Assessing Subjective Fall Concerns in Residential Living Seniors: Development of the Activities-Specific Fall Caution Scale

Article Outline

• Abstract
• Methods
  • Phase I: Conceptualization and Item Generation
    • Step 1: Exploration and item generation
    • Step 2: Item verification and reduction
    • Step 3: Pilot-testing
  • Phase II: Psychometric Testing and Refinement
  • Data Analysis
• Results
  • Phase I: Conceptualization and Item Generation
  • Phase II: Psychometric Testing and Refinement
    • Iteration 1
    • Iteration 2
    • Iteration 3
• Discussion
• Conclusions
• References
• Copyright

Abstract 

Pearce NJ, Myers AM, Blanchard RA. Assessing subjective fall concerns in residential living seniors: development of the Activities-specific Fall Caution Scale.

Objectives

To examine subjective fall concerns of seniors in residential care and to develop a tool applicable to both nursing home and assisted living settings.

Design

Used focus groups with residents and staff for construct examination and item generation; surveyed staff and interviewed residents for item verification; and conducted psychometric testing using Rasch analysis for scale refinement.

Setting

Seventeen residential care facilities in Ontario, Canada.

Participants

Convenience samples totaling 57 staff and 234 residents.

Interventions

Not applicable.

Main Outcome Measure

The Activities-specific Fall Caution (AFC) Scale, developed inductively with residents and staff, with items pertaining to residential living (eg, moving around a room full of people, furniture, or walkers).

Results

Resident terms (being cautious or careful) and qualifications (whether alone and proximity of gait aids) guided tool development. Rasch analysis showed that the final 13-item AFC Scale was hierarchic and unidimensional, with good person (.86) and item (.95) reliability.

Conclusions

The AFC scale is a promising new tool for assessing subjective fall concerns in residential care residents. This tool can be administered via interview in about 10 minutes to most residents with Mini-Mental State Examination scores of 12 or greater, using practice questions to determine understanding and a 4-point color response card similar to a traffic light to facilitate responding.

Key Words: Accidental falls, Assisted living facilities, Fear, Nursing homes, Rehabilitation

THERE IS AMPLE EVIDENCE that psychologic fall concerns are prevalent in community living older adults (even in nonfallers) and contribute to activity restriction, physical decline, depression, future falls, and possibly nursing home admission.¹, ², ³, ⁴ Fortunately, this psychologic construct is modifiable.⁵, ⁶, ⁷, ⁸, ⁹ Numerous tools, ranging from single fear-of-falling questions to multi-item, self-efficacy scales, have been used to assess this phenomenon.¹⁰ Dichotomous variables (presence, absence) fail to detect varying levels of such fear⁴, ¹⁰ and, although they are related, fear of falling and falls efficacy are distinct constructs.⁷ The multi-item, self-efficacy measures—notably, the Falls Efficacy Scale (FES)⁴ and the Activities-specific Balance Confidence Scale¹¹—have the advantage of theoretical grounding¹⁰ based on Bandura’s social cognitive theory.¹² Although vital to ensure conceptual relevance, only 2 developers¹¹, ¹³ have involved older adults in this process,¹⁰ and only a few have used scalogram¹¹ or Rasch analysis¹³, ¹⁴ to substantiate the structural properties of psychologic fall concern scales.

Seniors in residential care settings are particularly prone to falling,¹⁵, ¹⁶ yet little is known about related psychologic concerns in this population. Such concerns may be different and tools developed for community seniors may not capture this construct.⁷, ¹⁷ To date, only a few studies have empirically examined fall concerns in nursing home populations.¹⁸, ¹⁹, ²⁰, ²¹, ²² One study found that fear of falling in mobile residents was a significant, independent predictor of functional decline at 24 months,¹⁸ whereas another found fear of falling was present in both ambulatory and nonambulatory residents.¹⁹ Both these studies used a single item to assess fear and concern about falling, respectively. The other studies,²⁰, ²¹, ²² meanwhile, modified the 10-item FES⁴ for use with nursing home seniors, by changing or deleting items (eg, shopping, meal preparation). Apart from internal consistency,²² no other psychometric support was provided for these modifications. And, although the sample had little cognitive impairment, the authors noted that many residents still had difficulty responding to the FES confidence rating format.²²

Clearly, there is a need to further investigate the relevance of psychologic fall concerns in residential living seniors and the extent to which persons with various degrees of cognitive impairment can meaningfully respond to such measures. The validity of self-reports depends on the respondent’s awareness of the particular subjective state and his/her ability to comprehend the task.²³ Brod et al²³ developed a multi-item, quality of life tool for persons with dementia (the Dementia Quality of Life [DQoL] Instrument), beginning with focus groups of such people and care providers. Using interview administration, screening questions (to determine understanding), simple item stems and response options on an enlarged card, many seniors with mild to moderate dementia were able to respond, either verbally or by pointing to the cards.²³ Using similar techniques, another study showed that many nursing home residents were able to rate the severity and impact of constipation symptoms,²⁴ further supporting the feasibility of directly assessing subjective states in this population.

The present study sought to determine whether subjective fall concerns were relevant to seniors in residential settings, and if so, develop a tool specific to this context and suitable for administration to residents. An inductive, sequential and mixed-methods approach¹¹, ²³ was used for tool development. Guided by Bandura’s self-efficacy framework of domain and situation-specificity,¹² a priori we sought to develop a measure of progressively challenging items pertaining to standing balance, postural change, and possible perturbation,¹¹ applicable to nursing home and assisted living environments. Although assisted living facilities provide less nursing and personal care, both constitute supervised, residential living with shared environments (hallways, dining, and activity rooms), congregate dining, and other services.²⁵, ²⁶ Although assisted living residents tend to be higher functioning (particularly cognitively), mobility problems are common, often precipitating the transition from community living.²⁵, ²⁶ In our jurisdiction, older adults constitute the primary population of both nursing home and assisted living facilities; adults with complex medical needs (both young and old) are cared for in continuing care facilities.

Back to Article Outline

Methods 

The first phase of the study consisted of 3 sequential steps: construct exploration and item generation, item verification and reduction, and pilot-testing; and the second phase involved psychometric testing and scale refinement (through 3 iterations or versions). All procedures were approved by the ethics review board at the University of Waterloo and informed consent was obtained from all participants.

We recruited different convenience samples for each step using the same procedure. Staff assisted in identifying and approaching English-speaking residents they considered able to take part in simple group discussions or face-to-face interviews, similar to Frank et al.²⁴ Selection criteria regarding ambulation status was also lenient, that is, able to weight bear and walk 3m (10ft) (allowing for gait aid use and personal assistance if needed). Prior to data collection, a researcher met individually with interested residents to obtain study consent (written or verbal), background information, and permission to access their records. Resident sample characteristics for each step are shown in table 1.

Table 1. Characteristics of Resident Samples

Characteristic	Phase I			Phase II
	Focus Groups (n=14)	Ratings Sample (n=26)	Pilot Sample (n=17)	Field Sample I (n=68)	Field Sample II (n=101)
Facility type
Nursing home	10(71.5)	13(50.0)	9(52.9)	37(54.4)	52(51.5)
Assisted living	4(28.5)	13(50.0)	8(47.1)	31(45.6)	49(48.5)
Women	11(78.6)	13(88.5)	16(94.1)	46(67.6)	62(61.4)
Use gait aid	11(78.6)	22(84.6)	14(82.4)	56(82.4)	71(70.3)
Transfer assistance⁎	4(28.5)	4(15.4)	0(0.0)	8(11.9)	14(14.1)
Walking assistance⁎	4(28.5)	4(15.4)	0(0.0)	16(23.8)	3(2.9)
Falls†	11(78.6)	14(54.0)	11(64.7)	16(47.1)‡	40(40.0)§
Age (y)	80.2±8.4	78.3±8.8	81.6±7.5	83.2±8.3	82.6±7.3
MMSE score	ND	ND	ND	24.6±4.8(9−30)	24.2±5.2(5−30)

NOTE. Values are n (%) or mean ± standard deviation (range).

Abbreviations: MMSE, mini-mental State Examination; ND, no data.

Phase I: Conceptualization and Item Generation 

The primary objective was to obtain a better understanding of subjective fall concerns in residential care environments from the perspective of both residents and staff. We conducted 6 separate focus groups, 3 with residents and 3 with staff, at 2 nursing homes and 1 assisted living facility. Participating staff members (n=25) were directly involved in resident care (5 registered and 3 practical nurses, 12 personal support workers, 2 kinesiologists, 3 others). All but 1 was female, with experience working in residential care facilities ranging from 1 to 30 years (average, 12.4±8.3).

We deliberately kept the sessions short (≈30min) to sustain resident interest and minimize disruption of staff and resident routines. Semi-structured discussion scripts for both staff and resident groups centered on the meaning of fall concerns, situations or activities that evoked such concerns, and resulting action. Staff discussions additionally explored the extent of fall concerns, situations which appeared to evoke such concerns, and resident manifestations (verbal or behavioral) of this phenomenon.

Items that emerged from the situational analysis (as relevant to the residential care context, involving balance or postural change, and taking place in various areas of the facility), as well as those used in prior nursing home studies,²¹ were then subjected to ratings by residents and staff from 5 facilities. Prior to the ratings, items were grouped according to “zones” or areas where such mobility behaviors typically occur: (1) within the resident’s room; (2) outside the room, within the unit; (3) the rest of the facility; and (4) outside the facility, using the classification system from the Nursing Home Life-Space Diameter.²⁷

Using a checklist, we asked a new sample of 26 staff members to independently rate each of 21 items or activities with respect to: level of fall risk (low, moderate, high), relevance to falls (not, somewhat, very) and proportion of residents they felt would be worried or concerned in each situation (none, some, most). This sample was also predominantly women (all but 2), with occupations similar to staff in step 1, and had worked in residential care from 1 to 28 years.

Via structured interview, we asked a new sample of 26 residents (characteristics shown in table 1) whether they ever found themselves in the situation or doing the activity (yes, no) and the extent to which they were worried or concerned they might fall (not at all or never, sometimes, very, always). Additionally, both groups were asked about item clarity, whether any items should be added, and for feedback concerning possible response options—ranging from a 3-point scale to the 10-point FES format (1 [no confidence] to 10 [extreme confidence]) used previously in residential care.²¹, ²²

Reported difficulty by our sample with these numeric scales led to the development of a color-coded, visual analog response card similar to a traffic light (described in table 2). Explained in the results, this approach evolved from the residents’ own descriptions of how they reacted to different situations when concerned or worried about falling.

Table 2. Practice Questions,^⁎ Items, and Instructions for Each Scale Iteration

NOTE. Additional items for iteration 2 are in bold.

Similar to prior work,²³ 2 practice questions were developed to determine task understanding. The practice questions, color response card and preliminary set of items (determined from the ratings) were then pilot-tested with a new sample of 17 residents (described in table 1). The tool was administered via interview, item by item, using probing techniques.²³, ²⁴

Phase II: Psychometric Testing and Refinement 

The purpose of phase II was to subject the new tool—named the Activities-specific Fall Caution (AFC) Scale—to broader field testing. Based on the psychometric findings, the scale underwent 3 revisions or iterations, as described below.

The first iteration of the AFC Scale (see table 2) was administered to 73 residents (described in table 1) recruited from 12 different facilities (6 nursing home, 6 assisted living). In this phase, residents’ charts were also examined for documentation of dementia, balance disorders, and falls. Staff was asked whether each person required any personal assistance with walking and transfers (including supervision or cuing, as well as weight-bearing support). Each person was shown the laminated, 7-point response card and asked if he/she could see the various colors and corresponding labels. The 2 practice questions were then administered and the AFC Scale attempted, followed by administration of the Mini-Mental State Examination (MMSE).²⁸

Using the same protocol, we administered the second iteration of the AFC Scale to a new sample of 104 residents. To further substantiate self-awareness, during the interview this sample was asked the following questions: (1) whether they were “generally afraid or worried about falling” (not at all, a little, very); (2) what worried them most about falling; (3) whether they (or someone they knew) had had a bad fall; and (4) if so, how this had affected them. The third iteration of the scale constituted tool refinement, following Rasch analysis with this AFC dataset.

Data Analysis 

We used thematic analysis of focus group data to address construct relevance, and content analysis was used to determine situational relevance and identify preliminary items. Each group was analyzed separately prior to comparing staff versus residents, and assisted living versus nursing home residents. Quantitative analyses were conducted using SPSS^a for Windows. Preliminary scale examination looked at number of missing items, normality of distribution, item-total correlations, and internal consistency (via the Cronbach α).

Similar to 2 community fall concern tools,¹³, ¹⁴ we used Rasch analysis to examine the structural properties of the AFC Scale, namely: unidimensionality; hierarchicality; floor or ceiling effects; possible item redundancy or gaps; and adequacy of response options. Using the Winsteps computer program,^29,b raw scores were converted into interval-based, log-odds metric (or logits) and mean square standardized residuals were used to identify “misfit” items—that is, those that fail to discriminate between people with different expressions of the construct or produce erratic or variant responses. For rating scales, mean square values from 0.6 to 1.4, with associated standardized z values of less than 2.0, are considered acceptable.³⁰

In Rasch analysis, the person reliability index indicates how well the measure discriminates respondents, whereas the item reliability index denotes how well items discriminate from one another. Reliability estimates can range between 0 and 1, with higher values (closer to one) preferred.³¹ Mapping is used to detect floor or ceiling effects and probability curves illustrate adequacy of response options.³¹

Back to Article Outline

Results 

Phase I: Conceptualization and Item Generation 

The focus group data substantiated the salience of fall-related concerns in this population, but also revealed several important caveats. Although some residents openly admitted to being fearful or afraid, most disliked these terms and were much more likely to say they were “careful,” “watchful,” or “cautious.” Particularly prominent (used in all focus groups) was the phrase, “being careful.” Residents talked about their concerns in relation to actual fall experiences (themselves and others) and specific events or situations. For some scenarios, people said “I just go, don’t give it much thought”; whereas in other situations, they said: “I would go carefully or more slowly,” “I stop when I see a crowd in the hall,” or “I wait till someone comes to help me.” Two key considerations, which emerged from both men and women and across all groups, were whether they were alone and if their walker or “something to grab onto” (eg, furniture) was handy.

Although staff from both nursing home and assisted living facilities agreed that subjective fall concerns affected resident mobility behavior and feelings of safety or control, they thought that such worries were most salient for ambulatory and cognitively aware residents. Staff sometimes heard residents say they were afraid of doing certain things or witnessed manifestations of fall concerns (eg, “they stiffen up”). Although some staff thought residents felt safest in their own rooms, several residents mentioned that they were often alone in their room with no one around to help if needed. In addition, staff thought that some residents were particularly concerned about falling during transfers or bathing, but residents themselves seemed more comfortable in situations where staff provided assistance.

Content analysis revealed a number of situations, activities, and circumstances that evoked concerns about falling. The pool of 21 potential items was verified and reduced through ratings by residents and staff, in step 2, as previously described. We computed frequencies for the total sample and by facility type (nursing home vs assisted living), and we ranked items according to mean ratings or proportions (eg, frequency of doing the activity). Items with low rankings (eg, personal grooming, getting in and out of the dining room or a dining room chair) or considered ambiguous (eg, “walk around the nursing home”) were removed. The items with the highest ratings (relevance, fall risk) by both residents and staff from both types of facilities were retained for further examination.

Feedback from both staff and residents verified the importance of using terms residents were comfortable with (ie, careful versus fearful), qualifying if residents were alone and could use their gait aid, and acknowledging that for certain situations (such as getting into and out of a car) someone may assist them. Prior to pilot-testing, 2 practice questions (shown in table 2) were developed to assess residents’ ability to meaningfully respond to the tool. The first scenario (unrelated to falls) was used to determine appropriateness of responses to the rating card, and the second (postural change) was used to determine task understanding. Step 3, pilot-testing, supported the utility of the practice questions and color response card. Three quarters of the pilot sample could clearly see the 7 color gradations and written labels; the remainder (with severe vision problems) were able to respond verbally. In addition to clarification of instructions and item wording, pilot-testing was used to develop standardized prompts. For instance, if a person responded that he/she did not do a particular activity, we asked the reasons why. The person was then prompted, “If you really wanted to do something or get somewhere (say to the bathroom or the dining room for lunch), what would you do?” If the person said that they would wait for assistance or until the situation changed (eg, hallway becoming less crowded) the item was scored stop/avoid. Conversely, if the person still said they would not do the activity (eg, could not get up at night because of bed rails) the item was scored as not applicable and the interviewer proceeded with the next item.

Phase II: Psychometric Testing and Refinement 

The first iteration of the AFC Scale (shown in table 2) was administered to 73 residents. The majority, 68 (93%), were able to respond appropriately to the 2 practice questions and subsequently to all of the items. The 5 people unable to complete the practice questions could not complete the scale. Although these 5 had lower average MMSE scores (18±5; range, 11−23), some residents with comparable scores were able to respond. Characteristics of the 68 respondents are shown in table 1.

Compared with the nursing home group, the assisted living group had significantly higher MMSE scores (27.6±2.8 vs 22±4.7, P<.001). A greater percentage of nursing home residents were more likely to have documented falls in the past year (48% vs 33%), although this difference was not significant. Although use of gait aids (primarily walkers) was comparable, nursing home residents were more likely to use wheelchairs on an intermittent basis (P<.05) and require assistance with transfers (P<.01) and walking (P<.001).

Coded from 0 (go) to 6 (stop/avoid), scores on this version of the AFC Scale could potentially range from 0 to 78, with a theoretical mean of 39. Actual scores ranged from 3 to 68 (39.9±13.2). The Cronbach α indicated good internal consistency for the total sample (.85), and each subgroup (nursing home, .83; assisted living, .87). Item 7 (see a spill or wet floor sign) evoked the most caution (highest item mean score), whereas item 2 (getting in and out of bed) evoked the least (lowest mean score). All items correlated at least .33 with the total, none above .68.

Rasch analysis proceeded in stepwise fashion, first identifying misfit (erratic) items. Only 1 item (ie, item 13 [using a regular toilet]) had a mean square value (1.9; standardized z=4.2) outside the acceptable range.³⁰ This item, which also had the lowest item-total correlation (.33), may be capturing concerns other than falling (eg, some residents said that public toilets were dirty). When this item was dropped, and the mean square recalculated, the remaining 12 items all fell within the acceptable range.

Probability curves showed that respondents did not reliably distinguish among the 7 response options. Collapsing the middle response categories, as illustrated in figure 1, produced a more logical sequence (from lowest to highest levels of concern increasing along the x axis), distinct peaks, and adjacent categories crossing at or near the 50% probability line.³¹ The resulting 12-item, 4-category scale showed good person (.83) and item (.92) reliabilities, with mean square values between 0.7 and 1.4. The graphic relationship among person ability and item difficulty, however, showed some gaps and a ceiling effect (ie, no corresponding items for 11 respondents or 16% of the sample).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Probability curve for 4-category collapsed scale. Graphic representation of the probability of response for each of the 4 categories (1, go; 2, little, some, moderate; 3, very, extremely; 4, stop/avoid). The y axis represents the probability of responding to one of these categories (0−1.0) and the x axis represents the person measure minus the item measure in logits.

In an attempt to reduce gaps in the hierarchy, 5 items from the original item pool (items 4, 5, 6, 7, 17; see table 2 bold) were added. Although rated as relevant (in step 2), these items were not selected initially due to discrepant rankings by residents and staff (eg, turning and reaching were rated higher by staff versus residents, respectively). The instructions were clarified to more logically correspond with the response options and the 4-point response card modified to more closely resemble an actual traffic light, that is, presented vertically (as opposed to horizontally) and using colored circles (versus bars). The label stop/avoid was changed to wait or stop, based on resident feedback that certain activities cannot be avoided. In such instances, residents said that they either proceeded carefully or waited for assistance or more favorable circumstances (eg, dry floor).

This revised 17-item version of the AFC Scale (shown in table 2) was then administered to a sample of 104 residents. Three nursing home residents with low MMSE scores (2, 6, 7), and a documented diagnosis of Alzheimer’s disease, were unable to respond to the practice questions, and subsequently, to the AFC Scale. Characteristics for the 101 residents who completed the tool are shown in table 1.

Assisted living respondents had higher MMSE scores than nursing home respondents (25.8±4.4 vs 22.8±5.4, P<.01). Charts further indicated that only 12% of the assisted living group, as opposed to 39% of the nursing home group, had a documented diagnosis of dementia (P<.01). The use of walkers, wheelchairs, need for walking assistance, as well as the number of documented and self-reported falls, was comparable in the 2 samples. Compared with assisted living residents, nursing home residents were more likely to have balance disorders (37% vs 8%, P<.001) and to receive staff assistance with transfers (25% vs 2%, P<.001).

In response to the single-item fear-of-falling question, 51% of residents said they were “generally afraid or worried about falling” (either a little or very). Almost half of these residents (46%) were documented fallers, and 69% said they themselves had had a bad fall. When asked what worried them most about falling, the most frequent responses pertained to possible injuries (breaking a bone or hitting their head) and not being able to get up after a fall. Several of those who recalled a bad fall (55%), as well as some of those who knew someone who fell recently (36%), said they were now more careful as a result (whether or not an injury was sustained).

Coded from 0 (go) to 3 (wait/stop), scores could potentially range from 0 to 51, with a theoretical mean of 25.5. The actual mean was 16.9±9.6 (range, 0−37). Although the Cronbach α was high (.91), Rasch analysis indicated that 4 items (item 9 [walk down hallway], item 14 [take a bath or shower], item 15 [walk outside], item 17 [walk up/down a ramp]) had a mean square or standardized z value outside the acceptable range. Item 17 also had the most not applicable values (10 people).

Analysis proceeded with the third and final iteration of the tool as shown in table 2. After removing the 4 erratic items (specified above), the resulting 13-item, 4-category scale showed good person (.84) and item (.95) reliabilities and unidimensionality, with mean square values between 0.8 and 1.5. Total AFC Scale scores could now potentially range from 0 to 39 (theoretical mean, 19.5), with higher scores denoting greater caution. Twenty people had at least 1 not applicable response (most often item 3), but no more than 2. In these cases, the person’s median score based on his/her responses to the items immediately above and below the one in question (when ordered hierarchically via Rasch analysis) was used to replace the not applicable value.

Sample scores ranged from 0 to 29 (mean, 13.1±7.80). The Cronbach α was high for the sample as a whole (.90), as well as each subgroup (nursing home, .91; assisted living, .88). Item-total correlations ranged from .57 to .77.

Figure 2 shows that the AFC Scale has a hierarchic structure and good relationships between person ability and item difficulty. No items appear to be redundant (ie, on the same line); however, there are a few gaps (eg, between items 8 and 10) and a ceiling effect. Viewed hierarchically (in order of progressive difficulty), item 3 (putting on a sweater) was associated with the least caution, and item 10 (see spill or wet floor sign) was associated with the most. For tool administration, however, items are presented to residents in a logical order, beginning with situations in their own rooms and bathrooms (as shown in table 2).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  Person and item map for iteration 3. Person ability is shown on the left and item difficulty on the right (measured in logits or equal intervals). Abbreviations: M, mean; S, 1 standard deviation; T, 2 standard deviations. Legend: X is 1 respondent (located as having a 50% probability of endorsing the item).

Back to Article Outline

Discussion 

The present samples were restricted to convenience samples of English-speaking residents. Other study limitations include possible selection bias (staff may have approached higher functioning residents) and incomplete fall data. Although researchers were not blind to individual scores, residents’ charts were not reviewed until after focus groups, interviews or tool administration was completed and results for each step were not analyzed until all the data had been collected. Research in residential care generally requires staff assistance with recruitment, time to establish rapport, and patience to keep residents on track, because they often want to converse with “visitors.”²⁴ It is important to minimize distractions during tool administration and, for some people, repetition or further prompting may be required.²³, ²⁴

Increasingly, residential care facilities are implementing fall risk assessment procedures, policies to minimize restraint use and programs to mobilize residents and improve functioning.¹⁵, ¹⁶, ¹⁸, ¹⁹, ²⁰, ²¹, ²² The challenge is to increase activity and independence, while minimizing the incidence and severity of falls. For instance, use of hip protectors can lead to improved falls-related efficacy, which in turn may enhance activity.⁶ Although there is ample evidence that psychologic fall concerns are important and modifiable in community seniors, attempts to assess this construct in residential care have been limited to single-item fear-of-falling measures,¹⁸, ¹⁹ or unsupported, multi-item efficacy measures (the latter only with residents with little or no cognitive impairment).²⁰, ²¹, ²²

Our study supports prior findings that many older adults with mild to moderate cognitive impairment are still aware of their feelings and can reliably respond to subjective multi-item measures, provided such measures are simple and appropriate.²³, ²⁴ In order not to exclude potential respondents, practice or screening questions should be used to determine task comprehension.²³ Without exception, residents unable to respond appropriately to these questions were unable to complete the scale itself. Generally persons with MMSE scores of 12 or higher were able to successfully respond, as Brod et al²³ found with the DQoL Instrument. As noted by staff, more severely impaired residents may not be cognizant of fall concerns, or other subjective states. Although such people may still be able to respond to a single fear-of-falling question, there is no way to verify the validity of responses.

As in the community,² it appears that many seniors in residential care (both fallers and nonfallers) have subjective concerns about falling. Similar to the present findings, Franzoni et al¹⁸ found that nearly half their sample of mobile nursing home residents reported general fear of falling (using a single item), of whom half had fallen in the past 6 months.

Although the utility of single-item measures on general fear of falling is questionable,⁴, ¹⁰, ¹¹, ¹³ community efficacy scales cannot simply be adapted for residential care seniors, relying solely on face validity. Several items in the modified FES used in nursing homes²⁰, ²¹ were either not mentioned by our focus groups or viewed as confusing or not particularly relevant by our raters. Involving the intended target audience in the tool development process from the outset is essential to ensure conceptual relevance and content validity.¹⁰, ¹¹, ²³ Study participants discussed fall concerns logically, providing evidence of self-awareness. For instance, knowing someone who had fallen appears to contribute to fall concerns in residential living, as well as community living, seniors.² Although residents also worried about injury or getting up after a fall, other consequences of falling (eg, feeling embarrassed, causing a nuisance, losing independence) voiced by community seniors¹⁷ were not as salient in this sample.

Community seniors have also been found to respond negatively to the question “Are you afraid of falling?” saying that they used common sense and avoided risky situations if possible.¹⁴ Although some of our residents openly admitted to being afraid, most talked about being more “careful” when worried about falling, implying a desire for personal control over the situation. Similarly, nursing home residents in a psychosocial falls intervention group mentioned strategies such as taking one’s time, not rushing, walking slower, being more alert, having enough light, asking for help, and trying to avoid obstacles such as slippery floors,²⁰ further supporting the present scale. In general, our residents were far less concerned when their gait aid was handy and when in the presence of others. Falling when alone is a legitimate concern because unobserved falls are common in residential care.¹⁵ The AFC Scale assesses the resident’s perception of the extent to which he/she is cautious or careful due to fall-related concern or worry in specific situations commonly encountered in residential living. Whether the tool is tapping self-efficacy is unknown, because this construct is complex and confidence rating scales were previously found to be too difficult for nursing home residents, even those with little or no cognitive impairment.²² Our findings do support the premise that fall concerns of residential care residents are both context (being alone and proximity of gait aid) and situation-specific (eg, whether rooms or halls are crowded), consistent with Bandura’s theory.¹² According to Bandura’s theory, one’s perceived capabilities are a better predictor of behavior than actual abilities. People who are not confident or comfortable doing specific types of activities will try to avoid these situations if possible and are less likely to persist in the face of obstacles.¹²

The content of the AFC Scale is relevant to the residential care environment (both nursing home and assisted living) and representative of progressively challenging (hierarchically ordered) situations pertaining to standing balance, postural change (eg, turning, bending, getting in or out of a car), and possible perturbation (in their own words “being bumped” in a crowded hallway or room), in various locations.²⁷ Similar to community tools⁸, ¹⁴ the AFC Scale has a ceiling effect. Attempts to include more difficult items (eg, walking up and down a ramp), however, resulted in erratic or nonresponses. More challenging mobility activities contained in community fall-related tools (eg, stairs or escalators) are simply not relevant in most residential care facilities.

The structure of rating scales must also be empirically examined to provide meaning to the numeric values and connection with the tool’s descriptive content and underlying construct.¹⁴ Unlike internal consistency and factor analysis, Rasch modeling addresses the basic tenet of measurement, namely, additivity.³¹ Rasch analysis was used to locate items on a linear continuum, calibrated on a true interval scale. In the present case, the metric represents the level of fall concern expressed by the respondent, as well as the challenge of the item (ie, how much caution the situation evokes). Rasch analysis was also used to examine adequacy of response options. Too few options can restrict the respondent’s ability to communicate their experience of the latent construct, but too many options can cause confusion,³¹ a particular concern with this population. Prior nursing home samples have reportedly had difficulty with numeric rating scales.²², ²⁴ Our sample preferred the color-coded response options over numerical scales. The traffic light analogy evolved from comments by residents themselves and appeared to facilitate comprehension and ease of responding. Our sample, however, was not able to reliably differentiate among 7 response choices, similar to prior findings with community seniors.¹⁴ Residents responded more predictably after collapsing these options to a 4-point scale.

Back to Article Outline

Conclusions 

We report on the systematic development and structural properties of a new tool to assess subjective fall concerns of seniors living in residential care environments. This is the first study on subjective fall concerns to include both assisted living and nursing home residents. As expected, the assisted living sample was higher functioning,²⁵, ²⁶ although both groups had mobility problems and relevant concerns about falling. The next steps were to examine the test-retest reliability of AFC scores, relationships with functional performance measures and mobility patterns, as well as the feasibility and reliability of staff (versus researcher) administration and scoring.³² Prospective studies, similar to those in the community, are needed to determine whether scores on the AFC Scale are predictive of self-imposed activity restriction and functional decline (consistent with Bandura’s theory) and sensitive to change as the result of interventions such as walking and exercise programs, restraint reduction, or use of hip protectors. Studies with larger samples are also necessary to investigate facility differences and the impact of environmental modifications (eg, improved lighting, surface transitions, hallway design) on both falls and related subjective concerns.

Until now, research in this area has been hampered by the lack of appropriate tools for seniors in residential care. The AFC Scale, which can be administered via interview in about 10 minutes, provides a promising new tool for further investigation of subjective fall-related concerns in this population. The practice questions provide a means to determine comprehension without needlessly excluding residents with cognitive impairment, and the simple, color coded response card (similar to a familiar traffic light) appears to facilitate responding.

Suppliers

Back to Article Outline

References 

1. Cumming RG, Salkeld G, Thomas M, Szonyi G. Prospective study of the impact of fear of falling on activities of daily living, SF-36 scores, and nursing home admission. J Gerontol A Biol Sci Med Sci. 2000;55:M299–M305
   • View In Article
   • MEDLINE
   • CrossRef
2. Howland J, Lachman ME, Peterson EW, Cote J, Kasten L, Jette A. Covariates of fear of falling and associated activity curtailment. Gerontologist. 1998;38:549–555
   • View In Article
   • MEDLINE
   • CrossRef
3. Mendes de Leon CF, Seeman TE, Baker DI, Richardson ED, Tinetti ME. Self-efficacy, physical decline, and change in functioning in community-living elders: a prospective study. J Gerontol B Psychol Sci Soc Sci. 1996;51:S183–S190
   • View In Article
   • MEDLINE
4. Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. J Gerontol. 1990;45:P239–P243
   • View In Article
   • MEDLINE
5. Brouwer BJ, Walker C, Rydahl SJ, Culham EG. Reducing fear of falling in seniors through education and activity programs: a randomized trial. J Am Geriatr Soc. 2003;51:829–834
   • View In Article
   • MEDLINE
   • CrossRef
6. Cameron ID, Stafford B, Cumming RG. Hip protectors improve falls self-efficacy. Age Ageing. 2000;29:57–62
   • View In Article
   • MEDLINE
   • CrossRef
7. Li F, Fisher KJ, Harmer P, McAuley E. Falls self-efficacy as a mediator of fear of falling in an exercise intervention for older adults. J Gerontol B Psychol Sci Soc Sci. 2005;60:P34–P40
   • View In Article
   • MEDLINE
   • CrossRef
8. Myers AM, Fletcher PC, Myers AH, Sherk W. Discriminative and evaluation properties of Activities-specific Balance Confidence (ABC) Scale. J Gerontol A Bio Sci Med Sci. 1998;53:M287–M294
   • View In Article
9. Tennstedt S, Howland J, Lachman M, Peterson E, Kasten L, Jette A. A randomized, controlled trial of a group intervention to reduce fear of falling and associated activity restriction in older adults. J Gerontol B Psychol Sci Soc Sci. 1998;53:P384–P392
   • View In Article
   • MEDLINE
10. Jørstad EC, Hauer K, Becker C, Lamb SE. Measuring the psychological outcomes of falling: a systematic review. J Am Geriatr Soc. 2005;53:501–510
    • View In Article
    • MEDLINE
    • CrossRef
11. Powell LE, Myers AM. The Activities-specific Balance Confidence (ABC) Scale. J Gerontol A Biol Sci Med Sci. 1995;50:M28–M34
    • View In Article
12. Bandura A. Social foundations of thought and action: a social cognitive theory. Englewood Cliffs: Prentice-Hall; 1986;
    • View In Article
13. Lusardi MM, Smith EV. Development of a scale to assess concern about falling and applications to treatment programs. J Outcome Meas. 1997;1:34–55
    • View In Article
    • MEDLINE
14. Velozo CA, Peterson EW. Developing meaningful fear of falling measures for community dwelling elderly. Am J Phys Med Rehabil. 2001;80:662–673
    • View In Article
    • MEDLINE
    • CrossRef
15. Rubenstein LZ, Josephson KR, Osterweil D. Falls and fall prevention in the nursing home. Clin Geriatr Med. 1996;12:881–902
    • View In Article
    • MEDLINE
16. van Doorn C, Gruber-Baldini AL, Zimmerman S, et al. Dementia as a risk factor for falls and fall injuries among nursing home residents. J Am Geriatr Soc. 2003;51:1213–1218
    • View In Article
    • MEDLINE
    • CrossRef
17. Yardley L, Smith H. A prospective study of the relationship between feared consequences of falling and avoidance of activity in community-living older people. Gerontologist. 2002;42:17–23
    • View In Article
    • MEDLINE
    • CrossRef
18. Franzoni S, Rozzini R, Boffelli S, Frisoni GB. Fear of falling in nursing home patients. Gerontology. 1994;40:38–44
    • View In Article
    • MEDLINE
    • CrossRef
19. Schnelle JF, Mac Rae PG, Simmons SF, et al. Safety assessment for the frail elderly: a comparison of restrained and unrestrained nursing home residents. J Am Geriatr Soc. 1994;42:586–592
    • View In Article
    • MEDLINE
20. Gentleman B, Malozemoff W. Falls and feelings: description of a psychosocial group nursing intervention. J Gerontol Nurs. 2001;27:35–39
    • View In Article
    • MEDLINE
21. Schoenfelder DP. A fall prevention program for elderly individuals (Exercise in long-term care settings). J Gerontol Nurs. 2000;26:43–51
    • View In Article
    • MEDLINE
22. Schoenfelder DP, Rubenstein LM. An exercise program to improve fall-related outcomes in elderly nursing home residents. Appl Nurs Res. 2004;17:21–31
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (103 KB)
    • CrossRef
23. Brod M, Stewart AL, Sands L, Walton P. Conceptualization and measurement of quality of life in dementia: the Dementia Quality of Life Instrument (DQoL). Gerontologist. 1999;39:25–35
    • View In Article
    • MEDLINE
    • CrossRef
24. Frank L, Flynn J, Rothman M. Use of a self-report constipation questionnaire with older adults in long-term care. Gerontologist. 2001;41:778–786
    • View In Article
    • MEDLINE
    • CrossRef
25. Hawes C, Phillips CD, Rose M. A national survey of assisted living facilities. Gerontologist. 2003;43:875–882
    • View In Article
    • MEDLINE
    • CrossRef
26. Munroe DJ. Assisted living issues for nursing practice. Geriatr Nurs. 2003;24:99–105
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (142 KB)
    • CrossRef
27. Tinetti ME, Ginter SF. The nursing home life space diameter: a measure of extent and frequency of mobility among nursing home residents. J Am Geriatr Soc. 1990;38:1311–1315
    • View In Article
    • MEDLINE
28. Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading the cognitive status of patients for the clinician. J Psychiatr Res. 1975;12:189–198
    • View In Article
    • Full-Text PDF (635 KB)
    • CrossRef
29. Linacre JM, Wright BD. A user’s guide to Winsteps, Bigsteps, Ministeps: Rasch-model computer programs. Chicago: Mesa Pr; 1999;
    • View In Article
30. Wright B, Linacre M. Reasonable item mean-square fit values. Rasch Meas Trans. 1994;8:370
    • View In Article
31. Bond T, Fox C. Applying the Rasch model: fundamental measurement in the human sciences. Mahwah: Lawrence Erlbaum Associates; 2001;
    • View In Article
32. Blanchard RA, Myers AM, Pearce NJ. Reliability, construct validity, and clinical feasibility of the Activities-specific Fall Caution Scale for residential living seniors. Arch Phys Med Rehabil. 2007;88:732–739
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (131 KB)
    • CrossRef

• a Version 12.0; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
• b Winsteps, PO Box 811322, Chicago, IL 60681-1322.