- Email: [email protected]
Epidemiology of Falls in Residential Aged Care: Analysis of More Than 70,000 Falls From Residents of Bavarian Nursing Homes
JAMDA 13 (2012) 187.e1e187.e6
JAMDA journal homepage: www.jamda.com
Original Study
Epidemiology of Falls in Residential Aged Care: Analysis of More Than 70,000 Falls From Residents of Bavarian Nursing Homes Kilian Rapp MD, MPH a, b, *, Clemens Becker MD a, Ian D. Cameron MB BS, PhD c, Hans-Helmut König MD, MPH d, Gisela Büchele PhD, MPH b a
Department of Clinical Gerontology, Robert-Bosch-Hospital, Stuttgart, Germany Institute of Epidemiology and Medical Biometrie, Ulm University, Ulm, Germany c Rehabilitation Studies Unit, Sydney Medical School, University of Sydney, Ryde, NSW, Australia d Department of Medical Sociology and Health Economics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany b
a b s t r a c t Keywords: Epidemiology falls residential facilities
Objective: Falls and fall-related injuries are leading problems in residential aged care facilities. The objective of this study was to provide descriptive data about falls in nursing homes. Design/Setting/Participants: Prospective recording of all falls over 1 year covering all residents from 528 nursing homes in Bavaria, Germany. Measurements: Falls were reported on a standardized form that included a facility identification code, date, time of the day, sex, age, degree of care need, location of the fall, and activity leading to the fall. Data detailing homes’ bed capacities and occupancy levels were used to estimate total person-years under exposure and to calculate fall rates. All analyses were stratified by residents’ degree of care need. Results: More than 70,000 falls were recorded during 42,843 person-years. The fall rate was higher in men than in women (2.18 and 1.49 falls per person-year, respectively). Fall risk differed by degree of care need with lower fall risks both in the least and highest care categories. About 75% of all falls occurred in the residents’ rooms or in the bathrooms and only 22% were reported within the common areas. Transfers and walking were responsible for 41% and 36% of all falls respectively. Fall risk varied during the day. Most falls were observed between 10 AM and midday and between 2 PM and 8 PM. Conclusion: The differing fall risk patterns in specific subgroups may help to target preventive measures. Published by Elsevier Inc. on behalf of the American Medical Directors Association, Inc.
Nursing homes are locations with particularly high risks for falls. Therefore, falls and their consequences, such as fractures and soft tissue injuries, are leading problems in residential care settings.1 Fall rates in nursing homes were analyzed by several historical studies, mainly from the 1970s and 1980s. A summary of these studies calculated a mean fall rate of 1.7 falls per person-year (range 0.6e3.6),2 which is considerably higher than the fall rate observed in older people living in the community (mean 0.65, range 0.3e1.6).3 In previous years, data about falls in nursing homes were reported by a few observational studies4e8 and by fall prevention studies,9e12 confirming the magnitude of the fall rate in residential care.
The evaluation of the study was funded by the Bundesministerium fur Bildung und Forshung (Forderkennzeichen: 01EL0702, 01EL0717, 01EL0718). All authors declare that they have no conflicts of interest. * Address correspondence to Kilian Rapp, MD, Department of Clinical Gerontology, Robert-Bosch-Hospital, Auerbachstr. 110, 70376 Stuttgart, Germany. E-mail address: [email protected] (K. Rapp).
Some of the studies also give information about circumstances associated with falls, such as place, activity, time of day, weekday, or season. There is an agreement, for example, that the vast majority of falls occur in the residents’ rooms. Information about the predominant activity before the fall, such as walking or performing a transfer, is inconsistent and may be dependent on the setting and the case-mix of the observed population. Some of the studies suggested a variation of the fall rate over the day,4,7,8,13,14 whereas others did not.15 However, the absolute number of falls in these studies is usually relatively low, which compromises the validity and reliability of the results. Therefore, stratified analyses for example by sex and age have been difficult to perform and additional information about the residents’ functional status was not available. The objective of our study was to report descriptive data about falls and fall rates with detailed information about time and place of the fall and the activity leading to the fall. We used a database with more than 70,000 falls collected in 2008 in 528 nursing homes in Bavaria, Germany. Furthermore, data about the degree of care need/functional status were available for each resident
1525-8610/$ - see front matter Published by Elsevier Inc. on behalf of the American Medical Directors Association, Inc. doi:10.1016/j.jamda.2011.06.011
187.e2
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
experiencing a fall. This allowed us to perform additional analyses stratified by the residents’ functional status. To our knowledge, this evaluation is by far the largest descriptive analysis of falls in residents of nursing homes. Exact information about specific fall risks of specific subgroups may improve our understanding of falls in residential care and may help to target preventive measures more specifically. Methods Bavaria is a federal state with 12.5 million inhabitants in the south of Germany. In Bavaria, more than 12% of the 551,000 citizens aged 80 years or older are residents of one of about 1400 nursing homes. Persons live in a nursing home of their own choice. In 2007, Germany’s largest health insurance company (AOK) started to implement a statewide fall prevention program (The Bavarian Fall and Fracture Prevention Study) in nursing homes in Bavaria. Participating nursing homes committed to document each fall of each resident. Falls were reported on a standardized form, which included a facility identification code, date, time, sex, four age categories, four levels of care (see the following section), location of fall (resident’s room, bathroom, common area, outside), and activity leading to the fall (walking, transfer, others). The anonymous fall report forms were sent to the health insurance company’s data center by fax and read there by a data scanner. If a resident had more than one fall on one day, only the first fall or the most serious fall was reported to the data center. For each nursing facility, the number of beds was recorded. Level of Care Most residents who live in a German nursing home fulfill the requirements of the mandatory long-term care insurance.16 This insurance was introduced in 1995. The insurance is compulsory for all citizens. Depending on the amount of care required, recipients are categorized into 1 of 3 levels after an assessment by a physician (levels 1, 2, and 3 requiring basic care such as washing, feeding, or dressing for at least 0.75, 2.00 and 4.00 hours per day, respectively). Those residents whose assessment showed that they needed basic care for less than 0.75 hour per day are categorized as level of care “0” and represent a less frail population. Cognition is not part of the assessment process. People living in a nursing home and categorized as level of care “0” are often cognitively impaired, needing continuous supervision despite a low need for basic care. Our analyses are based on the falls reported by the 528 nursing homes that were included in the program in 2008. The dataset comprised all falls that occurred between 1 January 2008 and 31 December 2008. To analyze fall rates, person-years of exposure were estimated. The bed capacities of all participating homes were multiplied with an average degree of occupancy (factor 0.866), which was derived from the Statistical Bureau of the Federal Government and States.17 To determine the distribution of sex, age, and level of care within the observed population, data from a survey of the same year were used. In this survey, one nursing scientist visited a randomly drawn subset of the participating nursing homes (48 of 528 homes; 9%) and collected basic information (sex, age, and level of care) from all residents living in these facilities.18 The estimated person-years of exposure are shown in Table 1. There is evidence from the literature that fall rates in German nursing homes are clearly more than 1 fall per person-year.10 Some of the participating homes had implausibly low fall rates. Therefore, a sensitivity analysis was performed calculating fall rates after exclusion of homes with fall rates below one fall per person-year in homes with more than 50 beds. The small homes were not
Table 1 Summary of Number of Participating Nursing Homes and Estimated Number of Person-Years, Including by Age Group and Level of Care, for Women and Men Total Nursing homes, n Number of beds Median (Quartile range) Total exposure time, py (%) Age group, py (%) <70 70e79 80e89 90 Level of care, py (%) 0 1 2 3
Women
Men
42,843 (100)
33,503 (78.2)
9340 (21.8)
2536 7569 22,834 9904
(5.9) (17.7) (53.3) (23.1)
1082 4999 18,765 8657
(3.2) (14.9) (56.0) (25.8)
1454 2570 4068 1247
(15.6) (27.5) (43.6) (13.4)
6128 13,823 14,528 8398
(14.3) (32.3) (33.9) (16.6)
4456 10,788 11,324 6969
(13.3) (32.2) (33.8) (20.8)
1672 3035 3204 1429
(17.9) (32.3) (34.3) (15.3)
528 88 (64e115)
py, exposure time in person years.
excluded, because their variance in fall rate is high compared with their absolute number of falls. If excluded, this could have led to an exclusion of homes with correct documentation with low fall rates in 2008 and an overrepresentation of small homes with high fall rates. Statistics Fall rates were calculated by dividing the number of falls by the total number of person years (PY) of risk (or per 100 PY, respectively). Confidence intervals with 95% confidence probability were determined. Distribution of falls (in percent) by time of the day, location of falls, or activity at the time of the fall are presented in the figures. The upper panel of these figures refers to the total number of falls, the lower panel shows percentile distributions within level-of-care categories. Fall rates were different between women and men; however, an additional stratification by sex was not performed because of the very similar percentile distributions in women and men. Results Falls were documented over one calendar year in 528 nursing homes with a median of 88 beds. The estimated person-years of exposure were 33,503 years in women and 9340 years in men. The distribution over different age groups and in different levels of care is shown in Table 1. The fall rate was higher in men (2.18 falls per person-year) than in women (1.49 falls per person-year) (Table 2). The pattern of the Table 2 Number of Falls and Fall Rates, Total and by Age Group and Level of Care, for Women and Men Women
Number Age group <70 70e79 80e89 90 Level of care 0 1 2 3
Men
n
Rate per Person-Year (95% CI)
n
Rate per Person-Year (95% CI)
49,864
1.49 (1.48, 1.50)
20,332
2.18 (2.15, 2.21)
2044 6673 28,501 12,646
1.89 1.33 1.52 1.46
(1.81, (1.30, (1.50, (1.44,
1.97) 1.37) 1.54) 1.49)
2887 4897 9505 3043
1.99 1.91 2.34 2.44
(1.92, (1.86, (2.30, (2.35,
2.07) 1.97) 2.39) 2.53)
4549 21,040 20,042 4233
1.02 1.95 1.77 0.61
(0.99, (1.92, (1.75, (0.59,
1.05) 1.98) 1.79) 0.63)
2024 7841 8469 1998
1.21 2.58 2.64 1.40
(1.16, (2.53, (2.50, (1.34,
1.26) 2.64) 2.70) 1.46)
CI, confidence interval; n, absolute number of falls.
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
187.e3
Time of day
Frequency in %
0-2 a.m. 2-4 a.m. 4-6 a.m. 6-8 a.m. 8-10 a.m. 10-12 a.m. 12-2 p.m. 2-4 p.m. 4-6 p.m. 6-8 p.m. 8-10 p.m. 10-12 p.m.
5.18 5.35 6.27 7.98 7.78 10.54 8.49 11.91 11.84 11.37 6.18 7.12
Level of care 0-2 a.m.
2-4 a.m.
4-6 a.m.
6-8 a.m.
8-10 a.m.
10-12 a.m.
12-2 p.m.
2-4 p.m.
4-6 p.m.
6-8 p.m.
8-10 p.m.
10-12 p.m.
0 1 2 3
5.81 5.81 4.77 3.43
0 1 2 3
6.15 5.96 4.92 3.64
0 1 2 3
6.24 6.90 6.04 4.43
0 1 2 3
8.35 8.72 7.64 5.75
0 1 2 3
8.12 7.87 7.52 8.14
0 1 2 3
10.28 10.04 10.49 13.30
0 1 2 3
8.23 7.88 8.76 10.38
0 1 2 3
10.13 10.85 12.88 14.30
0 1 2 3
11.35 10.74 12.44 14.68
0 1 2 3
10.70 11.00 11.73 12.15
0 1 2 3
6.79 6.46 6.07 4.81
0 1 2 3
7.85 7.78 6.74 4.98
0
1
2
3
4
5
6
7
8
9
10
11
12
Frequency in % Fig. 1. Distribution of falls by the time of day and level of care.
13
14
15
187.e4
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
Location of falls
Frequency in %
Resident's room
62.26
Bathroom
12.65
Common area
22.05
Outside
3.05
Level of care
Resident's room
Bathroom
Common area
Outside
0
63.03
1
65.08
2
61.32
3
52.66
0
13.53
1
13.86
2
12.08
3
8.68
0
18.10
1
17.77
2
24.16
3
36.37
0
5.34
1
3.29
2
2.45
3
2.29
0
10
20
30
40
50
60
70
Frequency in % Fig. 2. Distribution of falls by the location of falls and level of care.
age-dependent fall rate differed between the genders. In women, fall rates decreased slightly with increasing age categories, whereas the highest fall rates in men were observed in the 2 highest age categories. The level of care was strongly associated with fall risk in women and men. Particularly high fall rates were observed in residents with low and medium care need (level of care 1 and 2) (Table 2). There were 85 nursing homes reporting implausibly low falls rates (fewer than 1 fall per person-year in homes with more than 50 beds). The sensitivity analysis excluding these homes showed a higher fall rate (1.70 and 2.47 falls per person-year in women and men, respectively). Figure 1 demonstrates the distribution of falls by the time of day. Two peaks were observed. One in the later morning between 10 AM and midday and another in the afternoon and the early evening between 2 PM and 8 PM. The rates in the afternoon were statistically still significantly higher than the rates in the morning (10 to 12 AM). In a nursing home with 100 beds, for example, 9 and 12.6 falls can be expected in women and men within 1 year in the early afternoon (2 PM to 4 PM), respectively (Supplement Table 1). The relative contribution to the 2 described peaks increased with increasing level of care, whereas the relative contribution of time-dependent fall rates of residents with less care need was higher during the night (8 PM until 8 AM, see Figure 1).
About 75% of all falls occurred in the residents’ rooms or in the (usually) adjoining bathrooms and only 22% were reported from the common areas (Figure 2). That means that in a nursing home with 100 beds, about 112 and 161 falls can be expected in the residents’ rooms and bathrooms compared with only 33 and 48 falls in the common area in women and men, respectively (Supplement Table 2). The functionally most limited residents had the highest relative contribution to falls in the common area. Not unexpectedly, the relative contribution to falls outside the building decreased with increasing care need (level of care) (Figure 2). Most falls occurred when performing a transfer (41%), 36% occurred when walking, and 23% could either not be classified or occurred during another activity (eg, when sitting). In contrast to the category “transfers,” the relative contribution to falls when walking decreased strongly with increasing care need (level of care) (Figure 3 and Supplement Table 3). The day of the week (eg, weekend compared with weekday) or the season had no effect on fall rates (data not shown).
Discussion The study presents detailed descriptive data about falls and fall rates in nursing homes. We found lower fall rates in women than in
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
187.e5
Fall occurred by
Frequency in %
Walking
36.30
Transfer
41.13
Others
22.58
Level of care
Walking
Transfer
Others
0
43.27
1
39.47
2
33.50
3
27.07
0
35.83
1
40.09
2
43.40
3
41.12
0
20.90
1
20.45
2
23.10
3
31.81
0
10
20
30
40
50
Frequency in % Fig. 3. Distribution of falls by activity before the fall and level of care.
men and decreasing fall rates with increasing care need (level of care). Prevalence of falls peaked in the morning between 10 AM and midday and in the afternoon and the early evening between 2 PM and 8 PM. Most of the falls occurred in the residents’ rooms and only less than a quarter were observed in the homes’ common areas. Residents with less functional limitations contributed more to falls during the night, to falls in their own room, and to falls when walking, whereas those with more functional limitations (level of care 3) contributed relatively more to falls during the day, to falls in the common area, and to falls when performing a transfer. The reported fall rate in this study is probably an underestimation of the true fall rate for several reasons. First, a maximum of 1 fall per resident per day was reported. Therefore, some falls from frequent fallers may have been missed. Second, some of the falls may not have been observed by the staff and therefore were not reported. This may be particularly the case for outdoor falls with a selective reporting of injurious falls. Third, the data were assessed as part of an intervention to reduce falls, which may have influenced the fall rate. Fourth, falls may have been observed but not reported by the staff because of high workload or other causes. This was the reason why we performed a sensitivity analysis excluding nursing homes with unrealistically low numbers of falls. Therefore, the higher fall rate calculated by sensitivity analysis may be closer to the true fall rate. A Finnish study reported a 1.6 times higher fall incidence in men than in women,8 which comes close to the relative incidence observed in our data. The difference is partly because men are
categorized more often in level of care 0 (high fall rates) and less often in level of care 3 (low fall rates). This may reflect differing patterns of disability in men and women, owing to women experiencing longer periods of severe disability before death. However, even within the same levels of care, fall rates are significantly higher in men than in women. The reason for the observed difference is not clear. The male nursing home population may be somewhat different from the female population in the way that men are frequently cared for at home by their wives as long as possible, whereas similar support is less common for women with disabilities. Therefore, the reason and the functional status at admission to a nursing home may be somewhat different between women and men, which may be only incompletely expressed by the categorization in the different levels of care. However, this is not a sufficient explanation for the observed difference. It also remains unclear why fall rates decreased slightly with increasing age categories in women, whereas the opposite was the case in men. Although the assessment of care need used in this study was based on a qualitative judgment made by a physician or a nurse, the method has been shown to have good levels of inter-rater reliability.19 The need for care is strongly associated with functional impairment, in particular mobility and comorbidity.20 It was not unexpected that the risk was lowest in residents with low care need (level 0) and with high care need (level 3). The first group is the less frail group with probably the fewest walking disorders, whereas the latter group is mainly bedridden and therefore less exposed to risky situations that result in falls.
187.e6
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
The time-dependency of the falls may be because of an increased activity during specific time periods. Our observation of a maximum of falls in the afternoon was also reported by other authors.4,7,21 Why fall rates in the afternoon exceed those in the morning remains unclear. Potential reasons could be that residents are more tired later in the day, higher activity patterns, or less availability of the nursing staff, for example because of shift changes during the afternoon. The lowest fall rates were observed during the night. However, time of exposure during the night, such as from going to and returning from the bathroom, is extremely short. Therefore, fall rates per step are very high, which means that activities during the night are probably high-risk situations. The vast majority of falls occurred in the residents’ rooms. This is in line with results from previous studies.4,14 It is interesting that people with high care needs have a higher relative risk to fall during daytime in common areas. An explanation may be that these very frail residents are brought to common areas to allow them to participate in the facility’s daily life but cannot be supervised continuously. In the analyzed population, the importance of transfers as an activity leading to falls is even higher than that related to walking. This has implications for prevention. Specific training of transfers may be worthwhile because of the very high risk of falls in nursing home residents who are able to stand but then are unable to maintain an upright posture.22 Another problem may be orthostatic hypotension in this population, with a high proportion of autonomic dysfunction and polypharmacy.2 Therefore, drugs influencing blood pressure should be prescribed very cautiously to residents of nursing homes. Because falls in the resident’s room are of greatest relevance, environmental aspects in the residents’ private space can be another target for preventive measures. Individual aspects such as walking patterns or transfer heights could be considered to reduce person-environment mismatches on an individual level. Efforts to improve incontinence, such as scheduled nocturnal prompted voiding or medical interventions for urge incontinence, may be particularly effective in reducing falls during the night.1 Our data show that the resident’s functional status influences his or her fall risk. Therefore, fall rates depend also on the case mix of the observed nursing home populations. The type of institutional care differs between different countries, which influences the case mix and may compromise the external validity of our reported fall rates. The calculated fall rates, however, fit very well to previously reported fall rates from nursing homes from different countries and continents.2,5,10e12,23 The total number of person-years under exposure could only be estimated. The analyzed residents, however, represented about 43% of all person-years of residential care in older people in Bavaria.24 Information about each home’s capacity and valid data regarding the overall occupancy in Bavarian nursing homes were available. Furthermore, representative data from a subgroup of 48 nursing homes about the distribution of sex, age, and the level of care were available. Therefore, we believe that our estimation of person-years is very close to the real number of person-years under exposure. Our study is the largest reported descriptive evaluation of falls and fall rates in residential care. The large number of falls allowed us to stratify the analyses in a detailed way for different variables of interest like sex, age, and even for very close time intervals. An additional strength of the study was the availability of the degree of care need, which allowed us to perform additional analyses stratified by the residents’ functional status. In summary, we examined falls and fall rates in residents from nursing homes. In this setting, falls and fall rates differed between women and men, between different places, between different
activities, between different degrees of care need, and within the course of the day. Acknowledgments We thank Regina Merk-Bäuml, Ralf Brum, Markus Gindl, and Stefanie Büttner from the Allgemeine Ortskrankenkasse (AOK), and Ulrich Rissmann from the Robert-Bosch-Hospital Stuttgart. Furthermore, we thank Andrea Kleiner from the Institute of Epidemiology and Medical Biometry, Ulm University, for the preparation of the figures. Supplementary Data Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jamda.2011.06.011 References 1. Becker C, Rapp K. Fall prevention in nursing homes. Clin Geriatr Med 2010;26: 693e704. 2. Rubenstein LZ, Josephson KR, Robbins AS. Falls in the nursing home. Ann Intern Med 1994;121:442e451. 3. Rubenstein LZ. Falls in older people: Epidemiology, risk factors and strategies for prevention. Age Ageing 2006;35:ii37eii41. 4. Lester P, Haq M, Vadnerkar A, et al. Falls in the nursing home setting: Does time matter? J Am Med Dir Assoc 2008;9:684e686. 5. Gibson RE, Harden M, Byles J, et al. Incidence of falls and fall-related outcomes among people in aged-care facilities in the Lower Hunter region, NSW. N S W Public Health Bull 2008;19:166e169. 6. Heinze C, Halfens RJ, Dassen T. Falls in German in-patients and residents over 65 years of age. J Clin Nurs 2007;16:495e501. 7. Jensen J, Lundin-Olsson L, Nyberg L, et al. Falls among frail older people in residential care. Scand J Public Health 2002;30:54e61. 8. Nurmi I, Luthje P. Incidence and costs of falls and fall injuries among elderly in institutional care. Scand J Prim Health Care 2002;20:118e122. 9. Jensen J, Lundin-Olsson L, Nyberg L, et al. Fall and injury prevention in older people living in residential care facilities. A cluster randomized trial. Ann Intern Med 2002;136:733e741. 10. Becker C, Kron M, Lindemann U, et al. Effectiveness of a multifaceted intervention on falls in nursing home residents. J Am Geriatr Soc 2003;51:306e313. 11. Kerse N, Butler M, Robinson E, et al. Fall prevention in residential care: A cluster, randomized, controlled trial. J Am Geriatr Soc 2004;52:524e531. 12. Dyer CA, Taylor GJ, Reed M, et al. Falls prevention in residential care homes: A randomised controlled trial. Age Ageing 2004;33:596e602. 13. Cacha CA. An analysis of the 1976 incident reports of the Carillon Nursing Home. J Am Health Care Assoc 1979;5:29e33. 14. Jantti PO, Pyykko VI, Hervonen AL. Falls among elderly nursing home residents. Public Health 1993;107:89e96. 15. Ashley MJ, Gryfe CI, Amies A. A longitudinal study of falls in an elderly population II. Some circumstances of falling. Age Ageing 1977;6:211e220. 16. Becker C, Leistner K, Nikolaus T. Introducing a statutory insurance system for long-term care (Pflegeversicherung) in Germany. In: Michel JP, Rubenstein LZ, Vellas BJ, et al., editors. Geriatric Programs and Departments Around the World. Paris: Serd-Springer; 1998. p. 55e64. 17. Statistische Ämter des Bundes und der Länder: Statistik über die Empfänger von Pflegeleistungen, Pflegebedürftige nach Leistungsart und Geschlecht, Stichtag, regionale Tiefe: Reg.-Bez./Stat. Available at: https://www.regionalstatistik.de/ genesis/online. Accessed December 17, 2010. 18. Klenk J, Kurrle S, Rissmann U, et al. Availability and use of hip protectors in residents of nursing homes. Osteoporos Int 2011;22:1593e1598. 19. Kliebsch U, Brenner H. [Interrater reliability of expert assessment by the health insurance medical service in determining eligibility for disability benefits]. Gesundheitswesen 1995;57:638e644. 20. Fried LP, Guralnik JM. Disability in older adults: Evidence regarding significance, etiology, and risk. J Am Geriatr Soc 1997;45:92e100. 21. Svensson ML, Rundgren A, Larsson M, et al. Accidents in the institutionalized elderly: injuries and consequences. Aging (Milano ) 1992;4:125e133. 22. Lord SR, March LM, Cameron ID, et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided. J Am Geriatr Soc 2003;51:1645e1650. 23. Rubenstein LZ, Robbins AS, Josephson KR, et al. The value of assessing falls in an elderly population. A randomized clinical trial. Ann Intern Med 1990;113: 308e316. 24. Bayerisches Landesamt für Statistik und Datenverarbeitung: Statistische Berichte. Pflegeeinrichtungen und Pflegegeldempfänger in Bayern. Ergebnisse der Pflegestatistik. Available at: http://www.statistik.bayern.de/statistik/. Accessed December 31, 2007.
JAMDA journal homepage: www.jamda.com
Original Study
Epidemiology of Falls in Residential Aged Care: Analysis of More Than 70,000 Falls From Residents of Bavarian Nursing Homes Kilian Rapp MD, MPH a, b, *, Clemens Becker MD a, Ian D. Cameron MB BS, PhD c, Hans-Helmut König MD, MPH d, Gisela Büchele PhD, MPH b a
Department of Clinical Gerontology, Robert-Bosch-Hospital, Stuttgart, Germany Institute of Epidemiology and Medical Biometrie, Ulm University, Ulm, Germany c Rehabilitation Studies Unit, Sydney Medical School, University of Sydney, Ryde, NSW, Australia d Department of Medical Sociology and Health Economics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany b
a b s t r a c t Keywords: Epidemiology falls residential facilities
Objective: Falls and fall-related injuries are leading problems in residential aged care facilities. The objective of this study was to provide descriptive data about falls in nursing homes. Design/Setting/Participants: Prospective recording of all falls over 1 year covering all residents from 528 nursing homes in Bavaria, Germany. Measurements: Falls were reported on a standardized form that included a facility identification code, date, time of the day, sex, age, degree of care need, location of the fall, and activity leading to the fall. Data detailing homes’ bed capacities and occupancy levels were used to estimate total person-years under exposure and to calculate fall rates. All analyses were stratified by residents’ degree of care need. Results: More than 70,000 falls were recorded during 42,843 person-years. The fall rate was higher in men than in women (2.18 and 1.49 falls per person-year, respectively). Fall risk differed by degree of care need with lower fall risks both in the least and highest care categories. About 75% of all falls occurred in the residents’ rooms or in the bathrooms and only 22% were reported within the common areas. Transfers and walking were responsible for 41% and 36% of all falls respectively. Fall risk varied during the day. Most falls were observed between 10 AM and midday and between 2 PM and 8 PM. Conclusion: The differing fall risk patterns in specific subgroups may help to target preventive measures. Published by Elsevier Inc. on behalf of the American Medical Directors Association, Inc.
Nursing homes are locations with particularly high risks for falls. Therefore, falls and their consequences, such as fractures and soft tissue injuries, are leading problems in residential care settings.1 Fall rates in nursing homes were analyzed by several historical studies, mainly from the 1970s and 1980s. A summary of these studies calculated a mean fall rate of 1.7 falls per person-year (range 0.6e3.6),2 which is considerably higher than the fall rate observed in older people living in the community (mean 0.65, range 0.3e1.6).3 In previous years, data about falls in nursing homes were reported by a few observational studies4e8 and by fall prevention studies,9e12 confirming the magnitude of the fall rate in residential care.
The evaluation of the study was funded by the Bundesministerium fur Bildung und Forshung (Forderkennzeichen: 01EL0702, 01EL0717, 01EL0718). All authors declare that they have no conflicts of interest. * Address correspondence to Kilian Rapp, MD, Department of Clinical Gerontology, Robert-Bosch-Hospital, Auerbachstr. 110, 70376 Stuttgart, Germany. E-mail address: [email protected] (K. Rapp).
Some of the studies also give information about circumstances associated with falls, such as place, activity, time of day, weekday, or season. There is an agreement, for example, that the vast majority of falls occur in the residents’ rooms. Information about the predominant activity before the fall, such as walking or performing a transfer, is inconsistent and may be dependent on the setting and the case-mix of the observed population. Some of the studies suggested a variation of the fall rate over the day,4,7,8,13,14 whereas others did not.15 However, the absolute number of falls in these studies is usually relatively low, which compromises the validity and reliability of the results. Therefore, stratified analyses for example by sex and age have been difficult to perform and additional information about the residents’ functional status was not available. The objective of our study was to report descriptive data about falls and fall rates with detailed information about time and place of the fall and the activity leading to the fall. We used a database with more than 70,000 falls collected in 2008 in 528 nursing homes in Bavaria, Germany. Furthermore, data about the degree of care need/functional status were available for each resident
1525-8610/$ - see front matter Published by Elsevier Inc. on behalf of the American Medical Directors Association, Inc. doi:10.1016/j.jamda.2011.06.011
187.e2
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
experiencing a fall. This allowed us to perform additional analyses stratified by the residents’ functional status. To our knowledge, this evaluation is by far the largest descriptive analysis of falls in residents of nursing homes. Exact information about specific fall risks of specific subgroups may improve our understanding of falls in residential care and may help to target preventive measures more specifically. Methods Bavaria is a federal state with 12.5 million inhabitants in the south of Germany. In Bavaria, more than 12% of the 551,000 citizens aged 80 years or older are residents of one of about 1400 nursing homes. Persons live in a nursing home of their own choice. In 2007, Germany’s largest health insurance company (AOK) started to implement a statewide fall prevention program (The Bavarian Fall and Fracture Prevention Study) in nursing homes in Bavaria. Participating nursing homes committed to document each fall of each resident. Falls were reported on a standardized form, which included a facility identification code, date, time, sex, four age categories, four levels of care (see the following section), location of fall (resident’s room, bathroom, common area, outside), and activity leading to the fall (walking, transfer, others). The anonymous fall report forms were sent to the health insurance company’s data center by fax and read there by a data scanner. If a resident had more than one fall on one day, only the first fall or the most serious fall was reported to the data center. For each nursing facility, the number of beds was recorded. Level of Care Most residents who live in a German nursing home fulfill the requirements of the mandatory long-term care insurance.16 This insurance was introduced in 1995. The insurance is compulsory for all citizens. Depending on the amount of care required, recipients are categorized into 1 of 3 levels after an assessment by a physician (levels 1, 2, and 3 requiring basic care such as washing, feeding, or dressing for at least 0.75, 2.00 and 4.00 hours per day, respectively). Those residents whose assessment showed that they needed basic care for less than 0.75 hour per day are categorized as level of care “0” and represent a less frail population. Cognition is not part of the assessment process. People living in a nursing home and categorized as level of care “0” are often cognitively impaired, needing continuous supervision despite a low need for basic care. Our analyses are based on the falls reported by the 528 nursing homes that were included in the program in 2008. The dataset comprised all falls that occurred between 1 January 2008 and 31 December 2008. To analyze fall rates, person-years of exposure were estimated. The bed capacities of all participating homes were multiplied with an average degree of occupancy (factor 0.866), which was derived from the Statistical Bureau of the Federal Government and States.17 To determine the distribution of sex, age, and level of care within the observed population, data from a survey of the same year were used. In this survey, one nursing scientist visited a randomly drawn subset of the participating nursing homes (48 of 528 homes; 9%) and collected basic information (sex, age, and level of care) from all residents living in these facilities.18 The estimated person-years of exposure are shown in Table 1. There is evidence from the literature that fall rates in German nursing homes are clearly more than 1 fall per person-year.10 Some of the participating homes had implausibly low fall rates. Therefore, a sensitivity analysis was performed calculating fall rates after exclusion of homes with fall rates below one fall per person-year in homes with more than 50 beds. The small homes were not
Table 1 Summary of Number of Participating Nursing Homes and Estimated Number of Person-Years, Including by Age Group and Level of Care, for Women and Men Total Nursing homes, n Number of beds Median (Quartile range) Total exposure time, py (%) Age group, py (%) <70 70e79 80e89 90 Level of care, py (%) 0 1 2 3
Women
Men
42,843 (100)
33,503 (78.2)
9340 (21.8)
2536 7569 22,834 9904
(5.9) (17.7) (53.3) (23.1)
1082 4999 18,765 8657
(3.2) (14.9) (56.0) (25.8)
1454 2570 4068 1247
(15.6) (27.5) (43.6) (13.4)
6128 13,823 14,528 8398
(14.3) (32.3) (33.9) (16.6)
4456 10,788 11,324 6969
(13.3) (32.2) (33.8) (20.8)
1672 3035 3204 1429
(17.9) (32.3) (34.3) (15.3)
528 88 (64e115)
py, exposure time in person years.
excluded, because their variance in fall rate is high compared with their absolute number of falls. If excluded, this could have led to an exclusion of homes with correct documentation with low fall rates in 2008 and an overrepresentation of small homes with high fall rates. Statistics Fall rates were calculated by dividing the number of falls by the total number of person years (PY) of risk (or per 100 PY, respectively). Confidence intervals with 95% confidence probability were determined. Distribution of falls (in percent) by time of the day, location of falls, or activity at the time of the fall are presented in the figures. The upper panel of these figures refers to the total number of falls, the lower panel shows percentile distributions within level-of-care categories. Fall rates were different between women and men; however, an additional stratification by sex was not performed because of the very similar percentile distributions in women and men. Results Falls were documented over one calendar year in 528 nursing homes with a median of 88 beds. The estimated person-years of exposure were 33,503 years in women and 9340 years in men. The distribution over different age groups and in different levels of care is shown in Table 1. The fall rate was higher in men (2.18 falls per person-year) than in women (1.49 falls per person-year) (Table 2). The pattern of the Table 2 Number of Falls and Fall Rates, Total and by Age Group and Level of Care, for Women and Men Women
Number Age group <70 70e79 80e89 90 Level of care 0 1 2 3
Men
n
Rate per Person-Year (95% CI)
n
Rate per Person-Year (95% CI)
49,864
1.49 (1.48, 1.50)
20,332
2.18 (2.15, 2.21)
2044 6673 28,501 12,646
1.89 1.33 1.52 1.46
(1.81, (1.30, (1.50, (1.44,
1.97) 1.37) 1.54) 1.49)
2887 4897 9505 3043
1.99 1.91 2.34 2.44
(1.92, (1.86, (2.30, (2.35,
2.07) 1.97) 2.39) 2.53)
4549 21,040 20,042 4233
1.02 1.95 1.77 0.61
(0.99, (1.92, (1.75, (0.59,
1.05) 1.98) 1.79) 0.63)
2024 7841 8469 1998
1.21 2.58 2.64 1.40
(1.16, (2.53, (2.50, (1.34,
1.26) 2.64) 2.70) 1.46)
CI, confidence interval; n, absolute number of falls.
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
187.e3
Time of day
Frequency in %
0-2 a.m. 2-4 a.m. 4-6 a.m. 6-8 a.m. 8-10 a.m. 10-12 a.m. 12-2 p.m. 2-4 p.m. 4-6 p.m. 6-8 p.m. 8-10 p.m. 10-12 p.m.
5.18 5.35 6.27 7.98 7.78 10.54 8.49 11.91 11.84 11.37 6.18 7.12
Level of care 0-2 a.m.
2-4 a.m.
4-6 a.m.
6-8 a.m.
8-10 a.m.
10-12 a.m.
12-2 p.m.
2-4 p.m.
4-6 p.m.
6-8 p.m.
8-10 p.m.
10-12 p.m.
0 1 2 3
5.81 5.81 4.77 3.43
0 1 2 3
6.15 5.96 4.92 3.64
0 1 2 3
6.24 6.90 6.04 4.43
0 1 2 3
8.35 8.72 7.64 5.75
0 1 2 3
8.12 7.87 7.52 8.14
0 1 2 3
10.28 10.04 10.49 13.30
0 1 2 3
8.23 7.88 8.76 10.38
0 1 2 3
10.13 10.85 12.88 14.30
0 1 2 3
11.35 10.74 12.44 14.68
0 1 2 3
10.70 11.00 11.73 12.15
0 1 2 3
6.79 6.46 6.07 4.81
0 1 2 3
7.85 7.78 6.74 4.98
0
1
2
3
4
5
6
7
8
9
10
11
12
Frequency in % Fig. 1. Distribution of falls by the time of day and level of care.
13
14
15
187.e4
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
Location of falls
Frequency in %
Resident's room
62.26
Bathroom
12.65
Common area
22.05
Outside
3.05
Level of care
Resident's room
Bathroom
Common area
Outside
0
63.03
1
65.08
2
61.32
3
52.66
0
13.53
1
13.86
2
12.08
3
8.68
0
18.10
1
17.77
2
24.16
3
36.37
0
5.34
1
3.29
2
2.45
3
2.29
0
10
20
30
40
50
60
70
Frequency in % Fig. 2. Distribution of falls by the location of falls and level of care.
age-dependent fall rate differed between the genders. In women, fall rates decreased slightly with increasing age categories, whereas the highest fall rates in men were observed in the 2 highest age categories. The level of care was strongly associated with fall risk in women and men. Particularly high fall rates were observed in residents with low and medium care need (level of care 1 and 2) (Table 2). There were 85 nursing homes reporting implausibly low falls rates (fewer than 1 fall per person-year in homes with more than 50 beds). The sensitivity analysis excluding these homes showed a higher fall rate (1.70 and 2.47 falls per person-year in women and men, respectively). Figure 1 demonstrates the distribution of falls by the time of day. Two peaks were observed. One in the later morning between 10 AM and midday and another in the afternoon and the early evening between 2 PM and 8 PM. The rates in the afternoon were statistically still significantly higher than the rates in the morning (10 to 12 AM). In a nursing home with 100 beds, for example, 9 and 12.6 falls can be expected in women and men within 1 year in the early afternoon (2 PM to 4 PM), respectively (Supplement Table 1). The relative contribution to the 2 described peaks increased with increasing level of care, whereas the relative contribution of time-dependent fall rates of residents with less care need was higher during the night (8 PM until 8 AM, see Figure 1).
About 75% of all falls occurred in the residents’ rooms or in the (usually) adjoining bathrooms and only 22% were reported from the common areas (Figure 2). That means that in a nursing home with 100 beds, about 112 and 161 falls can be expected in the residents’ rooms and bathrooms compared with only 33 and 48 falls in the common area in women and men, respectively (Supplement Table 2). The functionally most limited residents had the highest relative contribution to falls in the common area. Not unexpectedly, the relative contribution to falls outside the building decreased with increasing care need (level of care) (Figure 2). Most falls occurred when performing a transfer (41%), 36% occurred when walking, and 23% could either not be classified or occurred during another activity (eg, when sitting). In contrast to the category “transfers,” the relative contribution to falls when walking decreased strongly with increasing care need (level of care) (Figure 3 and Supplement Table 3). The day of the week (eg, weekend compared with weekday) or the season had no effect on fall rates (data not shown).
Discussion The study presents detailed descriptive data about falls and fall rates in nursing homes. We found lower fall rates in women than in
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
187.e5
Fall occurred by
Frequency in %
Walking
36.30
Transfer
41.13
Others
22.58
Level of care
Walking
Transfer
Others
0
43.27
1
39.47
2
33.50
3
27.07
0
35.83
1
40.09
2
43.40
3
41.12
0
20.90
1
20.45
2
23.10
3
31.81
0
10
20
30
40
50
Frequency in % Fig. 3. Distribution of falls by activity before the fall and level of care.
men and decreasing fall rates with increasing care need (level of care). Prevalence of falls peaked in the morning between 10 AM and midday and in the afternoon and the early evening between 2 PM and 8 PM. Most of the falls occurred in the residents’ rooms and only less than a quarter were observed in the homes’ common areas. Residents with less functional limitations contributed more to falls during the night, to falls in their own room, and to falls when walking, whereas those with more functional limitations (level of care 3) contributed relatively more to falls during the day, to falls in the common area, and to falls when performing a transfer. The reported fall rate in this study is probably an underestimation of the true fall rate for several reasons. First, a maximum of 1 fall per resident per day was reported. Therefore, some falls from frequent fallers may have been missed. Second, some of the falls may not have been observed by the staff and therefore were not reported. This may be particularly the case for outdoor falls with a selective reporting of injurious falls. Third, the data were assessed as part of an intervention to reduce falls, which may have influenced the fall rate. Fourth, falls may have been observed but not reported by the staff because of high workload or other causes. This was the reason why we performed a sensitivity analysis excluding nursing homes with unrealistically low numbers of falls. Therefore, the higher fall rate calculated by sensitivity analysis may be closer to the true fall rate. A Finnish study reported a 1.6 times higher fall incidence in men than in women,8 which comes close to the relative incidence observed in our data. The difference is partly because men are
categorized more often in level of care 0 (high fall rates) and less often in level of care 3 (low fall rates). This may reflect differing patterns of disability in men and women, owing to women experiencing longer periods of severe disability before death. However, even within the same levels of care, fall rates are significantly higher in men than in women. The reason for the observed difference is not clear. The male nursing home population may be somewhat different from the female population in the way that men are frequently cared for at home by their wives as long as possible, whereas similar support is less common for women with disabilities. Therefore, the reason and the functional status at admission to a nursing home may be somewhat different between women and men, which may be only incompletely expressed by the categorization in the different levels of care. However, this is not a sufficient explanation for the observed difference. It also remains unclear why fall rates decreased slightly with increasing age categories in women, whereas the opposite was the case in men. Although the assessment of care need used in this study was based on a qualitative judgment made by a physician or a nurse, the method has been shown to have good levels of inter-rater reliability.19 The need for care is strongly associated with functional impairment, in particular mobility and comorbidity.20 It was not unexpected that the risk was lowest in residents with low care need (level 0) and with high care need (level 3). The first group is the less frail group with probably the fewest walking disorders, whereas the latter group is mainly bedridden and therefore less exposed to risky situations that result in falls.
187.e6
K. Rapp et al. / JAMDA 13 (2012) 187.e1e187.e6
The time-dependency of the falls may be because of an increased activity during specific time periods. Our observation of a maximum of falls in the afternoon was also reported by other authors.4,7,21 Why fall rates in the afternoon exceed those in the morning remains unclear. Potential reasons could be that residents are more tired later in the day, higher activity patterns, or less availability of the nursing staff, for example because of shift changes during the afternoon. The lowest fall rates were observed during the night. However, time of exposure during the night, such as from going to and returning from the bathroom, is extremely short. Therefore, fall rates per step are very high, which means that activities during the night are probably high-risk situations. The vast majority of falls occurred in the residents’ rooms. This is in line with results from previous studies.4,14 It is interesting that people with high care needs have a higher relative risk to fall during daytime in common areas. An explanation may be that these very frail residents are brought to common areas to allow them to participate in the facility’s daily life but cannot be supervised continuously. In the analyzed population, the importance of transfers as an activity leading to falls is even higher than that related to walking. This has implications for prevention. Specific training of transfers may be worthwhile because of the very high risk of falls in nursing home residents who are able to stand but then are unable to maintain an upright posture.22 Another problem may be orthostatic hypotension in this population, with a high proportion of autonomic dysfunction and polypharmacy.2 Therefore, drugs influencing blood pressure should be prescribed very cautiously to residents of nursing homes. Because falls in the resident’s room are of greatest relevance, environmental aspects in the residents’ private space can be another target for preventive measures. Individual aspects such as walking patterns or transfer heights could be considered to reduce person-environment mismatches on an individual level. Efforts to improve incontinence, such as scheduled nocturnal prompted voiding or medical interventions for urge incontinence, may be particularly effective in reducing falls during the night.1 Our data show that the resident’s functional status influences his or her fall risk. Therefore, fall rates depend also on the case mix of the observed nursing home populations. The type of institutional care differs between different countries, which influences the case mix and may compromise the external validity of our reported fall rates. The calculated fall rates, however, fit very well to previously reported fall rates from nursing homes from different countries and continents.2,5,10e12,23 The total number of person-years under exposure could only be estimated. The analyzed residents, however, represented about 43% of all person-years of residential care in older people in Bavaria.24 Information about each home’s capacity and valid data regarding the overall occupancy in Bavarian nursing homes were available. Furthermore, representative data from a subgroup of 48 nursing homes about the distribution of sex, age, and the level of care were available. Therefore, we believe that our estimation of person-years is very close to the real number of person-years under exposure. Our study is the largest reported descriptive evaluation of falls and fall rates in residential care. The large number of falls allowed us to stratify the analyses in a detailed way for different variables of interest like sex, age, and even for very close time intervals. An additional strength of the study was the availability of the degree of care need, which allowed us to perform additional analyses stratified by the residents’ functional status. In summary, we examined falls and fall rates in residents from nursing homes. In this setting, falls and fall rates differed between women and men, between different places, between different
activities, between different degrees of care need, and within the course of the day. Acknowledgments We thank Regina Merk-Bäuml, Ralf Brum, Markus Gindl, and Stefanie Büttner from the Allgemeine Ortskrankenkasse (AOK), and Ulrich Rissmann from the Robert-Bosch-Hospital Stuttgart. Furthermore, we thank Andrea Kleiner from the Institute of Epidemiology and Medical Biometry, Ulm University, for the preparation of the figures. Supplementary Data Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jamda.2011.06.011 References 1. Becker C, Rapp K. Fall prevention in nursing homes. Clin Geriatr Med 2010;26: 693e704. 2. Rubenstein LZ, Josephson KR, Robbins AS. Falls in the nursing home. Ann Intern Med 1994;121:442e451. 3. Rubenstein LZ. Falls in older people: Epidemiology, risk factors and strategies for prevention. Age Ageing 2006;35:ii37eii41. 4. Lester P, Haq M, Vadnerkar A, et al. Falls in the nursing home setting: Does time matter? J Am Med Dir Assoc 2008;9:684e686. 5. Gibson RE, Harden M, Byles J, et al. Incidence of falls and fall-related outcomes among people in aged-care facilities in the Lower Hunter region, NSW. N S W Public Health Bull 2008;19:166e169. 6. Heinze C, Halfens RJ, Dassen T. Falls in German in-patients and residents over 65 years of age. J Clin Nurs 2007;16:495e501. 7. Jensen J, Lundin-Olsson L, Nyberg L, et al. Falls among frail older people in residential care. Scand J Public Health 2002;30:54e61. 8. Nurmi I, Luthje P. Incidence and costs of falls and fall injuries among elderly in institutional care. Scand J Prim Health Care 2002;20:118e122. 9. Jensen J, Lundin-Olsson L, Nyberg L, et al. Fall and injury prevention in older people living in residential care facilities. A cluster randomized trial. Ann Intern Med 2002;136:733e741. 10. Becker C, Kron M, Lindemann U, et al. Effectiveness of a multifaceted intervention on falls in nursing home residents. J Am Geriatr Soc 2003;51:306e313. 11. Kerse N, Butler M, Robinson E, et al. Fall prevention in residential care: A cluster, randomized, controlled trial. J Am Geriatr Soc 2004;52:524e531. 12. Dyer CA, Taylor GJ, Reed M, et al. Falls prevention in residential care homes: A randomised controlled trial. Age Ageing 2004;33:596e602. 13. Cacha CA. An analysis of the 1976 incident reports of the Carillon Nursing Home. J Am Health Care Assoc 1979;5:29e33. 14. Jantti PO, Pyykko VI, Hervonen AL. Falls among elderly nursing home residents. Public Health 1993;107:89e96. 15. Ashley MJ, Gryfe CI, Amies A. A longitudinal study of falls in an elderly population II. Some circumstances of falling. Age Ageing 1977;6:211e220. 16. Becker C, Leistner K, Nikolaus T. Introducing a statutory insurance system for long-term care (Pflegeversicherung) in Germany. In: Michel JP, Rubenstein LZ, Vellas BJ, et al., editors. Geriatric Programs and Departments Around the World. Paris: Serd-Springer; 1998. p. 55e64. 17. Statistische Ämter des Bundes und der Länder: Statistik über die Empfänger von Pflegeleistungen, Pflegebedürftige nach Leistungsart und Geschlecht, Stichtag, regionale Tiefe: Reg.-Bez./Stat. Available at: https://www.regionalstatistik.de/ genesis/online. Accessed December 17, 2010. 18. Klenk J, Kurrle S, Rissmann U, et al. Availability and use of hip protectors in residents of nursing homes. Osteoporos Int 2011;22:1593e1598. 19. Kliebsch U, Brenner H. [Interrater reliability of expert assessment by the health insurance medical service in determining eligibility for disability benefits]. Gesundheitswesen 1995;57:638e644. 20. Fried LP, Guralnik JM. Disability in older adults: Evidence regarding significance, etiology, and risk. J Am Geriatr Soc 1997;45:92e100. 21. Svensson ML, Rundgren A, Larsson M, et al. Accidents in the institutionalized elderly: injuries and consequences. Aging (Milano ) 1992;4:125e133. 22. Lord SR, March LM, Cameron ID, et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided. J Am Geriatr Soc 2003;51:1645e1650. 23. Rubenstein LZ, Robbins AS, Josephson KR, et al. The value of assessing falls in an elderly population. A randomized clinical trial. Ann Intern Med 1990;113: 308e316. 24. Bayerisches Landesamt für Statistik und Datenverarbeitung: Statistische Berichte. Pflegeeinrichtungen und Pflegegeldempfänger in Bayern. Ergebnisse der Pflegestatistik. Available at: http://www.statistik.bayern.de/statistik/. Accessed December 31, 2007.