PREDICTING WALKING ABILITY AFTER AMPUTATION 127
Similar to the original Sansam et al. SR, a standardized
CINAHL
Embase
checklist was used to extract each report’s methods, (n = 107) (n = 130) Cochrane (n = 216)
PubMed
(n = 23)
population, outcome measures, and predictive factors Identification
(5). Additionally, the UK National Service Frame- Records retrieved from searches
work for Long-term Conditions (3,9) was used to (n = 476)
assess the quality of each study, as it allows assess-
ment of quality in non-randomized cohort studies. Screening Records retrieved from Duplicates removed
searches
The reports and data extracted were verified by at (n = 476) (n = 157)
least two independent authors who agreed on final
Articles’ titles and
scoring and data extraction. The International Clas- abstracts screened Records excluded
sification of Functioning, Disability and Health (4) Eligibility (n = 319) (n = 220)
was used to present the predictive factors identified Full-text articles
Full-text articles
from these studies. Following study evaluation and assessed for eligibility excluded, after full
data extraction, factors predictive of walking ability (n = 99) article review
(n = 78)
following LEA were aggregated and compared nar- Studies included in
ratively with the findings of the original Sansam et Included quantitative synthesis
(meta-analysis)
al. SR. (n = 21)
RESULTS Figure 1. PRISMA 2009 flow diagram. Reprinted with permis-
sion from PLoS Medicine (Moher D, Liberati A, Tetzlaff J, Al-
Number of Identified Studies tman DG, The PRISMA Group. Preferred Reporting tems for
Systematic Reviews and Meta-Analyses: The PRISMA Statement.
A total of 319 unique studies were identified PLoS Med. 2009;6(7):e1000097), copyright 2009. For more infor-
through the electronic search. Of these, 298 were mation, visit www.prisma-statement.org.
eliminated, leaving a total of 21 for full evaluation
(Figure 1). (BMI) of 30.2 kg/m (median: 31.3 (IQR: 4); range:
2
27.6 to 31.6 kg/m ). Within the described control
2
Description of Sample group of subjects with LEA, the reported etiology was
The original SR from Sansam et al. included a typically PVD with comorbid diabetes mellitus. Their
total (n) of 9,080 subjects (5). Conclusions from this mean age was 61.8 years (median 66.7 (IQR: 11.8);
updated study are drawn from a total recruited sam- range: 46.0 to 67.7 years) and mean BMI was 29.6 kg/
2
ple (n) of 15,207 subjects. A total of 12,410 subjects m ± 6.3. Age and BMI were not different (p > 0.05)
completed the respective studies (18% attrition). between experimental and control subjects with LEA.
There was incomplete and inconsistent reporting of Finally, there was a smaller group of non-amputee,
anthropometric, demographic, and etiologic data; of otherwise healthy control subjects described whose
those studies sufficiently reporting this information, mean age was 49.0 years (median 59.2 years (IQR:
the lower extremity limb loss had the following dis- 35.6); range: 26.1 to 61.7 years) and mean BMI was
2
tribution: 37% peripheral vascular disease (PVD), 25.7 kg/m (Table 1).
27% trauma, 17% diabetic, 12% cancer, 6% infec-
tion, and 2% congenital. Three sub-groups of subjects Settings, Study Designs, and Independent
were included: an experimental group of subjects Variables
with LEA, a control group of subjects with LEA, and The predominant setting for these studies was the
another control group of otherwise healthy non-am- rehabilitation center. These were in varied organiza-
putee controls. Within the experimental group of LEA tions, including university medical centers, Veterans
patients, the subjects described had a mean age of 57.3 Administration hospitals, private sector hospitals,
years (median 60.9 (interquartile range (IQR): 8.5); and skilled nursing facilities. In addition to these,
range: 48.1 to 69.8 years) and a mean body mass index data were also collected from military treatment
128 KAHLE ET AL.
Quality Medium High Medium High Medium Medium High High Medium High High High High Medium High High High High High High High
Table 1. Diabetes Mellitus (DM), Cancer (CA), Infection (I), Peripheral Vascular Disease (PVD), Trauma (T), Congenital (CONG)
Results Older HDAs in good health & w/ low comorbidity prevalence able to successfully walk w/prosthesis in a community. Physical fitness of ≈60% VO2max necessary for older HAD patients to successfully walk. Rehab in CIRU improved mobility success for Veterans undergoing LEA for PVD or DM. No relation to differences in baseline mobility, demographic, psychosocial factors, or # reh
prosthetic fitting. successful prosthetic rehab. outcome when faced w/ subsequent LEA. candidates. during exercise. outcome after TTA. ambulation. Summary of studies included in literature review. Using the UK National Service Framework for Long-term Conditions.(3, 4, 9) Articles were scored out of 10, with up to 2 points awarded for each of the following 5 it
Walking Ability Measures Oxygen uptake AMP, etiology FIM, 6MWT, Stress Arm Ergometer Healing, mortality, 2MWT, TUG, Step activity Modified Houghton Scale, 2 groups: unlimited walking capacity (score=20) & incapacity to walk, regardless of severity (score <20) Patients who could walk ≥ 100 m w/ at most 1 cane = “successful prosthetic users”; otherwise, “failed users”
Total n 7 199 63 160 831 70 64 57 37 72 4727 4357 3691 55 58 58 151 71 118 215 87
Population CA, I PVD, DM PVD PVD, DM, T PVD, NR PVD, T, I, CA, CONG PVD, NR DM PVD, T, CA T, NR PVD, DM, T PVD, I, DM PVD PVD, DM, CA, I T, PVD T, PVD DM PVD T T PVD, DM
First Author Chin (10) Czerniecki (27) Erjavec (11) Erjavec (33) Fortington (30) Gremeaux (24) Hamamura (17) Landry (26) O’Neil (25) Raya (32) Stineman (22) Stineman (21) Sukow (23) van Eijk (20) Wezenberg (28) Wezenberg (14) Wong (19) Artwert (29) Gaunard (31) Linberg (12) Webster (18)
PREDICTING WALKING ABILITY AFTER AMPUTATION 129
facilities, trauma centers, private sector prosthetic and that new factors would potentially emerge as
practices, and university laboratories. Fifty percent important in candidacy determinations. This hypoth-
of the included studies were prospective, 38% were esis was confirmed, as all but five of the previous
retrospective, and 3% were SRs. Cohort and cross-sec- predictive factors were reported in the updated arti-
tional designs were the most common designs, and cles, with 15 of the same predictive characteristics
only two experimental studies were included. The from the original Sansam article recurring. Three
predominant independent variable was LEA. In addi- new predictive factors were identified in this review
tion to this, prosthetic rehabilitation was commonly that were not previously identified in the original
included as treatment. Sansam review (Table 2).
Since the original Sansam et al. article, the follow- This literature review spans the seven years (2007-
ing factors were each supported by a single reference: 2015, 21 studies) following the original Sansam et al.
BMI, motivation, social support, smoking, and phan- article, whereas the original search included 57 years
tom limb condition. The following predictive factors of literature (1950-2007, 57 studies). This updated
were moderately supported (i.e., two references): study increases the size of the original Sansam et al.
independence in activities of daily living (ADL), time report by including 137% more subjects for a total of
to rehabilitation, race, and vascular intervention. The 21,490 between the two articles. However, the authors
following predictive factors were more strongly sup- caution that, due to poor reporting, it is not clear
ported (i.e., three to five references): ability to stand at times if patients are repeat counted in multiple
on one leg, cognition and mood disturbance, gender, publications. Nevertheless, in terms of prosthetic
pre-amputation living status, and cause of amputa- studies, this is a considerably large study relative
tion. Race, vascular intervention, and pre-amputation to other SRs, which tend to include much smaller
living status were newly identified in this report and samples. For example, a recent comprehensive SR
not identified in the original Sansam et al. article. of microprocessor knees based conclusions on 625
The most strongly supported factors (i.e., ≥6 refer- subjects (15). The patients in this SR had predomi-
ences) emerging from the search when considering nantly lost their lower extremities due to PVD, which
prosthetic candidacy were: amputation level, physical is consistent with epidemiologic data (16). Therefore,
fitness, age, and comorbidities. There is increasing it is plausible that the results of this SR would have
agreement that these identified predictive factors are high generalizability to clinical practice. Given the
important when contemplating prosthetic candidacy predominant setting was the rehabilitation center
and walking ability. or major medical centers, results may be particularly
Meta-analysis was not possible, as the studies of relevant within these types of settings.
like outcome measures did not observe the same Predictive Factors in a Single Study in This
homogeneous patient characteristics; mainly, level, Literature Review
etiology, and mean ages were heterogeneous among
these studies (10-14). BMI
Linberg et al. found demographics (i.e., height,
DISCUSSION weight) did not affect the six-minute walk test
The purpose of this study was to extend the body of (6MWT) (12). This is consistent with previous reports
knowledge—using the same search strategy originally in finding that, when adjusting for medical comorbid-
completed in the Sansam et al. article—of predicting ities, age, and sex, BMI was not a significant predictor
walking ability following lower limb amputation. This of walking ability (5).
SR identifies predictive factors of walking ability and Motivation
updates the findings to include current literature. We Hamamura et al., in a high quality study, found
hypothesized that most factors previously identified significance in motivation as a predictive factor for
as important or predictive in determining prosthetic successful prosthetic ambulation among geriatric
candidacy and walking ability would be reinforced subjects (17). This is consistent with a previous report
130 KAHLE ET AL.
Table 2. Predictive Factors Investigated by Included Studies
Table 2. Predictive Factors Investigated by Included Studies
Cause of amputation Amputation Level Stump factors and pain Cognition/ Mood BMI -Height, weight Physical Fitness Ability to Stand on 1 Leg Comorbidities Gender (sex) Social support Time to rehabilitation Pre-morbid living status Vascular intervention/
First Author (Etiology) Disturbance Motivation (SLS) IADLs Age Smoking Race disease
Chin (10) 1 1
Czerniecki
(27) 1
Erjavec (11) 1
Erjavec (33) 1
Fortington 1
(30)
Gremeaux 1 1 1
(24)
Hamamura
(17) 1 1 1 1 1 1 1
Landry (26) 1
O'Neil (25) 1
Raya (32) 1
Stineman
(22) 1 1 1 1 1 1
Stineman 1 1 1 1 1 1
(21)
Suckow (23) 1 1
van Eijk( 20) 1 1 1 1 1 1 1
Wezenberg 1 1
(28)
Wezenberg
(14) 1 1
Wong (19) 1 1 1 1 1 1 1
Artwert (29)
Gaunard (31) 1 1 1
Linberg (12) 1 1 1
Webster (18) 1 1 1 1 1 1 1
Total 3 13 1 3 1 5 1 3 2 10 6 4 1 2 1 4 2 2
Predictive factors in common with Sansam et al, and this systematic review are grey color filled. Predictive factors identified in studies
in the Sansam et al. article, but not found in studies included in this extension are: pre-rehab motor function, employment and sport,
Predictive factors in common with Sansam et al., and this systematic review are grey color filled.
hemiparesis, psychological factors, and self-efficacy.(5) Predictive factors exclusive to this review are white color filled.
Predictive factors identified in studies in the Sansam et al. article, but not found in studies included
in this extension are: pre-rehab motor function, employment and sport, hemiparesis, psychological
factors, and self-efficacy (5). Predictive factors exclusive to this review are white color filled.
PREDICTING WALKING ABILITY AFTER AMPUTATION 131
finding a statistically significant association between independent negative predictor of walking ability up
patient “motivation” and the ability to learn to walk to 18 years after surgery. A significant association
with a prosthesis (5). between post-operative ADLs and walking ability
after rehabilitation with a prosthesis has also been
Social Support described (5). Therefore, it seems the preponderance
Webster et al. identified greater levels of baseline of evidence suggests that independence in completion
social support were associated with more hours of of ADLs is a factor worth considering when deter-
prosthetic walking and identified a need to build in mining prosthetic candidacy.
social support structures for patients that have lim-
itations in this regard (18). Greater perceived social Time to Rehabilitation
support as a predictor of higher mobility was not a Stineman et al., in two separate high quality stud-
well-supported predictive factor in previous studies ies, reported that patients who had early rehabilitation
(5). initially made higher motor gains than those individ-
Smoking uals who had later rehabilitation (21). Additionally,
Wong et al. reported smoking was associated with patients who received acute postoperative inpatient
rehabilitation, compared to those with no evidence of
significantly poorer outcomes in diabetic transtibial inpatient rehabilitation, had an increased likelihood
amputation (TTA) patients (19). Czerniecki et al. of one-year survival and home discharge. Prosthetic
reported smoking status and reported a likely rela- limb procurement did not differ significantly between
tionship between smoking and a more proximal level groups (22). Sansam et al. found a shorter time inter-
amputation, which is discussed later. Although smok- val between surgery and admission for rehabilitation
ing is implicated in the etiology of many amputations, is related to better walking potential. Similarly, the
Sansam et al. reported that it is unlikely to have a length of time taken from surgery to fitting for a
significant impact on mobility outcome (5).
prosthesis is significantly associated with outcome,
Stump Factors and Pain with those waiting longer having poorer walking
van Eijk et al., in a medium quality article, reported ability at one year (5).
that the presence of phantom pain was significantly Race
associated with the ability to use a prosthesis (20). In two high quality studies, race was reported
Phantom limb pain specifically was not reported as a as not significant in affecting the outcome of TTA
predictive factor outcome measure in previous stud- patients or being a predictive factor (18,19). Race was
ies.
not reported as a predictive factor outcome measure
Predictive Factors Supported with Two in previous studies.
References in This Literature Review
Vascular Intervention
Independence in Activities of Daily Living In a high quality study, Suckow et al. reported it is
van Eijk et al., in a medium quality article, reported possible, based on preoperative patient characteris-
preoperative Barthel Index (BI) as a measure of one’s tics, to identify patients undergoing lower extremity
ability to perform basic ADLs. It is reported as evalu- bypass surgery who are most or least likely to achieve
ation of the functional status at baseline. In addition, good functional outcomes even if a major amputation
the preoperative BI was estimated based on history is ultimately required. These findings may assist in
taking and was significantly positively associated patient education and surgical decision making in
with prosthetic use (20). Further, BI was a signifi- patients who are poor candidates for lower extremity
cant indicator of the ability to complete the Timed bypass (23). Wong et al. found indicators of poor
Up and Go (TUG) test. However, Wong et al., in a vascularity, such as absence of popliteal pulse and
high quality study, found no correlation between low Ankle Brachial Index (ABI), were significantly
independence in ADLs and outcome in TTA patients associated with poor clinical outcomes (19). Vascular
(19). This is inconsistent with the previous report of intervention was not reported as a predictive factor
dependency for self-care prior to amputation as an outcome measure in previous studies.
132 KAHLE ET AL.
Predictive Factors Supported with Three to Five Pre-Amputation Living Status
References in This Literature Review Several reports in this literature review collected
Ability to Stand on One Leg data on factors related to living status, such as marital
In a medium quality study, Grameaux et al. iden- status and independence. However, none of them
reported correlations on living status being a predic-
tified the one-leg balance test as appropriate for early tive characteristic of walking ability following lower
evaluation in the immediate follow-up of the ampu- extremity amputation (18,21,22,26).
tation in order to establish a prognosis for success
in prosthetic ambulation (24). Hamamura et al., in Cause of Amputation
a high quality study, identified the patient’s ability to Some authors reported cause of amputation as a
stand on one leg on the non-amputated limb as one factor in prosthetic candidacy; however, there was
of the most valuable factors contributing to successful no association of cause as a predictive characteristic
prosthetic ambulation in geriatric amputees, report- of walking candidacy (11,21,27). Hamamura et al.
ing significance in successful prosthetic walkers (17). reported on cause of amputation, yet no significant
van Eijk et al., in a medium quality article, reported difference between the two groups (successful and
that one-leg balance was significantly positively asso- unsuccessful ambulators) was observed (17). Wezen-
ciated with prosthetic use (20). These reports are berg et al. found that the presence of an amputation
consistent with previous reports identifying the ability as a result of vascular deficiency was significantly
to stand on one leg as indicative of better walking associated with a lower VO2 peak of 29.1%, whereas a
potential after unilateral lower limb amputation (5). traumatic amputation was not significantly associated
Cognition and Mood Disturbance with a difference in VO2 peak compared with controls.
Webster et al. reported, in a study of high quality, Having a vascular amputation was associated with a
26.4% decrease in VO2 peak compared with having an
that management of depression and promotion of amputation because of trauma (28). In a second later
social support may have a positive effect on pros- article, Wezenberg et al. found traumatic amputees
thetic use (18). O’Neill et al., in a report of medium walked at the same VO2 as able-bodied controls but
quality, concluded that identified cognitive deficits did so at a lower walking speed. Vascular amputees
may indicate augmented rehabilitation or provision walked at an even slower speed with a substantially
of alternative mobility rehabilitation (25). Stineman higher VO2. Both amputee groups chose a walking
et al. reported ADL function and mobility (physi- speed that was lower than the most efficient walking
cal functioning) using gains in motor FIM™ scores speed. Consideration of peak aerobic capacity is an
achieved by rehabilitation discharge. Psychosis was important factor when aiming to improve walking
associated with lower motor FIM™ gains, but no direct ability in amputees (14). Finally, peak aerobic capac-
correlation between psychosis and prosthetic candi- ity is an important determinant for walking ability.
dacy was reported (21). In the previous Sansam et al.
article, cognitive ability was consistently found to be Sansam et al. also reported an association between
a significant predictor of walking ability following the cause of amputation and walking potential in five
studies, with subjects undergoing an amputation for
rehabilitation, with a superior outcome reported in dysvascularity achieving a poorer outcome than those
those with better cognitive ability (5).
due to trauma or other non-vascular causes (5).
Gender
Four high quality studies found no association Predictive Factors Previously Reported Supported
between gender and walking ability following lower With Six or More References in This Literature
limb amputation, suggesting it is not a predictive Review
determinant (19-22). Previous studies also found no Amputation Level
consistent association with gender (5).
Artwert et al., in a high quality article, reported
that functional use of a prosthesis in TTA patients
PREDICTING WALKING ABILITY AFTER AMPUTATION 133
is related to residual-limb quality in general as well intact knee joint for providing the TTA patient with
as for specific bony aspects of the residual limb. A the ability to return to high-level mobility activities
tibial length of 12 to 15 cm from the knee joint line following rehabilitation (31). Stineman et al. reported
correlated with a better functional outcome than ADLs and mobility (physical functioning) using gains
shorter tibial lengths. Surgical technique and attention in motor FIM™ scores achieved by rehabilitation dis-
to the residual limb length ensures a more likely good charge. TFA was associated with lower motor FIM™
functional outcome after TTA (29). In a high quality gains, but no direct correlation between amputation
article, Suckow et al. reported TTA and transfemoral level and prosthetic candidacy was reported. There
amputee (TFA) patients were equally likely to ambu- was no statistically significant difference in prosthetic
late independently or with assistance (within groups) limb procurement for the group who received impa-
at hospital discharge. Between groups, however, there tient rehabilitation compared to the group who did
were significant differences based on level of ampu- not (21). van Eijk et al., in a medium quality article,
tation. Patients who underwent a minor amputation reported amputation level (low versus high) was sig-
were more likely to ambulate with or without assis- nificantly positively associated with prosthetic use
tance but less so than patients who did not have an (20). Further, van Eijk et al. found level as a predictor
amputation after lower extremity bypass (23). Linberg for ability to complete the TUG test. Webster et al.,
et al., in a high quality article, reported a significant in a high quality article, reported TFA patients were
difference in 6MWT performance between bilateral significantly less likely to achieve prosthetic fitting
TTA patients and TFA patients, with bilateral TTA success at one year (18). Wezenberg et al. reported
patients walking further (12). Chin et al. reported, in that the level of amputation was not associated with
a medium quality study, solely on the hip disarticu- VO2 peak (28). Sansam et al. previously found that
lation amputee (HDA), where older HDA patients in the majority of studies reported better walking ability
good physical condition and with a low prevalence and greater ability to achieve ADLs after distal and
of comorbidities were able to successfully walk with unilateral amputations compared with more proxi-
a prosthesis in a community setting (10). Czerniecki mal or bilateral amputations. At this time, it seems
et al. reported the rates of success were similar: 35%, the preponderance of evidence suggests that level
31%, and 33% of amputees with transmetatarsal (TM), of amputation is a factor in determining prosthetic
TTA, and TFA, respectively, achieved mobility success ability but not a preclusion from candidacy. Finally,
when seen in a comprehensive inpatient rehabilitation having more intact joints (i.e., having an intact knee
unit (27). Fortington et al. reported poorer perfor- compared with not having it) is consistent with a
mance by people with a TFA versus TTA. Slower higher potential level of function in TTA patients.
five-meter walk tests and fewer steps taken per day Moreover, the longer the transtibial residual limb,
were reported one year after amputation. One year the greater potential there is for increased functional
after discharge, people with TFA or TTA increased level. It is noteworthy, however, that having a long
the number of steps taken per day from 570 steps at TTA is not requisite for achieving community ambu-
discharge to 1314 steps and were able to maintain lation, as those with HDA can achieve community
this level in the second year (30). Grameaux et al. ambulation as well.
did not find a statistically significant worse result
in bioenergetic efficiency after TFA but did find a Physical Fitness
reduction in walking speed. Only when age was taken Raya et al., in a high quality article, reported
into account in a multiple regression model did the hip strength and balance were significant fac-
impact of the level of amputation become statistically tors impacting 6MWT scores in individuals with
significant (24). Hamamura et al. reported, in a high LLA. The 6MWT can identify impairments of the
quality study, no significant difference between the musculoskeletal system that can affect ambulation
successful and unsuccessful ambulator groups when ability such as weakness in the muscles that support
considering amputation level (17). Guanard et al., in a ambulation (32). Chin et al. reported that when older
high quality article, reported on the importance of an HDA patients are in good physical condition, they
134 KAHLE ET AL.
are successfully able to walk with a prosthesis in a age was associated with lower motor FIM™ gains,
community setting (10). Hamamura et al. found, in but no direct correlation between age and prosthetic
a high quality article, that the successful ambulatory candidacy was reported (21). Wong et al., in a high
group had a significantly higher VO2 max compared quality article, reported that age was not a signif-
with the failure group (17). van Eijk et al., in a medium icant predictive factor in affecting the outcome in
quality article, reported that functional ambulation TTA patients (19). van Eijk et al., in a medium qual-
(Functional Ambulation Catagories (FAC) score, ity article, reported that age was not significantly
which measures the independency of gait) was associated with prosthetic use (20). Sansam et al.
significantly positively associated with prosthetic previously reported that, in most studies, older age
use (20). Guanard et al., in a high quality article, at the time of amputation had an adverse effect on
reported rehabilitation-related factors, such as lower walking potential, with six studies reporting a much
limb strength, standing and dynamic balance, and stronger dependence of walking ability on age than
ability to displace the center of mass over the base on comorbidity. There is disagreement on the extent
of support, were all significantly related to ability to to which age is a factor in determining prosthetic
perform high-level mobility activities (31.) Sansam candidacy or success. The majority of the studies,
et al. previously reported that two medium to high including the higher quality evidence, suggests that
quality studies with the same first author have looked age does have a role in prosthetic and functional
at physical fitness and its relationship to walking abil- determinations but that it should not restrict candi-
ity following unilateral above knee amputation. Both dacy.
studies concluded that a %VO2 max of at least 50%
could be regarded as a guideline value for the level Comorbidities
of fitness required for successful ambulation with Chin et al. reported that, when there is a low prev-
an above-knee prosthesis. This provides agreement alence of commodities, older HDA patients were able
that fitness parameters, particularly cardiorespiratory to successfully walk with a prosthesis in a community
fitness, can be an important factor in determining setting (10). Hamamura et al. reported a significant
prosthetic candidacy and success.
difference between the successful and unsuccessful
Age ambulator groups in the number of comorbidities,
Erjavec et al. found that age was a key determining with fewer comorbidities having a positive effect
predictive factor in combination with the results of the (17). Webster et al., in a high quality article, reported
exercise stress test, a 6MWT at admission, the FIM™ that a history of dialysis was associated with signifi-
at admission, and consideration of gender, allowing cantly less prosthetic ambulation (18). Wong et al.,
the successful discrimination between patients who in a high quality article, reported that amputation
were fit with a prosthesis and those who were not due to diabetes, high total white count, erythrocyte
(11). Grameaux et al. reported a statistically signifi- sedimentation rate, C-reactive protein, urea, creati-
cant worse result in bioenergetic efficiency related to nine, neutrophils, absence of pulses, low ABI, and Toe
age when taken into account with amputation level Brachial Index were significantly associated with poor
(24). Hamamura et al. reported, in a high quality clinical outcomes in TTA patients (19). van Eijk et al.,
study, no significant difference between the suc- in a medium quality article, reported multimorbidity
cessful and unsuccessful ambulator groups when was not significantly associated with prosthetic use
considering age (17). Webster et al., in a high quality (20). Sansam et al. previously reported that the effect
article, reported that increased age was associated of comorbid conditions on walking outcome is not
with significantly less prosthetic ambulation (18). clear. In the majority of studies investigating the role
Stineman et al. reported ADL function and mobility of comorbidities, the conclusions have been drawn
(physical functioning) using gains in motor FIM™ from secondary analyses or in conjunction with other
scores achieved by rehabilitation discharge. Advanced factors. At this time, the majority of the evidence is
PREDICTING WALKING ABILITY AFTER AMPUTATION 135
unclear on the extent to which multiple comorbidi- ACKNOWLEDGMENTS
ties impacts prosthetic candidacy and use. However, Authors would like to acknowledge Tyler Kle-
some studies clearly demonstrate a negative effect of now, MSOP, CPT-ACSM, for his contributions to the
comorbidities on successful outcomes and a positive project. Contents of this manuscript represent the
effect of fewer comorbidities on successful outcomes. opinions of the authors and not necessarily those of
Moreover, no studies have shown positive outcomes the U.S. Department of Defense, U.S. Department of
associated with the presence of comorbidities. the Army, U.S. Department of Veterans Affairs, or any
academic or health care institution. Authors declare
Limitations no conflicts of interest. This project was funded by the
Several publications in this review were written by AOPA RFP 04012015 and the National Institutes of
the same or similar author teams. Further, the study Health Scholars in Patient Oriented Research (SPOR)
topics are also commonly related. The study samples grant 1K30RR22270.
vary slightly at times among these seemingly related
publications, making it difficult to be able to sum total REFERENCES
number of subjects. Caution is advised related to the 1. CARF International. CMS DMEPOS accredi-
interpretation of the results of total sample size. tation program standards manual with survey
preparation questions. 3 ed. Washington (DC):
rd
CONCLUSION CARF International; 2011.
In this literature review, body mass index, motiva- 2. CARF International. Medical rehabilitation pro-
tion, gender, and smoking are among the factors with gram descriptions. Washington (DC): CARF
minimal support to assist in determining candidacy International; 2013.
for a prosthesis. Moderately supported factors for 3. Royal College of Physicians, National Council for
prosthetic candidacy included cognition/mood dis- Palliative Care, British Society of Rehabilitation
turbance, etiology, physical fitness, ability to stand on Medicine. Long-term neurological conditions:
one leg, and pre-amputation living status. The most management at the interface between neurol-
strongly supported factors for considering prosthetic ogy, rehabilitation and palliative care. Concise
candidacy were amputation level, age, physical fitness, Guidance to Good Practice series, No 10. London
and comorbidities. These factors are mostly supported (UK): RCP; 2008.
in an earlier literature review and should be strongly 4. [WHO] World Health Organization. International
considered in a complete history and physical exam- Classification of Functioning, Disability and
ination by a multidisciplinary team. Predictive factor Health (ICF). Geneva (Switzerland): WHO;
data is likely available in a complete patient health 2001.
record. Prosthetic candidacy and functional level 5. Sansam K, Neumann V, O’Connor R, Bhakta B.
determinations could be predicted prior to prosthetic Predicting walking ability following lower limb
prescription based on a comprehensive review of the amputation: a systematic review of the literature.
aforementioned factors. Additionally, these factors J Rehabil Med. 2009;41(8):593-603.
seem to remain important later in rehabilitation when 6. Dobson A, El-Gamil A, Shimer M, DaVanzo JE.
determining the appropriate prosthetic prescription Economic value of prosthetic services among
and functional level potential. Prosthetic over-pre- Medicare beneficiaries: a claims-based retro-
scription is preferred to under-prescription because spective cohort study. Mil Med. 2016;181(2
the cost of a less active and less healthy patient, limited Suppl):18-24.
by technologies and therapies that adversely impacts 7. Samuelsson KA, Toytari O, Salminen AL, Brandt
their function and mobility, far outweighs the cost of A. Effects of lower limb prosthesis on activity,
a higher level prosthesis. participation, and quality of life: a systematic
review. Prosthet Orthot Int. 2012;36(2):145-58.
136 KAHLE ET AL.
8. Cunningham BP, Harmsen S, Kweon C, Patterson 17. Hamamura S, Chin T, Kuroda R, Akisue T, Iguchi
J, Waldrop R, McLaren A, McLemore R. Have T, Kohno H, Kitagawa A, Tsumura N, Kurosaka
levels of evidence improved the quality of M. Factors affecting prosthetic rehabilitation
orthopaedic research? Clin Orthop Relat Res. outcomes in amputees of age 60 years and over.
2013;471(11):3679-86. J Int Med Res. 2009;37(6):1921-7.
9. Department of Health Long-term Conditions 18. Webster JB, Hakimi KN, Williams RM, Turner
NSF Team. The national service framework AP, Norvell DC, Czerniecki JM. Prosthetic fit-
for long-term conditions. London (UK): ting, use, and satisfaction following lower-limb
Department of Health; 2005. Annex 2 research amputation: a prospective study. J Rehabil Res
and evidence; p. 88. Dev. 2012;49(10):1493-504.
10. Chin T, Kuroda R, Akisue T, Iguchi T, Kurosaka 19. Wong KL, Nather A, Liang S, Chang Z, Wong
M. Energy consumption during prosthetic TT, Lim CT. Clinical outcomes of below knee
walking and physical fitness in older hip dis- amputations in diabetic foot patients. Ann Acad
articulation amputees. J Rehabil Res Dev. Med Singapore. 2013;42(8):388-94.
2012;49(8):1255-60. 20. van Eijk MS, van der Linde H, Buijck B, Geurts A,
11. Erjavec T, Presern-Strukelj M, Burger H. The Zuidema S, Koopmans R. Predicting prosthetic
diagnostic importance of exercise testing in devel- use in elderly patients after major lower limb
oping appropriate rehabilitation programmes for amputation. Prosthet Orthot Int. 2012;36(1):45-
patients following transfemoral amputation. Eur 52.
J Phys Rehabil Med. 2008;44(2):133-9. 21. Stineman MG, Kwong PL, Xie D, Kurichi JE,
12. Linberg AA, Roach KE, Campbell SM, Stoneman Ripley DC, Brooks DM, Bidelspach DE, Bates BE.
PD, Gaunaurd IA, Raya MA, Gomez-Orozco Prognostic differences for functional recovery
C, Gailey RS. Comparison of 6-minute walk after major lower limb amputation: effects of
test performance between male Active Duty the timing and type of inpatient rehabilitation
soldiers and servicemembers with and without services in the Veterans Health Administration.
traumatic lower-limb loss. J Rehabil Res Dev. PM R. 2010;2(4):232-43.
2013;50(7):931-40. 22. Stineman MG, Kwong PL, Kurichi JE, Prvu-
13. Raya MA, Gailey RS, Fiebert IM, Roach KE. Bettger JA, Vogel WB, Maislin G, Bates BE, Reker
Impairment variables predicting activity limita- DM. The effectiveness of inpatient rehabilitation
tion in individuals with lower limb amputation. in the acute postoperative phase of care after
Prosthet Orthot Int. 2010;34(1):73-84. transtibial or transfemoral amputation: study
14. Wezenberg D, van der Woude LH, Faber WX, de of an integrated health care delivery system.
Haan A, Houdijk H. Relation between aerobic ArchPhys Med Rehabil. 2008;89(10):1863-72.
capacity and walking ability in older adults with a 23. Suckow BD, Goodney PP, Cambria RA, Bertges
lower-limb amputation. Arch Phys Med Rehabil. DJ, Eldrup-Jorgensen J, Indes JE, Schanzer A,
2013;94(9):1714-20. Stone DH, Kraiss LW, Cronenwett JL, Vascular
15. Highsmith MJ, Kahle JT, Bongiorni DR, Sutton Study Group of New England. Predicting func-
BS, Groer S, Kaufman KR. Safety, energy tional status following amputation after lower
efficiency, and cost efficacy of the C-Leg for extremity bypass. Ann Vasc Surg. 2012;26(1):67-
transfemoral amputees: a review of the literature. 78.
Prosthet Orthot Int. 2010;34(4):362-77. 24. Gremeaux V, Damak S, Troisgros O, Feki A,
16. Ziegler-Graham K, MacKenzie EJ, Ephraim PL, Laroche D, Perennou D, Benaim C, Casillas JM.
Travison TG, Brookmeyer R. Estimating the Selecting a test for the clinical assessment of bal-
prevalence of limb loss in the United States: 2005 ance and walking capacity at the definitive fitting
to 2050. Arch Phys Med Rehabil. 2008;89(3):422- state after unilateral amputation: a comparative
9. study. Prosthet Orthot Int. 2012;36(4):415-22.
PREDICTING WALKING ABILITY AFTER AMPUTATION 137
25. O’Neill BF, Evans JJ. Memory and executive ficiency. J Rehabil Res Dev. 2007;44(5):717-22.
function predict mobility rehabilitation outcome 30. Fortington LV, Rommers GM, Geertzen JH,
after lower-limb amputation. Disabil Rehabil. Postema K, Dijkstra PU. Mobility in elderly peo-
2009;31(13):1083-91. ple with a lower limb amputation: a systematic
26. Landry GJ, Silverman DA, Liem TK, Mitchell EL, review. JAm Med Dir Assoc. 2012;13(4):319-25.
Moneta GL. Predictors of healing and functional 31. Gaunaurd IA, Roach KE, Raya MA, Hooper
outcome following transmetatarsal amputations. R, Linberg AA, Laferrier JZ, Campbell SM,
Arch Surg. 2011;146(9):1005-9. Scoville C, Gailey RS. Factors related to high-
27. Czerniecki JM, Turner AP, Williams RM, Hakimi level mobility in male servicemembers with
KN, Norvell DC. The effect of rehabilitation in traumatic lower-limb loss. J Rehabil Res Dev.
a comprehensive inpatient rehabilitation unit 2013;50(7):969-84.
on mobility outcome after dysvascular lower
extremity amputation. Arch Phys Med Rehabil. 32. Raya MA, Gailey RS, Gaunaurd IA, Ganyard
H, Knapp-Wood J, McDonough K, Palmisano
2012;93(8):1384-91.
28. Wezenberg D, de Haan A, Faber WX, Slootman T. Amputee mobility predictor-bilateral: a per-
HJ, van der Woude LH, Houdijk H. Peak oxy- formance-based measure of mobility for people
gen consumption in older adults with a lower with bilateral lower-limb loss. J Rehabil Res Dev.
limb amputation. Arch Phys Med Rehabil. 2013;50(7):961-8.
2012;93(11):1924-9. 33. Erjavec T, Vidmar G, Burger H. Exercise testing
29. Arwert HJ, van Doorn-Loogman MH, Koning as a screening measure for ability to walk with
J, Terburg M, Rol M, Roebroeck ME. Residual- a prosthesis after transfemoral amputation due
limb quality and functional mobility 1 year after to peripheral vascular disease. Disabil Rehabil.
transtibial amputation caused by vascular insuf- 2014;36:1148-55.
ISSN 1949-8241 • E-ISSN 1949-825X
Technology and Innovation, Vol. 18, pp. XX-XX, 2016 ISSN 1949-8241 • E-ISSN 1949-825X
Technology and Innovation, Vol. 18, pp. 139-150, 2016
http://dx.doi.org/10.21300/18.1.2016.5
Printed in the USA. All rights reserved. http://dx.doi.org/10.21300/18.2-3.2016.139
Printed in the USA. All rights reserved.
Copyright © 2016 National Academy of Inventors. www.technologyandinnovation.org
www.technologyandinnovation.org
Copyright © 2016 National Academy of Inventors.
EFFECTS OF THE GENIUM KNEE SYSTEM ON FUNCTIONAL LEVEL,
STAIR AMBULATION, PERCEPTIVE AND ECONOMIC OUTCOMES
IN TRANSFEMORAL AMPUTEES
M. Jason Highsmith , Jason T. Kahle , Matthew M. Wernke , Stephanie L. Carey ,
6
1-3
7
4,5
Rebecca M. Miro , Derek J. Lura , and Bryce S. Sutton 9
1
8
1 School of Physical Therapy & Rehabilitation Sciences, University of South Florida, Tampa, FL, USA
2 Extremity Trauma & Amputation Center of Excellence (EACE), U.S. Department of Veterans Affairs, Tampa, FL, USA
th
3 319 Minimal Care Detachment, U.S. Army Reserves, Pinellas Park, FL, USA
4 OP Solutions, Inc., Tampa, FL, USA
5 Prosthetic Design + Research, Tampa, FL, USA
6 WillowWood, Mt. Sterling, OH, USA
7 Department of Mechanical Engineering, University of South Florida, Tampa, FL, USA
8 Department of Bioengineering and Software Engineering, Florida Gulf Coast University, Ft. Myers, FL, USA
9 Center of Innovation on Disability and Rehabilitation Research (CINDRR-TPA), James A. Haley Veterans’ Hospital, Tampa, FL, USA
Compared to non-microprocessor knees, the C-Leg microprocessor knee (MPK) is bioen-
ergentically and economically more efficient and safer for transfemoral amputation (TFA)
patients. The Genium MPK has demonstrated improvements in perceived function, knee
kinematics, and physical functional performance compared to C-Leg. Clinical and health
economic analyses have not been conducted with the Genium knee system. The purpose of
this study was to determine if laboratory determined benefits of Genium are detectable using
common clinical assessments and if there are economic benefits associated with its use. This
study utilized a randomized AB crossover study with 60 d follow-up including cost-effectiveness
analysis. Twenty TFA patients tested with both knees in mobility and preference measures.
Incremental cost-effectiveness ratios (ICER) were calculated based on performance measures.
Stair Assessment Index scores improved with Genium. Mean stair completion times and descent
stepping rate were not different between knees. Stair ascent stepping rate for C-Leg was greater
compared with Genium (p = 0.04). Genium use decreased Four square step test completion time
and increased functional level and step activity (p ≤ 0.05). Further, Genium use improved (p ≤
0.05) function and safety in three out of five Activities of Daily Living (ADL) survey domains.
Finally, more subjects preferred Genium following testing. Functional measures were used to
calculate ICERs. ICER values for Genium fall within established likely-to-accept value ranges.
Compared with C-Leg, Genium use improved stair walking performance, multi-directional
stepping, functional level, and perceived function. In this group of community ambulators with
TFA, Genium was preferred, and, while more costly, it may be worth funding due to significant
improvements in functional performance with ADLs.
Key words: Amputee mobility predictor; Four square step test; Gait; Incremental cost-effec-
tiveness ratio; Physical therapy; Preference; Rehabilitation; Stair assessment index; StepWatch
_____________________
Accepted July 1, 2016.
Address correspondence to M. Jason Highsmith, Extremity Trauma & Amputation Center of Excellence (EACE), 8900 Grand Oak Circle (151R), Tampa, FL
33637-1022, USA. Tel: +1 (813) 558-3936; Fax: +1 (813) 558-3990; E-mail: [email protected]
139
140 HIGHSMITH ET AL.
INTRODUCTION fied of each subject’s assigned condition via telephone
Compared to non-microprocessor prosthetic knee on the day of the subject’s knee fitting. All fittings
systems (NMPK), the C-Leg microprocessor knee and adjustments were performed by the same study
system (MPK; Otto Bock Healthcare, Duderstadt, prosthetist, who was state-licensed and certified by
Germany) is more efficient in terms of gait bioener- the American Board for Certification in Orthotics,
gentics and health-economic measures as well as safer Prosthetics, and Pedorthics as well as by Ottobock
for persons with transfemoral amputation (TFA) (1). Healthcare for fitting both C-Leg and Genium MPK
Recently, the Genium MPK system has demonstrated systems. Subjects’ prosthetic sockets and suspension
improvements in perceived functional measures systems were not changed for the experiment’s dura-
(2), knee kinematics (3), and physical functional tion to reduce confounding effects from fitting and
performance (4) compared to C-Leg. Clinimetric acclimation issues. Subjects were fit with an Ottob-
assessment and health economic analysis have not yet ock Trias (standard height) or Axtion (low profile)
been conducted. Therefore, the purpose of this study prosthetic foot, based on limb length, for use over
was to determine if laboratory determined benefits the study duration. Manufacturer specifications were
of the Genium are detectable using common clinical used to set componentry alignment and were verified
outcome assessments. A second purpose was to deter- using the LASAR (Ottobock, Duderstadt, Germany)
mine if there are health-economic benefits associated alignment system.
with use of the Genium knee system.
Fitting and Accommodation Periods
METHODS After enrollment, anthropometric and demo-
The study was approved by the University of South graphic data and the study foot were recorded. Knee
Florida’s Institutional Review Board and listed in a fittings and alignment were conducted and settings
federal clinical trial registry. Subjects gave informed recorded. Subjects were invited to return to the study
consent prior to participation in the study. prosthetist or physical therapist for adjustment, align-
Study Design Overview ment, and training as many times as they wished to
optimize fit, comfort, and function and to mirror real
A randomized experimental crossover design, clinical practice and component prescription. Visits
where TFA patients used Genium and C-Leg MPK were counted and reasons for each visit recorded. All
systems, was used. Subjects tested on both knee subjects, regardless of the knee system with which
systems in random order separated by an accommo- they began the study, received an initial training
dation period of >2 weeks to <3 months, depending session from the study physical therapist for each
upon when they determined their readiness to test. knee system for training in transitional movements,
Subjects were assessed in a university clinical labora- obstacle crossing, ramps, stairs, speed variation, and
tory setting using common clinical outcome measures variable surfaces. Portions of the study’s training tech-
as described below.
niques have been previously published (5,6).
Randomization, Eligibility, and Interventions The minimum accommodation period was two
Subjects had to be unilateral TFA patients from weeks. After this, subjects were contacted weekly
any etiology and not have impairments that adversely to determine their ability to walk without personal
impacted their gait beyond their amputations (e.g., assistance on 1) level ground, 2) inclines, 3) declines,
cardiopulmonary, orthopedic impairments). Addi- 4) up & down stairs, and 5) on uneven ground. Sub-
tionally, subjects had to be C-Leg users for ≥1 year jects could contact investigators at any time after
prior to enrollment. An electronic random number the two-week minimum to declare their readiness
generator was used to assign subjects (off site) to to physically demonstrate they had accommodated
either continue with the C-Leg or be fit with a Genium to their currently assigned knee and study foot. Sub-
MPK at recruitment. The study prosthetist was noti- jects were considered accommodated after verbally
GENIUM KNEE: MOBILITY & ECONOMIC OUTCOMES 141
acknowledging and physically demonstrating their squares on the floor. In older adults, scores of ≥12 s
ability to ambulate independently on all five of the are associated with fall risk, whereas in unilateral TTA
previous terrains (7,8). Following accommodation, patients, ≥24 s are associated with fall risk (12,13).
subjects were scheduled for initial data collection Instructions were consistent with previous applica-
(phase A testing). Following initial testing, knee units tions. Briefly, subjects were instructed to complete
were switched and the process repeated for the sec- the test as quickly as possible but not to hop or jump
ond data collection (phase B testing). Following the over the canes. Three repeated trials were conducted
second data collection (phase B test), subjects were and averaged to represent the subject’s final score. If a
switched back into their C-Leg (original, pre-study subject’s foot touched a cane, the trial was stopped and
knee). At this time, a third and final data collection repeated until three successful trials were completed.
(phase C testing) was administered 60 d following
the second data collection. The third data collection Amputee Mobility Predictor (AMP)
(i.e., the 60 d follow-up) was administered via U.S. The AMP is a 21-item test of functional mobility
mail to subjects at their home addresses. Return of used to determine a lower limb amputee’s ability to
this survey marked the formal conclusion of subjects’ ambulate. AMP was shown to have moderate to strong
involvement in the study. concurrent validity with the six-minute walk test and
the Amputee Activity Survey (14). Specific details of
Testing and Outcomes each item and test administration of the AMP have
been described previously (14). The following is a syn-
Objective Measures opsis of the mobility functions assessed by the AMP
Stair Assessment Index (SAI) (14). Items 1 and 2 test the ability to maintain sitting
Subjects were asked to ascend and descend stairs balance. Items 3 through 7 test the ability to maintain
compliant with the Americans with Disabilities Act balance while performing tasks of transferring from
(9) (four steps 17cm high x 28cm long x 91cm wide chair to chair and standing unchallenged. Items 8
leading up to a platform with railing on both sides). through 13 test more challenging standing balance
Three trials ascending and three trials descending at activities. Items 14 through 20 evaluate quality of
a self-selected speed were timed using a stopwatch gait and the ability to negotiate specific obstacles.
and video recorded. Stair assessment index (SAI) Item 21 accounts for the use of particular assistive
scores were determined later by two independent devices. Most AMP items offer three scoring choices:
reviewers viewing the video recording using criteria 0 indicates inability to perform the task, 1 indicates
outlined previously (7). The SAI is a stair gait evalu- minimal level of achievement or that some assistance
ation instrument using a 13-point scale to determine was required in completing the task, and 2 indicates
gross motor pattern implementation of the subject complete independence or task mastery. The AMP
and use of assistive devices (7,10). The SAI was found test requires approximately 10 to15 min to administer
to have excellent inter-rater and intra-rater reliability and was administered by the study’s licensed physical
for assessing both stair ascent and descent (11). Time therapist a single time at each data collection.
to complete the test was recorded at data collection
and SAI scores obtained from video review. Stepping Step Activity Derived Functional Level (SAD-FL)
rate was calculated by dividing the number of steps The Galileo cloud was accessed at https://galileo.
completed by time to complete the test. orthocareinnovations.com (Orthocare Innovations,
Mountlake Terrace, WA, USA), and subjects were
Four Square Step Test (4SST) registered in the cloud and the StepWatch device was
The four square step test is a timed assessment of programmed to start recording step activity. Subjects
multi-directional stepping. Subjects step forward, wore the StepWatch on the prosthetic limb’s lateral
backward, and to each side while stepping over side just proximal to the approximate location of the
canes oriented in a cross configuration to create four anatomical ankle. The StepWatch recorded subjects’
142 HIGHSMITH ET AL.
step activity for a two-week period immediately Activities of Daily Living Survey
prior to laboratory data collections. At the end of A questionnaire developed for a previous study
the two-week recording period, StepWatch data was (16) was used to survey subjects in activities of daily
uploaded into the Galileo cloud server. At the point of living (ADL) tasks (45 total items) divided into five
data upload, each subject’s body mass and the study activity categories: Personal Care and Dressing (four
physical therapist’s opinion of the subject’s functional activities), Family and Social Roles, Leisure Time
level (i.e., Medicare K-level) was also entered into the Activities (12 activities), Mobility and Transporta-
Galileo software. Following upload, a report of step tion (19 activities), Health-related Exercise (four
activity and functional level is generated. The report activities), and Other Activities (six activities). The
provides an estimated K-level (i.e., functional level) individual ADL items are listed elsewhere (16). The
based on multiple factors, including cadence variabil- survey has a portion for each comparative MPK
ity, potential to ambulate, ambulation requirement, system where subjects first rate the importance of
and the clinician’s observation of functional level. ADLs and then rate the perceived difficulty and safety
Cadence variability includes the proportion of steps with the respective knee system. This was asked for
taken at low, medium, and high cadence rates. Poten- each MPK at the respective test sessions for each
tial to ambulate relates to the intensity of walking knee. Finally, there is a third portion of the survey
during minutes of activity with the highest number in which subjects are asked to subjectively compare
of steps taken. Ambulation requirement considers the perceived difficulty and safety of performing
the energy exerted during walking. Additionally, the the same 45 ADLs between the comparative MPK
clinically observed K-level is included and all of these systems. This comparative survey was administered
K-levels are averaged into an overall K-level that has at a 60 d follow-up after the second (phase B) data
one-tenth level precision. The report’s K-level was collection. In this analysis, only the comparison of
intended to be rounded to achieve the final K-level. difficulty and safety were evaluated using an ordinal
scale and non-parametric analysis consistent with
Perceptive Measures survey’s authors (16). This provided insight into
The significance of patient input on prosthetic the difference in both functional ADL performance
prescription, knee selection, and fabrication and its and safety between the two MPK systems from the
influence on successful outcomes is established (15). patients’ perspective.
To circumvent some of the pitfalls associated with
subjective data collection yet still include it because Preference
of the value of capturing participant input and pref- In order to capture true patient preference and
erence, two methods were considered. The first was exclude potential novelty effects or glitz bias (15), at
to use a functional survey previously deployed in this the study’s conclusion (phase C test), subjects were
population (16). The second method was to directly asked which knee mechanism they preferred and
query participants regarding their component prefer- would actually wish to continue using following the
ence, a method which has previously been described study’s conclusion. This measure was used to identify
as having the ability to strengthen or refute other true subject preference regardless of the performance
study findings (15). Finally, these subjective measures data. Subjects were asked four questions conducive
were administered at a 60 d follow-up after subjects to completing a 2 × 2 contingency table. Questions
were returned to their pre-study knee, which was the were: “Do you prefer and would you like to keep the
C-Leg, and followed the initial two data collections C-Leg?” and “Do you reject and wish to stop using
(one for each knee for the crossover). This is partic- the C-Leg?” These were repeated for the Genium.
ularly important given that, historically, MPK studies Asked this way, subjects could ultimately choose to
have not offered a follow-up assessment. keep or reject either or both MPKs. Finally, subjects
were asked, “If you could only keep either the Genium
GENIUM KNEE: MOBILITY & ECONOMIC OUTCOMES 143
or the C-Leg, which would you prefer to keep as a purchasing discounts based on the size of their prac-
permanent part of your prosthesis?” tice, and, further, federal sector practitioners may also
receive discounted component costs. Uncertainty
Cost-Effectiveness with the difference in performance was addressed
A previous report of the current randomized by only using the domain score differences that were
control trial analyzed function in activities of daily both improved and statistically significant from our
living using the Continuous Scale-Physical Functional previously published work (4). This was important
Performance-10 assessment (CS-PFP-10), a measure given that Genium improved all domains of the
shown to be valid, reliable, and sensitive to change CS-PFP-10, but not all domains were significantly
in multiple diagnostic populations, including TFA improved. In terms of time horizon, ICERs were
patients (4,17-19). Incremental cost effectiveness calculated based on the findings of the randomized
ratios (ICER) were calculated using differences in clinical trial from which patients accommodated
scores for each of three domains (balance & coor- over a period <90 d. Therefore, ICER values were
dination (BAL), upper body flexibility (UBF), and then projected over five years to project value over a
endurance (END)) that were significantly improved common life expectancy for an MPK then amortized
(p ≤ 0.05) when subjects used Genium compared with across this five-year period.
C-Leg on the CS-PFP-10 as source data for effects Statistical Analyses
(denominator). Calculations for ICERs were con-
ducted using the following equation: Statistical analyses were performed with IBM SPSS
(v21, Armonk, NY, USA). Data were compiled into a
database, assessed for completeness, and descriptive
analyses were performed (i.e., frequency, central ten-
dency, variance). The Shapiro-Wilk test was used to
determine if data were normally distributed. Between-
knee comparisons were made for each dependent
The payor’s perspective was taken in order to under- variable. Normally distributed continuous data
stand if Genium use (new strategy) is cost-effective were assessed using dependent samples t tests (i.e.,
compared to the C-Leg (current practice) given the 4SST times). For ordinal data or data that were not
recent challenges with the reimbursement of advanced normally distributed, a Related-Sample Wilcoxon
microprocessor prosthetic technologies. Private and Signed Rank test was used. This test evaluates the
federal sector prosthetic practitioners (i.e., expert distribution of the difference between related sam-
opinion) were queried in terms of the differences ples rather than the difference between means. The
reimbursed between the two study interventions. a priori level of significance was p ≤ 0.05. Cohen’s d
Because private sector practitioners commonly accept was then calculated to represent the magnitude of
healthcare reimbursement from multiple payors, a effect size between knee conditions for continuously
considerable range of cost resulted. The range of dif- scaled data when statistically significant differences
ferences in cost used for calculations was $30,000 and were present. Cohen’s d was interpreted as d = 0.2
$55,000. ICERs were calculated in $5,000 increments representing a small effect, 0.5 representing a medium
across this range of reimbursement differences where effect, and 0.8 representing a large effect (20). Investi-
the Genium is the more costly strategy. Assumptions gators adopted the “last observation carried forward”
used for ICER calculations included generalizability or “next observation carried backward” methods as
of the reported reimbursement across the U.S. given the study’s a priori intention-to-treat plan (21,22).
that practitioners were only queried from Florida. To determine if there were significant differences for
Another assumption is that discounting is built into preference, chi-squares/Fishers exact tests were used
the range of cost differences. An example is that some (i.e., categorical variables (prefer/not prefer, accept/
practitioners within the private sector receive volume reject)).
144 HIGHSMITH ET AL.
RESULTS descent using C-Leg were 3.8 (1.1), 3.8 (1.0), and
3.7 (1.0) for trials one, two, and three, respectively.
Subject Demographics
Corresponding times, in seconds, for Genium were
Twenty TFA subjects (n = 20) participated and 3.8 (1.0), 3.7 (0.8) and 3.6 (0.8). Mean differences for
completed all study tasks with complete data from completion times were not significantly different.
both MPK systems. Most subjects were male (80%) With C-Leg, subjects ascended stairs at a mean
with a mean (SD) age of 46.5 years (14.2) and BMI of (SD) rate of 1.1 steps/s (0.5) compared to a mean
26.4 kg/m (4.2). The majority were employed (55%), rate of 0.8 steps/s (0.2) with Genium. The stepping
2
25% were governmentally classified as “disabled,” rate was significantly increased when using C-Leg (p
and the remaining 20% were students or retired. = 0.04; medium effect size). For descent, the mean
All subjects were independent, unlimited commu- (SD) stepping rate with C-Leg was 1.2 steps/s (0.4)
nity ambulators (Medicare functional classification compared to 1.1 steps/s (0.2) when using Genium.
level 3). Mean time since amputation was 17.7 years The stepping rate for descent was not significantly
(15.6), and amputation etiology was predominantly different between knee systems.
traumatic (70%) followed by malignancy (20%) and
peripheral vascular disease (10%). Mean relative 4SST, AMP, SAD-FL
residual limb length (SD) was 70% (30%) of the sound Results for the 4SST, AMP, and SAD-FL are shown
side femur, and the mean hip flexion contracture was in Table 1. Use of the Genium decreased time to
12.8° (7.7) as measured with a manual goniometer in complete the 4SST by 1.1 s. Functional performance
the Thomas test position. A variety of prosthetic sock- increased by two points with Genium use as measured
ets (e.g. ischial ramus containment, ischial support, by the AMP and also by 0.2 points as measured by step
subischial, quadrilateral) and suspension systems activity derived functional level. These differences
(e.g. locking liners, suction, elevated vacuum) were were all significantly improved (p ≤ 0.05) following
used. Sagittal knee alignment was not different (p > accommodation and use of the Genium MPK.
0.05) between knee systems.
SAI Table 1. Objective Measures: 4SST, AMP, and SAD-FL
Mean and median SAI scores for trials one, two, Table 1. Objective Measures: 4SST, AMP, and SAD-FL
and three using C-Leg during stair ascent were 5.6, C-Leg Genium
6.0, and 6.3 and 5, 6, and 6, respectively. Correspond-
ing scores for Genium during stair ascent were 9.7, Test Central Central p value
9.9, and 10.1 with a median score of 11 for all trials. Variance Variance
These results were significantly different (p = 0.001) Tendency Tendency
between knee systems. Mean and median SAI scores
for trials one, two, and three using C-Leg during stair
descent were 11.0, 10.3, and 11.2 with a median of 4SST 12.2 3.3 11.1 3.4 0.04
11 for all trials. Corresponding scores for Genium
during stair descent were 12, 11.7, and 11.8 with a AMP 42 33 to 45 44 39 to 46 ≤0.001
median score of 11 for all trials. These results were
also significantly different between interventions (p SAD-FL 3.4 1.8 to 4.0 3.6 2.0 to 4.3 0.01
= 0.04).
Mean (SD) times, in seconds, for stair ascent using
C-Leg were 4.4 (1.5), 4.4 (1.4), and 4.5 (1.5) for tri- 4SST is a 4 square step test. AMP is amputee mobility predictor.
als one, two, and three, respectively. Corresponding SAD-FL is step activity derived functional level. Central Ten-
4SST is 4 square step test. AMP is amputee mobility predictor. SAD-FL is step activity derived functional
dency is mean (SD) for 4SST and is median (range) for all other
level. Central Tendency is mean (SD) for 4SST and is median (range) for all other tests. Statistical
times, in seconds, for Genium were 5.0 (1.0), 5.2 (1.6), tests. Statistical Significance is p ≤0.05. Effect size for the 4SST
Significance is p ≤ 0.05. Effect size for the 4SST was small (d = 0.33).
and 4.7 (1.3). Mean (SD) times, in seconds for stair was small (d = 0.33).
GENIUM KNEE: MOBILITY & ECONOMIC OUTCOMES 145
Activities of Daily Living Survey respectively) assuming a $30,000 difference in cost
Results for the ADL survey are in Table 2. Genium (Figure 1). Using these same measures of function,
use resulted in improvements (p < 0.05) in perceived incremental cost-effectiveness of Genium vs. C-Leg
function and safety in three of five of the domains ranged from $11,000 to $11,957 per unit increase
queried in the ADL survey. Improved domains in function (END & UBF and BAL respectively)
included: 1) Family and Social Roles, Leisure Time assuming a $55,000 difference in cost. If these costs
Activities, 2) Mobility and Transportation, and 3) were divided across five equal annual payments (i.e.,
Other Activities. No differences were identified in the five-year service life), the annual cost per increase
remaining two domains of the ADL survey: Personal in function is $1,200 to $1,304 (assuming $30,000
Care and Dressing and Health-related Exercise. increased cost) or up to $2,200 to $2,391. The total
one-time reimbursed difference in cost divided into
Preference five equal annual payments (i.e., one payment per year
There was a significant difference (p < 0.001; Fish- for 5 years with no interest) is $6,000/year ($30,000
er’s Exact Test given cell counts <5) between those cost difference) or $11,000/year ($55,000 cost differ-
who selected accept Genium (80.0% responses or ence). These costs are not adjusted for inflation or
16/20 vs. 4/20 C-Leg accept) and those who selected discounting and are expressed in 2016 U.S. dollars.
reject C-Leg (65% or 13/20 vs. 1/20 who replied reject These costs also presume that each unit of functional
Genium). Some subjects seemed to be undecided or increase occurs in isolation. However, this is not the
had some preference for both devices. case, as each of these units of functional gain occurred
Cost-Effectiveness simultaneously (4), thereby increasing value.
Subjects achieved significantly higher (p < 0.05)
function with Genium use (4). Using these mea- DISCUSSION
sures of function, incremental cost-effectiveness This study’s hypothesis was that laboratory deter-
of Genium vs. C-Leg ranged from $6,000 to $6,522 mined benefits of Genium use previously identified
per unit increase in function (END & UBF and BAL (2,3,23-25) would translate into measurable improve-
Table 2. Activities of Daily Living Survey
Table 2. Activities of Daily Living Survey
Domain Functional Improvement Safety
Personal Care and Dressing ND ND
Family and Social Roles, Leisure Time Genium Genium
Activities
Mobility and Transportation Genium Genium
Health-related Exercise ND ND
Other Activities Genium Genium
ND is not different between MPK systems. When an MPK is listed (Genium or C-Leg), it was identified by sub-
jects to be either a functional improvement or safer when used with the activities of daily living in the respective
ND is not different between MPK systems. When an MPK is listed (Genium or C-Leg), it was identified by
domain (p ≤ 0.05).
subjects to be either a functional improvement or safer when used with the activities of daily living in the
respective domain (p ≤ 0.05).
146 HIGHSMITH ET AL.
Figure 1. Incremental cost effectiveness ratios for differences in cost between microprocessor knees based on
functional improvements.
Incremental cost effectiveness ratios were determined based on the difference in cost of the interventions (Y
axis), as a function of the ratio of the relative cost per unit of functional increase in three respective domains:
balance (BAL), upper body flexibility (UBF), and endurance (END).
ments in common clinical outcome assessments Medicaid Local Coverage Determination verbiage as
as well. A second hypothesis was that, due to the a suggested factor in functional level determination
increased functional benefits associated with Genium with the use of a cane limiting a prosthetic user to the
use, the technology could meet established thresholds K2 level, which does not allow for reimbursement of
acceptable to merit third-party reimbursement. MPK components (27).
Although median SAI scores were identical
SAI, 4SST, AMP, and SAD-FL between knees for stair descent, Genium allowed all
Genium use resulted in significantly improved subjects to perform a step-over-step descent without
SAI scores and decreased stepping rate (medium assistive device use. The maximum score for C-Leg
effect size) while ascending stairs in a similar time. was 11 during descent with a minimum score of three.
Further, Genium use improved SAI scores for stair This range of scores likely contributed to the statisti-
descent with similar descent times and stepping rate. cal difference between knees but suggests improved
With Genium’s stair ascent mode, most subjects were consistency with the Genium system. Although there
able to use a step-over-step pattern as opposed to was no aggregate difference in time or stepping rate
a skipping step pattern and were able to decrease for descent between MPK systems, the gross motor
handrail usage. The step-over-step pattern is more pattern implemented (i.e., SAI scores) and decreased
symmetrical and utilizes kinematic patterns more assistive device use with Genium further suggests that
like those of non-amputees. Typical stair climbing engineering advancements included in the Genium
patterns have demonstrated improved physiological allowed for increased stability and balance compared
costs relative to alternative stair gait patterns (26). to C-Leg as has been previously identified in other
Reduced energy demand resulting from an improved functional activities.
stair gait pattern is potentially corroborated by the The decreased time to complete the 4SST (small
significant reduction in stepping rate with use of the effect size), a test of multi-directional stepping, is
Genium. The reduction in handrail use is also clin- also consistent with the notion of improved balance
ically relevant, as it was used in recent Medicare/ suggested here by improved stair climbing abilities
GENIUM KNEE: MOBILITY & ECONOMIC OUTCOMES 147
resulting from Genium use. Beyond this, previous is desirable and yielded positive changes in function
tests of function in ADLs demonstrated improve- as previously noted in stair and multi-directional
ments in multi-directional stepping (4). Most notably, stepping function. Further supporting true change
rearward directed steps and small steps were improved is the fact that the AMP test also measured a signif-
(4,25). In a previous test involving moving laundry icant increase in functional level using parameters
from a washer to a dryer, a task involving small side other than step activity. While this was true for both
steps and rearward steps, significant improvements functional level measurements, neither functional
were observed (4), which are consistent with the 9% measure changed to a higher functional level. Instead,
improvement in 4SST times recorded in this study. subjects’ function increased significantly within their
Two means were used to assess functional level respective functional level. For instance, the group
in this study. One was the AMP test, a test of transi- did not change from limited community to unlimited
tional movement, stepping and balance (14) and the community ambulation.
second was a functional level based on step activity.
Tasks improved on the AMP test associated with ADL Survey and Preference
Genium use included stair gait, obstacle crossing, In addition to the objective functional measures
and variable cadence. Conversely, most subjects were described above, perceptive measures were increased
unable to achieve single-limb balance on the pros- or unchanged. Subjectively, there were no decreased
thesis regardless of which MPK they were using. The functional measures or decreased perceptions of
2.0 point increase observed with Genium use in this safety associated with use of the Genium. It is note-
study did not reach the minimal detectable change worthy that Genium increased safety and function in
(MDC) value of 3.4 points for the AMP reported areas requiring community engagement (e.g., family,
by Resnik and Borgia (28). Because the difference social and leisure roles, mobility and transportation)
in AMP scores did not reach the MDC, it could be because this is consistent with the type of subject
argued that true change may not have occurred rela- enrolled in the study who was (on average) in their
tive to this specific test. However, Resnik and Borgia’s mid-forties and active in the community. The fact that
sample was of mixed amputation level. Further, AMP there was no difference in self-care suggests C-Leg
scores were statistically significantly different, and may be meeting many of these needs. Conversely, the
key functions were improved with Genium use (e.g., lack of difference in exercise function could be that
stair gait, variable cadence, and obstacle crossing). both components still leave room for improvement.
Additional issues worthy of consideration include Interestingly, there was a significant difference in
the test’s use of interval level data and the wide range preference among this group of unlimited commu-
of task difficulty. nity ambulators. The magnitude of difference was
Previous study has shown that TFA patients tend comparable to that previously observed when a more
to not alter their usual living patterns and step activity functionally diverse subject group was asked their
based solely on receipt of a new knee system (29). preference between NMPKs and the C-Leg (15). It
However, in this study, there was a significant increase is unclear if preference would be so strong in more
in functional performance based on step activity. functionally diverse patient groups or patients more
These changes in activity took place while participants homogeneously located at higher or lower ends of
were in their community or home environments, as the functional spectrum.
lab steps were not counted. Thus, subjects had suffi-
ciently increased steps, steps/bout, or step intensity to Cost-Effectiveness
elevate their step activity derived functional levels. It In this study, using functional measures, the ICER
is possible that subjects took these additional steps or for reimbursing Genium vs. C-Leg ranged from $6,000
changed their stepping routine in part due to partici- to $6,522 per unit of functional increase assuming a
pation in the study while attempting to accommodate $30,000 intervention cost difference. Assuming the
with the study knee. From an activity perspective, this component would not require replacement for five
148 HIGHSMITH ET AL.
years, this cost would likely amortize over the five- CONCLUSION
year service period (i.e., five equal annual payments Accommodation and use of the Genium knee
with no interest). Thus, the estimated annual cost for system compared with C-Leg improved stair walking
this functional increase is $1,200 to $1,304. Alterna- performance, multi-directional stepping, functional
tively, depending upon discounting and other factors, level, and perceived function. Genium was also
if the cost difference between the interventions is preferred compared to C-Leg in this group of high-
higher, at $55,000, the ICER for reimbursing Genium functioning community ambulators with unilateral
vs. C-Leg ranges from $11,000 to $11,957 per unit of transfemoral amputation. Finally, Genium is a more
functional increase. Again this amortizes to $2,200 to costly microprocessor knee system but, in this group
$2,391 annually over a five-year service period. When of patients, is worth funding due to significant dif-
the C-Leg was initially introduced as an alternative to ferences in functional performance with activities of
NMPKs, the initial cost was notably higher. In time, it daily living.
was proven that while C-Leg was more expensive, it
was worth funding for numerous reasons, including ACKNOWLEDGMENTS
reducing falls, reducing lost time for prosthetic main- Contents of this manuscript represent the opin-
tenance, and others (1). In the case of the Genium, ions of the authors and not necessarily those of the
in higher-functioning patients, it is clear that higher U.S. Department of Defense, U.S. Department of the
function in ADLs, quality of life, and functional level Army, U.S. Department of Veterans Affairs, or any
are all domains that realize gains that are above what academic or health care institution. Authors declare
the C-Leg can provide (2-4,16,24,25). In this sense, no conflicts of interest. This project was funded by:
our estimates of incremental cost-effectiveness may
be conservative, as our ICERs are expressed in terms 1. The Florida High Tech Corridor/USF Connect
of a single functional measure. (Grant #FHT 10-26).
2. Otto Bock Healthcare (USF Grant
Limitations #6140101200).
This study lacked blinding, which is difficult to 3. The National Institutes of Health Scholars
incorporate in rehabilitation research for safety and in Patient Oriented Research (SPOR) grant
ethical reasons. The findings are based upon the expe- (1K30RR22270).
riences and performance of patients who may not
be representative of the entire TFA population, as REFERENCES
these subjects were unlimited community ambulators 1. Highsmith MJ, Kahle JT, Bongiorni DR, Sutton
who lost their limbs predominantly due to trauma BS, Groer S, Kaufman KR. Safety, energy
and malignancy. Regarding the cost effectiveness efficiency, and cost efficacy of the C-Leg for
portion of the study, the patient and societal perspec- transfemoral amputees: a review of the literature.
tives were not considered in the economic analysis. Prosthet Orthot Int. 2010;34:362-77.
Future studies need to carefully consider the impact 2. Highsmith MJ, Kahle JT, Miro RM, Mengelkoch,
of improvement in functional measures on produc- LJ. Perceived differences between the Genium
tivity. Our cost-effectiveness analysis only considered and the C-Leg microprocessor prosthetic knees
improvement in specific functional measures indi- in prosthetic-related function and quality of life.
vidually and not improvement simultaneously in Technol Innov 2014;15(4):369-75.
these measures, which likely yielded conservative 3. Lura DJ, Wernke MM, Carey SL, Kahle JT, Miro
estimates. Future studies will need to consider overall RM, Highsmith MJ. Differences in knee flexion
health-related quality of life and quality-adjusted life between the Genium and C-Leg microprocessor
years. knees while walking on level ground and ramps.
Clin Biomech (Bristol, Avon). 2015;30:175-81.
GENIUM KNEE: MOBILITY & ECONOMIC OUTCOMES 149
4. Highsmith MJ, Kahle JT, Miro RM, Cress ME, falling older adults. Arch Phys Med Rehabil.
Lura DJ, Quillen WS, Carey SL, Dubey RV, 2002;83:1566-71.
Mengelkoch LJ. Functional performance differ- 14. Gailey RS, Roach KE, Applegate EB, Cho B,
ences between the Genium and C-Leg prosthetic Cunniffe B, Licht S, Maguire, Nash MS. The
knees and non-Amputees. J Rehabil Res Dev. amputee mobility predictor: an instrument to
Forthcoming. assess determinants of the lower-limb ampu-
5. Highsmith MJ, Kahle JT, Lewandowski AL, Kim tee’s ability to ambulate. Arch Phys Med Rehabil.
SH, Mengelkoch LJ. A method for training step- 2002;83:613-27.
over-step stair descent gait with stance yielding 15. Kahle JT, Highsmith MJ, Hubbard SL.
prosthetic knees. J Prosthet Orthot. 2012;24:10-5. Comparison of nonmicroprocessor knee mech-
6. Highsmith MJ, Kahle JT, Lura DJ, Lewandowski anism versus C-Leg on Prosthesis Evaluation
AL, Quillen WS, Kim SH. Stair ascent and ramp Questionnaire, stumbles, falls, walking tests, stair
gait training with the Genium knee. Technol descent, and knee preference. J Rehabil Res Dev.
Innov. 2014;15(4):349-58. 2008;45:1-14.
7. Hafner BJ, Willingham LL, Buell NC, Allyn KJ, 16. Kannenberg A, Zacharias B, Mileusnic M, Seyr
Smith DG. Evaluation of function, performance, M. Activities of daily living: Genium bionic pros-
and preference as transfemoral amputees transi- thetic knee compared with C-Leg. J Prosthet
tion from mechanical to microprocessor control Orthot. 2013;25:110-7.
of the prosthetic knee. Arch Phys Med Rehabil. 17. Cress ME, Buchner DM, Questad KA, Esselman
2007;88:207-17. PC, deLateur BJ, Schwartz RS. Continuous-scale
8. Highsmith MJ. Microprocessor knees: consid- physical functional performance in healthy
erations for accommodation and training. J older adults: a validation study. Arch Phys Med
Prosthet Orthot. 2013;25:60-4. Rehabil. 1996;77:1243-50.
9. US Department of Justice. ADA standards 18. Cress ME, Petrella JK, Moore TL, Schenkman
for accessible design. Washington (DC): US ML. Continuous-scale physical functional
Department of Justice; 2010. performance test: validity, reliability, and sen-
10. Buell NC, Willingham LL, Allyn KJ, Hafner BJ, sitivity of data for the short version. Phys Ther.
Smith DG. Evaluation of gait style to ascend 2005;85:323-35.
and descend stairs for lower limb amputees. 19. Highsmith MJ, Kahle JT, Miro RM, Cress ME,
In: Boone D, editor. Proceedings of the 11th Quillen WS, Carey SL, Dubey RV, Mengelkoch
World Congress of the International Society of LJ. Concurrent validity of the continuous scale
Prosthetics and Orthotics; 2004 Aug 1-6 Hong physical functional performance-10 (CS-PFP-10)
Kong. Hong Kong: Hong Kong National Society test in transfemoral amputees. Technol Innov.
of the International Society for Prosthetics and Forthcoming 2016 Aug.
Orthotics; 2004. p 367. 20. Cohen J, editor. Statistical power analysis for
11. Highsmith MJ, Kahle JT, Kaluf B, Miro RM, the behavioral sciences. 2 ed. Hillsdale (NJ):
nd
Mengelkoch LJ, Klenow TD. Psychometric Erlbaum; 1988.
evaluation of the HAI and SAI in transfemoral 21. Buhi ER, Goodson P, Neilands TB. Out of sight,
amputees using a microprocessor knee system. not out of mind: strategies for handling missing
Technol Innov. Forthcoming 2016 Aug. data. Am J Health Behav. 2008;32:83-92.
12. Dite W, Connor HJ, Curtis HC. Clinical identi- 22. Kenward MG, Molenberghs G. Last observation
fication of multiple fall risk early after unilateral carried forward: a crystal ball? J Biopharm Stat.
transtibial amputation. Arch Phys Med Rehabil. 2009;19:872-88.
2007;88:109-14. 23. Highsmith MJ, Kahle JT, Lura DJ, Dubey RV,
13. Dite W, Temple VA. A clinical test of stepping Carey SL, Quillen WS, Mengelkoch LJ. Short
and change of direction to identify multiple and mid-distance walking and posturography
150 HIGHSMITH ET AL.
with a novel microprocessor knee. Technol Innov Local Coverage Determination (LCD): Lower
2014;15(4):359-68. limb prostheses (DL33787). Hingham (MA):
24. Bellmann M, Schmalz T, Ludwigs E, Blumentritt NHIC, Corp; 2015. Contract #16003. [accessed
S. Stair ascent with an innovative microproces- 2016 Aug 12]. http://www.amputee-coalition.org/
sor-controlled exoprosthetic knee joint. Biomed content/documents/medicare-alert-2015/ doc_
Tech. 2012;57:435-44. LLP_Draft_LCD_071615.pdf.
25. Bellmann M, Schmalz T, Ludwigs E, Blumentritt 28. Resnik L, Borgia M. Reliability of outcome mea-
S. Immediate effects of a new micropro- sures for people with lower-limb amputations:
cessor-controlled prosthetic knee joint: a distinguishing true change from statistical error.
comparative biomechanical evaluation. Arch Phys Ther. 2011;91:555-65.
Phys Med Rehabil. 2012;93:541-9. 29. Klute GK, Berge JS, Orendurff MS, Williams RM,
26. Shiomi T. Effects of different patterns of stair- Czerniecki JM. Prosthetic intervention effects on
climbing on physiological cost and motor activity of lower-extremity amputees. Arch Phys
efficiency. J Hum Ergol. 1994;23:111-20. Med Rehabil. 2006;87:717-22.
27. NHIC, Corp. A CMS Contractor. Proposed/Draft
Technology and Innovation, Vol. 18, pp. 151-157, 2016 ISSN 1949-8241 • E-ISSN 1949-825X
Printed in the USA. All rights reserved. http://dx.doi.org/10.21300/18.2-3.2016.151
Copyright © 2016 National Academy of Inventors. www.technologyandinnovation.org
EFFECTS OF THE GENIUM MICROPROCESSOR KNEE SYSTEM ON
KNEE MOMENT SYMMETRY DURING HILL WALKING
1-3
M. Jason Highsmith , Tyler D. Klenow , Jason T. Kahle , Matthew M. Wernke ,
7
5,6
4
Stephanie L. Carey , Rebecca M. Miro , and Derek J. Lura 9
8
1
1 School of Physical Therapy & Rehabilitation Sciences, University of South Florida, Tampa, FL, USA
2 Extremity Trauma & Amputation Center of Excellence (EACE), U.S. Department of Veterans Affairs, Tampa, FL, USA
3 319 Minimal Care Detachment, U.S. Army Reserves, Pinellas Park, FL, USA
th
4 Prosthetics and Sensory Aids Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA
5 OP Solutions, Inc., Tampa, FL, USA
6 Prosthetic Design and Research, Tampa, FL, USA
7 WillowWood, Mt. Sterling, OH, USA
8 Department of Mechanical Engineering, University of South Florida, Tampa, FL, USA
9 Department of Bioengineering and Software Engineering, Florida Gulf Coast University, Ft. Myers, FL, USA
Use of the Genium microprocessor knee (MPK) system reportedly improves knee kinematics
during walking and other functional tasks compared to other MPK systems. This improved
kinematic pattern was observed when walking on different hill conditions and at different speeds.
Given the improved kinematics associated with hill walking while using the Genium, a similar
improvement in the symmetry of knee kinetics is also feasible. The purpose of this study was to
determine if Genium MPK use would reduce the degree of asymmetry (DoA) of peak stance
knee flexion moment compared to the C-Leg MPK in transfemoral amputation (TFA) patients.
This study used a randomized experimental crossover of TFA patients using Genium and
C-Leg MPKs (n = 20). Biomechanical gait analysis by 3D motion tracking with floor mounted
force plates of TFA patients ambulating at different speeds on 5° ramps was completed. Knee
moment DoA was significantly different between MPK conditions in the slow and fast uphill
as well as the slow and self-selected downhill conditions.In a sample of high-functioning TFA
patients, Genium knee system accommodation and use improved knee moment symmetry
in slow speed walking up and down a five degree ramp compared with C-Leg. Additionally,
the Genium improved knee moment symmetry when walking downhill at comfortable speed.
These results likely have application in other patients who could benefit from more consistent
knee function, such as older patients and others who have slower walking speeds.
Key words: Amputee; Biomechanics; Gait; Knee kinetics; Physical therapy; Ramps; Rehabil-
itation; Slope; Transfemoral
_____________________
Accepted July 1, 2016.
Address correspondence to M. Jason Highsmith, Extremity Trauma & Amputation Center of Excellence (EACE), 8900 Grand Oak Circle (151R), Tampa, FL
33637-1022, USA. Tel: +1 (813) 558-3936; Fax: +1 (813) 558-3990; E-mail: [email protected]
151
152 HIGHSMITH ET AL.
INTRODUCTION reflective markers and an 8-camera Vicon (Oxford,
The Genium microprocessor knee (MPK) sys- United Kingdom) motion analysis system. Subjects
tem (Ottobock Healthcare, Duderstadt, Germany) walked up and down 5° ramps at slow, normal, and
reportedly implements several engineering advance- fast walking speeds. The task variations (walking
ments over previously manufactured MPK systems speeds, ascent/descent) were intended to evaluate
(1-5). Study of these advancements has resulted in knee component performance under diverse gait
a recent body of literature showing improved per- conditions seen during community ambulation.
formance of the Genium even when compared to
the C-Leg (Ottobock Healthcare, Duderstadt, Ger- Randomization and Interventions
many), the most studied commercially available All subjects entering the study were C-Leg users
MPK system (1-5). Compared to C-Leg, Genium for ≥1 year prior to enrollment. An electronic random
use resulted in improved symmetry in peak knee flex- number generator was used to assign subjects (off site)
ion angles during stance and swing when ascending to either continue with their C-Leg or to be fit with a
and descending ramps in a sample of transfemo- Genium knee system at recruitment. The study pros-
ral amputation (TFA) patients (6). This improved thetist was notified of the subject’s assigned condition
kinematic pattern was more similar to non-amputee via telephone on the day of the subject’s knee fitting.
controls when walking on different hill conditions All fittings and adjustments were performed by the
and at different speeds. Given the improved kine- study prosthetist, who was state-licensed and certified
matics associated with hill walking while using the by the American Board for Certification in Orthotics,
Genium, a similar improvement in the symmetry of Prosthetics, and Pedorthics as well as by Ottobock
knee kinetics is also feasible. Healthcare for fitting both C-Leg and Genium MPK
Due to the increased kinematic symmetry of the systems. The subjects’ prosthetic sockets and suspen-
knee previously reported following Genium accom- sion systems were not changed for the duration of the
modation and use, the hypothesis of this study was experiment to reduce the confounding effects from
that the Genium MPK system would increase sym- fit and acclimation issues. All subjects were fit with
metry in peak knee flexion moment during ramp an Ottobock Trias (standard height) or Axtion (low
gait. Thus, the study’s purpose was to determine if profile) prosthetic foot, based on limb length, for use
use of the Genium MPK would reduce the degree over the study duration. Manufacturer specifications
of asymmetry (DoA) of peak stance knee flexion were used to set componentry alignment and were
moment compared to the C-Leg MPK in persons verified using the LASAR (Ottobock, Duderstadt,
with transfemoral amputation. Germany) alignment system.
METHODS Fitting and Accommodation Periods
The study was approved by the University of South After enrollment, anthropometric data, random-
Florida’s Institutional Review Board and listed on ization order, and the study foot were recorded. Knee
clinicaltrials.gov (#NCT01473992). Subjects gave fittings and alignment were conducted and recorded.
informed consent prior to participation in the study. All subjects were invited to return to the study pros-
thetist or physical therapist for adjustment, alignment,
Study Design Overview and training as many times as they wished to optimize
This study was a randomized experimental their fit, comfort, and function and to mirror real clin-
crossover of TFA patients using Genium and C-Leg ical practice and component prescription. Visits were
MPKs. Each subject tested on both knee conditions counted, and reasons for each visit were recorded. All
in a random order separated by an accommodation subjects, regardless of which knee system they began
period of >2 weeks to <3 months, depending upon the study with, received an initial training session
when subjects determined their readiness to test. from the study physical therapist for each knee sys-
Subjects’ gait patterns were recorded using passive tem for training in transitional movements, obstacle
KNEE MOMENT SYMMETRY DURING HILL WALKING 153
crossing, ramps, stairs, speed variation, and variable ramp downhill, 5) fast speed (i.e., hurried, late for
surfaces. Portions of the training techniques used in appointment) on 5° ramp uphill, and 6) fast speed on
this protocol have been previously published (1,7). 5° ramp downhill. Subjects were given the opportu-
The minimum accommodation period was two nity to rest between tests. A total of six different trial
weeks. After this, subjects were contacted weekly types were assessed, and each type was completed
to determine their ability to walk without personal twice to record data over the force platform for both
assistance on 1) level ground, 2) inclines, 3) declines, the prosthetic and anatomical leg.
4) up & down stairs, and 5) on uneven ground. Sub-
jects could contact investigators at any time after Data Processing
the two-week minimum to declare their readiness Tracking segments were defined using the sur-
to physically demonstrate they had accommodated face markers, and redundancy was included in the
to their currently assigned knee and study foot. Sub- marker set to compensate for marker drop out and
jects were considered accommodated after verbally to increase data consistency and reliability. Frames
acknowledging and physically demonstrating their were tracked by least squares minimization of the
ability to ambulate independently on all five of the in-segment marker reconstruction error (12). A
previous terrains. This study accommodation test was combination of anatomical markers and calculated
adapted from Hafner et al. (8,9). Following accom- joint centers were then used to define anatomical
modation, subjects were scheduled for initial testing.
Following initial testing, knee units were switched, segments. Segment axes definitions were based on
and the process was repeated for follow-up testing. the recommendations of the International Society of
Biomechanics (13). Joint angles were calculated from
Testing Euler angle transformations of anatomical segments.
Although the marker set used enabled tracking of
Passive marker-based 3D motion tracking was the entire lower body, knee flexion moment was the
selected to record and analyze knee movements measure of interest that addressed the a priori hypoth-
because it is valid, reliable, and considered to be the eses, and therefore was the only measure included
gold standard measurement technique for gait anal- in this report. The gait cycle was defined from heel
ysis (10). Specifically, an 8-camera Vicon motion
analysis system was used to collect knee kinematic strike to heel strike of the involved foot. Peak knee
flexion moment was defined as the maximum knee
data of subjects performing hill gait tasks. Passive
reflective markers were attached to subjects using flexion moment from 0% to 30% of the gait cycle to
neoprene straps and double side adhesive collars as assure peak stance flexion in the loading response.
previously described (6). Force platforms (AMTI, The remainder of the gait cycle (30% to 100%) was
Watertown, MA, USA) were embedded in the gait considered swing phase and was not included in
platform and ramps and used to record ground reac- this report. These definitions were used to prevent
tion forces and to time events. For instance, heel strike subjects who used the ‘ride-down’ strategy on the
and toe off were identified with force plate data; how- ramp descent from inflating the stance flexion angles
ever, the subsequent heel strike was recorded with after the loading response and further increasing
kinematic approximation (11). In accordance with deviations in subject moment data. Vertical ground
manufacturer specifications and recommendations, reaction forces (GRF) were determined from the
the Vicon cameras were calibrated and force platforms force plates. A 2D kinematic model was written in
zeroed before each session and after adjusting the Vicon Bodybuilder software to determine joint angles,
ramp to the slope condition. forces, and moments of the knee in the sagittal plane.
All subjects completed the hill walking tasks in DoA between sound and prosthetic side knee flex-
the same order: 1) normal speed on 5° ramp uphill, ion moments of the same trial type was calculated
2) normal speed on 5° ramp downhill, 3) slow speed using the following equation:
(i.e., casual) on 5° ramp uphill, 4) slow speed on 5°
154 HIGHSMITH ET AL.
or retired. All subjects were independent, unlimited
community ambulators (Medicare functional clas-
sification level 3). The mean time since amputation
In this equation, “S” represents the sound side was 17.7 years (15.6), and amputation etiology was
knee peak stance flexion moment and “P” represents predominantly traumatic (70%) followed by malig-
the prosthetic side knee peak stance flexion moment. nancy (20%) and peripheral vascular disease (10%).
As a result of this calculation, a positive value would Mean relative residual limb length (SD) was 70%
indicate a greater moment on the sound side, a neg- (±30%) of the sound side femur, and the mean hip
ative value would indicate a greater moment of the flexion contracture was 12.8° (7.7) as measured with
prosthetic side, and a value of zero would indicate a manual goniometer in the Thomas test position. A
perfect symmetry between the two sides. variety of prosthetic sockets (i.e., ischial ramus con-
tainment, ischial support, subischial, quadrilateral)
Statistical Analyses and suspension systems (i.e., locking liners, suction,
Statistical analyses were performed with IBM SPSS elevated vacuum) were utilized by these subjects.
(v21, Armonk, NY, USA). Data were compiled into a
database, assessed for completeness, and descriptive Overall
analyses were performed (i.e., central tendency). The Knee moment DoA was statistically different
Shapiro-Wilk test was used to determine if data were between MPK conditions in the slow and fast uphill
normally distributed. Normally distributed data were as well as the slow and self-selected downhill condi-
assessed using dependent samples t tests for both peak tions (Table 1). The Genium improved knee moment
stance and swing flexion moment by knee type per DoA in all of these conditions except in the fast uphill
walking condition. For data that were not normally condition, where there was no mean difference. In the
distributed, a Related-Sample Wilcoxon Signed Rank latter condition, however, the increased variance, as
test was used. This test evaluates the distribution of shown with the increased SD of C-Leg compared with
the difference between related samples rather than Genium, resulted in the statistical difference. Effect
the difference between means. The a priori level of size within uphill conditions was medium for slow
significance was p ≤ 0.05. Cohen’s d was then calcu- walking speed and small for the fast walking speed.
lated to represent the magnitude of effect size between Effect size within the downhill condition was small for
knee conditions with regard to DoA of peak knee the slow speed and large for the self-selected walking
flexion moment. Cohen’s d was interpreted as d = 0.2 speed. The Genium MPK also displayed lower vari-
representing a small effect, 0.5 representing a medium ance in peak knee stance flexion moment symmetry,
effect, and 0.8 representing a large effect (14). Investi- as shown by SD, in all walking speeds in the uphill
gators adopted the “last observation carried forward” conditions and the self-selected downhill condition.
or “next observation carried backward” methods as The C-Leg displayed less variance than the Genium
the study’s a priori intention-to-treat plan (15,16). in the slow and fast downhill conditions.
RESULTS DISCUSSION
Subject Demographics Due to engineering advancements incorporated
Twenty TFA subjects (n = 20) participated and into the Genium MPK, including a sagittal knee
completed all study tasks with complete data from moment sensor, angle sensor, and gyroscope, we
both MPK systems. The majority of subjects were hypothesized Genium use would increase symme-
male (80%) with a mean (SD) age of 46.5 years (14.2) try in peak stance knee flexion moment in TFA
and BMI of 26.4 kg/m2 (4.2). The majority were patients compared to the C-Leg. This hypothesis
employed (55%), 25% were governmentally classified was supported, as a majority (3 of 4) of the signifi-
as “disabled,” and the remaining 20% were students cant differences resulted in improved knee moment
KNEE MOMENT SYMMETRY DURING HILL WALKING 155
Table 1. Degree of Asymmetry for Knee Moment during Hill Walking
Table 1. Degree of Asymmetry for Knee Moment during Hill Walking
Walking Speed Slow Self-selected Fast
Hill Measurement
Condition Parameter C-Leg Genium C-Leg Genium C-Leg Genium
M
S
Uphill DoA (mn ± sd) 0.03 (0.15) 0.00 (0.04)* 0.01 (0.23) -0.01 (0.07) NA 0.00 (0.11) 0.00 (0.00)*
Downhill DoA (mn ± sd) 0.03 (0.06) 0.01 (0.10)* 0.07 (0.10) 0.00 (0.01)* 0.03 (0.04) 0.08 (0.29)
NA
L
S
DoA is degree of Asymmetry. *significantly different from C-Leg (p ≤ 0.05). When differences are significant, effect sizes are S for small,
M for medium and L for large. When differences are not significant, effect size is NA (not available).
DoA is degree of Asymmetry. *significantly different from C-Leg (p ≤ 0.05). When differences are
significant, effect sizes are S for small, M for medium and L for large. When differences are not
symmetry with Genium use. The fourth signifi- groups (24). This reduced gait speed may be related to
significant, effect size is NA (not available).
cant finding yielded no mean difference but, rather, a fear of falling, which could be intensified in ampu-
decreased variance in knee moment with Genium tees, thus contributing to notably reduced walking
use, which was a common trend in the results when speed (21-23,25). Results of this work suggest that
compared to C-Leg. These findings indicate accom- the Genium MPK system may be a desirable inter-
modation and use of the Genium MPK enabled vention for patients with transfemoral limb loss who
increased prosthetic gait control as reflected by more range in function from high-functioning community
favorable DoA values and variability in knee flexion ambulators to older amputees who walk slower and
moment in stance phase during ramp ascent and require more consistent knee performance. Current
descent for unilateral TFA patients capable of unlim- U.S. healthcare regulations mandate the ability to
ited community ambulation. vary cadence in order to be considered for MPK eli-
Previous work has shown improved stability and gibility. Elderly persons with TFA are thus at risk of
knee flexion with C-Leg use compared to non-MPK denial of MPK consideration due to this requirement.
(17,18) and MPK systems alike (19). Use of the However, limited community ambulators have been
Genium MPK has been shown to provide a more con- shown to gain as many (or more) of the benefits of
sistent stance flexion knee angle in a variety of walking MPK technology than their unlimited community
speeds and terrains compared with C-Leg (6). The ambulating peers (26).
results of the present study corroborate the previously
identified kinematic symmetry, as maximum stance CONCLUSION
flexion moments were markedly symmetrical and Accommodation and use of the Genium knee sys-
minimally varied with Genium MPK use compared tem, compared with C-Leg, improved knee moment
with C-Leg use in this sample of high-functioning symmetry in slow speed walking up and down a five
persons with TFA. degree ramp. Additionally, the Genium improved
Although these findings were derived from gen- knee moment symmetry when walking downhill at
erally young, healthy, and highly-active amputees, comfortable speed. At fast walking speed, variance
clinical application exists for elderly amputees who in knee moment symmetry was lower when using
may display high variation in gait parameters, which Genium. These results were found in a sample of
leads to falls (20). Prosthetic components that reduce high functioning persons with unilateral transfemoral
biomechanical gait variability may be an effective amputation; however, the results likely have applica-
means to facilitate independence for amputees at tion in other patients who could benefit from more
increased fall risk (21-23). consistent knee function, such as older patients and
A notable finding from this work is the improved others who have slower walking speeds.
knee moment symmetry and reduced variance in
knee kinetics at slow walking conditions among ACKNOWLEDGMENTS
MPKs. Older individuals are consistently found to Contents of this manuscript represent the opin-
have reduced gait speed relative to those in other age ions of the authors and not necessarily those of the
156 HIGHSMITH ET AL.
U.S. Department of Defense, U.S. Department of the prosthetic knees. J Prosthet Orthot. 2012;24:10-5.
Army, U.S. Department of Veterans Affairs, or any 8. Hafner BJ, Willingham LL, Buell NC, Allyn KJ,
academic or health care institution. Authors declare Smith DG. Evaluation of function, performance,
no conflicts of interest. This project was funded by: and preference as transfemoral amputees transi-
1. The Florida High Tech Corridor/USF CON- tion from mechanical to microprocessor control
NECT (USF Grant #FHT 10-26). of the prosthetic knee. Arch Phys Med Rehabil.
2. Otto Bock Healthcare (USF Grant #6140101200). 2007;88:207-17.
3. The National Institutes of Health Scholars 9. Highsmith MJ. Microprocessor knees: consid-
in Patient Oriented Research (SPOR) grant erations for accommodation and training. J
(1K30RR22270). Prosthet Orthot. 2013;25:60-4.
10. Hillman SJ, Donald SC, Herman J, McCurrach
E, McGarry A, Richardson AM, Robb JE.
REFERENCES Repeatability of a new observational gait score
1. Highsmith MJ, Kahle JT, Lura DJ, Lewandowski for unilateral lower limb amputees. Gait Posture.
AL, Quillen WS, Kim SH. Stair ascent and ramp 2010;32:39-45.
gait training with the Genium knee. Technol 11. O’Connor CM, Thorpe SK, O’Malley MJ,
Innov. 2014;15(4):349-58. Vaughan CL. Automatic detection of gait events
2. Highsmith MJ, Kahle JT, Lura DJ, Dubey RV, using kinematic data. Gait Posture. 2007;25:469-
Carey SL, Quillen WS, Mengelkoch LJ. Short and 74.
mid-distance walking and posturography with 12. Cappozzo A, Cappello A, Della Croce U,
a novel microprocessor knee. Technol Innov. Pensalfini F. Surface-marker cluster design cri-
2014;15(4):359-68. teria for 3-D bone movement reconstruction.
3. Highsmith MJ, Kahle JT, Miro RM, Lura D, IEEE Trans Biomed Eng. 1997;44:1165-74.
Dubey RV, Carey SL, Quillen WS, Mengelkoch 13. Wu G, Siegler S, Allard P, Kirtley C, Leardini
LJ. Perceived differences between the Genium A, Rosenbaum D, Whittle M, D’Lima DD,
and the C-Leg microprocessor prosthetic knees Cristofolini L, Witte H, Schmid O, Stokes I,
in prosthetic-related function and quality of life. Standardization and Terminology Committee
Technol Innov. 2014;15(4):369-75. of the International Society of Biomechanics. ISB
4. Bellmann M, Schmalz T, Ludwigs E, Blumentritt recommendation on definitions of joint coordi-
S. Stair ascent with an innovative microproces- nate system of various joints for the reporting
sor-controlled exoprosthetic knee joint. Biomed of human joint motion--part I: ankle, hip, and
Tech. 2012;57:435-44. spine. International Society of Biomechanics. J
5. Bellmann M, Schmalz T, Ludwigs E, Blumentritt Biomech. 2002;35:543-8.
S. Immediate effects of a new micropro- 14. Cohen J, editor. Statistical power analysis for
cessor-controlled prosthetic knee joint: a the behavioral sciences. 2 ed. Hillsdale(NJ):
nd
comparative biomechanical evaluation. Arch Erlbaum; 1988.
Phys Med Rehabil. 2012;93:541-9. 15. Buhi ER, Goodson P, Neilands TB. Out of sight,
6. Lura DJ, Wernke MM, Carey SL, Kahle JT, Miro not out of mind: strategies for handling missing
RM, Highsmith MJ. Differences in knee flexion data. Am J Health Behav. 2008;32:83-92.
between the Genium and C-Leg microprocessor 16. Kenward MG, Molenberghs G. Last observation
knees whil walking on level ground and ramps. carried forward: a crystal ball? J Biopharm Stat.
Clinical Biomech (Bristol, Avon). 2015;30:175- 2009;19:872-88.
81. 17. Kaufman KR, Levine JA, Brey RH, Iverson BK,
7. Highsmith MJ, Kahle JT, Lewandowski AL, Kim McCrady SK, Padgett DJ, Joyner MJ. Gait and
SH, Mengelkoch LJ. A method for training step- balance of transfemoral amputees using pas-
over-step stair descent gait with stance yielding sive mechanical and microprocessor-controlled
KNEE MOMENT SYMMETRY DURING HILL WALKING 157
prosthetic knees. Gait Posture. 2007;26:489-93. 22. Miller WC, Speechley M, Deathe B. The preva-
18. Segal AD, Orendurff MS, Klute GK, McDowell lence and risk factors of falling and fear of falling
ML, Pecoraro JA, Shofer J, Czerniecki JM. among lower extremity amputees. Arch Phys
Kinematic and kinetic comparisons of transfem- Med Rehabil. 2001;82:1031-7.
oral amputee gait using C-Leg and Mauch SNS 23. Miller WC, Speechley M, Deathe AB. Balance
prosthetic knees. J Rehabil Res Dev. 2006;43:857- confidence among people with lower-limb ampu-
70. tations. Phys Ther. 2002;82:856-65.
19. Thiele J, Westebbe B, Bellmann M, Kraft M. 24. Bohannon RW. Comfortable and maximum
Designs and performance of microproces- walking speed of adults aged 20-79 years: ref-
sor-controlled knee joints. Biomed Tech (Berl). erence values and determinants. Age Ageing.
2014;59:65-77. 1997;26:15-9.
20. Hollman JH, McDade EM, Petersen RC. 25. Miller WC, Deathe AB. A prospective study
Normative spatiotemporal gait parameters in examining balance confidence among individ-
older adults. Gait Posture. 2011;34:111-8. uals with lower limb amputation. Disabil Rehabil.
21. Miller WC, Deathe AB, Speechley M, Koval 2004;26:875-81.
J. The influence of falling, fear of falling, and 26. Kannenberg A, Zacharias B, Probsting E. Benefits
balance confidence on prosthetic mobility and of microprocessor-controlled prosthetic knees
social activity among individuals with a lower to limited community ambulators: systematic
extremity amputation. Arch Phys Med Rehabil. review. J Rehabil Res Dev. 2014;51:1469-96.
2001;82:1238-44.
Technology and Innovation, Vol. 18, pp. 159-165, 2016 ISSN 1949-8241 • E-ISSN 1949-825X
Printed in the USA. All rights reserved. http://dx.doi.org/10.21300/18.2-3.2016.159
Copyright © 2016 National Academy of Inventors. www.technologyandinnovation.org
BIOENERGETIC DIFFERENCES DURING WALKING AND RUNNING
IN TRANSFEMORAL AMPUTEE RUNNERS USING ARTICULATING
AND NON-ARTICULATING KNEE PROSTHESES
M. Jason Highsmith , Jason T. Kahle , Rebecca M. Miro , and Larry J. Mengelkoch 6
4,5
1
1-3
1 School of Physical Therapy & Rehabilitation Sciences, University of South Florida, Tampa, FL, USA
2 Extremity Trauma & Amputation Center of Excellence (EACE), U. S. Department of Veterans Affairs, Tampa, FL, USA
3 319 Minimal Care Detachment, U.S. Army Reserves, Pinellas Park, FL, USA
th
4 OP Solutions, Inc., Tampa, FL, USA
5 Prosthetic Design + Research, Tampa, FL, USA
6 Doctor of Physical Therapy Program, University of St. Augustine for Health Sciences, St. Augustine, FL, USA
Transfemoral amputation (TFA) patients require considerably more energy to walk and run
than non-amputees. The purpose of this study was to examine potential bioenergetic differ-
ences (oxygen uptake (VO2), heart rate (HR), and ratings of perceived exertion (RPE)) for
TFA patients utilizing a conventional running prosthesis with an articulating knee mechanism
versus a running prosthesis with a non-articulating knee joint. Four trained TFA runners
(n = 4) were accommodated to and tested with both conditions. VO2 and HR were significantly
lower (p ≤ 0.05) in five of eight fixed walking and running speeds for the prosthesis with an
articulating knee mechanism. TFA demonstrated a trend for lower RPE at six of eight walking
speeds using the prosthesis with the articulated knee condition. A trend was observed for
self-selected walking speed, self-selected running speed, and maximal speed to be faster for TFA
subjects using the prosthesis with the articulated knee condition. Finally, all four TFA partic-
ipants subjectively preferred running with the prosthesis with the articulated knee condition.
These findings suggest that, for trained TFA runners, a running prosthesis with an articulating
knee prosthesis reduces ambulatory energy costs and enhances subjective perceptive measures
compared to using a non-articulating knee prosthesis.
Key words: Above-knee amputee; Energy costs; No-knee running prosthesis; Oxygen uptake;
Physical therapy; Rehabilitation
INTRODUCTION improve energy costs and ambulatory performance
Transfemoral amputation (TFA) patients require would be functionally important to persons with TFA.
considerably more energy to ambulate than non-am- Mengelkoch et al. (2) recently reported on the
effects of prosthetic foot components on energy costs
putees. Studies by Genin at al. (1) and Mengelkoch and ambulatory performance for TFA patients during
et al. (2) have reported that the energy costs (oxygen walking and running. In this study, all TFA subjects
uptake (VO2)) during walking for TFA patients are were tested using three prosthetic feet conditions: a
30% to 78% greater than for non-amputee control conventional solid ankle cushioned heel (SACH) foot;
subjects. Thus, any prosthetic component that could a general-purpose energy storing and return (ESAR)
_____________________
Accepted July 1, 2016.
Address correspondence to M. Jason Highsmith, Extremity Trauma & Amputation Center of Excellence (EACE), 8900 Grand Oak Circle (151R), Tampa, FL
33637-1022, USA. Tel: +1 (813) 558-3936; Fax: +1 (813) 558-3990; E-mail: [email protected]
159
160 HIGHSMITH ET AL.
foot, which utilized a carbon fiber keel and ankle; study compared VO2 peak attained during running
and a running-specific ESAR foot, which utilized a for two TFA runners utilizing both a conventional
carbon fiber C-shaped keel but was heelless. During running prosthesis with an articulating knee mech-
walking for TFA, at both fixed speeds and self-selected anism and a prosthesis that had a non-articulating
walking speeds (SSWS), no significant differences knee joint (3). Results were mixed in that VO2 peak
were observed for energy costs (VO2, gait economy, was higher for one subject using the prosthesis with
gait efficiency (GE)) among the three prosthetic an articulating knee mechanism and one subject using
feet conditions. However, at SSWS, TFA patients the prosthesis that had the non-articulating knee
demonstrated significantly improved speed with the joint. However, both subjects were able to run longer
general-purpose ESAR foot and running-specific and attained faster speeds using the prosthesis that
ESAR foot compared to the SACH foot (7% and 9% had the non-articulating knee joint. Based on their
respectively). results, these researchers suggested that a prosthe-
Studies reporting the effects of prosthetic foot sis with a non-articulating knee joint may be more
components on energy costs and ambulatory per- energy efficient for TFA runners. However, this study
formance for TFA patients during running are very had several limitations. It utilized only two subjects,
limited. In their study, Mengelkoch et al. (2) reported did not specify an accommodation period for TFA
that TFA patients were not safely able to utilize the patients to utilize each type of prosthesis, performed
SACH foot during running. They observed that TFA the maximal exercise tests for both prostheses with
participants were able to run at speeds up to their only a 30 min rest between tests, and did not include
self-selected running speeds (SSRS) using the gen- information concerning ratings of perceived exertion
eral purpose ESAR foot and the running-specific during testing or subjective preference for running
ESAR foot. At SSRS, the speed deemed comfort- with each type of running prosthesis.
able for sustained distance running, GE was similar Given the limitations in the study by Wening et al.
for TFA subjects using the general-purpose ESAR (3), the purpose of this study was to further examine
foot and the running-specific ESAR foot. However, potential bioenergetic differences for TFA patients
a functional difference was that SSRS was signifi- utilizing a conventional running prosthesis with an
cantly slower using the general-purpose ESAR foot articulating knee mechanism versus a running pros-
(13%) compared to the running-specific ESAR foot. thesis that has a non-articulating knee joint.
Another important observation from this study was
that TFA participants were only able to run at speeds METHODS
greater than SSRS using the running-specific ESAR
foot. These researchers recommended that clinicians Subjects
should recommend and prescribe a running-specific Two male and two female (n = 4) unilateral TFA
ESAR foot for TFA runners interested in performing runners with amputation due to non-vascular causes
more vigorous distance-type running (i.e., for exercise were recruited (Table 1). Participants were healthy
and running competition). recreational runners (K4, Medicare Functional Clas-
It has been observed that some TFA distance sification Level), age ≤45 years, who performed run
-1
-1
runners prefer to run with a prosthesis that has a training 3 to 5 d·week for ≤30 min·d for ≥1 year.
non-articulating knee joint (i.e., a no-knee condition, The study was conducted in accordance with ethical
in which a straight pylon attaches to the prosthetic standards recommended by the Belmont Report (4).
socket and foot components). Anecdotally, it has The study protocol was approved by the University of
been suggested that, during running, increased South Florida’s Institutional Review Board, and each
energy may be required for TFA patients to control study participant provided written informed consent.
the prosthetic articulating knee to prevent it from
buckling, compared to a prosthesis with a non-ar-
ticulating knee joint (3). Previously, a preliminary
KNEE VS. NO-KNEE TRANSFEMORAL RUNNING 161
Table 1. Physical Characteristics of Transfemoral Amputee were then given a one-month accommodation period
Table 1. Physical Characteristics of Transfemoral Amputee Participants
Participants to train and exercise with the non-articulating knee
Gender n = 2 male, n = 2 female condition prior to assessment. To ensure the assign-
Age (y) 28.5 ± 4.2 ment of the order of testing for the two prosthetic
Height (cm) 173.6 ± 6.2
Weight (kg) 68.5 ± 23.4 knee conditions was balanced and randomized, a
2-1
Body Mass Index kg*m 22.5 ± 7.0 block randomization method was used (5,6). Subjects
acclimated to both conditions then tested with each
All amputees were non-dysvascular.
All amputees were non-dysvascular. prosthetic configuration on separate days in random
order.
Study design
Exercise Testing Procedures
Figure 1. Running prostheses: (a) articulated knee prosthesis
and (b) non-articulated knee prosthesis. For exercise testing, participants reported to the
laboratory in the morning following a minimum 8
h fasting period and having refrained from exercise
for approximately 48 h. Participants performed peak
effort exercise testing for each test condition using
an incremental treadmill (Quinton TM65™, Cardiac
Science, Waukesha, WI, USA) walking and running
protocol. Testing began at 0.67 m·s at a 0% grade.
-1
-1
Speed increased every 2 min by 0.233 m·s . Approxi-
mately 48 to 72 h prior to testing, participants came to
the laboratory for a treadmill familiarization session.
At familiarization, individual SSWS & SSRS were
determined for the given prosthetic knee condition
and programmed into the subjects’ respective exercise
tests.
Measurements
Heart rate (HR) and VO2 were measured con-
tinuously by telemetry and breath-by-breath gas
exchange analysis (COSMED K4b ™, Rome, Italy).
2
Calibration was performed immediately prior to test-
ing according to manufacturer specifications. Flow
The study utilized a two-period repeated measures volume measures were calibrated using a 3 L syringe
crossover experimental design. Each TFA participant and gas analyzers were calibrated to known gas mix-
was tested with two prosthetic knee conditions (Fig- tures. Body weight measurements without prosthesis
-1
-1
ure 1). Condition 1: The participant’s usual running were used for VO2 (ml O2·kg ·min ) measurements
prosthesis was used with an articulating knee mecha- relative to body weight. During each minute of exer-
nism. All TFA runners utilized the same articulating cise testing and at peak exercise, participants rated
knee mechanism, and all TFA participants utilized perceived exertion (RPE) using the Borg scale (6 to
a running-specific ESAR foot, but the manufacturer 20) (7). Upon concluding exercise testing with both
differed among subjects (Table 2). Condition 2: The prosthetic knee conditions, participants were asked
participant’s usual running prosthesis fitted with a to subjectively rank the two prosthetic conditions by
pylon (non-articulating knee condition, also called which was most preferred.
no-knee condition) of sufficient length to replace
their preferred articulating knee mechanism. Subjects
162 HIGHSMITH ET AL.
Table 2. Characteristics of the Two Types of Running Prostheses
Table 2. Characteristics of the Two Types of Running Prostheses
Articulated Knee Non-Articulated Knee
Socket n = 2, ischial containment; n = 2, n = 2, ischial containment; n = 2,
sub-ischial. sub-ischial.
Suspension n = 2, elevated vacuum; n = 2, n = 2, elevated vacuum; n = 2,
suction. suction.
a.
Knee n = 4, Total Knee 2000® (Ossur, n = 4, Pylon
Reykjavek, Iceland)
Foot b. n = 2, Flex Run® (Ossur, n = 2, Flex Run® (Ossur,
Reykjavek, Iceland); Reykjavek, Iceland);
n = 2, Nitro® (Freedom n = 2, Nitro® (Freedom
Innovations, Irvine, CA, USA) Innovations, Irvine, CA, USA)
Weight of Prosthesis (kg) c. 3.65 ± 0.40 3.05 ± 0.40
a. The Total Knee 2000 utilizes a mechanical hydraulic knee system.
b. The Flex Run and Nitro prosthetic feet are running-specific, energy storing and return feet.
a. The Total Knee 2000 utilizes a mechanical hydraulic knee system.
c. No significant difference in the weight of the articulated knee prosthesis vs. the non-articulated knee prosthesis.
b. The Flex Run and Nitro prosthetic feet are running-specific, energy storing and return feet.
c. No significant difference in the weight of the articulated knee prosthesis vs. the non-articulated
Data Analysis indicating the non-articulating knee condition cost
knee prosthesis.
Data were verified for accuracy, completeness, and more energy to use at most speeds. Mean RPE was
normality. Parametric tests were selected and applied not significantly different between the two pros-
when appropriate; otherwise, non-parametric equiva- thetic knee conditions. However, as seen in Figure
lent tests were used to compare responses between the 4, there was a trend in which RPE was higher for
two prosthetic knee conditions. It was expected that, the non-articulating knee condition at six of eight
during running, TFA participants would have variable speeds, which suggests more effort was needed at
speed/stage end-points of exercise tolerance for each most speeds with the non-articulating knee condition.
prosthetic knee condition. Thus, some missing data Differences in gait speeds between the two prosthetic
for the TFA participants for the two prosthetic knee knee conditions are shown in Figure 5. There were
conditions was anticipated. We selected, a priori, the no significant differences between the two prosthetic
“last observation carried forward” method as our knee conditions for SSWS, SSRS, or maximal speed
intention-to-treat strategy for imputation of missing attained. However, a trend emerged whereby use
data (8). Statistical analyses were performed using of the articulating knee condition resulted in faster
IBM SPSS software (v22, Armonk, NY, USA). For SSWS, SSRS, and maximal speed. All four TFA par-
all procedures, statistical significance was p < 0.05. ticipants subjectively ranked the prosthesis with the
Values are reported as means ± standard deviation articulated knee condition as their most preferred
(SD). running prosthesis.
RESULTS DISCUSSION
Mean VO2 for five of eight speeds, represented The results in this study differ from those reported
as the shaded region (speeds 1.12 to 2.01 m·sec ) in by Wening et al. (3). In that study, they tested two TFA
-1
Figure 2, were significantly greater (p ≤ 0.05) for the runners and reported only on their end of exercise
non-articulating knee (no-knee) condition, indicat- data. They reported VO2 peak was higher for one
ing the non-articulating knee condition cost more subject using the prosthesis with an articulating knee
energy to use at these speeds. Mean HR for five of mechanism and one subject using the prosthesis that
eight speeds, represented as the shaded region (speeds had the non-articulating knee joint. However, both
1.34 to 2.24 m·sec ) in Figure 3, were significantly subjects were able to run longer and attained faster
-1
greater for the non-articulating knee condition, also speeds using the prosthesis that had the non-articu-
KNEE VS. NO-KNEE TRANSFEMORAL RUNNING 163
Figure 2. Differences in oxygen uptake (VO2) during walking & running for TFA using a
non-articulated knee prosthesis (no-knee) & an articulated knee prosthesis (knee).
Transfemoral amputees (TFA). VO2 at speeds 1.12 – 2.01 m·sec (shaded region), were significantly
-1
greater (p < 0.05) for the non-articulating knee (no-knee) condition.
Figure 3. Differences in heart rate during walking & running for TFA using a non-articulated knee
prosthesis (no-knee) & an articulated knee prosthesis (knee).
Transfemoral amputees (TFA). Heart rate at speeds 1.34 – 2.24 m·sec (shaded region), were signifi-
-1
cantly greater (p < 0.05) for the non-articulating knee (no-knee) condition.
164 HIGHSMITH ET AL. KNEE VS. NO-KNEE TRANSFEMORAL RUNNING 165
Figure 4. Differences in rating of perceived exertion (RPE) during walking & running for TFA lating knee joint. In the current study, we compared REFERENCES
using a non-articulated knee prosthesis (no-knee) & an articulated knee prosthesis (knee). VO2, HR, and RPE data at eight fixed ambulation 1. Genin JJ, Bastien GJ, Franck B, Detrembleur C,
speeds (walking & running) and SSWS and SSRS. We Willems PA. Effect of speed on the energy cost of
observed significant differences at most fixed speeds walking in unilateral traumatic lower limb ampu-
for VO2 and HR, suggesting that energy costs were tees. Eur J Appl Physiol. 2008;103(6):655-63.
lower using the prosthesis with the articulated knee 2. Mengelkoch LJ, Kahle JT, Highsmith MJ. Energy
condition. For RPE, we observed a trend wherein, at
most fixed speeds, RPE was lower using the prosthesis costs & performance of transfemoral amputees
with the articulated knee condition, suggesting that & non-amputees during walking & running: a
less effort was required using that prosthesis. We also pilot study. Prosthet Orthot Int. Forthcoming.
observed that there was a trend for SSWS, SSRS, and 3. Wening J, Stockwell M. Oxygen consumption and
maximal speed attained to be faster for TFA subjects prosthetic moments for two transfemoral ampu-
using the prosthesis with the articulated knee con- tees running with and without a knee. Paper
dition. Finally, all four TFA participants preferred presented at: AAOP 2012. American Academy
th
ambulating with the prosthesis with the articulated of Orthotists & Prosthetists 38 Academy
knee condition. Annual Meeting and Scientific Symposium;
The primary limitation of this study was the small 2012 Mar 21-24; Atlanta, GA.US Department
sample size and thus the generalizability; these find- of Health. HCFA Common Procedure Coding
ings may be limited to TFA runners with similar System (HCPCS) 2001. Springfield (VA): US
characteristics. Moreover, more thorough demo- Department of Commerce, National Technical
Transfemoral amputees (TFA). No significant differences in RPE between knee conditions. graphic (i.e., time since amputation), anthropometric Information Service; 2001.
(i.e., limb length), and history (i.e., exercise history) 4. US Department of Health, Education, and
data could be gathered to facilitate better understand- Welfare. The Belmont report: ethical principles
Figure 5. Differences in self-selected walking speeds (SSWS), self-selected running speeds ing regarding to whom the results would apply. and guidelines for the protection of human
(SSRS), & maximal speeds (MAX) attained for TFA using a non-articulated knee prosthesis (no-
knee) & an articulated knee prosthesis (knee). CONCLUSION subjects of research. Washington (DC): US
These findings suggest that, for trained TFA run- Government Printing Office; 1979 [accessed
2016 Jul 7]. http://www.hhs.gov/ ohrp/human-
ners, a running prosthesis with an articulating knee subjects/guidance/belmont.html.
prosthesis reduces ambulatory energy costs and
enhances subjective perceptive measures compared 5. Kang M, Ragan BG, Park JH. Issues in outcomes
to using a non-articulating knee prosthesis. research: an overview of randomization tech-
niques for clinical trials. J Athl Train. 2008;43:
ACKNOWLEDGMENTS 215-221.
Contents of this manuscript represent the opin- 6. Doig GS, Simpson F. Randomization and
ions of the authors and not necessarily those of the allocation concealment: a practical guide for
U.S. Department of Defense, U.S. Department of the researchers. J Crit Care. 2005;20:187-191.
Army, U.S. Department of Veterans Affairs, or any 7. Borg GA. Psychophysical bases of perceived exer-
academic or health care institution. Authors declare tion. Med Sci Sports Exerc. 1982;14(5):377-81.
no conflicts of interest. This project was funded by 8. Kenward MG, Molenberghs G. Last observation
the National Institutes of Health Scholars in Patient carried forward: a crystal ball? J Biopharm Stat.
Oriented Research (SPOR) grant (1K30RR22270). 2009;19(5):872-8
Transfemoral amputees (TFA). No significant differences in SSWS, SSRS and MAX between knee
conditions.
KNEE VS. NO-KNEE TRANSFEMORAL RUNNING 165
lating knee joint. In the current study, we compared REFERENCES
VO2, HR, and RPE data at eight fixed ambulation 1. Genin JJ, Bastien GJ, Franck B, Detrembleur C,
speeds (walking & running) and SSWS and SSRS. We Willems PA. Effect of speed on the energy cost of
observed significant differences at most fixed speeds walking in unilateral traumatic lower limb ampu-
for VO2 and HR, suggesting that energy costs were tees. Eur J Appl Physiol. 2008;103(6):655-63.
lower using the prosthesis with the articulated knee 2. Mengelkoch LJ, Kahle JT, Highsmith MJ. Energy
condition. For RPE, we observed a trend wherein, at
most fixed speeds, RPE was lower using the prosthesis costs & performance of transfemoral amputees
with the articulated knee condition, suggesting that & non-amputees during walking & running: a
less effort was required using that prosthesis. We also pilot study. Prosthet Orthot Int. Forthcoming.
observed that there was a trend for SSWS, SSRS, and 3. Wening J, Stockwell M. Oxygen consumption and
maximal speed attained to be faster for TFA subjects prosthetic moments for two transfemoral ampu-
using the prosthesis with the articulated knee con- tees running with and without a knee. Paper
dition. Finally, all four TFA participants preferred presented at: AAOP 2012. American Academy
th
ambulating with the prosthesis with the articulated of Orthotists & Prosthetists 38 Academy
knee condition. Annual Meeting and Scientific Symposium;
The primary limitation of this study was the small 2012 Mar 21-24; Atlanta, GA.US Department
sample size and thus the generalizability; these find- of Health. HCFA Common Procedure Coding
ings may be limited to TFA runners with similar System (HCPCS) 2001. Springfield (VA): US
characteristics. Moreover, more thorough demo- Department of Commerce, National Technical
graphic (i.e., time since amputation), anthropometric Information Service; 2001.
(i.e., limb length), and history (i.e., exercise history) 4. US Department of Health, Education, and
data could be gathered to facilitate better understand- Welfare. The Belmont report: ethical principles
ing regarding to whom the results would apply.
and guidelines for the protection of human
subjects of research. Washington (DC): US
CONCLUSION Government Printing Office; 1979 [accessed
These findings suggest that, for trained TFA run- 2016 Jul 7]. http://www.hhs.gov/ ohrp/human-
ners, a running prosthesis with an articulating knee subjects/guidance/belmont.html.
prosthesis reduces ambulatory energy costs and
enhances subjective perceptive measures compared 5. Kang M, Ragan BG, Park JH. Issues in outcomes
to using a non-articulating knee prosthesis. research: an overview of randomization tech-
niques for clinical trials. J Athl Train. 2008;43:
ACKNOWLEDGMENTS 215-221.
Contents of this manuscript represent the opin- 6. Doig GS, Simpson F. Randomization and
ions of the authors and not necessarily those of the allocation concealment: a practical guide for
U.S. Department of Defense, U.S. Department of the researchers. J Crit Care. 2005;20:187-191.
Army, U.S. Department of Veterans Affairs, or any 7. Borg GA. Psychophysical bases of perceived exer-
academic or health care institution. Authors declare tion. Med Sci Sports Exerc. 1982;14(5):377-81.
no conflicts of interest. This project was funded by 8. Kenward MG, Molenberghs G. Last observation
the National Institutes of Health Scholars in Patient carried forward: a crystal ball? J Biopharm Stat.
Oriented Research (SPOR) grant (1K30RR22270). 2009;19(5):872-8
Technology and Innovation, Vol. 18, pp. 167-173, 2016 ISSN 1949-8241 • E-ISSN 1949-825X
Printed in the USA. All rights reserved. http://dx.doi.org/10.21300/18.2-3.2016.167
Copyright © 2016 National Academy of Inventors. www.technologyandinnovation.org
THE EFFECT OF TRANSFEMORAL INTERFACE DESIGN
ON GAIT SPEED AND RISK OF FALLS
Jason T. Kahle , Tyler D. Klenow , William J. Sampson , M. Jason Highsmith 5-7
1,2
3
4
1 OP Solutions, Tampa, FL, USA
2 Prosthetic Design + Research, Tampa, FL, USA
3 Prosthetics and Sensory Aids Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA
4 Sampson’s Prosthetic & Orthotic Laboratories, Schenectady, NY, USA
5 School of Physical Therapy & Rehabilitation Sciences, University of South Florida, Tampa, FL, USA
6 Extremity Trauma & Amputation Center of Excellence (EACE), U.S. Department of Veterans Affairs, Tampa, FL, USA
th
7 319 Minimal Care Detachment, U.S. Army Reserves, Pinellas Park, FL, USA
Falls and diminished walking capacity are impairments common in persons with transfemoral
amputation (TFA). Reducing falls and optimizing walking capacity through such means as
achieving a more normal gait speed and community ambulation should be considered when
formulating the prosthetic prescription. Because walking capacity and balance confidence are
compromised with TFA, these outcomes should be considered when evaluating interfaces
for transfemoral prosthetic users. The purpose of this study was to compare the effect of TFA
interface design on walking capacity and balance confidence A retrospective cohort design
was utilized involving unilateral TFA patients who used ischial ramus containment (IRC)
and High-Fidelity (HiFi) interfaces (independent variables). Dependent variables included
the Activity-specific Balance Scale (ABC) and the two-minute walk test (2MWT). Complete
records were available for 13 patients (n = 13). The age range was 26 to 58 years. Three patients
functioned at the K4 activity level, whereas all others functioned at the K3 level. Mean ABC
scores were significantly different (p ≤ 0.05) at 77.2 (±16.8; 35.6 to 96.9) for IRC and 90.7 (±5.7;
77.5 to 98.7) for HiFi. The mean distance walked on the 2MWT was 91.8 m (±22.0, 58.3 to
124.7) for IRC compared to 110.4 m (±28.7; 64.7 to 171.1) for the HiFi socket (p ≤ 0.05). Al-
ternative transfemoral interface design, such as the HiFi socket, can improve walking capacity
and balance confidence in higher-functioning TFA patients.
Key words: Above the knee amputee; Activity balance confidence; Compression release socket;
High-fidelity interface; Ischial containment; Prosthetic socket; Walking tests
_____________________
Accepted July 1, 2016.
Address correspondence to Jason T. Kahle, OP Solutions, 12206 Bruce B. Downs Blvd., Tampa, FL 33612, USA. Tel: +1 (813) 971-1100; Fax: +1 (813) 971-
9300; E-mail: [email protected]
167
168 KAHLE ET AL.
INTRODUCTION METHODS
Falling is prevalent among the amputee population A retrospective cohort design was utilized. All data
particularly when comorbidities exist. Developing were collected in accordance with the Declaration of
and studying fall prevention is a priority for reducing Helsinki. In order to be included in the record review,
adverse effects and costs. Falls cost $23.3 billion annu- subjects had to meet the following eligibility criteria:
ally. Their mean cost ranges from $3,476 per faller to 1. Unilateral TFA
$10,749 per injurious fall to $26,483 per fall requiring 2. 18 to 60 years of age
hospitalization. Approximately 20% of falls require 3. K2 or higher activity level
medical attention (1,2). Annually, 52.4% of lower limb 4. History of prosthetic ambulation without an
amputees report falling, whereas 49.2% report fear of assistive device ≥2 years
falling (3,4). Falling and fear of falling are common 5. No other comorbidities
among patients with transfemoral amputation (TFA), 6. History of use of both socket conditions of
who have yearly fall rates as high as 66% (5,6). Further interest
studies are required to characterize the mechanisms 7. Complete outcome assessment for both socket
of falling and to develop appropriate prevention strat- conditions
egies (7). Rehabilitation of amputee Service Members,
Veterans, and civilians using prosthetic devices is a Subjects were excluded if they did not meet all inclu-
priority of the Department of Defense (DoD), the sion criteria.
Veterans Health Administration (VA), and Centers Independent Variable: Interfaces
for Medicare and Medicaid Services (CMS), respec- The Standard of Care (SOC) interface is the ischial
tively. Fall prevention programs have been established ramus containment (IRC). The experimental interface
and evaluated, including exercise programs designed is the High-Fidelity Interface™ (HiFi, biodesigns, inc.,
to improve function and balance (1,2). However, the Westlake Village, CA,USA) TFA interface design (12)
role of the prosthetic interface in maximizing balance (Figure 1).
confidence is poorly studied. Exploring interventions,
including prosthetic interfaces that may prevent falls
and reduce healthcare costs, is a priority in amputee
care (8,9).
In addition to reducing falls, walking capacity
is another important factor in determining func-
tion for TFA patients (10). Poor walking capacity,
as evidenced by lower gait speed, is associated with
increased comorbidity, falls, and mortality (11).
Optimizing walking capacity through such means as
achieving a more normal gait speed and community
ambulation should be considered when formulating
the prosthetic prescription. Because walking capacity
and balance confidence are compromised with TFA,
these outcomes should be considered when evalu-
ating interfaces for transfemoral prosthetic users.
Therefore, the purpose of this study was to compare
the effect of TFA interface design on walking capacity
and balance confidence.
F
Figure 1. Subischial Transfemoral HiFi Interface. igure 1. Subischial transfemoral HiFi interface.
TRANSFEMORAL SOCKET IMPACT ON GAIT & FALLS 169
Dependent Variables: Outcome Measures Data Collection Timeline
The following outcome measures were routinely Patients included in the review were initially fit
utilized during clinical evaluation of new prosthetic into an IRC interface and assessed following at least
fittings: 30 d of accommodation. All data were documented
in the clinical prosthetic progress notes. Following
The Activity-specific Balance Confidence Scale eventual rejection or failure, patients were subse-
(ABC): The ABC Scale is a 16-item self-report quently fit with a HiFi interface. Again, following at
measure of balance confidence in performing least 30 d accommodation, patients were assessed and
various activities of daily living (ADL) without results recorded in their prosthetic clinic records.
falling. Items are scored on a rating scale from
0 to 100, with higher scores reflecting higher Statistical Analyses
levels of balance confidence. An average score is Data were entered into a database and evaluated
calculated by adding all item scores and dividing for completeness and accuracy. Central tendency and
by the total number of items. The ABC can be variance were calculated. Parametric tests were used
administered in 10 to 20 min and is appropriate when appropriate considering data dependency and
for use in the clinical environment (13). normality; otherwise, equivalent non-parametric tests
Two-Minute Walk Test (2MWT): The 2MWT were used. Statistical significance was set a priori at p
was used to determine walking capacity. The ≤ 0.05. IBM SPSS (v21, Armonk, NY, USA) was used
six-minute walk test (6MWT) is highly regarded for statistical analysis.
in clinical care; however, Reid et al. determined
the 2MWT is as predictive as the 6MWT in RESULTS
determining an amputee’s ability to ambulate Complete records were available for eleven males
in the community. The 2MWT was used given and two females (n = 13). The age range was 26 to
its comparable performance to the 6MWT and 58 years. Three patients functioned at the K4 activity
because it takes less time to administer within level and the remaining 10 at the K3 level. Mean body
the clinical environment (14). mass was 85.2 kg (57.7 to 137.7). Subjects’ mean
Table 1. Sociodemographic Data
Table 1. Sociodemographic Data
Subject Prosthetic Gender Amputated Activity Height Weight
History Side Level (Ft. In) (lbs.)
(y)
1 3 M L K3 5.5 190
2 13 M L K3 5.9 222
3 5 M R K3 5.1 165
4 5 M R K3 5.9 187
5 6 M R K3 5.8 303
6 20 M L K4 5.9 185
7 16 M L K3 5.9 168
8 33 F L K4 5.3 127
9 40 M L K3 5.7 180
10 8 M L K3 5.9 156
11 25 M L K3 5.9 145
12 3 M L K3 6 220
13 10 F R K4 5.7 190
Table 1.
170 KAHLE ET AL.
Figure 2. Activity-specific Balance Confidence Scale. Subjects’ mean ABC score was 77.2 (±16.8; 35.6 to
96.9) for the IRC and 90.7(±5.7; 77.5 to 98.7) for the HiFi. The difference was statistically significant (p =
0.02).
Figure 3. Two-Minute Walk Test. The aggregated mean distance walked on the 2MWT was 91.8 m (±22.0,
58.3 to 124.7) on the IRC compared to 110.4 m (±28.7; 64.7 to 171.1) for the HiFi socket (p = 0.0001).
TRANSFEMORAL SOCKET IMPACT ON GAIT & FALLS 171
prosthetic use was 14.4 years (3 to 40) (Table 1). MDC has not been established in the TFA population.
Subjects’ mean ABC score was 77.2 (±16.8; 35.6 In this study, the mean difference was a 23.5 point
to 96.9) for the IRC and 90.7 (±5.7; 77.5 to 98.7) for improvement (p = 0.02) for the users of the HiFi
the HiFi. The difference was statistically significant (90.7) versus the IRC (77.2) sockets. An ABC score of
(p = 0.02) (Figure 2). The aggregated mean distance 68 is associated with falling post-stroke (18). In this
walked on the 2MWT was 91.8 m (±22.0, 58.3 to study, both groups were above this fall risk threshold.
124.7) on the IRC compared to 110.4 m (±28.7; 64.7 Our hypothesis that the improved prosthetic control
to 171.1) for the HiFi socket (p = 0.0001) (Figure 3). would yield increased ABC scores was supported.
Further, the HiFi group reported ABC scores higher
DISCUSSION than other patient populations, such as those suffering
We hypothesized that the less cumbersome walls from Parkinson’s or stroke and the elderly. Conversely,
and subischial trimlines afforded by the HiFi, com- when using the IRC socket, ABC scores tended to be
pared with SOC interfaces, would offer improved similar to scores in these populations (18-20).
freedom of movement. This freedom of movement The socket has been identified as the most import-
may enable improved walking capacity and improved ant prosthetic element; however, TFA prosthetic
prosthetic control as evidenced by improved ABC socket fit is problematic using SOC IRC prosthetic
scores. The minimum detectable change (MDC) for socket interface. A subischial compression and tissue
the 2MWT has been previously reported as 34.3 m release design such as the HiFi may improve comfort
(15). This threshold was not reached in this study. and increase user control by utilizing femoral control.
However, in three previous studies comparing lower This could lead to improved walking capacity and
extremity amputees using the 2MWT, distances confidence, which could reduce falls. Discomfort
walked reached up to 140 m (14). The majority of and lack of control have been associated with poor
amputees tested in these studies had transtibial level socket fit. This reduces function of the prosthesis
amputation. In this study, only transfemoral level for the amputee. Although common, the SOC IRC
amputees were evaluated. TFA patients using the interface design has potential limitations, such as
HiFi socket achieved distances similar to transtibial limiting range of motion, decreasing comfort, and
amputees in previous studies. Patients’ walking capac- interfering with urogenital function, and potential
ity was significantly improved with use of the HiFi fitting complications affecting overall quality of life
interface relative to the SOC alternative, supporting (21,22). Benefits of a novel TFA interface such as
this portion of the hypothesis. Reduced gait speed, an the HiFi could potentially address areas identified as
indicator of walking capacity, has been associated with problematic and lead to alternative interface designs
falling (16). This is of particular relevance in the TFA that improve quality of life among TFA patients.
population. Maintaining optimal walking capacity, A common conclusion among the aforementioned
including normal or near normal gait speed, should studies is that falls in amputees can be mitigated with
be of primary concern in prosthetic design. Prosthetic training programs and alternative interventions. To
elements, including the socket, must be considered begin a strength and walking program, the use of an
as potential factors in assisting a patient to achieve effective prosthesis is imperative. Collectively, prior
variable cadence and faster gait speeds. In this study, authors also agreed that ongoing research is required
the HiFi interface design was determined to have a to develop appropriate intervention strategies to ame-
significant effect on improving walking capacity. liorate fall risk (6,23-25). Such interventions should
Deathe et al. assessed 17 outcome measures used consider alternative prosthetic interfaces.
in amputee clinical trials to assess mobility (17). The
ABC scale was recognized as being valid and widely Limitations
used. The MDC for the ABC has been previously A limitation of this study was the small patient
reported as 11 and 13 in studies of other pathologies. population exposed to both socket designs. Socket
172 KAHLE ET AL.
studies are expensive, and it can be difficult to control 3. Miller WC, Speechley M, Deathe B. The preva-
attrition. Interface studies are particularly challeng- lence and risk factors of falling and fear of falling
ing due to the socket being an intimate part of the among lower extremity amputees. Arch Phys
prosthesis that can require weeks of accommodation Med Rehabil. 2001;82(8):1031-7.
prior to effective use. A paucity of funding has created 4. Miller WC, Speechley M, Deathe AB. Balance
a void in the understanding of a prosthetic socket confidence among people with lower-limb ampu-
interface’s role in affecting falls and walking capacity tations. Phys Ther. 2002;82(9):856-65.
(26). 5. Gooday HM, Hunter J. Preventing falls and
stump injuries in lower limb amputees during
CONCLUSION inpatient rehabilitation: completion of the audit
Prosthetic clinical documentation and outcome cycle. Clin Rehabil. 2004;18(4):379-90.
measure implementation can be effective means of 6. Pauley T, Devlin M, Heslin K. Falls sustained
demonstrating changes and improvements in clin- during inpatient rehabilitation after lower limb
ical interventions. The Activity-specific Balance amputation: prevalence and predictors. Am J
Confidence Scale and two-minute walk test are valid Phys Med Rehabil. 2006;85(6):521-32.
measures that can be used to determine differences 7. Yu WY, Hwang HF, Hu MH, Chen CY, Lin MR.
among intervenions in the transfemoral amputee Effects of fall injury type and discharge place-
population. Alternative transfemoral interface design, ment on mortality, hospitalization, falls, and
such as the HiFi socket, can improve walking capacity ADL changes among older people in Taiwan.
and balance confidence in higher functioning patients Accid Anal Prev. 2013;50:887-94.
with transfemoral amputation. 8. Davis JC, Donaldson MG, Ashe MC, Khan KM.
The role of balance and agility training in fall
ACKNOWLEDGMENTS reduction. A comprehensive review. Eura Med-
icophys. 2004;40(3):211-21.
Contents of this manuscript represent the opinions 9. Dyer D, Bouman B, Davey M, Ismond KP. An
of the authors and not necessarily those of the U.S. intervention program to reduce falls for adult
Department of Defense, U.S. Department of the Army, in-patients following major lower limb ampu-
U.S. Department of Veterans Affairs, or any academic tation. Healthc Q. 2008;11(3 Spec No.):117-21.
or health care institution. Authors declare no conflicts 10. [CMS] Centers for Medicare and Medicaid Ser-
of interest. This project was partially supported by vices. Healthcare Common Procedure Coding
the National Institutes of Health Scholars in Patient System. Springfield (VA): U.S. Department of
Oriented Research (SPOR) grant (1K30RR22270). Commerce, National Technical Information
Service; 2007.
REFERENCES 11. Fritz S, Lusardi M. White paper: “walking
1. Gillespie LD, Robertson MC, Gillespie WJ, Lamb speed: the sixth vital sign”. J Geriatr Phys Ther.
SE, Gates S, Cumming RG, Rowe BH. Interven- 2009;32(2):46-9.
tions for preventing falls in older people living 12. Alley RD, Walley, Albuquerque MJ, Altobelli DE.
in the community. Cochrane Database Syst Rev. Prosthetic sockets stabilized by alternating areas
2012 [accessed 2016 Jul 30]. http://onlinelibrary. of tissue compression and release. J Rehabil. Res.
wiley.com/doi/10.1002/ 14651858.CD007146. Dev. 2011;48(6):679-96.
pub2/abstract;jsessionid=BA68AD47531B- 13. Powell LE, Myers AM. The activities-specific
F7024E46E7B7A696A558.f04t03 balance confidence (ABC) scale. J Gerontol A
2. Tinetti ME, Mendes de Leon CF, Doucette JT, Biol Med Sci. 1995;50A(1):M28-34.
Baker DI. Fear of falling and fall-related efficacy 14. Reid L, Thomson P, Besemann M, Dudek N.
in relationship to functioning among communi- Going places: does the two-minute walk test pre-
ty-living elders. J Gerontol. 1994;49(3):M140-7. dict the six-minute walk test in lower extremity
TRANSFEMORAL SOCKET IMPACT ON GAIT & FALLS 173
amputees? J Rehabil Med. 2015;47(3):256-61. with vacuum-assisted suspension comparison
15. Resnik L, Borgia M. Reliability of outcome mea- of hip kinematics, socket position, contact pres-
sures for people with lower-limb amputations: sure, and preference: ischial containment versus
distinguishing true change from statistical error. brimless. J Rehabil Res Dev. 2013;50(9):1241-52.
Phys Ther. 2011;91(4):555-65. 22. Kahle JT, Highsmith MJ. Transfemoral interfaces
16. Montero-Odasso M, Schapira M, Soriano ER, with vacuum assisted suspension comparison
Varela M, Kaplan R, Camera LA, Mayorga LM. of gait, balance, and subjective analysis: ischial
Gait velocity as a single predictor of adverse containment versus brimless. Gait Posture.
events in healthy seniors aged 75 years and older. 2014;40(2):315-20.
J Gerontol A Biol Med Sci. 2005;60(10):1304-9. 23. Vrieling AH, van Keeken HG, Schoppen T, Hof
17. Deathe AB, Wolfe DL, Devlin M, Hebert JS, AL, Otten B, Halbertsma JP, Postema K. Gait
Miller WC, Pallaveshi L. Selection of out- adjustments in obstacle crossing, gait initiation
come measures in lower extremity amputation and gait termination after a recent lower limb
rehabilitation: ICF activities. Disabil Rehabil. amputation. Clin Rehabil. 2009;23(7):659-71.
2009;31(18):1-19. 24. Vanicek N, Strike S, McNaughton L, Polman R.
18. Botner EM, Miller WC, Eng JJ. Measurement Postural responses to dynamic perturbations in
properties of the Activities-specific Balance Con- amputee fallers versus nonfallers: a comparative
fidence Scale among individuals with stroke. study with able-bodied subjects. Arch Phys Med
Disabil Rehabil. 2005;27(4):156-63. Rehabil. 2009;90(6):1018-25.
19. Huang TT, Wang WS. Comparison of three 25. Vanicek N, Strike S, McNaughton L, Polman R.
established measures of fear of falling in commu- Gait patterns in transtibial amputee fallers vs.
nity-dwelling older adults: psychometric testing. non-fallers: biomechanical differences during
Int J Nurs Stud. 2009;46(10):1313-9. level walking. Gait Posture 2009;29(3):415-20.
20. Mak MK, Pang MY, Mok V. Gait difficulty, 26. Highsmith MJ, Kahle JT, Miro RM, Orendurff
postural instability, and muscle weakness are MS, Lewandowski AL, Oriolla JJ, Sutton B, Ertl
associated with fear of falling in people with Par- JP. Prosthetic interventions for transtibial ampu-
kinson’s disease. Parkinsons Dis. 2012 [accessed tees: a systematic review and meta-analysis of
2016 Jul 30]. doi: 10.1155/2012/901721. high-quality prospective literature and systematic
21. Kahle JT, Highsmith MJ. Transfemoral sockets reviews. J Rehabil Res Dev. 2016;53(2):157-84.
Technology and Innovation, Vol. 18, pp. 175-183, 2016 ISSN 1949-8241 • E-ISSN 1949-825X
Printed in the USA. All rights reserved. http://dx.doi.org/10.21300/18.2-3.2016.175
Copyright © 2016 National Academy of Inventors. www.technologyandinnovation.org
COMPARATIVE EFFECTIVENESS OF AN ADJUSTABLE
TRANSFEMORAL PROSTHETIC INTERFACE ACCOMMODATING
VOLUME FLUCTUATION: CASE STUDY
Jason T. Kahle , Tyler D. Klenow , M. Jason Highsmith
4-6
3
1,2
1 OP Solutions, Tampa, FL, USA
2 Prosthetic Design + Research, Tampa, FL, USA
3 Prosthetics and Sensory Aids Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA
4 School of Physical Therapy & Rehabilitation Sciences, University of South Florida, Tampa, FL, USA
5 Extremity Trauma & Amputation Center of Excellence (EACE), U.S. Department of Veterans Affairs, Tampa, FL, USA
th
6 319 Minimal Care Detachment, U.S. Army Reserves, Pinellas Park, FL, USA
The socket-limb interface is vital for functionality and provides stability and mobility for the
amputee. Volume fluctuation can lead to compromised fit and function. Current socket tech-
nology does not accommodate for volume fluctuation. An adjustable interface may improve
function and comfort by filling this technology gap. The purpose of this study was to compare
the effectiveness of the standard of care (SOC) ischial ramus containment to an adjustable
transfemoral prosthetic interface socket in the accommodation of volume fluctuation. A
prospective experimental case study using repeated measures of subjective and performance
outcome measures between socket conditions was employed. In the baseline volume condition,
the adjustable socket improved subjective and performance measures 19% to 37% over SOC,
whereas the two-minute walk test demonstrated equivalence. In the volume loss condition, the
adjustable socket improved all subjective and performance measures 22% to 93%. All aggregated
data improved 16% to 50% compared with the SOC. In simulated volume gain, the SOC socket
failed, while the subject was able to complete the protocol using the adjustable socket. In this
case study, the SOC socket was inferior to the comparative adjustable transfemoral amputation
interface in subjective and performance outcomes. There is a lack of clinical trials and evidence
comparing socket functional outcomes related to volume fluctuation.
Key words: Amputee; Ischial containment; Lower extremity amputee; Limb loss; Rehabilita-
tion; Socket
_____________________
Accepted July 1, 2016.
Address correspondence to: Jason T. Kahle, OP Solutions, 12206 Bruce B. Downs Blvd., Tampa, FL 33612, USA. Tel: (813) 971-1100; Fax: (813) 971-9300;
E-mail: [email protected]
175
176 KAHLE ET AL.
INTRODUCTION fluctuations require an iterative process involving
In the U.S., approximately two million people numerous trips to the prosthetist for socket adjust-
live with limb loss (1). It is estimated that by 2050, ments. Poor fit can lead to prosthetic abandonment
nearly 3.6 million Americans will be living with (9). RL volume management is a common issue for
lower extremity (LE) limb loss (1). Of the two mil- prosthetic users, especially during the intermedi-
lion American amputees, approximately 86% are ate recovery stage of amputee rehabilitation when
individuals living with lower limb loss and 18.5% the most rapid volume fluctuation occurs (7). It has
have transfemoral amputation (TFA) (2). In spite of been shown that limb volume decreases 17% to 35%
this increasing amputee population, there is limited over the first 160 d post amputation, 7% to 10% in
prosthetic research, resulting in healthcare service the 12-month post-operative period and approxi-
gaps, excess hospital utilization, and increased cost mately 2% on a daily basis thereafter, thus requiring
to patients and payors (3). Addressing these issues is patient-provider coordination (7,10). In addition,
of critical importance since rehabilitation care, fitting chronic volume change may continue for up to 12
of prostheses, and adjustment of devices alone were to 18 months post amputation due to tissue atrophy
the fifteenth most expensive condition treated in U.S. and indefinite diurnal volume fluctuations. Poor vol-
hospitals in 2011, with a total cost of more than $5.4 ume management can result in a variety of secondary
billion (4). adverse effects of prosthetic use, including ulcers,
The socket-limb interface is vital for functionality verrucous hyperplasia, and osteomyelitis (7). These
and provides stability and mobility for the ampu- effects may lead to further amputation and re-hospi-
tee. An inadequate fit may lead to skin breakdown, talization, which contributes to the annual $8 billion
thereby limiting mobility and requiring additional expenditure on amputee hospital care (11).
clinician time, replacement components, and a pos- Traditional rigid sockets do not accommodate vol-
sible remaking of the prosthesis altogether (5). As a ume fluctuations. Poor fit can cause skin ulcerations
result, Medicare data shows that 45% of the overall and infection and may lead to revision amputation
$750 million in Medicare expenditures on prosthetic (12). Furthermore, socket discomfort is common
among LE amputees and may delay prosthetic use,
technology each year were for socket-related codes.
Successful socket fitting reduces this economic prevent return to normal function, compromise
burden and increases prosthetic usage. Amputees patient outcomes, and increase healthcare costs. The
encounter multiple challenges during their recovery, primary cause for failure of amputee prostheses is
user dissatisfaction associated with poor socket fit
rehabilitation, and reintegration into their homes and and comfort (9,12-14). In addition to the unmet
communities. Learning and adopting new strategies need in addressing comfort, there is a considerable
for basic mobility, personal hygiene, and activities of technology gap in the area of socket fabrication and
daily living with a prosthesis is difficult (6). Compli- access. Therefore, the objective of this prospective
cating this process, the residual limb (RL) naturally experimental clinical case study is to compare the
goes through a period of volume fluctuation post effectiveness of the standard of care (SOC) ischial
amputation that impacts fit (7). Newly amputated ramus containment (IRC) to an adjustable transfemo-
limbs commonly undergo reduction in size, shape, ral prosthetic interface socket in the accommodation
and volume (7,8). This progression occurs in two of volume fluctuation by observing both functional
phases: 1) rapid, acute shrinkage immediately fol- and subjective outcomes.
lowing amputation and 2) progressive stabilization of
volume one year post amputation. These changes are METHODS
dependent on individual lifestyle, activity level, and Methods were in accordance with the Declaration
weight. Moreover, amputees experience daily volume of Helsinki, and the subjects provided informed con-
fluctuations influenced by multiple factors, including sent.
diet, environment, and weather conditions. These