Very Frequent Fallers and Future Fall Injury

Abstract

Objectives: To examine the relationship between falls history, especially those with frequent recent falls, and future injurious falls. Method: Resident Assessment Instrument for Home Care records of 167,162 home care recipients in Ontario, Canada, were linked to emergency department records recording an injurious fall. Diagnosis codes further informed the nature of the injuries. Results: Persons with a high number of recent falls tended to be younger, and more likely to have Parkinson’s or multiple sclerosis. Odds ratios for a future injurious fall, compared with zero recent falls, were as follows: 1.58 (1 fall), 1.91 (2 or 3 falls), 2.54 (4-8 falls), 3.07 (9 or more falls). Injuries among those with multiple recent falls were more likely to be head injuries with an open wound. Discussion: Persons reporting high number of recent falls were at the greatest risk of a future injurious fall and should receive the greatest attention in care planning.

Keywords

falls frequent faller injury resident assessment instrument home care

Introduction

Falls among community-based older persons have a well-established relationship with injury, hospitalization, and death (World Health Organization, 2007). Numerous risk factors have been studied as potential predictors of future falls (Gillespie et al., 2012). However, none is more predictive of a future fall than a history of falls (Deandrea et al., 2010), and history of repeated or numerous falls is associated with the greatest risk (Gassmann, Rupprecht, Freiberger, & IZG Study Group, 2009; Leclerc et al., 2008; Nevitt, Cummmings, Kidd, & Black, 1989). Studies of fall-related injuries show similar risk factors, including experiencing multiple falls (Bergland & Wyller, 2004; Divani, Vazquez, Barrett, Asadollahi, & Luft, 2009; Sambrook et al., 2007). Few, if any, studies have described those who have experienced a high number of recent falls (as many as nine or more).

This article draws on a sub-population of community residing elderly in Ontario, Canada: long-stay home care clients. The risk of fall-related injury is examined in relation to history of falls in a large provincial data set, to describe the relationship between increasing number of recent falls and the outcome of a fall-related injury and the nature of that injury.

Method

As part of normal clinical practice since 2002, Ontario adults receiving public home care services for more than 60 days and who are not palliative care clients are assessed with the Resident Assessment Instrument for Home Care (RAI-HC), a comprehensive assessment with good reliability and validity that documents important health domains to guide the plan of care (Fries et al., 2000; Landi et al., 2000; Morris et al., 1997). The RAI-HC assessment is completed around the time of home care entry, and reassessment occurs at approximately 6-month intervals (earlier if there is a significant clinical change). RAI-HC assessments are conducted by trained care coordinators who use all sources of information, including medical records, the assessor’s observations, and information provided by the care recipient and unpaid caregivers. The care coordinators continue to manage the person’s care after the assessment is completed. The RAI-HC appears frequently in the falls literature (Doran et al., 2013; Fletcher, Berg, Dalby, & Hirdes, 2009; Leung, Chi, Lou, & Chan, 2010; Yoo, 2011). The RAI-HC is the source of independent variables in this analysis.

Eligibility for receipt of long-stay public home care services in Ontario is based on the need to receive personal care, such as help with bathing or dressing, or another professional health service (usually nursing) to remain at home. Recipients are those with functional limitations and/or chronic health conditions. Although it is the case for some provinces in Canada, there is no co-payment required for these types of home care services in Ontario.

In this retrospective cohort study, the RAI-HC assessment determines the date at which independent variable measurements are recorded. The history of falls item in the RAI-HC records the number of falls experienced by the home care recipient in the 90 days prior to the assessment, using all available sources of information including recall of the client, caregiver, or others knowledgeable about the recent history, as well as falls documented in the clinical record. The cohort selected for these analyses included only assessments completed at least 90 days into the home care episode, because this represents a uniform service period and avoids bias toward cases where a fall within the 90-day look-back period but prior to home care initiation may have been the event leading to home care. Each individual contributed one record, and if more than one eligible RAI-HC assessment was available, the latest one was selected.

As part of a study on persons with neurological conditions (Cesar-Chavannes & MacDonald, 2013), all RAI-HC assessment records for Ontario home care clients who were active between April 2007 and September 2010 were linked to records from the National Ambulatory Care Reporting System (NACRS) by the Canadian Institute for Health Information. NACRS captures emergency department (ED) and ambulatory care encounters in hospital (Canadian Institute for Health Information, 2013). All encounters in NACRS for this RAI-HC assessed cohort were selected and both data sets assigned unique non-real-world identifiers to support individual-level linkage by the researchers. The NACRS records are the source of the dependent variable in this analysis: an unscheduled visit to an ED where a fall and an injury were both recorded. Note that although the broad study had as its focus those persons with neurological conditions, all clients regardless of diagnosis made up the analytic data set.

In NACRS, unscheduled ED visits were identified by their visit type code, with diseases and injuries coded using ICD-10-CA (the Canadian enhancement of the International Statistical Classification of Diseases and Related Health Problems 10th Revision [World Health Organisation, 1992]). All recorded ICD-10-CA entries were considered, including the main problem and up to nine additional entries. Falls were classified by any code beginning with “W0” or “W1” (all types of falls, including unspecified falls), and injuries by any code beginning with “S” (injury of a single body region), or “T00” through “T14” (injury of multiple body regions). These injury codes describe fractures, wounds, strains and sprains, and other presentations that are consistent with injuries likely to result from a fall. The injury codes “T15” and higher document events such as poisoning, burns, or complications of care not normally associated with a fall, and were not classified as injuries in these analyses. If more than one of these chosen injury codes were listed on the same record, the first one recorded was used for the subsequent break-down analysis.

Note that persons making the choice, or having the choice made for them, to seek assistance at an ED in the case of an injurious fall, adds a degree of imprecision in the detection of injurious falls. Especially for the least serious injuries, detection may be variable, with some seeking medical attention and others not. The care recipient and others aware of the injury may choose to call 911/emergency, or decide to travel on their own to the ED, or to manage the injury by themselves or through other professionals.

The dependent variable of interest was the first unscheduled ED visit where both a fall and an injury were coded, from the day after the RAI-HC was completed forwards for 90 days. The 90-day period was arbitrarily chosen. The timing of the RAI-HC administration is unlikely to be associated with the underlying risk of an injurious fall itself, especially because newer than 90-day cases are excluded, these being most likely to represent a post-acute situation where fall risk may be different than at other times in a home care episode. The independent variable of most interest, number of falls in previous 90 days was collapsed, based on distributional and practical considerations, to five groups: no falls, 1, 2 to 3, 4 to 8, and 9 or more. Other variables for the logistic regression were selected from RAI-HC items based on evidence in the literature. These include a measure of physical dependency using the Activities of Daily Living (ADL) Hierarchy Scale (Morris, Fries, & Morris, 1999) grouped low-mid-high for a possible U-shaped association, a measure of cognitive impairment using the Cognitive Performance Scale (Morris et al., 1994) collapsed as none versus any, having an unsteady gait, limiting going outside due to fear of falling, ADL decline in the last 90 days, not using stairs, using a walker or crutch as a primary assistive device for mobility, taking an antidepressant in the last 7 days, taking nine or more different prescribed or over-the-counter medications (based on a medication review by the assessor) in the last 7 days, being told in the last 90 days to cut down drinking of alcohol, any bladder incontinence in the last 7 days, and currently co-resides with a caregiver who provides ADL or instrumental ADL care. Disease diagnoses are documented on the RAI-HC when they are known to be recorded in the client’s medical record. Manual backward elimination of variables was performed in the logistic regression, with a test of eliminated variables in the resulting parsimonious model.

Figure 1 depicts the study design regarding timelines and data sources. Ethics clearance was granted from the University of Waterloo Office of Research Ethics (ORE# 17045).

Figure 1.

Data sources, measures, and time frames.

Results

Of 167,162 individuals in the analytic data set, 7,564 experienced one or more ED visits with an injurious fall in the 90-day period after the assessment date, a rate of 4.5%, ranging from 3.4% to 10.5% with increasing number of past falls. Table 1 shows selected characteristics by fall history. This group consists of mostly older females, with high rates of gait disturbance, cognitive impairment, and high number of medications taken. About half live with a person who provides unpaid care, whereas almost all others receive this type of care from someone who lives elsewhere.

Table 1.

Characteristics by Fall History.

Characteristics†	All	Past falls
Characteristics†	All	0	1	2-3	4-8	9+
n	167,162	119,151	25,678	15,092	4,958	2,283
Proportion		71.3%	15.4%	9.0%	3.0%	1.4%
Injurious fall in ED next 90 days	4.5%	3.4%	6.2%	7.7%	9.8%	10.5%
Age
18 to 64	16.3%	17.2%	12.0%	13.8%	17.2%	33.3%
65 to 74	14.4%	14.8%	12.7%	13.1%	15.8%	17.4%
75 to 84	34.8%	34.4%	36.3%	35.9%	34.5%	29.9%
85+	34.6%	33.7%	39.1%	37.2%	32.5%	19.5%
% Female	65.2%	65.9%	66.7%	61.6%	56.0%	52.3%
Physical impairment^a
Independent to limited assistance	85.2%	86.2%	85.2%	81.9%	76.4%	71.4%
Extensive assistance	10.8%	9.4%	11.8%	15.3%	19.7%	28.8%
Dependent	4.1%	4.4%	3.0%	2.8%	3.9%	4.8%
Cognitive impairment^b	57.8%	54.0%	62.5%	70.3%	75.3%	77.0%
Unsteady gait	62.9%	55.3%	76.1%	86.6%	91.7%	95.1%
Limits going outside, fear of falling	44.4%	39.2%	52.7%	61.5%	65.0%	68.6%
Recent ADL decline	35.5%	29.4%	45.3%	54.7%	61.7%	61.7%
Does not use stairs	48.0%	45.8%	52.2%	55.0%	56.9%	54.9%
Walker or crutch	34.9%	30.8%	42.5%	47.9%	49.0%	45.0%
Diagnosis of Alzheimer’s or other dementia	21.7%	20.7%	22.5%	25.5%	27.3%	25.2%
Parkinson’s	4.4%	3.2%	5.1%	7.6%	11.5%	19.0%
Multiple sclerosis	1.5%	1.4%	1.3%	1.8%	2.9%	6.0%
Osteoporosis	23.7%	23.2%	25.8%	24.9%	23.2%	19.8%
Antidepressant last 7 days	26.6%	24.2%	29.4%	34.1%	38.8%	44.9%
Nine or more medications last 7 days	52.4%	50.4%	56.8%	58.5%	59.2%	56.7%
Told to cut down drinking	1.1%	0.8%	1.3%	1.9%	2.5%	3.3%
Bladder incontinence	45.1%	40.9%	51.1%	58.4%	63.9%	66.1%
Co-resides with caregiver	53.2%	50.1%	52.9%	52.9%	55.7%	61.4%

Note. ED = emergency department.

†

Past fall characteristics differ for all listed characteristics, p < .0001.

ADL Hierarchy scale: 0 to 2, 3 and 4, 5 and 6.

Cognitive Performance Scale 1 or greater.

Individuals with more past falls tend to have more physical and cognitive impairment, unsteady gait, bladder incontinence, and recent ADL decline, along with higher rates of Parkinson’s, multiple sclerosis, and antidepressant use.

Table 2 shows the results of logistic regression for ED visit with an injurious fall as the dependent variable. Unadjusted characteristics associated with greater risk of an injurious fall include being female, older age, higher number of past falls, cognitive impairment, unsteady gait, fear of falling, ADL decline, not using stairs, walker/crutch, a diagnosis of Parkinson’s, dementia, or osteoporosis, antidepressant use, nine or more medications, told to cut down drinking, and bladder incontinence. Those highly dependent in their ADLs and those co-residing with a caregiver were at lower risk. In the adjusted model, increasing number of past falls brings additional risk of a future ED visit with an injurious fall, such that those with nine or more previous falls in the last 90 days have a three times greater adjusted odds of an injury-related ED visit compared with non-fallers.

Table 2.

Logistic Regression: Presentation in the ED in the Next 90 Days With an Injurious Fall.

Independent variable	Unadjusted OR (95% CI)	Adjusted OR (95% CI)
Female	1.26 [1.20, 1.32]	1.14 [1.08, 1.20]
Age (ref. = 18-64)
65 to 74	1.50 [1.35, 1.66]	1.35 [1.21, 1.50]
75 to 84	2.12 [1.94, 2.32]	1.74 [1.59, 1.92]
85+	2.77 [2.54, 3.02]	2.17 [1.97, 2.38]
Past falls (ref. = none)
One	1.84 [1.74, 1.96]	1.58 [1.49, 1.68]
Two to three	2.33 [2.17, 2.49]	1.91 [1.78, 2.05]
Four to eight	3.06 [2.78, 3.38]	2.54 (2.30, 2.82]
Nine or more	3.30 [2.87, 3.78]	3.07 [2.66, 3.55]
ADL impairment^a (ref. = independent to limited impairment)
Extensive assistance required	1.05 [0.97, 1.13]	0.82 [0.76, 0.88]
Dependent on others	0.37 [0.31, 0.45]	0.35 [0.29, 0.43]
Cognitive impairment^b	1.64 [1.56, 1.73]	1.28 [1.21, 1.35]
Unsteady gait	1.73 [1.64, 1.83]	1.22 [1.15, 1.30]
Limits going outside, fear of falling	1.34 [1.28, 1.40]	0.93 [0.88, 0.98]
Recent ADL decline	1.40 [1.34, 1.47]	1.09 [1.04, 1.14]
Does not use stairs	1.31 [1.26, 1.38]	1.07 [1.02, 1.13]
Walker or crutch	1.66 [1.59, 1.74]	1.15 [1.09, 1.21]
Parkinson’s	1.61 [1.47, 1.77]	1.32 [1.20, 1.46]
Diagnosis of Alzheimer’s or other dementia	1.40 [1.33, 1.48]	1.14 [1.08, 1.21]
Osteoporosis	1.36 [1.29, 1.43]	1.12 [1.06, 1.19]
Antidepressant last 7 days	1.32 [1.26, 1.39]	1.20 [1.14, 1.27]
Nine or more medications last 7 days	1.29 [1.23, 1.35]	1.10 [1.05, 1.15]
Told to cut down drinking	1.69 [1.41, 2.02]	1.66 [1.38, 1.99]
Bladder incontinence	1.45 [1.39, 1.52]	1.07 [1.02, 1.13]
Co-resides with caregiver	0.74 [0.70, 0.77]	0.85 [0.81, 0.89]

Note. c = .668.

ED = emergency department; OR = odds ratio; CI = confidence interval.

Activies of Daily Living Hierarchy Scale ref. = 0, 1, 2, compared with 3 and 4, and with 5 and 6.

Cognitive Performance Scale 1 or greater equivalent of Mini-Mental State Examination of 22 or less.

Table 3 shows the ICD-10-CA injury codes corresponding to body areas and types of injury by fall history. There are increasing proportions of head injuries among frequent past fallers, but decreasing proportions of injuries to the hip and thigh. In addition, increasing proportions of open wounds were found among those with more past falls, whereas fractures were a more common presentation among those injuries where few or no falls had been previously reported.

Table 3.

Body Area and Type of Injury Among ED-Presenting Injurious Falls (n = 7,564).

	All	Past falls
	All	0	1	2-3	4-8	9+
Body area*
Head	26.2%	24.0%	23.2%	30.5%	34.8%	47.1%
Neck	0.8%	0.7%	0.9%	0.7%	1.4%	0.8%
Thorax	7.0%	6.4%	7.8%	7.5%	7.6%	8.3%
Abdomen, pelvis	8.8%	9.1%	9.5%	8.6%	6.8%	4.6%
Shoulder, upper arm	7.9%	8.2%	8.8%	7.1%	6.4%	4.2%
Elbow, forearm, wrist, hand	10.4%	10.1%	11.1%	10.4%	11.1%	10.4%
Hip, thigh	24.2%	26.1%	24.1%	22.1%	19.5%	12.1%
Lower leg	7.6%	8.2%	7.6%	6.9%	5.3%	5.0%
Ankle foot	2.7%	3.0%	2.5%	2.4%	1.8%	2.5%
Multiple body regions	4.4%	4.4%	4.6%	3.9%	5.3%	5.0%
Type of injury**
Superficial	19.9%	19.0%	20.3%	19.8%	20.1%	20.4%
Open wound	15.2%	13.8%	17.7%	17.3%	18.0%	28.8%
Fracture	39.5%	42.0%	37.1%	38.5%	32.2%	22.1%
Dislocation, sprain	3.9%	4.3%	2.8%	2.5%	2.3%	4.2%
Other or unspecified	21.5%	20.8%	22.3%	21.9%	27.5%	24.6%

Note. ED = emergency department.

Past falls differ by body area, chi-square, p < .0001.

Past falls differ by type of injury, chi-square, p < .0001.

Discussion

Both the unadjusted and adjusted results show that individuals reporting high number of recent falls are the most likely to present in the ED with an injurious fall in the future. This is consistent with other findings that have reported that any fall or multiple falls are related to risk of future falls or injury (Gassmann et al., 2009; Leclerc et al., 2008; Nevitt et al., 1989; Pluijm et al., 2006; Stel et al., 2003); however, these findings extend that work to a population where recent fall counts of up to nine or more recent falls are recorded.

The logistic regression findings are generally consistent with other studies of older community-dwelling individuals, although availability and differences in measurements in multivariate models present some challenges in comparing findings. Females carry a slight additional risk as does older age, consistent with most reports (Gassmann et al., 2009; Kelly et al., 2003; Pluijm et al., 2006), although males are found to have higher risk in some studies (Fletcher & Hirdes, 2002; Grundstrom, Guse, & Layde, 2012). Functional limitations as measured by the activities of daily living scale may identify much higher levels of disability (and therefore less exposure to walking), which may explain the protective finding of the multivariate model, contrary to the predictive effect generally found (Langlois et al., 1995). Cognitive impairment is widely reported as a risk factor (Fletcher & Hirdes, 2002; Gassmann et al., 2009; Stel et al., 2003), and likely reduces judgment or processing speed or other factors related to executive function that results in falls. While limiting going outside due to fear of falling is predictive on its own, it is marginally protective in the multivariate model. The finding that a recent ADL decline is associated with increased risk, while ADL impairment itself is protective suggests that the adjustment period after a new physical limitation brings with it additional risk of an injurious fall. The finding that non-use of stairs is a risk factor may seem counterintuitive when one considers the dangers associated with stair use; however, these individuals likely have more gait disturbance making them avoid stair use. Use of a walker or crutch as an independent risk factor is consistent with other findings (Gassmann et al., 2009; Stel et al., 2003).

Parkinson’s, dementia, and osteoporosis have been described as being associated with either falling or injurious falls (Fletcher & Hirdes, 2002; Kelly et al., 2003; Nevitt et al., 1989), and their association with gait disturbance, cognitive impairment, and risk of fracture, align with the findings here. Antidepressant medication has been reported to be an independent risk factor (Kelly et al., 2003), as has polypharmacy (McMahon, Cahir, Kenny, & Bennett, 2014; Wu et al., 2013). Alcohol as a risk factor has been reported in the literature, with possible mechanisms including loss of motor control with high consumption, bone loss, and interaction with other medications (Grundstrom et al., 2012; Stel et al., 2003). Incontinence has also been commonly found in other studies (Gassmann et al., 2009; Kelly et al., 2003; Nevitt et al., 1989). Finally, the result that co-residing with one’s caregiver was protective is consistent with the previously reported finding that living alone can be predictive of falls (Gassmann et al., 2009; Nevitt et al., 1989). Persons left alone for periods of time may encounter risk in engaging in activities that would be safer if assistance or supervision were present.

The finding that individuals with high number of past falls are more likely to experience a head-area injury and open wounds, compared with those reporting few to no falls who report more hip/thigh injuries and fractures, has not been reported in the literature. Younger age (33% of those with 9+ recent falls are below 65) and less diagnosed osteoporosis may relate to being able to come out of a forceful fall impact to the hip/thigh area without injury, whereas older individuals may be more likely to fracture. Osteoporosis is undoubtedly underdiagnosed (Chesnut, 2001) and with older age and being female as risk factors, the reported values of osteoporosis in these analyses is likely more acute than the modest differences measured here (23-26% among those with 0 or 1 fall, 19% among those with 9+ falls).

Although the relative risk-adjusted rate of falls has been used as an indicator of the quality of community- and facility-based services (Jones et al., 2010; Morris, Fries, Frijters, Hirdes, & Steel, 2013), some have argued that falls resulting in injury or requiring medical attention should be considered as more important outcomes of interest (Burland, Martens, Brownell, Doupe, & Fuchs, 2013). These findings suggest that in this population of home care recipients, falls history is a strong prospective predictor of a fall-related injury serious enough to seek medical attention at the ED. Although reducing injuries from falls is always the goal, identifying those at greatest risk before the serious injury occurs is important in targeted interventions.

A limitation of this study is the possible imprecision of the main explanatory variable of interest, history of falls recorded in the RAI-HC assessment. Information to inform the assessor’s response for this item comes from all sources available, including medical records, the home care recipient themselves as well as any family member or care provider. Some degree of under-reporting may occur due to social desirability on the part of the care recipient or the family. The 90-day fall reporting period of the RAI-HC seeks to balance precision of recent events with a sufficiently long period to produce reasonable classifications based on a recent history of falls. There could be more under-reporting among care recipients living alone, as they are less likely to be observed by others, and the faller may not be forthcoming in sharing these events both with other family members or the RAI-HC assessor. That said, under-reporting is unlikely to lead to a reversal of the association between multiple fallers and future injuries because it would have to be combined with over-reporting by the multiple fallers.

There is undoubtedly some under-detection of injurious falls in the ED, as noted this may be especially true for the least serious of injuries where factors other than the injuries themselves dictate whether treatment in the ED is sought.

Strengths of this study include its large sample of individuals from a vulnerable community-dwelling population, the comprehensive range of covariate measures from the RAI-HC, and the strong outcome detection through ED records linked to the home care recipients.

These results are specific to a single provincial jurisdiction of public home care recipients in a health care system characterized by public funding with no user fees, where eligibility and receipt of services are based primarily on need. It has implications for understanding risk among a population of community-dwelling, mostly older individuals where similar types of services are directed at this population to support their continued aging in place.

The implication for practice is to take reports of any fall seriously, as it carries with it the risk of a future injury serious enough to seek medical assistance in an ED. Moreover, differences in the risk-adjusted rates of falls are reasonable to consider as quality indicators whether those falls by chance resulted in an injury or not. Once a fall is reported, it is important to probe to understand if this is a pattern of multiple falls, as this places an individual at additional risk.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Bergland

Wyller

(2004). Risk factors for serious fall related injury in elderly women living at home. Injury Prevention, 10, 308-313.

Burland

Martens

Brownell

Doupe

Fuchs

(2013). The evaluation of a fall management program in a nursing home population. Gerontologist, 53, 828-838.

Canadian Institute for Health Information. (2013, November). Retrieved from http://www.cihi.ca/CIHI-ext-portal/internet/en/document/types+of+care/hospital+care/emergency+care/NACRS_METADATA#

Cesar-Chavannes

MacDonald

(2013). National population health study of neurological conditions in Canada. Chronic Diseases and Injuries in Canada, 33, 188-191.

Chesnut

(2001). Osteoporosis, an underdiagnosed disease. Journal of the American Medical Association, 286, 2865-2866.

Deandrea

Lucenteforte

Bravi

Foschi

La Vecchia

Negri

(2010). Risk factors for falls in community-dwelling older people: A systematic review and meta-analysis. Epidemiology, 21, 658-668.

Divani

Vazquez

Barrett

Asadollahi

Luft

(2009). Risk factors associated with injury attributable to falling among elderly population with history of stroke. Stroke, 40, 3286-3292.

Doran

Hirdes

Blais

Baker

Poss

. . . McIsaac

(2013). Adverse events associated with hospitalization or detected through the RAI-HC assessment among Canadian home care clients. Healthcare Policy, 9, 76-88.

Fletcher

Berg

Dalby

Hirdes

(2009). Risk factors for falling among community-based seniors. Journal of Patient Safety, 5, 61-66.

10.

Fletcher

Hirdes

(2002, August). Risk factors for falling among community-based seniors using home care services. Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 57, M504-M510.

11.

Fries

Hirdes

Morris

Steel

Mor

Frijters

Jönsson

(2000, February). Integrated health information systems based on the RAI/MDS series of instruments. Healthcare Management Forum, 12, 30-40.

12.

Gassmann

Rupprecht

Freiberger

IZG Study Group. (2009, February). Predictors for occasional and recurrent falls in community-dwelling older people. Zeitschrift für Gerontologie und Geriatrie, 42, 3-10.

13.

Gillespie

Robertson

Gillespie

Sheerington

Gates

Cleemson

Lamb

(2012). Interventions for preventing falls in older people living in the community (Review). Cochrane Database of Systematic Reviews, CD007146. doi:10.1002/14651858.CD007146.pub3. Retrieved from http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD007146.pub3/full

14.

Grundstrom

Guse

Layde

(2012). Risk factors for falls and fall-related injuries in adults 85 years of age and older. Archives of Gerontology and Geriatrics, 54, 421-428.

15.

Jones

Hirdes

Poss

Kelly

Berg

Fries

Morris

(2010). Adjustment of nursing home quality indicators. BMC Health Services Research, 10, 96.

16.

Kelly

Pickett

Yiannakoulas

Rowe

Schopflocher

Svenson

(2003). Medication use and falls in community-dwelling older persons. Age and Ageing, 32, 503-509.

17.

Landi

Tua

Onder

Carrara

Sgadari

Rinaldi

Bernabei

(2000). Minimum data set for home care: A valid instrument to assess frail older people living in the community. Medical Care, 38, 1184-1190.

18.

Langlois

Smith

Nelson

Sattin

Stevens

DeVito

(1995). Dependence in activities of daily living as a risk factor for fall injury events among older people living in the community. Journal of the American Geriatrics Society, 43, 275-278.

19.

Leclerc

Bégin

Cadieux

Goulet

Leduc

Kergoat

. . . Lebel

(2008). Risk factors for falling among community-dwelling seniors using home-care services: An extended hazards model with time-dependent covariates and multiple events. Chronic Diseases in Canada, 28, 111-120.

20.

Leung

Chi

Lou

Chan

(2010). Psychosocial risk factors associated with falls among Chinese community-dwelling older adults in Hong Kong. Health & Social Care in the Community, 18, 272-281.

21.

McMahon

Cahir

Kenny

Bennett

(2014). Inappropriate prescribing in older fallers presenting to an Irish emergency department. Age and Ageing, 43, 44-50.

22.

Morris

J. N.

Fries

Frijters

Hirdes

Steel

(2013). interRAI home care quality indicators. BMC Geriatrics, 13, 127.

23.

Morris

J. N.

Fries

Mehr

Hawes

Phillips

Mor

Lipsitz

(1994). MDS Cognitive Performance Scale. Journal of Gerontology, 49, M174-M182.

24.

Morris

J. N.

Fries

Morris

(1999). Scaling ADLs with the MDS. Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 54, M546-M553.

25.

Morris

J. N.

Fries

Steel

Ikegami

Bernabei

Carpenter

. . . Topinkova

(1997). Comprehensive clinical assessment in community setting: Applicability of the MDS-HC. Journal of the American Geriatrics Society, 16, 259-269.

26.

Nevitt

Cummmings

Kidd

Black

(1989, May). Risk factors for recurrent nonsyncopal falls. A prospective study. Journal of the American Medical Association, 261, 2663-2668.

27.

Pluijm

Smit

Tromp

Stel

Deeg

Bouter

Lips

(2006). A risk profile for identifying community-dwelling elderly with a high risk of recurrent falling: Results of a 3-year prospective study. Osteoporosis International, 17, 417-425.

28.

Sambrook

Cameron

Chen

Cumming

Lord

March

. . . Simpson

(2007). Influence of fall related factors and bone strength on fracture risk in the frail elderly. Osteoporosis International, 603-610.

29.

Stel

Pluijm

Deeg

Smit

Bouter

Lips

(2003, October). A classification tree for predicting recurrent falling in community-dwelling older persons. Journal of the American Geriatrics Society, 51, 1356-1364.

30.

World Health Organisation. (1992).International statistical classification of diseases and related health problems, 10th Revision (ICD-10). Geneva: WHO.

31.

World Health Organization. (2007). WHO global report on falls prevention in older age. Geneva, Switzerland: Author.

32.

Chie

Yang

Liu

Kuo

Wong

Liaw

(2013, July). Factors associated with falls among community-dwelling older people in Taiwan. Annals of the Academy of Medicine, Singapore, 42, 320-327.

33.

Yoo

(2011). Recurrent falls among community-dwelling older Koreans: Prevalence and multivariate risk factors. Journal of Gerontological Nursing, 37, 28-40.