Professor, School of Rehabilitation, Faculty of Medicine, Universite´ de Montre´al, C.P. 6128, Succursale Centre-ville, Montreal, Quebec H3C 3J7, Canada. Tel: +514 343-6111 (ext. 0836). E-mail: johanne.filiatrault@ umontreal.ca
costs for fall-related injuries among seniors will be of $4.4 billion in Canada, according to projections [8]. These considerations, and the aging of the population, have led to the development of national plans for falls prevention among older people and the inclusion of fall prevention interventions among the continuum of care offered to seniors [5,9,10]. A widely recognized strategy to prevent falls among the elderly is to implement multifactorial fall prevention strategies [11]. Consistent with such strategies, the assessment of fall risks should consider the multiple factors involved in falls. While many instruments have been developed to assess intrinsic risk factors (e.g. alteration of balance and vision) and environmental risk factors (e.g. home hazards), behavioral fall risk factors have received relatively less attention in the fall prevention literature [5,12]. To the best of our knowledge, the Falls Behavioral (FaB) Scale is the only available valid instrument reported in the literature to assess behavioral fall risk factors among communitydwelling seniors.
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
DOI: 10.3109/09638288.2013.871073
Adaptation and validation of a scale to assess falls-related behaviors
The FaB was originally developed in Australia by Clemson, Cumming and Heard [13,14]. This questionnaire was designed to assess seniors’ awareness and practice of behaviors that could potentially protect against falling. It includes 30 items related to fall prevention behaviors in 10 dimensions, including cognitive adaptation, protective mobility, avoidance and awareness among others. Respondents are asked to indicate how much each item describes the actions they are in fact doing in their everyday life (e.g. I use a light if I get up during the night). Each item of the FaB is rated on a four-level response scale ranging from 1 (never) to 4 (always). The FaB was designed to be self-administered or administered by an interviewer. It usually takes between 5–10 min to complete. Six items of the FaB (items #7–10, #19 and #23) need to be recoded so that higher item scores reflect safer behaviors [14]. An average item score can be computed for the respondent after completion of the FaB. The FaB has been shown to be reliable and valid for community-dwelling seniors [13,15]. The initial psychometric study revealed that the FaB has good internal consistency (Cronbach a ¼ 0.84), excellent test–retest reliability (ICC ¼ 0.94; p50.01) and a high content validity index (0.93) [13]. In addition, the FaB scores were positively associated with increasing age and negatively associated with greater physical mobility, implying high construct validity for evaluating protective behaviors related to falls. The FaB also demonstrated its potential to discriminate fall-related behaviors between fallers and non-fallers (seniors who reported a fall in the previous year were more inclined to adopt safer behaviors than those who did not) [13]. A second study conducted by Clemson and colleagues further established the validity and robustness of the tool using Rasch analysis [15]. Thus, the FaB seems a relevant tool to use with community-dwelling seniors. However, until now, it was not available in French. This interfered with its use among a large proportion of Canadians. Thus, the goal of this study was to develop a French Canadian version of the FaB (FaB-FC) and to examine the psychometric properties of this adapted version.
Methods The development and validation of a French Canadian version of the FaB was done with the approval of Dr. Lindy Clemson, first author of the original scale. This study was submitted to the ethics research committee of the Institut universitaire de ge´riatrie de Montre´al and received its approval in November 2010. Development of the French Canadian version of the FaB The design used in this study is based on the guidelines proposed by Vallerand [16] for cross-cultural adaptation of questionnaires. In keeping with these guidelines, the original version of the FaB was translated in French by one professional translator and two bilingual health professionals (parallel translations) working in the field of geriatrics and whose mother tongue was French. As shown in Figure 1, the two translated versions were subsequently translated in English (parallel backward translations) by a second professional translator and another team of bilingual health professionals whose mother tongue was English. This backward translation step increases the quality of the adaptation process by ensuring the intercultural equivalence between the original and translated versions of a questionnaire [16,17]. All versions of the FaB (two French Canadian versions, two backward translated English versions and the original version) were examined by an expert committee. This committee was composed of three researchers (JF, LD and MP), one research professional, one professional translator and two health professionals involved in the translation process. After comparing the
1799
Original version of the FaB Translation 1
Translation 2
2 occupational therapists
1 professional translator
Backward translation 2
Backward translation 1
1 professional translator
2 physical therapists
Expert committee 3 researchers 1 research professional 1 occupational therapist 1 physical therapist 1 professional translator
Pilot test (n = 6)
Psychometric study of the FaB-FC (n = 64)
Figure 1. Diagram showing the steps followed for the cross-cultural adaptation of the FaB.
five versions of the FaB, the committee agreed on an optimal French version for each item of the FaB. This led to the elaboration of a pre-experimental version of the FaB-FC, which was pilot tested among six bilingual community-dwelling seniors, recruited among the research participants’ database of the Institut universitaire de ge´riatrie de Montre´al and among acquaintances. Participants were asked to compare each item of the French and original versions of the FaB and indicated if they thought there were discrepancies between versions. They were also asked to report on any ambiguous or incomprehensible terms or expressions in the French version of the FaB. Minor changes were brought to the questionnaire and involved making some items more precise (#1, #22, #26), more concise (#3, #28), substituting a word (#4, #24) and removing a word (#15). This pilot test led to the elaboration of a final version of the FaB-FC. Psychometric study of the FaB-FC Design and participants The study sample included 64 community-dwelling seniors who were primarily recruited from the research participants’ database of the Institut universitaire de ge´riatrie de Montre´al, from a retirement home in a suburb of Montreal and from communitybased organizations. For inclusion in the psychometric study, participants had to be aged 60 or older and be independent for walking (with or without an assistive device). Seniors with cognitive problems, a neurological condition such as Parkinson’s disease, and seniors with uncorrected sensory deficits were not eligible for the study. Half of the participants had to be bilingual (able to understand and speak English and French). The bilingualism level was determined with the questionnaire developed by Vallerand & Halliwell [18]. This tool is composed
1800
J. Filiatrault et al.
of four items measuring a person’s ability to read, write, understand and speak a given language. For each item of the questionnaire, persons are asked to indicate their language mastery level on a response scale ranging from 1 (low) to 4 (excellent). As suggested by Vallerand & Halliwell, a minimal score of 12 was required for both languages (English and French) to consider potential participants bilingual.
Disabil Rehabil, 2014; 36(21): 1798–1803
scores obtained on the FaB-FC and fear of falling, balance confidence scores, and fall status allowed an appreciation of the construct validity of the FaB-FC. This was respectively done using Kruskal–Wallis H test, Pearson correlation and Mann–Whitney U test. Finally, the internal consistency of the FaB-FC was determined by computing a Cronbach a value using the FaB-FC scores at T1. All analyses were performed with the SPSS software version 20 (Armonk, NY).
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Data collection procedures Data were collected at the research center or in the participating retirement home or community organizations. Each participant was interviewed twice in a room free from distraction. A 1-week interval was set between interviews to limit memory bias. During the first interview (T1), the research coordinator invited each person agreeing to participate in the study to sign a consent form. Next, data on various sociodemographic and health variables (age, sex, level of education, marital status, perceived health status, fear of falling, balance confidence and fall status) were collected. In order to examine the concurrent validity of the FaBFC, 31 bilingual participants were administered both the original and French Canadian versions of the FaB. Sixteen of them were administered the original version of the FaB at T1 while 15 were administered the FaB-FC first. The other version of the FaB was administered 1 week later (T2). Thirty-three unilingual participants were administered the FaB-FC at both instances (T1 and T2). This allowed an appreciation of the test–retest reliability of the FaB-FC. Measures Fear of falling was assessed with a single question (‘‘Are you afraid of falling?’’) and a four-level response scale (1 ¼ never; 4 ¼ very often). Balance confidence was evaluated using the French and simplified version of the Activities-specific Balance Confidence (ABC) scale, originally developed by Powell & Myers[19]. The simplified version includes 15 items evaluating the person’s confidence in his/her capacity to maintain balance when performing daily activities of various levels of difficulty [20]. A psychometric study of the simplified scale, or ABC-S, indicated that this scale has good internal consistency (reliability coefficient of 0.86) and good convergent validity demonstrated by significant associations between ABC-S scores and balance performance scores, fear of falling and fall status [20]. Fall status was assessed with the following question: ‘‘How many times have you fallen during the last year?’’. Participants were later categorized as fallers (one fall or more) or non-fallers (no fall). Data analysis Descriptive statistics for each variable were computed to provide a profile of study participants. The concurrent validity of the FaB-FC was examined by comparing its scores with those obtained on the original FaB. The test–retest reliability of the FaB-FC was examined by comparing scores obtained on the French questionnaire at T1 and T2. The test–retest reliability and concurrent validity of the FaB-FC were estimated using weighted kappa (kw) for individual items [21] and intraclass correlation coefficients (ICC2) for mean scores [22]. Our sample size was appropriate to calculate kw values for items with four-level response options [23] and ICC2 values [24]. For both statistics, values closer to 1 reflect higher level of stability in measures. A kw of 0.40 and an ICC of 0.70 were set as the minimal acceptable values. Agreement percentages (Pa) were also calculated for the tool’s items, as kappa calculations can lead to paradoxical results when some levels of the scale are used by few respondents [25]. An examination of associations between
Results Participants profile Table 1 provides information on the sociodemographic characteristics and health status for the entire sample (n ¼ 64), as well as for the two subsamples (bilingual and unilingual participants). The age range of participants was 61–92 years with a mean age of 77.3 years. The sample consisted of a majority of women (78%) and most of the participants lived alone (72%). In terms of mobility, 70% of participants were independent for walking without an assistive device, while nearly a third of participants used one or two types of assistive devices to walk. Almost 75% of the participants declared being afraid to fall, at least occasionally. Finally, about a third of the participants had fallen within the last 12 months. Participants in the bilingual subsample had a higher education level and higher socioeconomic conditions than participants in the unilingual subsample (p50.05). Psychometric results The mean average FaB-FC and FaB scores obtained across the bilingual subsample were quite similar (FaB-FC: 1.79, sd ¼ 0.51; FaB: 1.74, sd ¼ 0.49). The ICC2 value computed in this subsample was 0.94 (95%IC: 0.87–0.97). In the unilingual subsample, the mean FaB-FC scores obtained at T1 and T2 were also comparable (FaB-FCT1: 1.69, sd ¼ 0.52; FaB-FCT2: 1.73, sd ¼ 0.54). The ICC2 value computed in this subsample was 0.94 (95%IC: 0.88–0.97). Table 2 reports the weighted kappas (kw) and agreement percentages (Pa) computed for each item for the analyses of the concurrent validity and the test–retest reliability of the FaB-FC. For the concurrent validity, the kw values ranged from 0.16 to 0.72, but most of the items (63%) had a kw value superior to 0.40. The Pa values between each item of the FaB-FC and the FaB ranged from 33% to 90%, and a majority of items (70%) had a Pa value higher than 50%. With regards to the test–retest reliability, calculated kw values ranged from 0.38 to 0.73. In this case, 85% of the kw were superior to 0.40. The Pa values for each item ranged from 45% to 100%, and all items except one had a Pa higher than 50%. Table 3 reports associations between the FaB-FC scores and fear of falling, balance confidence scores and fall status reported at T1 for the 64 study participants. It shows significant associations between FaB-FC scores and balance confidence scores, as well as fear of falling, but no significant association with fall status. Finally, a Cronbach’s a value of 0.91 was computed for the FaB-FC.
Discussion The purpose of this study was to develop a French Canadian version of the Falls Behavioral Scale and to examine its psychometric properties. Concurrent validity of the FaB-FC The ICC2 value computed for the analysis of consistency between the average mean scores obtained for the FaB-FC and FaB scales is indicative of good concurrent validity of the FaB-FC according
Adaptation and validation of a scale to assess falls-related behaviors
DOI: 10.3109/09638288.2013.871073
1801
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Table 1. Participants profile. Entire sample (n ¼ 64)
Unilingual subsample (n ¼ 33)
Bilingual subsample (n ¼ 31)
Mean (s.d.) or %
Mean (s.d.) or %
Mean (s.d.) or %
77.3 (8.5) 78.1
77.5 (8.9) 81.8
77.2 (8.2) 64.2
7.9 32.9 59.4
15.1 36.4 48.5
0.0 29.0 70.9
39.0 48.4 12.5 71.9
27.3 63.6 9.1 63.6
51.7 32.3 16.1 80.6
70.3 17.2 6.3 6.3
75.8 15.2 3.0 6.1
64.5 19.4 9.7 6.5
26.6 51.6 12.5 9.4 34.7 (7.0) 32.8
27.3 57.6 6.1 9.1 34.9 (7.3) 39.4
25.8 45.2 19.4 9.7 34.3 (6.8) 25.8
A. Sociodemographic variables Age Sex (women) Education level* Elementary school High school College/university Socio-economic conditions* Very wealthy/wealthy Average Poor Living condition (live alone) B. Health variables Mobility No assistive device Cane Walker Cane and walker Fear of falling Never Sometimes Often Very often Balance confidence (ABC-S) Fall status (fallers)
*Significant group difference (p50.05).
Table 2. Weighted-kappa and agreement percentage for each item of the FaB-FC. Concurrent validity (n ¼ 31) Items
n
Kw
IC 95%
1. When I stand up, I pause to get my balance. 2. I do things at a slower pace. 3. I talk with someone I know about things I do that might help prevent a fall. 4. I bend over to reach something only if I have a firm handhold. 5. I use a walking stick or walking aid when I need it. 6. When I am feeling unwell, I take particular care doing everyday things. 7. I hurry when I do things. 8. I turn around quickly. 9. To reach something up high, I use the nearest chair, or whatever furniture is handy to climb on. 10. I hurry to answer the phone. 11. I get help when I need to change a light bulb. 12. I get help when I need to reach something very high. 13. When I am feeling ill, I take special care of how I get up from a chair and move around. 14. When I am getting down from a ladder or step stool, I think about the bottom rung/step. 15. I notice spills on the floor. 16. I use a light if I get up during the night. 17. I have made changes at home to make the lighting better. 18. I clean my spectacles. 19. When wearing bifocals or trifocals, I misjudge a step or do not see a change in floor level. 20. When I buy shoes, I check the soles to see if they are slippery. 21. When I walk outdoors, I look ahead for potential hazards. 22. I avoid ramps and other slopes. 23. I avoid going out on windy, icy or wet days. 24. When I go outdoors, I think about how to move around carefully. 25. I cross at traffic lights or pedestrian crossings whenever possible. 26. I hold onto a handrail when I climb stairs. 27. I avoid walking about in crowded places. 28. I keep shrubbery and plants trimmed back on the pathways to my front/back doors. 29. I carry groceries up the stairs only in small amounts. 30. I ask my pharmacist or doctor questions about side effects of my medications.
31 31 31
0.31 0.56 0.35
0.06; 0.57 0.30; 0.81 0.10; 0.61
31 31 30 31 31 31
0.41 * 0.23 0.54 0.49 0.16
0.16; * 0.04; 0.29; 0.24; 0.09;
31 31 31 27
0.53 0.72 0.72 0.43
0.28; 0.47; 0.46; 0.18;
18
*
31 31 31 28 18
* * 0.38 * *
31 31 31 31 31 31 30 31 3 14 28
Test–retest reliability (n ¼ 33)
Pa (%)
n
Kw
IC 95%
Pa(%)
41.9 58.1 48.4
33 33 33
0.52 0.65 *
0.45; 0.96 0.39; 0.90 *
66.7 66.7 66.7
0.55 0.80 0.75 0.41
48.4 90.3 43.3 58.1 54.8 54.8
33 32 33 33 33 33
0.73 0.78 0.41 0.38 0.49 0.64
0.47; 0.52; 0.15; 0.13; 0.23; 0.39;
0.98 1.03 0.66 0.64 0.74 0.89
69.7 87.5 54.5 54.5 54.5 69.7
0.79 0.97 0.97 0.68
64.5 67.7 74.2 48.1
33 33 33 33
0.72 0.73 0.64 0.68
0.47; 0.47; 0.38; 0.42;
0.98 0.98 0.89 0.93
69.7 72.7 63.6 72.7
66.7
27
*
*
66.7
* * 0.12; 0.63 * *
54.8 64.5 41.9 71.4 77.8
33 33 33 31 20
0.49 0.73 0.40 * *
0.23; 0.74 0.47; 0.98 0.15; 0.66 * *
60.6 72.7 54.5 67.7 45.0
0.53 * * 0.51 0.30 * 0.64 0.47 *
0.28; * * 0.25; 0.04; * 0.39; 0.22; *
67.7 83.9 61.3 58.1 64.5 77.4 76.7 54.8 33.3
33 33 33 33 33 33 33 33 4
* * * 0.55 0.56 * 0.46 0.57 *
* * * 0.30; 0.80 0.31; 0.82 * 0.20; 0.71 0.31; 0.82 *
87.9 75.8 66.7 54.5 75.8 78.8 63.6 54.5 100.0
0.55 *
0.30; 0.81 *
57.1 71.4
19 32
0.53 *
0.27; 0.78 *
57.9 65.6
0.67
*
0.79 0.76 0.55 0.89 0.73
*SPSS could not compute the Kw statistics because one or more response options were not chosen by any participant for this item.
1802
J. Filiatrault et al.
Disabil Rehabil, 2014; 36(21): 1798–1803
Table 3. Associations between FaB-FC scores and fall-related variables (n ¼ 64). Variables
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
Fear of falling Balance confidence Fall status
Statistics KW ¼ 8.9; p ¼ 0.03 r ¼ 0.56; p50.001 MW ¼ 441.5; p ¼ 0.89
to the expected criteria [26]. Indeed, for individual items, we found that the majority of kw values were between 0.41–0.72. This indicates that the level of agreement between the French Canadian and the original items was moderate to substantial for most of the items [27]. This confirms that most items have equivalent meaning in the original and French Canadian versions of the FaB. Surprisingly, the item 9 (To reach something up high, I use the nearest chair, or whatever furniture is handy to climb on) had a particularly low kw value (0.16) but a quite good Pa (54.8%). This apparently paradoxical result may be explained by the fact that some levels of the scale were used by very few respondents [25]. Test–retest reliability The ICC2 value computed for the mean average FaB-FC scores at T1 and T2 in the unilingual subsample also indicate that the FaB-FC has good test–retest reliability. The calculated kw values for each item of the FaB-FC were superior to 0.40 for 85% of the items, thus indicating that participants responded in a consistent manner from T1 and T2 for a majority of items. This result was further supported by moderate to high Pa values. Construct validity The FaB-FC was found to have good construct validity with balance confidence and fear of falling, as reflected by significant associations between the mean FaB-FC scores and the ABC-S scores and fear of falling measures. The directions of the measured relationships were in the expected direction, with more protective behaviors associated with increasing fear of falling and low balance confidence scores. In their original study, Clemson and colleagues also found a significant association between fall history and FaB scores (seniors who reported a fall in the previous year were more inclined to adopt safer behaviors than those who did not) [13]. In the current study, the association between FaB-FC scores and fall status was not significant, possibly because of our relatively small sample size (n ¼ 64) compared to the sample (n ¼ 418) in the study of Clemson and colleagues [13]. We nonetheless observed a positive trend; indeed, participants with a fall history in the previous 12 months generally had higher FaB-FC scores (data not shown). Finally, analyses also indicate that the FaB-FC has high internal consistency, reflected by a Cronbach alpha value of 0.91 [26]. Strengths and limits of the study Besides the rigorous methods used to develop the French Canadian version of the FaB [16], the reliance on a bilingual subsample (n ¼ 33) to examine the concurrent validity of the FaBFC is of particular value [28]. However, limitations of this study include its relatively small sample size (n ¼ 64) and the fact that a large proportion of study participants (25%) were recruited from a retirement home. Future studies should be conducted with a larger and more representative sample of community-dwelling seniors. Future research should also examine the sensitivity of the scale to change.
Conclusion Overall, findings of this psychometric study support that the FaB-FC is a valid and reliable tool for assessing fall-related behaviors among community-dwelling seniors, hence providing a complementary tool for clinicians and researchers in the detection of seniors who are at risk for falls. This tool could also be helpful in guiding fall prevention interventions (e.g. engaging discussion with seniors about their risky behaviors and raising awareness about fall prevention strategies) and measuring the impact of these interventions on seniors’ behaviors. Further studies are warranted to examine the tool’s psychometric properties and, more specifically, its sensitivity to change. As falls among community-dwelling seniors are recognized as being multifactorial, integrating a questionnaire on behavioral fall risk factors in the assessment process, such as the FaB or its French Canadian version, is undoubtedly relevant for fall prevention.
Acknowledgements The authors would like to thank seniors who participated in the study and the experts involved in the cross-validation of the FaB-FC. They would also like to acknowledge the assistance of Myriam Beaulieu who coordinated the first phases of the project, Anne-Marie Belley who collected data for the psychometric study, and Francine Giroux who guided the statistical analyses. The authors are also grateful to the Comite´ aviseur pour la recherche clinique (CAREC) of the Institut universitaire de ge´riatrie de Montre´al who financially supported this project.
Declaration of interest This project was financially supported by the Comite´ aviseur pour la recherche clinique (CAREC) of the Institut universitaire de ge´riatrie de Montre´al. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
References 1. Kannus P, Parkkari J, Niemi S, Palvanen M. Fall-induced deaths among elderly people. Am J Public Health 2005;95:422–4. 2. World Health Organization. Global report on falls prevention in older age. Geneva: World Health Organization; 2007. 3. Zecevic AA, Chesworth BM, Zaric GS, et al. Estimating the cost of serious injurious falls in a Canadian acute care hospital. Can J Aging 2012;31:139–47. 4. Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS) [online] 2010. Available from: http://www.cdc.gov/injury/wisqars/pdf/ 10LCID_All_Deaths_By_Age_Group_2010-a.pdf [last accessed 14 Jan 2013]. 5. Public Health Agency of Canada. Report on seniors’ falls in Canada. Ottawa, Canada: Public Health Agency of Canada – Division of Aging and Seniors; 2005. 6. Canadian Institute for Health Information. Health Care in Canada: a Focus on Seniors and Aging [online]. 2011. Available from: https://secure.cihi.ca/free_products/HCIC_2011_seniors_report_en. pdf [last accessed 14 Jan 2013]. 7. Lin PC, Lu CM. Hip fracture: family caregivers’ burden and related factors for older people in Taiwan. J Clin Nurs 2005;14:719–26. 8. Scott V, Wagar L, Elliott S. Falls & related injuries among older Canadians: Fall-related hospitalizations & intervention initiatives. Victoria, Canada: Victoria Scott Consulting [online] 2010. Available from: http://www.hiphealth.ca/media/research_cemfia_phac_epi_ and_inventor_20100610.pdf [last accessed 14 Jan 2013]. 9. Peel NM, Travers C, Bell RAR, Smith K. Evaluation of a health service delivery intervention to promote falls prevention in older people across the care continuum. J Eval Clin Pract 2010;16: 1254–61. 10. Ministe`re de la Sante´ et des Services sociaux. La pre´vention des chutes dans un continuum de services. Cadre de re´fe´rence. Quebec,
DOI: 10.3109/09638288.2013.871073
11. 12. 13. 14.
15. 16.
Disabil Rehabil Downloaded from informahealthcare.com by Gazi Univ. on 02/02/15 For personal use only.
17. 18.
Adaptation and validation of a scale to assess falls-related behaviors
Canada: Direction ge´ne´rale de la sante´ publique, ministe`re de la Sante´ et des Services sociaux; 2004. Chang JT, Morton SC, Rubenstein LZ, et al. Interventions for the prevention of falls in older adults: systematic review and metaanalysis of randomised clinical trials. BMJ 2004;328:680. Perell KL, Nelson A, Goldman RL, et al. Fall risk assessment measures: an analytic review. J Gerontol A Biol Sci Med Sci 2001; 56:M761–6. Clemson L, Cumming RG, Heard R. The development of an assessment to evaluate behavioral factors associated with falling. Am J Occup Ther 2003;57:380–8. Clemson L, Cumming RG, Heard R. The Falls Behavioural (FaB) Scale for the older person. Instruction manual. [online] 2003. Available from: http://sydney.edu.au/health-sciences/staff/docs/ lindy_clemson/FaB_manual_2003.pdf [last accessed 14 Jan 2013]. Clemson L, Bundy AC, Cumming RG, et al. Validating the Falls Behavioural (FaB) scale for older people: a Rasch analysis. Disabil Rehabil 2008;30:498–506. Vallerand R. Vers une me´thodologie de validation transculturelle de questionnaires psychologiques: implications pour la recherche en langue franc¸aise. Can Psychol 1989;30:662–80. Vallerand R, Hess U. Me´thode de recherche en psychologie. Montreal: Gae¨tan Morin e´diteur; 2000. Vallerand RJ, Halliwell WR. One method of trans-cultural validation of psychological questionnaires: implications for sports psychology. Can J Appl Sport Sci 1983;8:9–18.
1803
19. Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. J Gerontol A Biol Sci Med Sci 1995;50A:M28–34. 20. Filiatrault J, Gauvin L, Fournier M, et al. Evidence of the psychometric qualities of a simplified version of the activitiesspecific balance confidence scale for community-dwelling seniors. Arch Phys Med Rehabil 2007;88:664–72. 21. Cohen J. Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 1968;70: 213–20. 22. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. 23. Cicchetti DV. Assessing inter-rater reliability for rating scales: resolving some basic issues. Br J Psychiatry 1976;129:452–6. 24. Streiner D, Norman GR. Health measurement scales – a practical guide to their development and use. 3rd ed. New York: Oxford University Press; 2003. 25. Feinstein AR, Cichetti DV. High agreement but low kappa: I. The problems of two paradoxes. J Clin Epidemiol 1990;43:543–9. 26. Nunally JC, Bernstein IH. Psychometric theory. 3rd ed. New York: McGraw-Hill; 1994. 27. Landis J, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74. 28. Deutscher I. Asking questions cross-culturally: some problems of linguistic comparability. In: Warwick D, Osherson S, eds. Comparative research methods. New Jersey: Prentice Hall; 1973:163–88.
