INTERVENTIONS

Multifactorial/Multicomponent Interventions
to Address Identified Risk(s) and Prevent Falls

Background

Most falls result from interactions between long- and short-term factors within the host and precipitating factors in the environment. [Tinetti 1988, 1995]  Observational studies have shown that each of the following conditions or factors increases the subsequent risk of falling: arthritis; depressive symptoms; postural (orthostatic) hypotension; impaired cognition, vision, balance, gait, or muscle strength; use of psychoactive medications; and treatment with four or more prescription medications. Furthermore, the risk of falling has been shown to increase as the number of these risk factors increases. [Tinetti 1988, Nevitt, 1989, Robbins 1989]   In clinical trials, researchers have attempted to modify either a single risk factor or multiple risk factors, and both strategies have been shown to be effective in reducing the rate of falling. Targeting multiple risk factors appears to be effective only if efforts are made to ensure that the interventions are carried out. The reduction in fall risk may be associated with the number of risk factors improved or eliminated [Tinetti et al., 1996].

Two methods for reducing multiple risk factors have been tested in clinical trials. The first method, termed “multicomponent intervention” in this guideline, refers to a set of interventions offered to all participants in a program that addresses more than one intervention category. This method has been used most often in long-term care settings. In the second method, called “multifactorial intervention,” participants are offered only the tailored subset of interventions that target the risk factors that have been identified through a fall risk factor assessment. This targeted or tailored approach has been implemented primarily among community-dwelling older persons. [Hauer, 2006] There is a great deal of heterogeneity among the designs of the multifactorial and multicomponent studies and they can be differentiated in many dimensions (i.e., health care based vs. population based, high risk population vs general older adults, direct intervention vs. referral). Since differentiation of these approaches was beyond the scope of this guideline we therefore included trials with multifactorial or multicomponent approaches regardless of dimensions.

Most of the components included in multicomponent or tailored multifactorial interventions can be described under the broad headings of exercise and physical activity, medical assessment and management, medication adjustment, environmental modification, and education. These components represent distinct areas of expertise and clinical practice. Therefore, the interventions are often administered by several clinicians from various disciplines, presenting challenges of coordination.

In deciding which groups of older adults will benefit most from multicomponent or multifactorial interventions, it is helpful to review the evidence for relevant subgroups of older adults.  The main subgroup addressed in this section and throughout the guideline is the population of older persons residing in the community.  For the populations of older people residing in long-term care settings or of individuals with cognitive deficits, information is offered in Appendix A and B of this update.

Recommendations

13. A strategy to reduce the risk of falls should include multifactorial assessment of known fall risk factors and management of the risk factors identified.[A]
14. The components most commonly included in efficacious interventions were:
1. Adaptation or modification of home environment [A]
2. Withdrawal or minimization of psychoactive medications [B]
3. Withdrawal or minimization of other medications [C]
4. Management of postural hypotension [C]
5. Management of foot problems and footwear [C]
6. Exercise, particularly balance, strength, and gait training [A]
15. All older adults who are at risk of falling should be offered an exercise program incorporating balance, gait, and strength training. Flexibility and endurance training should also be offered, but not as sole components of the program. [A]
16. Multifactorial/multicomponent intervention should include an education component complementing and addressing issues specific to the intervention being provided, tailored to individual cognitive function and language. [C]
17. The health professional or team conducting the fall risk assessment should directly implement the interventions or should assure that the interventions are carried out by other qualified healthcare professionals. [A]

Rationale

An intervention strategy based on a multifactorial assessment of known fall risk factors and followed by linked interventions appears to be an effective approach for reducing the rate of falls among cognitively intact, community-living older people at risk of falling. However, to date, studies evaluating multifactorial interventions have not been designed to assess the contribution of each component. Therefore, we are unable to make strong recommendations concerning the benefits accrued from individual intervention components.  Recommendations are based on the most commonly included components of the effective interventions.

The multifactorial/multicomponent approach to interventions designed to prevent falls in older persons is supported by a significant body of evidence including two meta-analyses by the Cochrane Collaborative and by Chang et al (2004). Additional studies have been published since these meta-analyses which supplement the earlier evidence. The need for careful monitoring and follow-up is highlighted in several studies in which nine of ten that documented assessment and intervention processes that were carefully overseen and monitored proved to be beneficial. This contrasted with studies which provided only advice, knowledge or unmonitored referral in which only two of 12 were effective. These findings were also corroborated in recent meta-analyses. (Gates et al, 2008)  

Evidence Statements

Systematic reviews

Gillespie et al., 2003.  A meta-analysis of five randomized controlled trials found that "multidisciplinary, multifactorial, health/environmental risk factor screening and intervention programs" significantly reduced the number of participants falling and also reduced the incidence of falls among community-dwelling older people.

Chang et al., 2004. This meta-analysis of 40 randomized controlled trials investigated the effectiveness of multifactorial assessments plus various combinations of multiple interventions aimed at preventing falls in older adults. The results demonstrated a significant reduction in the risk of falling (risk ratio, 0.88) in the assessment and intervention groups compared to “usual care” or control groups. Monthly rate of falling was also significantly lower (incidence rate ratio, 0.80). Multifactorial assessment and management programs were the most effective component in reducing fall risk (incidence rate ratio, 0.82; number needed to treat, 11).

Hill, 2002. The effectiveness of strategies for preventing future falls was examined in this meta-analysis which pooled data from 12 studies of fall prevention. Eight of the studies included exercise (three offered exercise only) and three included comprehensive risk assessment and targeted interventions. The analysis found a 4% decrease in the rate of falls for individuals in treatment groups receiving various fall prevention interventions.  Exercise alone was not significantly effective. Exercise combined with other risk factor modifications was more effective in community-based programs compared to programs in residential institutions. Fall prevention programs showed greater effects when outcomes were measured for 12 months or longer.

Weatherall, 2004. This report estimated the effectiveness of fall prevention programs from the randomized controlled trials cited in the 2001 guideline and in another published guideline from 2000 (Feder et al., 2000).  The authors evaluated 17 studies available up to August, 2002. The analysis demonstrated that: a) exercise as a sole intervention may have a beneficial effect but the results are not conclusive; b) multiple intervention programs are more effective than exercise alone (number needed to treat: multiple interventions, 9.8 versus exercise alone, 19.5); c) a “visit and advice” intervention may be effective but has the largest number needed to treat estimate. The authors concluded that multiple intervention strategies were particularly effective for fall prevention. 

Targeted randomized controlled trials

Tinetti et al., 1994, utilized a multiple risk factor intervention strategy. Subjects had at least one risk factor for falling. After assessment, subjects in the intervention group received targeted interventions in the following areas: medication adjustments; home hazard review and adjustment; behavioral recommendations (such as advice regarding postural hypotension); and a home exercise program (balance and strength training). Control subjects received "usual care" plus social visits. During one year of follow-up, there was a significant reduction in time to first fall and proportion of fallers in the intervention group compared to the control, with 35% falling in the intervention group compared to 47% falling in the control group (P=0.04). Adjusted incidence-rate ratio for falling in the intervention group compared to controls was 0.69. The percentage of subjects with particular fall risk factors also declined significantly from baseline.

Close et al., 1999, focused on older people presenting to an emergency department after a fall. The study utilized a core assessment by medical and occupational therapy staff, with subsequent referral to other specialist services if required. After one year, the number of falls in the intervention group (183) was significantly lower than that in the control group (510; P=0.0002). Risk of falling was also significantly reduced with intervention (odds ratio, 0.39) as were the risk of recurrent falls (odds ratio, 0.33) and likelihood of admission to hospital (odds ratio, 0.61).

Clemson, 2004 evaluated a multicomponent community-based program called “Stepping On.” This program, which employs a small-group learning environment, is effective in reducing falls in at-risk people living at home. Key aspects of the program are based on evidence that falls can be prevented by a) improving lower limb strength and balance; b) optimizing environmental and behavioral home safety; c) conducting regular medication reviews; and d) undergoing regular vision screening. Interventions included cognitive behavioral learning strategies for self-efficacy and decision-making; education about risk management; a lower limb strength and balance exercise program; medication management; and home and community safety.  The “Stepping On” program was associated with a 31% reduction in falls compared to usual care.

Day et al., 2002.  Three interventions (group exercise, home hazard management, and vision improvement) were offered to older individuals living in the community. A significant fall prevention effect was demonstrated for group-based exercise, the most potent single intervention (rate ratio, 0.82). A significant effect was also found for combinations of interventions that involved exercise. Balance measures improved in association with exercise. Neither home hazard management nor treatment of poor vision were effective alone, but the strongest effect occurred with all three interventions combined (rate ratio, 0.67). With all three interventions, the annual fall rate decreased by 14% (number needed to treat, 7).   

Lightbody et al., 2002. This study evaluated a nurse assessment and management plan and care pathway development for older people discharged home from emergency departments after a fall. The intervention consisted of a fall risk assessment in the home that addressed modifiable risk factors (medication, electrocardiogram, blood pressure, cognition, visual acuity, hearing, vestibular dysfunction, balance, mobility, feet and footwear). The intervention was carried out by a trained nurse 2-4 weeks after the index fall. Identified risk factors were addressed using referral to existing services. Advice and education about safety in the home were also provided.  At the six-month follow-up, a non-significant trend towards lower falls was found in the intervention group compared to usual care. Fewer fall-related admissions (8 versus 10) and bed days (69 versus 233) were reported. 

Davison, 2005, conducted a randomized clinical trial in which conventional care was compared to a multifactorial assessment and intervention (medical, physiotherapy and occupational therapy) in individuals presenting with a fall or fall-related injury and at least one additional fall in the preceding year. Significantly fewer falls (36% reduction) occurred in the intervention group (relative risk, 0.64) although the proportion of subjects continuing to fall and the number of fall-related presentations and hospital admissions did not differ between groups. Duration of hospital admission was reduced and falls efficacy was better in the intervention group.

Untargeted randomized controlled trials

Steinberg et al., 2000, evaluated a multicomponent intervention aimed at major fall risk factors in reducing slips, trips and falls.  Volunteers were randomized to receive one of four interventions: a) an education program (oral presentation with pamphlet); b) the education program plus an exercise class once a month; c) education, exercise, plus a home safety assessment with financial and practical support for home modification; and d) education, exercise, home modification plus clinical assessment and advice on medical fall risk factors. At one year follow-up, a statistically significant reduction in the risk of slips and trips and a trend towards a reduction in the risk of falling was found in all intervention groups relative to the control group.

Whitehead et al., 2003. Patients presenting to the emergency department after a fall were randomized to usual care or to an intervention consisting of a falls risk assessment and an evidence-based prescription faxed to their physician. Fall rates as well as compliance with advice were monitored for six months. Patients in the intervention group were more likely to comply with preventative advice (odds ratio, 12.3) but there was no significant reduction in falls in the intervention group (odds ratio, 1.7).

Evidence Table

	Evidence	Source	LE	QE	SR
1	For community-living older adults at risk of falling, a multifactorial assessment followed by linked intervention is effective. Evidence suggests that the health professional or team conducting the fall risk assessment should assure that the interventions are carried out by health professionals to ensure effectiveness.  Multifactorial assessment without ensuring  intervention beyond advice and information provision is ineffective.	Campbell, 1999 Chang, 2004 Clemson, 2004 Close, 1999 Coleman, 1999 Davison, 2005 Day, 2002 Gallagher, 1996 Gillespie, 2003 Hogan, 2001 Hornbrook, 1994 Kingston, 2001 Lightbody, 2002 Newbury, 2001 Steinberg, 2000 Tinetti, 1994 VanHaastregt, 2000 Vetter, 1992 Wagner, 1994 Whitehead, 2003	I	Good	A
2	General medical or geriatric assessments and interventions that are not targeted at known fall risk factors do not appear to reduce fall rates or occurrence.	Coleman, 1999	I	Fair	B
3	The effective  multifactorial/multicomponent interventions included the following components: environmental adaptation and/or modification (9 studies out of 11); balance, strength, and gait training (7 out of 11); assistive devices; reducing psychoactive medications; reviewing and reducing other medications; managing vision problems;  managing orthostasis; and addressing cardiovascular and other medical problems.	Campbell, 1999 Chang, 2004 Clemson, 2004 Close, 1999 Davison, 2005 Day, 2002 Gillespie, 2003 Nikolaus, 2003 Steinberg, 2000 Tinetti, 1994 Wagner, 1994 Whitehead, 2003	I	Good	A
4	Risk factor assessment without direct intervention of the identified risk factors does not appear to be effective.	Gallagher, 1996 Hogan, 2001 Hornbrook, 1994 Kingston, 2001 Lightbody, 2002 Newbury, 2001 Rubenstein, 1990 VanHaastregt, 2000 Vetter, 1992	I	Good	E

LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.1. Minimize Medications 

Background

Medications have consistently been associated with increased risk of falls. Reasons for this include both direct effects (e.g., lowering of blood pressure, sedation) and side effects (e.g., fatigue, confusion, ataxia, dizziness). The strongest risk associations occur with psychotropic medications and polypharmacy (defined as more than an arbitrary number of different prescription medications, usually four or more). As a result, many multifactorial fall prevention programs have included medication reduction and simplification. All studied programs that have included such strategies have shown significant efficacy in fall prevention.

Recommendations

18. Psychoactive medications (including sedative hypnotics, anxiolytics, antidepressants) and antipsychotics (including new antidepressants or antipsychotics) should be minimized or withdrawn, with appropriate tapering if indicated. [B]
19. A reduction in the total number of medications or dose of individual medications should be pursued.  All medications should be reviewed, and minimized or withdrawn. [B]

Rationale

There is one published randomized controlled trial of medication manipulation as a separate intervention (as part of a 2 x 2 factorial design) to reduce falls. In addition, reduction of medications has been a prominent component of fall-reducing interventions in a large number of effective community-based and long-term care multifactorial/multicomponent studies.

Most of the trials of multifactorial interventions do not provide sufficiently detailed information to allow estimation of benefit attributable to medication minimization. Seven studies of multifactorial interventions have included medication modification. Of these, three focused on psychoactive agents and four on other types of medications. All demonstrated benefit overall.

The strongest evidence supports withdrawal of psychotropic medication, both as a single intervention and as a component of multifactorial/multicomponent intervention. If discontinuation of a particular high-risk medication is not possible due to medical conditions, dose reduction should be considered.

Evidence Statements

For all settings, including the community, long-term care, rehabilitation facilities, or hospital, there is a consistent association between psychotropic medication use (neuroleptics, sedative-hypnotics, anxiolytics, and antidepressants) and falls.

Observational studies: medication as a risk factor 

While some clinicians believe that selective serotonin reuptake inhibitors (SSRIs) are genrally safer to use in older adults than tricyclic antidepressants in terms of fall prevention, the data have not supported this. In fact, evidence is building that SSRIs increase fall risk as much as the older tricyclic antidepressants. (Leipzig, 1999; Arfken, 2001, Ensrud, 2002)

Leipzig et al., 1999, carried out systematic reviews of the effect of medications on falls, and identified a significant association between certain types of psychotropic, cardiovascular and analgesic medications and an increased risk of one or more falls in older adults. No randomized controlled trials were identified in this systematic review. Results were based on the pooling of data from cohort, case-control and cross-sectional studies. 

Ensrud et al., 2003, confirmed that the use of benzodiazepines, antidepressants, and anticonvulsant medication was associated with an increased risk of frequent falls in a large sample (N=8127) of community-dwelling older women. During this three-year study, most participants visited the clinic at least four times and the other participants were followed by home visits or questionnaires to determine whether current use of central nervous system-active medications increases the risk for subsequent falls. Those taking medications (8% benzodiazepines, 6% antidepressants, 6% anticonvulsants, 5% narcotics) were at increased risk of frequent falls. Benzodiazepine use (long-acting only) was associated with 34% greater likelihood for falls and antidepressant use was associated with 54% increased risk of frequent falls (marginal significance for one fall). Subjects taking anticonvulsants had 75% increased risk for one or more falls and were twice as likely to have frequent falls. There was no difference in risk of falls between narcotic and non-narcotic users.

Medication minimization alone and as a component of multifactorial intervention

Campbell et al., 1999. This randomized controlled trial investigated two interventions, withdrawal of psychotropic medication and a home-based exercise program, in older people taking psychotropic medication. Patients were randomly assigned to one of four groups in a 2 x 2 design: a) gradual withdrawal of medication over 14 weeks versus continuing to take medication; and b) a home-based exercise program versus no exercise. After 44 weeks, the withdrawal of psychotropic medication significantly reduced the risk of falling by 66%. No interaction effect was found between the two interventions. However, 47% of participants who ceased psychotropic use during the study had resumed taking their medication one month after completion of the study. The authors emphasized that support services, including counseling, relaxation therapy and sleep promotion resources, need to be considered for patients attempting to terminate psychotropic medications.

Although there have been no randomized controlled trials of medication manipulation as the sole intervention (other than in the Campbell 1999 study described above), reduction of medications has been a prominent component of effective multifactorial/multicomponent fall-reducing interventions in community-based and long-term care studies (Close, 1999; Tinetti, 1994; Wagner, 1994; Ray, 1997; Clemson 2004; Healey 2004, Jensen 2002).  While it is not possible to assess the relative value of medication reduction alone in the effectiveness of these controlled interventions, virtually all of the multifactorial fall-reduction programs that included medication minimization were significantly effective in reducing falls. Furthermore, several multifactorial/multicomponent studies that did not include medication reduction were not effective.

Tinetti et al., 1994, included medication reduction and non-pharmacological strategies in a multifactorial randomized controlled trial. They found that the number of subjects taking four or more medications declined by 23% relative to the control group after one year. The multiple intervention program also resulted in a significant reduction in time to first fall and in the proportion of subjects who fell during the study period compared to the control group.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Consistent association has been found between psychotropic medication use (i.e., neuroleptics, sedative-hypnotics, anxiolytics, and antidepressants) and falls.	Arfken, 2001 Ensrud,  2002 Leipzig, 1999	III	-	C
2	Reduction of psychotrophic medication as  a single intervention reduces rate of falls.	Campbell, 1999	I	Fair	I
3	Assessment, adjustment, and discontinuation  of medication regimens as part of multifactorial intervention reduces falls in old persons living in the community.	Campbell, 1999 Clemson, 2004 Close, 1999 Davison, 2005 Healey, 2004 Tinetti, 1994 Wagner, 1994	I	Fair	B

LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.2. Initiate an Individually-Tailored Exercise Program

Background

Exercise programs are a commonly used fall prevention strategy. There are a number of models of exercise delivery, such as group exercise and individualized home exercise programs. A range of exercise types have been investigated that can be used in isolation or in combination within a specific exercise program, including balance exercises, strength training, flexibility (muscle and joint stretching techniques), Tai Chi, and cardiovascular, endurance, and fitness training.

Numerous research studies have evaluated the types and quantity of exercise that help to reduce falls among older adults.  Having certain physical attributes such as weak legs, poor muscle strength, poor balance and stability, and limited mobility have been found to negatively impact gait and increase the risk of falling.  Since strength, muscle mass, gait, balance and stability are all closely interlinked, many of the exercise intervention programs have included strengthening exercise as well as balance and stability training.  Even among frail older adults who are relatively weak, strength training programs appear to increase muscle strength, core balance and gait.

Recommendations

20. Exercise should be included as a component of multifactorial interventions for fall prevention in community-residing older persons. [A]
21. An exercise program that targets strength, gait and balance, such as Tai Chi or physical therapy, is recommended as an effective intervention to reduce falls [A]
22. Exercise may be performed in groups or as individual (home) exercises, as both are effective in preventing falls. [B]
23. Exercise programs should take into account the physical capabilities and health profile of the older person, (i.e., be tailored) and be prescribed by qualified health professionals or fitness instructors. [I]
24. The exercise program should include regular review, progression and adjustment of the exercise prescription as appropriate.  [I]

Rationale and Evidence Statements

 A large body of evidence supports the recommendation that exercise, in the form of resistance (strength) training, and balance, gait and co-ordination training, is effective in reducing falls. The reduction in fall rate resulting from exercise is modest (approximately 16%). The best estimate of number needed to treat to prevent one fall is 16 people. (Chang, 2004).

Twenty-four studies have been conducted in community-dwelling populations evaluating exercise as a single intervention. Thirteen studies found that the exercise program was effective in reducing falls. In most of these positive trials, the duration of the exercise program was longer than 12 weeks with variable intensity ranging from once a week to 90 minutes three times per week. 

Exercise may be more effective when applied alongside other interventions. Exercise programs were associated with a reduction in falls in both multifactorial and multicomponent studies.   (Campbell, 1999; Steinberg, 2000; Tinetti, 1994; Clemson, 2004; Day, 2002). The Hogan et al. (2001) multifactorial  intervention included exercise and demonstrated that compared to the control group, the intervention group had significantly more time between falls.  In two other studies that combined exercise with other interventions, fall risk factors were reduced but the intervention did not prevent falls. (Whitehead, 2003; Lord, 2005) 

Exercise may be considered as a single intervention to reduce falls in selected groups (Gardner, 2000).  Initiating exercise programs should be done with caution as some studies have shown that exercise may increase the rate of falls in persons with limited mobility who are not used to exercising. Exercise is an important component of multifactorial fall prevention programs, and future research should address the possibility that, in some populations, exercise may be as effective as multifactorial fall prevention programs.

Recommendations limited to specific types of exercise cannot be made with complete confidence, but general principles may be distilled from the literature, despite the fact that many reports fail to provide adequate details of their interventions. Further research is needed to identify the most effective components of interventions. (Gardner, 2000)

Endurance (aerobic) training has not been widely tested as a falls prevention strategy. There is insufficient evidence to support the inclusion of endurance exercise in fall prevention exercise programs, although the broader health benefits of endurance training should be considered.

Some trials included a specific Tai Chi program (Gardner, 2000; Li, 2005; Liu-Ambrose, 2004; Lord, 2003; Suzuki, 2004; Wolf, 1996) as part of the intervention. Some of these have showed significant reduction in falls in addition to other benefits in gait, balance, and reduction in fear of falling.  Other forms of balance training have demonstrated similar results.

Recommendations concerning other settings and special populations, particularly long-term settings, are subject to considerable uncertainty.
To be of maximum benefit, future trials should standardize the populations from which they recruit, describe the interventions tested, and ensure adequate power and adherence. Studies to compare variations in exercise type and prescription will necessitate large sample sizes and should investigate whether alternative approaches are more effective, cost effective and/or acceptable than resistance, balance, gait and coordination training.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Exercise should be included as a component of multifactorial interventions designed to reduce falls.	Tinneti,  1994 Day, 2002 Whitehead, 2003	I	Good	A
2	A multimodal exercise program  should include a combination of strength, gait, and balance training.	Gillespie, 2003 Chang, 2004 Gardner,  2000	I	Good	A
3	Tai Chi may be effective in reducing falls.	Gardner, 2000 Li, 2005 Liu-Ambrose, 2004 Lord, 2003 Suzuki, 2004 Wolf, 1996	I	Fair	A
4	There is insufficient evidence to support the inclusion of endurance exercise in fall prevention exercise programs.				I
5	Exercise may be performed in groups or as individual (home) exercises.	(See table, Appendix C3)	I	Fair	B

LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.3. Treat Vision Impairment

Background

Aging is often associated with changes in visual acuity, development of cataracts, macular degeneration, glaucoma, and other conditions that would suggest an impact on risk of falling. Although correction of these conditions should intuitively improve fall risk, there is not enough data to support this intervention alone.

Recommendations

25. In older women in whom cataract surgery is indicated, surgery should be expedited as it reduces the risk of falling.  [B]
26. There is insufficient evidence to recommend for or against the inclusion of vision interventions within multifactorial fall prevention interventions. [I]
27. There is insufficient evidence to recommend vision assessment and intervention as a single intervention for the purpose of reducing falls. [D]
28. An older person should be advised not to wear multifocal lenses while walking, particularly on stairs. [C]

Rationale

A systematic review (Gillespie, 2003) found no evidence that referral for correction of vision in community-dwelling older people was effective in reducing the number of people falling. However this conclusion was based on a single randomized controlled trial. (Day, 2002)

Two randomized controlled trials assessing the effect of a cataract operation and waiting list time for surgery showed reductions in rate of falling for immediate versus delayed cataract surgery.  However, these studies did not address the larger question of the benefits of screening for cataracts in a general population.

Three studies included vision correction as part of a multifactorial assessment and intervention. The results were mixed. Combined interventions, which included vision correction, reduced the rate of falls, but it is difficult to ascertain whether the reduction is attributable to the vision correction.

One randomized trial looking at a vision assessment and follow-up intervention alone indicated that vision assessment and intervention actually increased risk of falling. This may be related to the effects of adjusting to new glasses. (Cumming, 2007)

Evidence Statements

Vision intervention alone

Harwood et al., 2005, studied 306 frail community-residing women over the age of 70 years with cataract. The intervention group was referred for cataract surgery at one month versus a referral at 12 months for the control group. Over the 12 months monitoring period, there was a 40% decrease in recurrent falls risk and a 34% reduction in fall rate with intervention (P=0.03). In terms of general health status, first cataract surgery improved activity levels, anxiety, depression, confidence, visual disability, and handicap compared with controls. Four participants in the operated group sustained fractures (3%) compared with 12 (8%) in the control group (p = 0.04). 

Foss et al., 2006, randomized 239 older (>70 years) community-domiciled women to a second cataract surgery or to a waiting list. The second eye cataract surgery reduced the rate of falling and improved visual function (especially stereopsis). The rate of falling was reduced by 32% in the operated group compared with the waiting list group, but the difference was not statistically significant.

Cumming et al, 2007, randomized 616 community-living older people to receive either a comprehensive vision examination followed by eyeglass provision and other indicated eye care or usual care.  Surprisingly, after the 12-month follow-up period, the intervention group had significantly more falls than the controls. The authors speculated that this unexpected result may have arisen from problems adjusting to new eyeglasses, the most common intervention. The intervention group was also slightly less frail, and therefore may have been more active.

Vision as a component of multifactorial interventions

Clemson et al., 2004, studied a multifactorial community program that included components designed to encourage regular visual screening and to help older persons adapt to low vision. Other intervention components included lower limb exercises, medication management, and education to improve safety in the home and community. Among the persons (N=310) who suffered a fall within the previous 12 months, or who had a fear of falling, interventions were associated with a significant 31% decrease in falls compared to the controls. The influence of vision training was not specified.

Day et al., 2002, assessed the effectiveness of vision testing and eye care education in healthy community-residing older people (age >70). The intervention was evaluated alone (N=139), in combination with home hazard assessment (N=137), or with all three combined (N=137).  Vision intervention alone did not have an impact on fall reduction.  Visual acuity remained unchanged in the intervention groups. When combined with both home assessment and exercise, the annual fall rate was reduced by 14%.

Dyer et al., 2004, demonstrated a modest but non-significant reduction in falls rates in the intervention group (N=102) receiving a comprehensive program to reduce falls. Optician assessment was included in the multifactorial program.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Fall reduction after cataract surgery	Harwood, 2005 Foss, 2006	I	Moderate Poor	B
2	Vision testing and intervention	Day, 2002 Cumming, 2007	I	Zero benefit Harmful	D
3	It is unclear whether vision is an essential component of  multifactorial  intervention. Only 4 out of 11 effective multifactorial studies provided details of vision interventions; the remaining 7 referred for vision assessment.	Clemson, 2004 Day, 2002 Davison, 2005 Wagner, 1994	I	Poor	I

      LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.4. Manage Postural Hypotension

Background

Postural hypotension is associated with an increased risk of falls. It results in loss of balance due to low blood pressure and consequent cerebral hypoperfusion. Postural hypotension most commonly occurs as a result of dehydration, concomitant medications and autonomic neuropathy.  Many multifactorial fall prevention programs have included medication reduction and simplification to modify postural blood pressure. Some have also included specific strategies for management of postural hypotension such as hydration, elastic stockings, abdominal binders and medications (i.e., fludrocortisone and midodrine).

Recommendation

29. Assessment and treatment of postural hypotension should be included as components of multifactorial interventions to prevent falls in older persons. [B]

Rationale

Multifactorial studies which incorporated assessment and management of postural hypotension, including modification and simplification of medications, have shown benefit for fall prevention. 

Evidence Statements 

Three randomized controlled trials have demonstrated a benefit associated with treatment of postural hypotension in addition to interventions for other traditional risk factors such as medication reduction, optimization of fluids, and behavioral intervention (Tinetti, 1994; Close, 1999; Davison, 2005). No adequate prospective study has been published that permits selection of a specific treatment for postural hypotension.  Also, no randomized controlled trials have been carried out that examines the benefit of treatment of postural hypotension as a single intervention for fall prevention.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Multifactorial intervention strategies which included management of orthostatic hypotension reduced falls in community-dwelling older persons.	Tinetti, 1994 Close, 1999 Davison, 2005	I	Fair	B

LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.5. Manage Heart Rate and Rhythm Abnormalities

Background

Cardiovascular factors are frequently cited as risk factors for falls. The common cardiovascular disorders associated with falls include carotid sinus hypersensitivity, vasovagal syndrome, bradyarrhythmias (such as sick sinus syndrome and atrioventricular block) and tachyarrhythmias. Two mechanisms have been proposed. The first is transient loss of consciousness with amnesia in which the patient has no recollection of short episodes of syncope. Such cases have been associated with postural hypotension and carotid sinus hypersensitivity (Parry, 2005). Given that up to 70% of falls in older persons are not witnessed, these patients may present with a report of a fall rather than syncope.  A second proposed mechanism is that of transient hypotensive episodes, either due to primary hypotension or hypotension secondary to arrhythmias, which cause a person with comorbid gait and balance instability to lose balance and fall without frank syncope.

For the subset of older adults who meet the necessary diagnostic criteria, dual chamber cardiac pacing for bradyarrhythmias (including carotid sinus hypersensitivity and conduction disorders) and treatment of tachyarrhythmia are components of a multifactorial intervention designed to reduce the risk for falls.

Recommendations

30. Dual chamber cardiac pacing should be considered for older persons with cardioinhibitory carotid sinus hypersensitivity who experience unexplained recurrent falls. [B]

Rationale

Symptomatic cardioinhibitory carotid sinus hypersensitivity is characterized by significant heart rate slowing. Cardiac pacing treats bradycardia. One randomized controlled trial of cardiac pacing in community-dwelling older people who had recurrent unexplained falls reported a significant reduction in fall rates at 12-month follow-up. (Kenny 2001)

Evidence Statements

Kenny, 2001, One study demonstrated a benefit for cardiac pacing as the sole intervention for falls in patients with cardioinhibitory carotid sinus hypersensitivity (N=175). In patients who were cognitively normal with recurrent non-accidental falls (>2 in the previous year), cardiac pacing reduced subsequent falls by 68% and injurious falls by 70% at 12-month follow-up. The control group reported 669 falls while paced patients had 216 falls (odds ratio, 0.42). These findings have not yet been confirmed in other trials.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Cardiac pacing reduced falls in community dwelling older people with carotid sinus hypersensitivity.	Kenny, 2001	I	Fair	B

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.6. Supplement Vitamin D

Background

Vitamin D deficiency is common among older people and when present impairs muscle strength and possibly neuromuscular function via pathways in the central nervous system. Several recent randomized controlled trials have shown a beneficial effect from vitamin D supplementation in the prevention of falls, perhaps from a direct effect on neuromuscular function and distinct from its beneficial effect on bone health. Moreover, some of these trials have shown a positive effect from vitamin D supplementation even among older persons with normal serum vitamin D levels at baseline.

Vitamin D deficiency should be treated if identified in order to reduce falls among older persons. Vitamin D supplementation at appropriate levels  should be considered for all older adults.

Recommendations

31. Vitamin D supplements of at least 800 IU per day should be provided to older persons with proven vitamin D deficiency. [A] 
32. Vitamin D supplements of at least 800 IU per day should be considered for people with suspected vitamin D deficiency or who are otherwise at increased risk for falls. [B]

Rationale

Given the low number needed to treat of 15 and the evidence of significant fall risk reduction (Bischoff-Ferrari, 2004) as well as the fact that vitamin D is safe and inexpensive, older persons with vitamin D deficiency should be routinely offered supplementation in order to reduce fall risk. Four randomized controlled trials provide evidence that vitamin D supplementation may reduce the risk of falls among older adults by more than 20%. Moreover, two studies from a meta-analysis and another recent randomized trial in residential care (Flicker et al., 2005) showed significant reduction in fall rates with vitamin D supplementation in populations with baseline deficiency. This strengthens the argument for supplementation in all older adults, especially those at increased risk of falls.

There is evidence that vitamin D deficiency is common among older people and that when present impairs muscle strength and possibly neuromuscular function via central-nervous system-mediated pathways.

Evidence Statements

Bischoff-Ferrari, 2004. A meta-analysis based on a systematic review of the literature from 1960 to 2004 yielded five high quality double-blinded randomized trials assessing the effect of supplemental vitamin D on fall prevention in older persons (age >60). Randomized controlled trials (Graafmans, 1996; Pfeifer, 2000; Gallagher, 2001; Bischoff, 2003; Dukas, 2004) form the basis of this meta-analysis. The pooled study data (N=1237) indicated that vitamin D reduced overall falls by 22% (corrected odds ratio, 0.78) compared with patients receiving calcium or placebo (number needed to treat, 15). Moreover, when five additional studies with less rigorous methodology were added to the analysis, the effect was smaller but remained significant.

The analysis did not distinguish whether the effect was greater in persons with vitamin D deficiency at baseline. However, because such a high prevalence of suboptimal vitamin D levels occurred in the populations of older adults studied, and because vitamin D is inexpensive and generally benign, the authors recommended that supplementation be strongly considered for all older adults. Moreover, two of the studies in the meta-analysis showed a significant reduction in rate of falling associated with vitamin D supplementation in populations without baseline deficiency. This strengthens the argument for supplementation in all older adults, especially for those at increased risk for falls.

Latham, 2003, evaluated 13 trials in a systematic review of supplemental vitamin D for ambulatory patients. Most of the trials were small and had methodological problems. In 10 trials, there was no evidence that vitamin D or vitamin D metabolites had any effect on falls or physical function. However, three trials showed a positive effect of vitamin D in combination with calcium. When available data from the four highest quality trials were pooled, there continued to be no evidence that vitamin D reduced the risk of falling, although a single trial of vitamin D plus calcium showed a positive effect.

Pfeifer, 2000. Older persons with vitamin D insufficiency (serum levels <50 IU) were randomized to calcium monotherapy or calcium plus vitamin D. Combination therapy was found to reduce sway, falls, and fallers by 43% after one year. 

Evidence Table

	Evidence	Source	LE	QE	SR
1	Supplemental vitamin D in patients  with deficiency reduces falls	Bischoff-Ferrari , 2004 Pfeifer, 2000	I	good	A
2	Supplemental vitamin D to all older persons	Bischoff-Ferrari, 2004 Dukas, 2004 Flicker, 2005 Gallegher, 2001 Porthouse, 2005	I	Fair	B

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

 

2.7. Manage Foot and Footwear Problems

Background

Foot problems are common in older people and are associated with impaired balance and performance in tests of function.  Serious foot problems (i.e., moderate or severe bunions, toe deformities, ulcers or deformed nails) have been shown to predispose older adults to falls. (Tinetti et al., 1988) Also, foot position awareness is significantly poorer in older persons.

The type and condition of footwear may also contribute to the risk of falling. Inappropriate footwear has been cited as a contributory factor in falls occurring at home. Footwear that fits poorly, has worn soles, has high heels, or is not laced or buckled when worn has been associated with a higher risk of falling. (Rubenstein 1988)

Recommendations

33. Identification of foot problems and appropriate treatment should be included in multifactorial fall risk assessments and interventions for older persons living in the community. [C]
34. Older people should be advised that walking with shoes of low heel height and high surface contact area may reduce the risk of falls. [C]

Rationale

Several authors have reported that suitable footwear is important in reducing the risk for falls in older people. Although most of the earlier information was anecdotal, a number of recent studies provide scientific data regarding the effects of footwear on balance in older adults. The results of these studies suggest that readily identifiable characteristics of footwear may provide a useful way to characterize safe shoes for older adults.

Some of the effective multifactorial studies have included attention to footwear, but no randomized controlled trials address this issue in isolation. One small study has suggested that anti-slip devices may be useful for reducing the risk of outdoor falls.

Most of the studies that implemented a multifactorial assessment for reducing the risk of falling included a foot assessment coupled with advice or referral for appropriate treatment if any foot problems were identified in the assessment.  (Lightbody, 2002; Lord, 2005; Rucker, 2006; Davison, 2005)  (See Multifactorial Intervention)  Assessment of footwear and recommendations for use of appropriate shoes were also included in studies implementing assessment and mitigation of home hazards. (Dyer, 2004; Cumming, 1999)

Evidence Statements

Robbins et al., 1995 found that wearing shoes resulted in impaired foot position sense in both younger and older adults.

Robbins et al., 1998, reported results from a randomized trial in which younger and older men had increased stability when wearing shoes with low resiliency soles and reduced stability with shoes containing soft, high resiliency soles.

Arnadottir et al., 2000; Lord, Bashford, 1996. These reports, the first from an observational study and the second a randomized controlled trial, showed that bare feet and walking shoes maximize balance in older women, whereas high-heeled shoes present an unnecessary hazard.

Lord et al., 1999, investigated the effects of shoe collar height and sole hardness on balance in a group of older women. The findings indicated that subjects were more balanced when wearing shoes with a high rather than a low collar, or when barefoot. No association was found between sole hardness and balance.

Tencer et al., 2004, investigated the biomechanical properties of shoes in a group of community-dwelling older adults who reported a fall during a surveillance period, and in controls from the same surveillance cohort.  Fall risk was nearly double for individuals wearing shoes with heel height >2.5 cm.  Heel collar height, forefoot stiffness/flexibility and foresole thickness had little association with fall risk.

Menz et al., 2001. This cross-sectional, retrospective study investigated the relationship between foot problems, balance, and functional ability in community-dwelling older people and determined that older people with a history of multiple falls exhibit greater foot impairment than those who have not fallen or who have fallen only once.

McKiernan et al., 2005, demonstrated, in a small randomized controlled trial, that anti-slip shoe devices are effective in reducing outdoor falls in slippery conditions.

Evidence Table

	Evidence	Source	LE	QE	SR
1	Shoes with low heel height and high surface contact area reduce risk for falling.	Robbins, 1998 Lord and Bashford, 1996 Arnadottir, 2000 Tencer, 2004	II	Moderate	B

            LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.8. Modify the Home Environment

Background

Falls within and around the home are common and often result from a complex interaction between environmental hazards and physical abilities or risk taking.  Environmental hazards are any objects or circumstances in the environment that increase an individual's risk of falling.  Examples include inadequate lighting, absence of handrails on stairs or grab bars in bathtubs, and clutter. Environmental hazards may be within the home and grounds (commonly termed “home falls hazards”), or away from the home (commonly termed “public falls hazards”). In addition to internal risk factors that may be a cause for falling, external factors in the environment often increase fall risk.

Identification and mitigation of environmental hazards has been a recommended feature of many successful fall prevention programs. The environmental hazards can be modified or eliminated if identified by an appropriate assessment.

Several effective multifactorial and multicomponent interventions include the domain of home safety as one of many (equally valued) program components. Targeted interventions featuring home safety programs have also demonstrated their effectiveness through randomized trials. Such interventions provide education to support an individual’s safe functioning in the home as well as prevention programs that include home hazard assessments by trained individuals, removal or modification of identified hazards, installation of safety devices such as handrails on stairs and grab bars on bathrooms, and improvements in lighting.

Screening of the home environment with follow-up for any needed modifications by a health care professional is effective targeted intervention for people with a previous fall history or other fall risk factors.

Recommendations

35. Home environment assessment and intervention carried out by a health care professional should be included in a multifactorial assessment and intervention for older persons who have fallen or who have risk factors for falling. [A]
36. The intervention should include mitigation of identified hazards in the home, and evaluation and interventions to promote the safe performance of daily activities. [A]

Rationale

The evidence supporting home environmental assessment and intervention has progressed since the 2001 Guidelines were published. Randomized controlled trials support the use of environmental interventions that address both hazard reduction and behavioral strategies for improving safety in the home. These have been shown to be effective when used as part of a multifactorial approach. Randomized trials suggest that environmental interventions are not appropriate for individuals with lower risk for falls. The evidence points to the importance of utilizing trained health professionals to ensure that high quality interventions are provided.

While evidence supporting the use of home environment assessment and intervention alone as a strategy to reduce falls among community-dwelling older adults is mixed (Stevens et al., 2001a and Day et al., 2002 are not supportive; Campbell et al., 2005, is supportive), evidence for home environment assessment and intervention as part of a multifactorial fall prevention program is strong. The Day et al., (2002) and Campbell et al., (2005) studies were unique in that they tested home safety interventions alone and in combination with other interventions. The interventions should be considered “targeted” since in each case, participant inclusion criteria included either a history of a fall or presence of one or more fall risk factors. In Day et al., (2002), professional home maintenance staff members were involved in the home hazard modifications, and in the other studies, a health care professional provided delivery of the intervention. Further insights regarding effectiveness of the interventions are gained through the meta-analysis conducted by Gillespie et al (2003).

Overall, the evidence supports the recommendation that community-dwelling older adults who have fallen or who have fall risk factors benefit from a multifactorial intervention that includes identification and mitigation of hazards in the home.

Evidence Statements

Home intervention alone or in combination with other behavioral strategies: randomized studies

Stevens et al., 2001a, used trained research nurses to assess and make recommendations for minimizing home fall hazards for individuals randomized to intervention. Persons in the control group received usual care.  Intervention included home hazard assessment, installation of free safety devices, and education about home hazard removal and modification.  At one year follow-up, 281 of the 421 falls in the home involved environmental hazards. No significant group differences were found in the rate of falls or falls injuries, the proportion of fallers or the frequency of falls involving an environmental hazard. In this report and a follow-up publication (Stevens et al., 2001b), the authors suggested several methodological issues that may have contributed to the non-significant results, including a) the intervention had a limited effect on the number of home falls hazards; b) subject attrition may have impacted on outcomes; c) low inter-rater reliability for many items on the home assessment form used; d) only a small sample of homes were formally re-assessed for modifications. Seventy-four percent of the control group reported taking more care to avoid falls and 16% reported taking action to reduce falls risk in their homes.

Day et al., 2002, reported that home hazard assessment and management did not result in a significant reduction in falls when conducted in isolation. In the intervention group, participants removed or modified hazards themselves or through a home maintenance program. Hazards were identified by the home maintenance staff who visited participant's homes, provided work estimates, free labor, and materials. There was a significant reduction in falls when the home hazard assessment and management program was combined with an exercise program, or with both exercise and vision interventions. However the effect appeared to be largely associated with the group exercise intervention.

Cumming et al., 1999, reported the results of a randomized controlled trial of home visits undertaken by an occupational therapist for assessment and modification of environmental hazards including recommendations for home modifications as well as considerations of general safety within the home and advice on strategies to minimize individual risk.  Cumming found that the intervention was not effective in reducing falls in the overall study population. A non-compliance rate of between 30-50% for most recommendations was documented at the 12-month follow-up. The intervention was effective in the subgroup who reported having had one or more falls during the year before recruitment into the study; the relative risk of at least one fall during follow-up was 0.64 in this subgroup. The effectiveness of the program in reducing risk of falls among these previous fallers was similar for falls within the home and away from the home.

Pardessus et al., 2002, investigated whether home visits by an occupational therapist reduce the risk of falling and improve the autonomy of older patients hospitalized for falling. Participants who were hospitalized for a fall and were returning home were randomized to usual care or to additional intervention consisting of a home visit by a physiatrist and an ergonomist-therapist; social worker support for problems; and information on home safety and social assistance. Homes were assessed for environmental hazards, and modifications were recommended. At 12-month follow-up, no statistical differences were evident between groups for rate of falls, hospitalization for falls, institutionalization, or death, although a significant loss of functional autonomy was reported for the control group only.

Nikolaus et al., 2003. This study evaluated the effect of a fall reduction intervention by a multidisciplinary team in frail older people discharged home from hospital after a fall. In this randomized controlled trial, the intervention group received a comprehensive geriatric assessment with a diagnostic home visit, assessment of environmental hazards, advice about changes, offer of facilities for necessary home modification, training in technical or mobility aids, and an additional home visit at three months. The control group received assessment with recommendations plus usual care. After one year, the intervention group had significantly fewer falls (31% reduction). Intervention group subjects with a falls history reduced their fall rate by 37%. The proportion of frequent fallers (>2 falls) did not differ significantly between groups. The authors recommended that intervention should target high-risk frail older subjects.

Modification of the home as component of a multifactorial intervention

Gillespie et al., 2003, conducted a meta-analysis in which data from 1176 subjects from five randomized studies (Close et al. 1999; Hogan et al. 2001; Kingston et al. 2001; Lightbody et al. 2002; and van Haastreght et al., 2000) were pooled. Their results demonstrated a significant reduction in the proportion of fallers in the intervention groups (pooled risk ratio, 0.86)

 
The cluster-randomized study conducted by Tinetti et al., (1994) (n=301) and the individually randomized study conducted by Gallagher (1996) (n=100) were not included in the Gillespie et al. (2003) analysis.  However it is noteworthy that Tinetti and colleagues reported that during the one-year follow up, 35% of the intervention group experienced falls, compared with 47% of the control group (P=0.004). The adjusted incidence-rate ratio for falling in the intervention group as compared to the control group was 0.69. Gallagher (1996) reported no statistically significant differences between intervention and control groups. The authors reported that the study was underpowered due to recruitment difficulties.  Additionally, only about half the intervention group subjects were compliant with recommendations.
 
Subsequent randomized controlled trials

Campbell et al., 2005, assessed the efficacy and cost-effectiveness of a home safety program and a home exercise program to reduce falls and injuries in older people with low vision.  The home safety program reduced falls and proved to be more cost-effective than the exercise program. The author suggested that careful targeting to specific population groups is important in preventing falls.

 

Evidence Table

	Evidence	Source	LE	QE	SR
1	Multifactorial assessment and intervention that include home environment assessment and modification reduce the risk of falls	Close, 1999 Clemson, 2004 Davison, 2005 Day,  2002 Gallagher and Brunt, 1996 Hogan, 2001 Hornbrook, 1994 Huang, 2004 Kingston, 2001 LaGrow, 2006 Lightbody, 2002 Steinberg, 2000 Tinetti,  1994 van Haastreght, 2000 Whitehead, 2003	I	Good	A
2	Home safety evaluation plus modification in combination with other behavioral strategies  in high risk populations  may be effective in reducing falls	Cumming, 1999  Campbell, 2005 Nikolaus, 2003 Pardessus, 2002	I	Fair	B
3	Home safety interventions, assessment and modification in persons with no history or unknown history of falls has  not been shown to reduce falls	Cummings, 1999 Day,  2002 Stevens, 2001a, b	I	Small- Zero benefit	D

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

2.9. Provide Education and Information

Background

All fall prevention programs include educational and health promotion components that are intended to raise the awareness of older people and/or health care workers about risk factors for falls and inform them about strategies to minimize risk. It is assumed that increased awareness will translate into behavioral changes. Education of both patient and caregiver can be considered both as a preventive measure to reduce the likelihood of first falls as well as an intervention for an older person who has fallen. In general, education is also important for the implementation and sustained use of falls prevention strategies and compliance with these strategies.  Many effective fall prevention programs include opportunities for older adults to access fall prevention resources (e.g., durable medical equipment, local exercise programs) and to take specific actions that maintain or improve health or build fall prevention skills (e.g., transferring safely into the bathtub, learning how to use mobility devices).

Recommendations

37. Education and information programs should be considered part of a multifactorial intervention for older persons living in the community. [C]
38. Education should not be provided as a single intervention to reduce falls in older persons living in the community. [D]

Rationale

Education is considered to be an important part of multicomponent intervention programs for fall prevention. Specific educational goals have been included in a number of studied programs as fundamental components of fall prevention interventions.  These goals include increasing older adults’ activity level, improving ability to identify and mitigate fall hazards in the home, and providing information to make good choices about footwear.  However, there is little evidence to permit us to determine either the incremental benefit of such educational input on fall rates in a multicomponent intervention or as a sole fall prevention intervention.

Evidence Statements

Education intervention alone

Rucker et al., 2006, investigated the effectiveness of an educational intervention, provided by an emergency department, in reducing fear of falling and preventing recurrent falls in community-dwelling patients who had sustained a fragility fracture.  The intervention group (N=47) received brief printed educational materials containing evidence-based recommendations concerning fall prevention. Almost half (48%) of all study patients reported an increased fear of falling at three months, but the difference between groups was not significant. There was a trend towards an increase in recurrent falls among intervention patients (17%) compared to control patients (5%).

Lord et al., 2005, conducted a multicomponent individualized fall prevention program comprised of exercise, visual intervention, and counseling. Individuals in the intervention group were compared to individuals in a control group who received no interventions and to those in a minimal intervention group who received only written advice on exercise, vision, and compensation for loss of peripheral sensation.  Fall rates in the intervention groups were not significantly different from the control group.  The intervention appeared to have a greater effect on those with a greater risk of falls.

Steinberg et al., 2000, included education and awareness presentations on fall risk factors in four intervention groups of a multifactorial randomized trial. Additional interventions (e.g., exercise, home safety, medical assessment and advice) were added to three of the groups. These were associated with significant reductions in slipping and tripping compared to the group that received the educational component only. Time to first fall (hazard ratio, 0.30) was not significant for any of the groups.

Brouwer et al., 2003, compared exercise versus educational interventions in older community residents (ages 67-87 years) who reported a fear of falling and activity restriction. The eight-week education program engaged participants in discussing their concerns about falling and focused on identifying and reducing risk factors for falls. Both the education and exercise groups demonstrated a significant improvement in balance confidence. The education group also showed improvement in perceived mental health, whereas the exercise group demonstrated improvement in perceived physical health.  The study was not designed to document incidence of falls.

Education as a component of multifactorial fall prevention programs

Kingston et al., 2001, reported that a health visitor intervention that provided information and education regarding risk factors, exercise, and home modifications did not have a significant impact on health status and was not effective in reducing falls. 

Lightbody et al., 2002. A similar study employed “fall nurses” for home visits to educate patients about home safety. No significant differences were found between intervention and usual care, although there were positive trends in the intervention group for fewer falls, fewer hospital attendances and less time spent in hospital.

Nikolaus et al., 2003, provided education about home hazards as a component of fall prevention interventions. The intervention included assessment of the home for environmental hazards, provision of information about possible changes, facilitation of any necessary modifications, and training in the use of technical and mobility aids. The intervention reduced the fall rate by 31% in a selected group of frail older subjects with a history of recurrent falling.

Pardessus et al., 2002, evaluated home visits and associated education from occupational therapists during hospitalization of older patients at risk for falling. The intervention helped to preserve the patient’s autonomy but had no significant effect on falls.

Stevens et al., 2001a, An educational strategy to empower subjects to remove or modify home hazards was one component of the interventions included in a randomized controlled trial that enrolled community-dwelling older people.  The intervention failed to achieve a reduction in the occurrence of falls. The study provides evidence that a one-time intervention program of education, hazard assessment, and home modification to reduce fall hazards in the homes of healthy older people is not an effective strategy for the prevention of falls in seniors.

Clemson et al., 2004. In this randomized controlled trial, “Stepping On”, a cognitive-behavioral falls prevention program designed for a small-group environment, was included as part of a multicomponent community-based program.  The educational interventions included seven weekly two-hour sessions that offered the participants self-efficacy strategies, a decision-making theory guide, and options for risk management. The educational interaction included a presentation by a content expert, sharing of accomplishments by the participants, and review of exercise.  Falls were reduced by 31% in the intervention group. 

Evidence Table

	Evidence	Source	LE	Benefit	SR
1	Educational intervention alone is no more effective than usual care in reducing fear of falling or falls in community-dwelling older patients.	Lord, 2005 Rucker, 2006 Steinberg, 2000	I	None	D

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

3. OLDER PERSONS IN LONG-TERM CARE FACILITIES

Falling is an even more frequent occurrence among ambulatory residents of long-term care facilities than among older persons residing in the community. About half of ambulatory long-term care residents experience at least one fall each year. The risk factors associated with falling among persons residing in this setting are similar to factors identified among community-living older adults and include impairments in strength, balance, gait, vision, and cognition; use of multiple medications, especially psychoactive medications; and environmental hazards.

Trials in long-term care facilities have addressed both single interventions administered alone as well multiple interventions administered together as described in the next section. (Oliver, 2007) Single interventions have included use of hip protectors, fall alarm devices, removal of physical restraints, medication review, and supplementation with calcium and vitamin D. Interventions that were studied only with observational methods or historical controls are excluded from this guideline. Only randomized controlled trials or cluster randomized trial results were used to develop the guideline.

Interpreting the evidence from randomized controlled trials in the long-term-care setting is complicated by several factors. First, long-term care facilities range from care homes in which residents are independent in many activities to skilled nursing facilities in which most residents are dependent in most of their activities of daily living. Second, the structure of care and terminology used to describe facilities varies among different countries. Third, many of these trials do not identify the cognitive or physical functioning level of participants. Finally, the content of the interventions are either not described in detail or vary from study to study.

---

 

Multicomponent Interventions

BACKGROUND

Multicomponent interventions, in which the same set of interventions addressing more than a single category is offered to everyone included in the trial, are the most commonly studied strategies in long-term care settings. “Targeted” or “tailored” multifactorial interventions have also been tested. Staff training and feedback, environmental adaptations, balance and gait training, strength training, training in the use of appropriate assistive devices, and decrease in psychotropic medications are interventions that have frequently been included in multicomponent intervention and multifactorial trials in the long-term care setting.

RECOMMENDATION

39. Multifactorial/multicomponent interventions should be considered in long-term care to reduce falls. [C]

RATIONALE

The approach to intervention to reduce falls in long-term care settings differs from the approach in the community, both in content and in implementation strategies.

Models of intervention in long-term care settings differ from those in the community, both in content and in implementation. The available literature varies in the quality of studies and methodological design. Many different types and numbers of interventions occur from study to study. Also, most studies employ complex interventions, making it more difficult to reach a clear conclusion regarding efficacy. While some studies are negative, several have shown efficacy, and there are implications that a well-designed intervention may be beneficial. Medication review has been studied as part of a multicomponent intervention to reduce falls in long-term care settings. The evidence is inconclusive as to whether assessment, adjustment, and discontinuation of medication regimens result in fewer falls in older persons living in such facilities. (Dyer, 2004; Jensen, 2002; Ray, 1997) There is no randomized controlled trial of medication review and minimization as a single intervention in this population, and most multiple intervention trials did not include enough detail to demonstrate benefit of medication adjustment or minimization in long-term care.

Six of the eight studies of multicomponent interventions in the long-term care setting included environmental components. Three studies were ineffective (Jensen, 2002; Kerse, 2004; Dyer, 2004) and three effective (Becker, 2003; Jensen, 2003; Ray 1997). At this time, we can only conclude that evidence for effectiveness in environmental interventions is uncertain in this population. Two randomized controlled trials incorporating multifactorial interventions and achieving significant reductions in falls both incorporated environmental assessment and modifications as one of the intervention components. Becker (2003) reported using a 76-item checklist, with the most common modifications including changes to lighting, chair and bed heights, reduced clutter in residents’ rooms, installation of extra rails in bathrooms, and maintenance of walking aids. Similarly, Jensen (2002) undertook modifications such as removal of loose carpets, bedding adjustments, provision of rails and improved lighting. Neither of these studies provided sub-analyses on the effectiveness of the environmental modifications alone.

The education of long-term care staff has resulted in mixed results, but probably contributes to reduction of falls in some large studies. There is some evidence to support the effectiveness of training the health care team in awareness of fall risk factors and prevention strategies, although several multifactorial studies failed to show significant reduction in falls.

EVIDENCE STATEMENTS

Additions to the evidence base since the last guideline make findings in relation to long-term care more uncertain, with some new studies demonstrating benefit, and others finding none.

SYSTEMATIC REVIEW

Oliver et al., 2007, evaluated the evidence for strategies to prevent falls or fractures in eight studies of multifaceted interventions among residents in long-term care homes. One of the studies reported results among participants with, and without, cognitive impairment. Components utilized in the multiple intervention strategies included various combinations of risk assessment, hip protectors, removal of restraints, exercise and/or physical therapy, nursing education and training, equipment and environmental modification, fall alarm devices, and medication review. The meta-analysis found no significant effect of intervention on falls (rate ratio, 0.80), fallers (relative risk, 0.92), or fractures (relative risk, 0.91), although some individual studies showed strongly positive results.

Jensen et al., 2002, carried out a cluster randomized, controlled, non-blinded trial enrolling 439 older persons (>65 years) living in nine residential care facilities in Sweden. The 11-week multifactorial intervention program incorporated general as well as resident-specific tailored strategies including staff education, environmental modification, exercise, provision of hip protectors and assistive devices, medication review, and post-fall problem-solving conferences. During the 34-week follow-up, the incidence of falls decreased 12% (from 56% to 44%; risk ratio, 0.78) associated with the interventions compared to controls. There was also a significant reduction in femoral fractures (risk ratio, 0.23).

Becker et al., 2003, evaluated the effectiveness of a multifaceted, non-pharmaceutical intervention on incidence of falls and fallers in a prospective, cluster randomized trial (N=981, age >60 years) comparing an intervention group from three long-term facilities to controls from three other facilities in Germany. The intervention included education for residents and staff on fall prevention, advice on environmental adaptations, written educational materials, progressive balance and resistance training, and hip protectors. The percentage of fallers in the intervention group (36.9%) was lower than that in the control group (52.3%; relative risk, 0.75) and the incidence density rate of frequent fallers also declined over two years (relative risk, 0.56).

Ray et al., 1997, evaluated an intervention program in high-risk nursing home residents in seven pairs of nursing homes. One facility in each pair was randomly assigned to intervention (N=482) and the other facility served as the control (N=261). Review of psychotropic drug use was included in the comprehensive structured assessment along with specific safety recommendations targeting environmental and personal safety, wheelchair use, and transferring and ambulation. In the year following the intervention, the facilities that carried out the intervention had a 19.1% reduction in the mean proportion of recurrent fallers compared to control facilities. There was no significant difference in injurious falls.

Shaw et al., 2003, is the only multifactorial trial that specifically enrolled participants with cognitive impairment and dementia. The participants (N=274; age>65) in this randomized controlled trial, 80% of whom resided in a long-term care setting, were randomized to assessment and intervention or to conventional care after presenting to the emergency department after a fall. Intervention was not effective in reducing falls in the year following the intervention (relative risk, 0.92). (See Appendix B.) Dyer et al., 2004, carried out a cluster randomized controlled trial involving 196 residents (age >60) of 20 residential care homes to determine the effect of risk factor modification and balance exercise on fall rates. The multifactorial intervention program included three months of gait and balance training, medication review, podiatry and optometry The intervention group demonstrated significantly reduced medication use after 3 months. Although this group had a mean of 2.2 falls per resident per year compared to 4.0 falls in the control group, this difference did not reach statistical significance (P=0.2)

Kerse et al., 2004, conducted a cluster randomized controlled trial in 14 facilities (N=628). Low-intensity interventions applied in the intervention group included staff and resident education, environmental hazard assessment in rooms and public areas and continued follow-up for 6 months. Significantly more fallers and multiple fallers were reported in the intervention group. There was no significant difference in injurious fall incidence between patients in intervention groups versus control groups. Fall prevention intervention did not reduce falls or injury from falls, and may be worse than usual care in persons who are independent.

	Evidence	Source	LE	QE	SR
1	Effectiveness of multicomponent studies in long-term care is uncertain. Three of the eight trials of multiple component interventions were effective in long-term care institutions.	Becker, 2003 Dyer, 2004 Jensen, 2002 Kerse, 2004 McMurdo, 2000 Ray, 1997 Rubenstein, 1990 Shaw, 2003	I	Fair	C

LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

 

3.1 Exercise

BACKGROUND

While exercise may provide certain benefits for long-term care patients, particularly in terms of quality-of-life parameters such as depression, mobility, appetite, behaviors, and sleep, there are currently no randomized clinical trials to recommend for or against the use of individually-tailored exercise programs to prevent falls in long-term care settings. Confounding variables, i.e., differences in frailty levels, cognitive function, prior falls history, and the small size of many studies mitigate against clearly defined conclusions.

RECOMMENDATION

40. Exercise programs should be considered for a variety of benefits to reduce falls in older persons living in long-term care settings (with caution regarding risk of injury); however their effect on fall risk in these settings is yet unproven (C);

Evidence Statements

Rosendahl et al., 2008. In a multicenter Swedish RCT of long-term care (LTC), 158 men (27%) and women, ages ≥65, (mean age 84) participated. The subjects were randomized to either a high-intensity functional tailored exercise program led by 2 physiotherapists per group of 3-9 participants, and consisting of five 45-minute sessions every 2 weeks for 13 weeks (29 sessions) with 6 month follow-up, or seated social activities. Exercises were weight-bearing, progressive, tasks integrated into ADL. The intervention did not significantly reduce the rate of falls or the proportion of participants who sustained a fall either during intervention or at the 6-month follow-up.

Faber et al., 2006 conducted a multicenter RCT in the Netherlands that was single blinded with two levels of block-wise randomization and included 208 frail (49%) and pre-frail (51%) male and female residents of 15 LTC homes (mean age 85). Two exercise interventions, a functional walking (FW) program (n = 54) and the In Balance (IB) intervention (n=70) were compared to a control group (n=84). FW consisted of 20 weeks of an exercise program of 10 exercises relating to balance, mobility, and transfer training. The IB program included Tai Chi principles, and seven therapeutic elements of Tai Chi (ankle ROM, proprioception, sensation, co-contractions, slow continuous motions, trunk rotation, weight shifting). There was a 52-week follow-up. Frailty was found to be a strong effect modifier, with interventions having opposite effects in the frail and pre-frail groups; Both FW and IB programs were effective in reducing fall risk by 61% and improving POMA and Physical Performance Scores in the pre-frail elderly group but not in the frail elderly.  Benefits were evident within 11 weeks. The risk of becoming a faller was significantly increased by the intervention in the pre-frail group, without any significant changes in physical performance measures.

McMurdo et al., 2000, reported a 6-month RCT of exercise in nine residential LTC facilities in the UK, which enrolled 133 men and women >70 years (mean age, 84 years) with a 7-17 month follow-up of falls monitoring (FOPANU study). The intervention consisted of 6 months of an exercise program (2 x week, 30 minutes) incorporating seated exercise of progressive intensity addressing balance, strength, and joint flexibility. Control subjects took part in seated social activities. No differences were found between groups for the number of falls. However, the drop-out rate was very high with only 68% completing the 6-month intervention and 64% completing the follow-up, and no significant differences were found between groups except for reduced prevalence of postural hypertension (39 to 9%), and reduction in poor visual acuity (63 to 46%) in the intervention group.

Norwalk et al., 2001, carried out an RCT of two individualized, exercise programs in LTC in two senior housing communities in the US. The 24-month study enrolled 110 men and women, age > 65 (mean age 84.7). The two interventions were a) FNBF (Fit NB Free) program (n=37), with individualized strength training and conditioning, 3 x week; and b) LL/TC (Living and Learning/Tai Chi) (n=38) consisting of behavioral, psychotherapeutic methods to reduce fear of falling 1 x month and Tai Chi classes 3 x week. Control subjects participated in social and music programs. No significant differences between groups were found. In the FNBF, falls rate was 72% while in the LL/TC it was 58%. Falls in the control were 75%. The fallers were significantly different at baseline and follow up. These subjects had a greater decline in MMSE, IADLs, greater increase in walking time, and decrease in ADLs. Adherence was low, with only 55.8% for the FNBF program, which was still significantly better than LL/TC at 24.2%. Potential benefits were masked by variability of participation (overall adherence = 40%).

Schoenfelder and Rubenstein, 2004. This exercise study in LTC was carried out in 10 facilities in the US. The RCT enrolled matched pairs of men and women (N = 81), 42 of whom were randomized to intervention and 39 to the control group (age >65, mean age 84.1). The 3-month intervention consisted of individually tailored, progressive ankle strengthening followed by supervised walking for up to 10 minutes, 3 x week for 15-20 minute sessions followed by a 6 month follow-up. The control group read books or socialized. The assistive device groups showed maintenance or improvement overtime with the semi-tandem stance which remained significant at 6 months (3 months after supervised exercise ended). This group also maintained the same level of fear of falling or experienced some improvement.

Shimada et al., 2004, undertook an RCT in Japan of 32 physically disabled men and women, mean age 82.4 (range 66-98) in a LTC facility. One group (n=18) received an intervention (treadmill exercise) while the control group (n = 14) carried out usual exercise. The intervention consisted of 6 months of gait training on a treadmill, using handrails, with 6 months follow up. The program was divided into eight phases, with the maximum walking speed measured at the beginning of each phase, and 50-70% of maximum set as training speed. Perturbation stimuli (decelerations) increased in magnitude up to 100%. The controls received physical therapy for pain, TENS, stretching, low and high resistance training, gait training on level surfaces, outdoor walking, balance training, stairs, and group exercise for lower limb function. Fifteen of the intervention subjects and 11 controls completed the study. No significant differences between groups were found although the intervention group showed improvements for one-leg standing time, functional reach, walking, and perturbed walking (33.3% fall rate compared to 54.5% fall rate for controls), and longer time to first fall.

Wolf et al., 2003. This 48-week RCT of Tai Chi in 20 congregate living facilities in the US enrolled 311 transitionally frail men and women (mean age 80.9; range 70-97) with at least 1 fall in the prior year. The intervention was an intense Tai Chi exercise program with an instructor, at 2 sessions per week progressing from 60 minutes to 90 minutes. Subjects in the control group received a Wellness education program for 1 hour/week. The fallers (1 or more falls) had a 47.6% fall rate with intervention while the control group had a 60.3% fall rate. This difference was not significant. Previous fall-related fractures and education were the only variables that modified the effect of the Tai Chi group significantly. Tai Chi subjects without previous fall fractures, or with no high school diploma, had significantly lower fall rates than control subjects. Participants with no high school degree were significantly less physically active at baseline.  Tai Chi subjects with significantly lower risk of falling had better SIP psychosocial scores and 4-12 months of intervention. The subjects who attended their sessions had a marginally significant lower risk of falling.

 

	Evidence	Source	LE	QE	SR
1	Effectiveness of exercise program  studies in long-term care.     Exercise in frail elderly groups may increase risk of falls.	Faber, 2006 McMurdo, 2000 Norwalk, 2001 Schoenfelder and Rubenstein, 2004. Shimada, 2004 Wolf, 2003 Faber, 2006	I	Fair	C

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

---

3.2 Vitamin D

Recommendation

41. Vitamin D supplements of at least 800 IU per day should be provided to older persons residing in long-term care settings with proven or suspected vitamin D insufficiency. [A] 
42. Vitamin D supplements of at least 800 IU per day should be considered in older persons residing in long-term care settings who have abnormal gait or balance or who are otherwise at increased risk for falls. [B]

Rationale

The use of combined calcium and vitamin D3 supplementation has been found to reduce fracture rates in older people in long-term care.  Two studies from a meta-analysis (Bischoff-Ferrari, 2004) as well as one recent randomized controlled trial support the use of vitamin D supplementation to prevent falls in long-term care residents.

Evidence Statements

Flicker et al., 2005,  conducted a two-year multicenter randomized controlled trial in 60 assisted living facilities and 89 nursing homes in Australia. Participants (N=625; mean age, 83.4 years) had serum 25-hydroxyvitamin D levels between 25 and 90 nmol/L. Vitamin D (ergocalciferol, initially 10,000 IU given once weekly, then 1,000 IU daily) was administered to test subjects All participants also received 600 mg calcium carbonate per day. Vitamin D supplementation was associated with an incident rate ratio for falling of 0.73. The odds ratio for ever falling was 0.82 and for ever fracturing was 0.69. Subjects who reported taking at least half the prescribed capsules (n=540) demonstrated an incident rate ratio for falls of 0.63, for ever falling of 0.70, and an odds ratio for ever fracturing of 0.68.

Broe et al., 2007, administered one of four doses of vitamin D (200 IU, 400 IU, 600 IU, or 800 IU) or placebo to 124 long-term nursing home residents (average age, 89 years) in a five-month, randomized, multiple-dose study. Outcomes measured were number of fallers and number of falls assessed using a facility tracking database. The proportion of fallers was 44% in the placebo group, 58% in the 200 IU group, 60% in the 400 IU group, 60% in the 600 IU group, and 20% in the 800 IU group. Residents in the highest-dose group also had a 72% lower adjusted-incidence rate ratio of falls than participants receiving placebo (rate ratio, 0.28).

 

	Evidence	Source	LE	QE	SR
1	Effectiveness of vitamin D supplementation in long-term care settings.	Broe, 2007 Flicker, 2005	I	Good	A
2	Vitamin D supplementation for older persons with abnormal gait or balance.	Bishoff ferrari, 2004	I	Fair	B

      LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

       

---

   

 

4.  OLDER PERSONS WITH COGNITIVE IMPAIRMENT

Background

Older people with cognitive impairment and dementia are at increased risk for falls, with an annual incidence of around 60% (twice that of cognitively normal older people).  (Tinetti, 1988; Van Dijk,1999).  Mobility problems experienced by elderly people with dementia are associated with falls, fractures and admission to long-term care.

Multifactorial assessment and intervention after a fall or single intervention in patients with cognitive impairment or dementia have not been shown to reduce falls

Recommendations

43. There is insufficient evidence to recommend for or against multifactorial or single interventions to prevent falls in older persons with known dementia living in the community or in long-term care facilities. [I]

Rationale

Cognitive impairment is an independent risk factor for falls. Nevertheless, older persons with cognitive impairment have been excluded from most of the successful falls prevention randomized controlled trials in the community setting. The only study that specifically investigated cognitive impairment in the community demonstrated lack of efficacy.

Based on the studies that have been conducted in long-term care facilities in older persons with cognitive impairment, the evidence is inconclusive. More studies with adequate sample size, sensitive and validated measurements, and higher specificity for the types of intervention targeting subgroups of patients with different degrees of cognitive impairment are required to allow for evidence-based recommendations.         

Evidence Statements

Hauer et al., 2006. This systematic review cites 11 randomized controlled trials that evaluated the effect of physical activity (exercise) on fall prevention in older persons with cognitive impairment. The review found conflicting evidence regarding the effect of physical training on motor performance and falls in older people with cognitive impairment. However, there was a large heterogeneity regarding methodology, sample size, type of intervention, study outcomes, and analyses, which hampered the evaluation of the effectiveness of training. The investigators concluded that randomized controlled trials reveal only limited effectiveness of physical training or exercise in patients with cognitive impairment.

Jensen, 2003, evaluated a multicomponent intervention program comprising staff education, environmental adjustment, exercise, drug review, aids, hip protectors, and post-fall problem-solving conferences. All consenting residents (N=402) were divided into a group of either lower or higher cognition based on the results of a mini-mental state examination (MMSE) (score=19 was the dividing point). The lower MMSE group was older and more functionally impaired and had a higher risk of falling (64% versus 36%) than the higher MMSE group.  A significant intervention effect on falls appeared in the higher MMSE group but not in the lower MMSE group (adjusted incidence rates ratio of falls P=.016 and P=.121; adjusted hazard ratio P<.001 and P=.420, respectively). 

Shaw et al., 2003, conducted a randomized controlled trial to determine the effectiveness of a tailored multicomponent intervention after multifactorial clinical assessment in older patients with cognitive impairment and dementia presenting to the emergency department after a fall. Interventions included optical correction, medical assessment, physiotherapy, occupational therapy and foot care. Of the 274 patients with cognitive impairment and dementia, 130 were randomized to assessment and intervention and 144 were randomized to assessment followed by conventional care. Intention-to-treat analysis showed no significant difference between intervention and control groups in the proportion of patients who fell during a one-year follow-up (74% and 80%; relative risk, 0.92.) No significant differences were found between groups for secondary outcome measures. 

Evidence Table

	Evidence	Source	LE	QE	SR
	There is insufficient evidence to recommend for or against single or multifactorial interventions in community-living older adults with known cognitive impairment.	Shaw, 2003 Hauer, 2006 Jensen, 2003	I	Poor	I

          LE = level of evidence; QE = quality of evidence; SR = strength of recommendation.

 

---