/clinical/,/clinical/cckm-tools/,/clinical/cckm-tools/content/,/clinical/cckm-tools/content/questionnaires/,/clinical/cckm-tools/content/questionnaires/related/,

/clinical/cckm-tools/content/questionnaires/related/name-97161-en.cckm

201606168

page

100

UWHC,UWMF,

Tools,

Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

A Validation Study of the MAHC Fall Risk Assessement Tool

A Validation Study of the MAHC Fall Risk Assessement Tool - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related


Home Health Care Management & Practice
XX(X) 1 –6
2012 SAGE Publications
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1084822312457942
http://hhcmp.sagepub.com
Introduction
Falls are a major health concern for community-based dwellers
above the age of 65. A fall may be defined as “an event that
results in a person coming to rest unintentionally on the ground
or other level, not as the result of a major intrinsic event or
overwhelming hazard.”
1-3
Falls are responsible for two thirds
of all unintentional injuries leading to death and are the leading
cause of preventable injury in older adults.
4
Thirty percent of
community dwelling people aged 65 and older fall at least once
a year while 15% sustain multiple falls.
5
Since these are com-
munity-based home dwellers, these figures are probably lower
than actuality because of underreporting. In the United States,
falls suffered by people above 65 years of age that required
medical attention resulted in more than $19 billion in medical
costs.
6
The average cost of a fall requiring hospitalization was
found to be $26,483 in a recent review.
6
Serious injuries from
falls include hip, spine, and other bone fractures as well as
traumatic brain injury.
7
Other effects include reduced mobility,
activity, and physical fitness due to fear of falling, which para-
doxically increases the likelihood of another fall, reducing
mobility and quality of life.
8
Because falls are a serious and expensive national health
problem, accurate fall risk assessment is a critical component
of the evaluation of the home health patient. The OASIS-C
Guidance Manual (Item M1910, Responses 1 and 2) states,
The multi-factor falls risk assessment must include at
least one standardized tool that 1) has been scientifically
tested in a population with characteristics similar to that
of the patient being assessed (for example, community-
dwelling elders, non-institutionalized adults with dis-
abilities, etc.) and shown to be effective in identifying
people at-risk for falls; and 2) include a standard
response scale. The standardized tool must be both
appropriate for the patient based on their cognitive and
physical status, and appropriately administered as indi-
cated in the instructions.
9
For a fall risk assessment tool to meet OASIS-C guide-
lines, it must meet three criteria. The tool must be
1. Multifactorial: The tool needs to include more than
one contributing factor to fall risk. This includes
but is not limited to balance issues, previous falls,
environmental concerns, medications, inconti-
nence, and cognitive deficits.
2. Standardized: The tool has to be administered
and scored the same way every time it is used and
according to the directions of the authors of the tool.
457942HHCXXX10.1177/1084822312457942Ho
me Health Care Management & PracticeCalys et al.
1
North Kansas City Hospital, North Kansas City, MO, USA
2
University of Kansas Medical Center, Kansas City, KS, USA
Corresponding Author:
Mary Calys, DPT, PT, Cancer Rehabilitation and Fatigue Management,
North Kansas City Hospital, 2800 Clay Edwards Drive, MPN 210, North
Kansas City, MO 64116, USA
Email: mary.calys@nkch.org
A Validation Study of the Missouri
Alliance for Home Care Fall Risk
Assessment Tool
Mary Calys, DPT, PT
1
, Kendra Gagnon, PhD, PT
2
, and
Stephen Jernigan, PhD, PT
2
Abstract
A retrospective review (n = 2,247) was conducted from July 1st to October 1st 2010 to determine validity of the Missouri
Alliance for Home Care Fall Risk Assessment Tool (MAHC-10). Fall risk was identified by a MAHC-10 score > 4. Two sample
t test, chi-square, logistic regression, and ROC curve analyses were performed. Fallers (6.35 ± 1.7, n = 195) and nonfallers
(5.70 ± 1.9, n = 2,052) had significantly different (p = .011) MAHC-10 scores. The MAHC-10 cutoff score of 4 demonstrated
96.9% sensitivity and 13.3% specificity; however, ROC curve analyses revealed a cutoff score of 6 maximized combined
sensitivity and specificity. The MAHC-10 is valid for fall risk screening in the home health setting; however, a cutoff score of 6
may more accurately predict fall risk.
Keywords
fall risk, fall risk assessment, home health care, validity, sensitivity, specificity, community dwelling elders
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from

2 Home Health Care Management & Practice XX(X)
3. Validated: The tool and the score it generates have
to be confirmed as having a real relationship to
fall risk by means of outside testing on the target
population. In the case of home health, that would
be community dwelling elders. There needs to be
research to support the tool findings.
10
As of today, the industry has yet to identify a single falls
risk assessment tool that satisfies all three criteria. As a
result, most agencies have incorporated several tools as an
alternative method to meet OASIS-C criteria. A single multi-
factorial, standardized, and validated tool would provide the
clinician with a quick and efficient way to assess fall risk in
the home health patient while meeting OASIS-C guidelines.
The Missouri Alliance for Home Care Fall Risk Assessment
Tool (MAHC-10) has the potential to fill that role.
The MAHC-10 Fall Risk Assessment provides an easy-
to-use, evidence-based starting point for home health clini-
cians to address fall prevention in community dwelling
patients. The MAHC-10 is multifactorial. It measures patient
fall risk by assessing 10 core elements including age, diag-
nosis, fall history, incontinence, visual impairment, impaired
functional mobility, environmental hazards, polypharmacy,
cognitive impairment, and pain. The MAHC-10 is standard-
ized. The MAHC-10 is administered in the same way for
each assessment using a standard form. MAHC-10 adminis-
trators must be trained in the 10 core elements and how they
are assessed. However, the MAHC-10 tool falls short of
OASIS-C guidelines because the validity of the tool has not
yet been established. Thus the purpose of this research was
to determine the validity of the MAHC-10 Fall Risk
Assessment.
Method
This study is a retrospective review of all fall risk assess-
ments collected by nine Missouri home health agencies over
a 4-month period per their normal procedure for participa-
tion in the MAHC Falls Reduction Benchmarking Study.
Inclusion criteria for this retrospective review were all home
health patients initially assessed for fall risk from July 1 to
October 31, 2010. Falls recorded for these subjects up to 60
days after their initial fall risk assessment (through December
31, 2010 for subjects initially assessed at the end of October)
were also reviewed. Hospice patients and mother/pediatric
patients were excluded from this review.
General procedures
As part of the normal procedure for the MAHC Falls Reduction
Benchmarking Study, all patients who were admitted to one of
the nine Missouri home health agencies participating in the
study were administered the MAHC-10 by a trained staff mem-
ber on admission or readmission. The patient’s response for
each of the 10 core elements and a total score for the MAHC-10
were recorded. Patients with a MAHC-10 score ≥4 were con-
sidered to be “at-risk” for falls, while patients with a score <4
were considered to be “not at-risk” for falls. Each fall within 60
days of the initial MAHC-10 assessment—including the date of
the fall, whether the fall required emergent care, and whether
the fall resulted in a hip fracture—was recorded for every
patient. Because falls were recorded for each patient for 60 days
postadmission, falls data through December 2010 were
reviewed and analyzed. For the purposes of this review, any
patient who experienced at least one fall within the 60-day
period after admission was considered a “faller,” while a sub-
ject who did not fall during that 60-day period was classified as
a “nonfaller.” If a subject was readmitted during the 4-month
review period, only data from his or her first admission and the
subsequent 60-day period were included in the analysis.
Each patient was assigned an identification number and all
data were recorded by patient number on a standard format-
ted Excel spreadsheet. On IRB approval of this retrospective
review, all fall risk assessment data were de-identified, coded,
and entered into a secure database for analysis.
Data analysis
Descriptive characteristics. Patient age and gender were
recorded at time of initial admission. A two-sample t test was
used to determine whether there was a significant difference
in the age of fallers and nonfallers while a chi-square test
was used to determine between-group gender differences.
Construct validity. Construct validity refers to the ability of
a test to measure what it is intended to measure. In this study,
construct validity of the MAHC-10 was assessed by deter-
mining whether raw scores and risk classifications differenti-
ated between fallers and nonfallers. A two-sample t test was
used to determine if there was a significant difference between
MAHC-10 scores of fallers and nonfallers. A chi-square test
was used to determine whether a significantly larger propor-
tion of fallers were classified as “at-risk” for falls using the
MAHC-10. Significance was set at p < .05.
Predictive validity, sensitivity, and specificity. To evaluate the
predictive validity of the MAHC-10, a chi-square test was
used to determine if a significantly larger proportion of at-risk
patients experienced at least one fall within 60 days of admis-
sion than not at-risk patients. Predictive validity was further
assessed by determining if the 10 core elements of the
MAHC-10 were significantly related to fall status. Logistic
regression was used to determine whether age, diagnosis, fall
history, incontinence, visual impairment, impaired functional
mobility, environmental hazards, polypharmacy, and/or pain
were significantly related to whether or not an individual was
a faller or nonfaller. Significance was set at p < .05.
Sensitivity is defined as a tool’s ability to obtain a true posi-
tive result, while specificity refers to a tool’s ability to obtain a
true negative result. For this study, a 2 × 2 contingency table
was constructed to determine whether the MAHC-10 was a
sensitive (classified a subject as “at-risk” who was actually a
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from

Calys et al. 3
faller) or specific (classified a subject as “not at-risk” who was
actually a nonfaller) tool. To determine the sensitivity and
specificity of the MAHC-10 for different cutoff scores, a
Receiver Operating Characteristic (ROC) curve was used. An
ROC curve is a graphical representation of sensitivity and
specificity for various cutoff scores associated with an assess-
ment tool, in this case, the MAHC-10.
Results
A total of 2,247 subjects met the inclusion criteria during the
4-month admission window. Eight hundred sixty-three (38.4%)
subjects were male and 1,385 (61.6%) were female. The aver-
age age was 71.9 years (range 18-103, standard deviation 15.2
years). There were 195 fallers (8.7%) and 2,056 nonfallers
(91.3%). The average MAHC-10 score for nonfallers was 5.70
(SD 1.95) and for fallers was 6.35 (SD 1.65). Descriptive char-
acteristics for fallers and nonfallers are included in Table 1.
Construct Validity
A two-sample t test revealed a significant difference between
MAHC-10 raw scores for fallers and nonfallers, as seen in
Table 1 (p = .011). Using cross-tabulation (Table 2) with a chi-
square test, subjects who were classified as at-risk for falls with
a score of 4 or greater on the MAHC-10 were significantly
more likely to be fallers than subjects who were classified as not
at-risk for falls with a MAHC-10 score of <4 (p < .001).
Predictive Validity, Sensitivity, and Specificity
Using backward logistic regression with faller/nonfaller as the
response variable, four items included in the MACH-10
assessment were significantly related to faller status (p < .10):
functional mobility (p = .03, odds ratio 1.98), three or more
coexisting diagnoses (p = .03, odds ratio 1.83), history of falls
within 3 months (p < .001, odds ratio = 2.07), and pain limit-
ing function (p = .08, odds ratio 0.76). Using a MAHC-10
cutoff score of 4, sensitivity was 96.9% and specificity was
13.3%. The ROC curve analysis indicated that different cutoff
scores resulted in different sensitivities and specificities
(Table 3). The ROC curve for the MAHC-10 Fall Risk
Assessment tool is depicted in Figure 1; the points on the
curve that correspond to the sensitivity and 1-specificity val-
ues associated with each possible score on the MAHC-10 are
labeled. A cutoff score of 6 on the MAHC-10 maximized
combined sensitivity (68.7%) and specificity (46.9%).
Discussion
The purpose of this study was to determine the validity of the
MAHC-10 Fall Risk Assessment tool for fall risk screening.
The results of this study contribute significantly to the litera-
ture regarding this tool and inform users about the tool’s
utility for fall risk screening in the home health care setting.
It has been suggested that a score of 4 or greater on the
MAHC-10 Fall Risk Assessment tool be used to indicate that
someone is “at-risk” for falls. Based on the current study, this
cutoff score resulted in a sensitivity of 96.9%, which means
that this same percentage of people who fell during the 60
days after admission were accurately identified as being “at-
risk” for falls. However, paired with this high sensitivity was a
low specificity of only 13.3%. This low specificity resulted in
1,780 of the 2,052 people who did not fall during the 60 days
after admission, being identified as having fall risk. While
high sensitivity is valued when screening for fall risk, such
low specificity may or may not be acceptable to a specific
Table 1. Descriptive Characteristics of Fallers and Nonfallers
Nonfaller (n = 2,052) Faller (n = 195) p value
Age (years)
X 71.6 74.6 .001*
SD 15.4 13.2
Range 18-103 38-97
Gender
Male 782 (38.1%) 81 (41.5%) 0.228
Female 1,270 (61.9%) 114 (58.5%)
MAHC-10 score
X 5.70 6.35 0.011*
SD 1.95 1.65
Note: Independent t tests were used to assess group differences for age and MACH-10 score. A chi-square test was used to assess group differences for
gender. X = mean. SD = standard deviation. MAHC-10 = Missouri Alliance for Home Care Fall Risk Assessment Tool.
Significance was set at p < .05.
Table 2. Cross-Tabulation: Risk Status × Faller Status
Nonfaller Faller Total
Not at-risk (MAHC-10 >4) 272 6 278
At risk (MAHC-10 ≤ 4) 1,780 189 1,969
Total 2,052 195 2,247
Note: MAHC-10 = Missouri Alliance for Home Care Fall Risk Assessment
Tool.
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from

4 Home Health Care Management & Practice XX(X)
clinic or agency depending on its goals. The disproportionate
ratio of sensitivity and specificity associated with this tool
may increase the cost, resources, and burden in home health
care to provide services to those identified as having fall risk
that perhaps do not need fall-related interventions. If an agency
has the goal of identifying all fallers, without concern for
overidentifying people as having fall risk who actually are not
at-risk, specificity may not be a concern. However, if an
agency wants to conduct a more thorough fall risk assessment
or implement a fall prevention program that requires signifi-
cant resources, based on the results of this screening tool, low
specificity may be undesirable. The goal associated with using
the MAHC-10 Fall Risk Assessment tool and the available
resources may significantly influence the utility of this tool for
a specific clinic or agency.
In fall risk assessment tool research, cutoff scores have been
modified to find a balance in sensitivity and specificity of a fall
risk assessment tool for specific populations of patients.
11-13
In
an effort to inform those who use the MAHC-10 Fall Risk
Assessment tool, resulting sensitivities and specificities for dif-
ferent cutoff scores were reported as part of this study (see
Table 3). As the cutoff score approached 10, the sensitivity
decreased and the specificity increased. At higher cutoff scores,
more people who fell were not identified as having fall risk;
there were more false negatives. However, at these higher cut-
off scores, there were also fewer false positives meaning that
fewer people were identified as having fall risk when they had
not fallen. A cutoff score of 6 maximized combined sensitivity
(68.7%) and specificity (46.8%) over any other cutoff score.
Again, depending on the purpose for which an agency or orga-
nization uses the MAHC-10 Fall Risk Assessment tool, differ-
ent cutoff scores may achieve different purposes.
The MAHC-10 has construct validity in that it is able to
discriminate between fallers and nonfallers, as defined by the
study. The difference in the average MAHC-10 scores between
the two groups is significant although this difference may not
be clinically meaningful. On the MAHC-10 scale, there is
1 point difference between each possible score and the differ-
ence in the average scores between the two groups is only
0.65. This difference in average scores between fallers and
nonfallers would not be detectable when using the MAHC-10.
Interestingly, the MAHC-10 score that falls between the group
averages is 6, which is also the score that maximized sensitiv-
ity and specificity for the tool using the ROC curves.
The logistic regression analysis revealed that 4 factors con-
tributed significantly to faller status, including a history of
falls within the previous 3 months, impaired functional mobil-
ity, three or more coexisting diagnoses, and pain affecting the
level of function. It has been reported previously that one of
the primary risk factors for falls is a history of falls.
14
Impaired
functional mobility on the MAHC-10 is defined broadly,
including the need for help with activities of daily living or
instrumental activities of daily living, gait or transfer prob-
lems, arthritis, pain, fear of falling, foot problems, impaired
sensation, impaired coordination, or improper use of assistive
devices. Many of these factors are known risk factors for falls
and therefore, this finding is consistent with current fall risk
literature.
14,15
Other studies have also indicated that the num-
ber of coexisting diagnoses or comorbidities can influence fall
risk although in the study by Lee and Stokic,
16
high fall risk
was noted only in those with nine or more comorbidities.
While pain affecting level of function was also significantly
related to faller status, we are concerned that this overlaps
with the “impaired functional mobility” item of the MAHC-
10, which was defined above. When interpreting logistic
regression, it is important that one understands that these
4 factors, if included with 6 other variables than what are
included in the MAHC-10, may not again be significantly
related to faller status. However, based on this sample of
patients, these 4 factors warrant further investigation. From a
fall risk perspective, it is important to note that impaired func-
tional mobility, pain, and coexisting diagnoses are potentially
modifiable and if treated, may influence a patient’s fall risk
although further studies would need to be conducted to con-
firm this possibility as well.
Table 3. Sensitivity and Specificity of Cutoff Scores for the
Missouri Alliance for Home Care Fall Risk Assessment Tool
(MAHC-10)
MAHC-10 cutoff score Sensitivity Specificity
1 100% 0.2%
2 100% 1.2%
3 100% 4.2%
4 96.9% 13.3%
5 85.6% 26.9%
6 68.7% 46.9%
7 45.1% 67.6%
8 24.6% 81.3%
9 11.3% 91.0%
10 2.6% 97.6%
1
4
5
6
7
8
9
10
0%
20%
40%
60%
80%
100%
0% 20% 40% 60% 80% 100%
S
e
n
s
i
t
i
v
i
t
y
1-Specificity
3
2
Figure 1. Receiver Operating Characteristic (ROC) curve for
the Missouri Alliance for Home Care Fall Risk Assessment Tool
(MAHC-10)
Note: This curve depicts scores (labeled) on the MAHC-10 and the cor-
responding sensitivity and 1-specificity percentages.
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from

Calys et al. 5
While this study provides information about the utility of
the MAHC-10, the results should be interpreted in light of the
study’s limitations. When applying study results to another
population, it should be noted that this study included people
of ages between 18 and 103. While this large age range might
be beneficial for generalizability, no subanalyses were con-
ducted to determine applicability to specific age groups (i.e.,
65 years and older) or those that are most often seen in the
home health care setting. Additional studies that investigate
the use of the MAHC-10 in specific populations or more
homogeneous groups of patients may be beneficial for
improved external validity. Although this study used prospec-
tive methods for classifying fallers and nonfallers, only a
60-day period was used to determine the classification. Many
fall risk studies, although most are retrospective, use longer
time periods (i.e., 6 months to 1 year) to determine fall sta-
tus.
11,13,17-19
It is possible that some of those who were classi-
fied as nonfallers could have fallen after the 60-day period;
these falls would not have been captured. This may help to
explain why many nonfallers were identified as having fall
risk, which relates to the low specificity. While the prospec-
tive design of this study is useful for determining predictive
validity of the MAHC-10 tool, additional longitudinal studies
that employ longer time frames need to be conducted to be
able to more accurately determine the tool’s predictive valid-
ity. Other future studies could also include developing a
shortened form of the MAHC-10 tool, in particular as it
relates to those items that consistently predict faller status
versus those that do not seem to be predictive, and studies to
determine whether or not the MAHC-10 is sensitive to
changes in fall risk status.
In conclusion, the MAHC-10 fall risk assessment tool
appears to be a useful fall risk screening tool with good sensi-
tivity at the currently recommended cutoff score of four; how-
ever, specificity is poor. When data is aggregated, the tool also
demonstrates construct validity. It is important for a clinic or
agency to consider how closely their patients match the sample
used in this study, to determine generalizability of these results.
In addition, because fall risk is multifactorial in nature, it is
imperative that clinicians do not use a fall risk screening tool,
such as the MAHC-10, as the end point for fall risk assessment,
especially when fall risk is identified. When fall risk is identi-
fied, a more comprehensive fall risk assessment needs to be
conducted to help guide appropriate interventions.
Acknowledgments
The authors thank Samuel Park, SPT, Emily Pierce, SPT, and
Allison Prewitt, SPT for their assistance with data management and
interpretation.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
References
1. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls
among elderly persons living in the community. N Engl J Med.
1988;319:1701-1707. http://www.ncbi.nlm.nih.gov/pubmed/
3205267. Accessed August 4, 2012.
2. Tilling LM, Darawil K, Britton M. Falls as a complication
of diabetes mellitus in older people. J Diabetes Complica-
tions. 2006;20:158-162. http://www.ncbi.nlm.nih.gov/pubmed/
16632235. Accessed August 4, 2012.
3. Visser M, Marinus J, Bloem B, et al. Clinical tests for the evaluation
of postural instability in patients with Parkinson’s disease. Arch
Phys Med Rehabil. 2003;84:1669-1674. http://www.sciencedirect
.com/science/article/pii/S0003999303003484. Accessed May 23,
2012.
4. Fortinsky R, Baker D, Gottschalk M, et al. Extent of imple-
mentation of evidence-based fall prevention practices for older
patients in home health care. J Am Geriatr Soc. 2008;56:737-743.
http://onlinelibrary.wiley.com/doi/10.1111/j.1532-5415.2007
.01630.x/full. Accessed May 23, 2012.
5. Leclerc BS, Bégin C, Cadieux E, et al. A classification and
regression tree for predicting recurrent falling among com-
munity-dwelling seniors using home-care services. Can J Pub-
lic Health. 2009;100:263-267. http://www.ncbi.nlm.nih.gov/
pubmed/19722338. Accessed August 4, 2012.
6. Hanley A, Silke C, Murphy J. Community-based health efforts
for the prevention of falls in the elderly. Clin Interv Aging.
2011;6(1):19-25. http://www.ncbi.nlm.nih.gov/pmc/articles/
PMC3066249/. Accessed May 23, 2012.
7. Centers for Disease Control. Falls among older adults: an over-
view. 2012. http://www.cdc.gov/ncipc/factsheets/adultfalls.htm.
Accessed May 20, 2012.
8. Evitt C, Quigley P. Fear of falling in older adults: a guide
to its prevalence, risk factors, and consequences. Rehabil
Nurs. 2004;29:207-210. http://www.ncbi.nlm.nih.gov/pubmed/
15597999. Accessed May 23, 2012.
9. Centers for Medicare and Medicaid Services. OASIS-C
Guidance Manual. 2011. http://www.oasisanswers.com/_blog/
Simplifying_the_Complex/post/CMS_posts_newly_revised_
OASIS-C_Guidance_Manual!_/. Accessed May 23, 2012.
10. Fazzi Associates. Fall risk assessment and OASIS-C. http://
www.fazzi.com/BE_fall_risk_assessment.html. Accessed May
20, 2012.
11. Dibble LE, Christensen J, Ballard DJ, Foreman KB. Diagno-
sis of fall risk in Parkinson disease: an analysis of individual
and collective clinical balance test interpretation. Phys Ther.
2008;88:323-332.
12. Shumway-Cook A, Baldwin M, Polissar NL, Gruber W. Pre-
dicting the probability for falls in community-dwelling older
adults. Phys Ther. 1997;77:812-819.
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from

6 Home Health Care Management & Practice XX(X)
13. Behrman AL, Light KE, Flynn SM, Thigpen MT. Is the func-
tional reach test useful for identifying falls risk among indi-
viduals with Parkinson’s disease? Arch Phys Med Rehabil.
2002;83:538-542. http://linkinghub.elsevier.com/retrieve/pii/
S0003999302640043. Accessed August 4, 2012.
14. Rubenstein LZ, Josephson KR. The epidemiology of falls and
syncope. Clin Geriatr Med. 2002;18:141-158.
15. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls
among elderly persons living in the community. N Engl J Med.
1988;319:1701-1707. http://www.ncbi.nlm.nih.gov/pubmed/
3205267. Accessed August 4, 2012.
16. Lee JE, Stokic DS. Risk factors for falls during inpatient reha-
bilitation. Am J Phys Med Rehabil. 2008;87:341-350; quiz 351,
422. http://www.ncbi.nlm.nih.gov/pubmed/18427218. Accessed
August 4, 2012.
17. Ashburn A, Stack E, Pickering RM, Ward CD. Predict-
ing fallers in a community-based sample of people with
Parkinson’s disease. Gerontology. 2001;47:277-281.
http://dx.doi.org/10.1159/000052812. Accessed August
4, 2012.
18. Arnold CM, Faulkner RA. The history of falls and the associa-
tion of the timed up and go test to falls and near-falls in older
adults with hip osteoarthritis. BMC Geriatr. 2007;7:17.
19. Thrane G, Joakimsen RM, Thornquist E. The association
between timed up and go test and history of falls: the Tromso
study. BMC Geriatr. 2007;7:1.
at UNIV OF WISCONSIN-MADISON on February 16, 2015hhc.sagepub.comDownloaded from