Within the next three decades, well over 1 million Americans will be living with an amputation of the lower limb.1 Prosthetic users report that one of the most important issues they face involves maintaining prosthetic fit in response to diurnal fluctuations in the volume of the residual limb.2 A study by Board et al.3 that used casting methods to assess limb volume before and after 30 minutes of treadmill walking observed changes of 6% to 7% as a result of the task and suggested that these changes were equivalent to the volume losses that occur during a work day.
Fluctuations in limb volume are often managed by donning socks, although this requires the user to carry socks of varying ply, know the proper overall ply to add, and know the frequency with which to do so. Too many ply can cause proximal tightness and loss of total contact with the socket, resulting in hematomas or verrucous hyperplasia; too few ply can promote sinking into the socket and excessive loading of the distal end of the limb. The need to rely on socks to maintain fit may be minimized through use of vacuum-assisted socket suspension (VASS), in which the socket is tightly anchored to the limb by removing air from the interface between the two using negative pressure. The reduction of air in this sealed chamber results in constant levels of high vacuum. In two separate studies, VASS has been shown to limit fluctuations in limb volume during gait.3,4 The use of VASS has also been shown to lower levels of peak socket pressures,5 reduce pistoning (movement of the limb within the socket, an indicator of poor fit),3 and reduce reported problems with blistering6 and, in a case study, was associated with cessation of contact dermatitis and hair-regrowth over the entire limb.7 In a novel application of VASS to a partial foot prosthesis, Arndt et al.8 showed compelling images (e.g., Figure 4 from that study) demonstrating healthier overall appearance of the skin of the plantar surface of the foot after 10 days of use, later followed by softening of the skin and apparent increases in hydration and vascularization. Although long-term use of VASS may improve limb health, the effects on function and mobility remain largely unknown. Moreover, the effects of acute loss of vacuum in current VASS users is unknown, despite the fact that many VASS users may periodically experience problems with their vacuum system such as leaks or malfunctioning pump.
With regard to function and mobility, VASS may help to reduce the energetic costs of walking. It is well established that the energetic costs of gait for individuals with amputation are significantly greater than those of able-bodied individuals,9 with costs increasing anywhere from 20% to >100% depending on level of amputation.10 This is not trivial as expending greater energy during gait can prevent the user from completing all desired activities of daily living limit social participation, and can contribute to subsequent disability and even mortality.11–13 The improved fit and reduction in pistoning reported with VASS should logically limit dissipation of energy during gait and in turn reduce the energetic costs of walking. Additionally, by providing a more secure link between the limb and socket, VASS may limit adaptive behaviors used to maintain fit, such as prolonged activation of hip muscles during preswing and throughout the swing phase of gait,14 which can directly contribute to energetic costs. Accordingly, it is quite possible that for current VASS users, the loss of active vacuum would result in gait adaptations that increase the energetic costs of walking,
By reducing peak intrasocket pressure, VASS use may also improve socket comfort. Accordingly, VASS users may experience immediate reductions in comfort after acute loss of vacuum. Moreover, given that socket discomfort is one of the leading factors limiting prosthetic use,15 if VASS improves comfort, then over time, prosthetic users might wear and use their prosthesis more with VASS. The presumably tighter linkage between the user and the prosthesis that VASS offers may also facilitate faster self-selected and maximum walking speed and better performance on demanding locomotor tasks (i.e., those other than steady-state, level-ground walking such as turning while walking or ascending stairs). Again, these effects could be expected to both acutely impair functional mobility, just after the loss of vacuum, and could similarly lead to improvements in functional mobility in current VASS versus non-VASS users. In the latter case, such improvements could help to improve overall well-being.
The purpose of this study was to quantify the extent to which VASS improved energetics of gait and both performance-based (timed up and go [TUG], 10-m walk test [10MWT], and self-selected speed [SSS]) and self-reported outcome measures related to function, mobility, and prosthetic use. We hypothesized that in current users of VASS, a loss of vacuum would: H1) increase the energetic costs of gait, quantified as the cost of transport at SSS; H2) worsen performance-based outcomes measures related to function and mobility, including TUG, 10MWT, and self-selected walking speed; and H3) reduce socket comfort. We also expected that H4) long-term use of VASS would improve self-reported outcomes measures related to prosthetic wear and use (measured by the Houghton scale16), locomotor capabilities (measured by the Locomotor Capabilities Index 5 [LCI5]17,18), and related to domains of the Prosthetic Evaluation Questionnaire (PEQ), most importantly the well-being scale. We also expected to validate the following previously reported findings: 1) VASS reduces asymmetries in step length (SL) and stance time (ST)3 and 2) VASS improves balance confidence as quantified by the Activities-Specific Balance Confidence (ABC) score.6
Thirty-nine participants provided written informed consent to participate in either a single laboratory data-collection session (study 1) and/or completion of online questionnaires (study 2). The University of Illinois at Chicago Institutional Review Board approved all protocols. Participant recruitment took place through flyers posted at Chicagoland area prosthetic clinics, during face-to-face visits with clinicians, or through an online advertisement posted on the Amputee Coalition website. Inclusion criteria included 1 year or more since amputation and ability to walk without assistance for 6 minutes. We recruited 19 current VASS users to the VASS group for participation in study 1 and study 2. These individuals were required to have used their current vacuum system for 6 months or more, with all but one participant having VASS for less than 1 year. The remaining subjects were recruited as part of the non-VASS group and these individuals participated only in study 2. To maximize sample size, we did not match the non-VASS and VASS group with regard to demographics.
Eighteen of the recruited VASS participants completed at least one aspect of either study 1 or study 2 and 18 of the non-VASS participants were responsive to the online surveys. Figure 1 summarizes subject recruitment and the number of data points entered into each aspect of the analysis, and Table 1 summarizes demographics for the 36 subjects (18 VASS and 18 non-VASS) who participated in one or more aspects of the study. Where available, data are presented for the suspension of non-VASS subjects. Data are not available for the socket-type of non-VASS subjects.
PROTOCOL: STUDY 1 (LABORATORY ASSESSMENT OF THE VACUUM-ASSISTED SOCKET SUSPENSION GROUP)
Study 1 consisted of three tasks presented in the following order: 1) measuring kinematics of gait, 2) measuring functional mobility outcomes, 3) measuring the cost of transport at SSS. Participants wore their current socket during all tasks. Three different suspension conditions were considered: 1) VASS; 2) suction—the prosthetists inactivated the vacuum of the VASS system and a one-way valve prevented air from entering the socket; and 3) sleeve—the prosthetists inactivated the vacuum of the VASS system and blocked the one-way valve. We randomized the order of suspension within a task. Although we did not provide a specific accommodation period after each alteration in suspension, the ordering of tasks allowed acclimation to each suspension (task 1) before collection of the primary outcome data (tasks 2 and 3). Participants completed all suspension conditions within a given task before moving to the next task. All tasks in study 1 were completed during a single testing session lasting 2.5 to 3 hours.
TASK 1: QUANTIFYING GAIT ASYMMETRY
Subjects traversed an 8-m walkway approximately 10 times while an eight-camera motion capture system (Motion Analysis, Satan Rosa, CA, USA) tracked 22 passive reflective markers placed on bony landmarks of the upper and lower limbs.19 Custom software was used to calculate SL, ST, and step time from the motion of these markers. The latter was included based on the established association between step time asymmetry and metabolic costs of walking.20 Step length was calculated as the distance in the sagittal plane between the centroids of the two feet at successive midstances, defined as the point in time at which the ankle marker of the swing limb passed that of the stance. Prosthetic (or sound) side SL was defined for steps taken with the respective limb. We defined ST as the time from heel-strike to ipsilateral toe-off and step time as the time from heel-strike to contralateral heel-strike. We calculated symmetry indices (SIs) for all variables based on the formula from Herzog21: 100 × [(prosthetic − sound)/½(prosthetic + sound)]. A smaller SI implies less asymmetry.
TASK 2: TIMED UP AND GO AND 10-M WALK TESTING
The TUG and timed walking tests such as 10MWT are valid and reliable measures of mobility, recommended as “gold standard measures” for persons with amputation.22,23 For 10MWT, the participant walked a 10-m path as fast as possible and repeated the task three times. A stopwatch recorded the time taken to walk the middle 6 m of the path. We calculated the average completion time for all trials, from which we determined maximum walking speed.
After the 10MWT, subjects completed TUG. For TUG, subjects rose from an armchair of standard height, walked 3 m, turned, and returned to the chair. A certified prosthetist used a stopwatch to measure time from a “go” cue to when the participant's buttocks touched the seat.
TASK 3: QUANTIFYING COST OF TRANSPORT AND COMFORT AT SSS
Participants walked for 6 minutes around a carpeted indoor track (1 lap = 90.58 m) with their prosthetic limb on the inside of the track. Before and throughout the walk, the rate of oxygen (O2) consumption (in mL O2/min) was measured on a breath-by-breath basis using a portable device (K4 b2; Cosmed, Rome, Italy). During the trial, an investigator walked slightly behind the subject while wheeling a distance wheel. After the 6 minutes of walking, subjects self-reported socket fit comfort score (SFCS). The SFCS is a validated assessment of comfort that asks, “If 10 represents the most comfortable socket fit you can imagine and 0 represents the most uncomfortable socket fit you can imagine, rate the comfort of your socket fit at this moment.”24 Before repeating the task with a different suspension, subjects rested until the rate of oxygen consumption was similar to that before the trial. Here, we report data only from those subjects who performed the task at an SSS; some subjects were asked to perform certain conditions at intentionally fast speeds that were greater than SSS (see Figure 1).
We calculated SSS offline based on the distance walked over the entire 6-minute trial. To process the raw oxygen consumption data, we calculated a five-breath running average using custom software (Matlab; Mathworks, Cambridge, MA, USA). To calculate the cost of transport, we normalized the mean rate of O2 consumption over the final 2 minutes of walking by the product of the mass (kg) and the mean speed during that period. The units of cost of transport are (mL O2/kg*m).
PROTOCOL: STUDY 2 (QUESTIONNAIRES)
Both VASS and non-VASS subjects were sent an email with a link to an online version of the ABC scale managed using Research Electronic Data Capture (REDCap) tools hosted at the University of Illinois at Chicago. REDCap is a secure Web-based application designed to support data capture for research studies.25 Participants first completed the ABC scale. The ABC scale is a 16-item survey that asks participants to rate how confident (0% to 100%) they feel in their ability to complete a variety of tasks without losing balance. Total score represents the average of all items. Upon submitting the ABC scale, subjects were sent links to three additional surveys: the LCI5, the Houghton Scale, and selected portions of the PEQ.
The LCI5 measures basic and advanced functional locomotor capabilities using 14 questions with five ordinal responses to reduce ceiling effects; higher scores indicate higher function (maximum score of 56).17,18 The survey includes sections that addresses basic locomotor functions (e.g., walking indoors) and one that addresses advanced functions (e.g., walking outdoors on sloped sidewalks).
The Houghton Scale measures function in terms of prosthetic wear and use and consists of four questions each scored 0 to 3, with a maximum score of 12 indicating high levels of prosthetic use and wear and is recommended for clinical use.16
The PEQ is organized into 10 different domain scales that are valid and reliable measures regarding prosthesis-related quality of life.26 Each scale is scored from 0 to 100. The score is the average of individual questions within a domain. In the original version, participants place a mark on a line with two anchor points and the distance from one end is measured and scaled from 0 to 100. In the current study, responses were provided by sliding a bar along a line with the corresponding value being displayed. We assessed the following PEQ domains: well-being, frustration, social burden, and perceived response.
Before testing the hypotheses, for each outcome variable in study 1, we ran three separate independent-samples t tests, one for each suspension condition, to determine if there was an effect of level of amputation. We observed no significant effects for any comparison and therefore excluded level of amputation as a covariate in the ensuing analysis. Lastly, for each outcome variable in study 2, we collapsed the data across groups and tested for an effect of unilaterality versus bilaterality. We observed an effect for ABC score, overall LCI5, and the advanced function section of the LCI5. To test for an effect of group on these variables, we ran a one-way analysis of variance (ANOVA) with unilaterality versus bilaterality as a covariate. For all other variables, to test for an effect of group, we used independent t tests.
To test H1, that in current users of VASS, a loss of vacuum would increase the energetic costs of gait, we ran a linear mixed model with subjects as a random variable, suspension type as a repeated, fixed factor, and cost of transport at SSS as the dependent variable. We used a mixed model, rather than an ANOVA, as it is able to account for missing data (Figure 1). In the event of a significant effect of suspension, post hoc tests were run using LSD corrections. A similar analysis was performed to test H2, that in current users of VASS a loss of vacuum would worsen performance-based outcomes measures related to function and mobility, with separate models for TUG, 10MWT, and SSS. To test H3, that in current users of VASS, a loss of vacuum would reduce SFCS, we ran a mixed model with SFCS as the dependent variable. Finally, to test H4, that long-term use of VASS would improve self-reported outcomes measures, we ran either independent-sample t tests or ANOVAs, as described above.
In an attempt to validate the work of Ferraro (2011), we used independent-samples t test to test for an effect of group on ABC scores. In an attempt to validate lower ST and SL symmetry with VASS and explore the effect of VASS on step time asymmetry, we used mixed models to test for an effect of suspension condition on each SI. We used IBM SPSS (IBM, Armonk, NY, USA) for all analyses with significance set at p ≤ 0.05.
There was no effect of loss of vacuum on cost of transport (p = 0.47). Table 2 summarizes values for the individual conditions. However, VASS did allow subjects to perform certain locomotor tasks more quickly. In particular, subjects completed the TUG almost 1 second faster with VASS compared with sleeve (p = 0.02 VASS vs. sleeve), although TUG was also performed more quickly with suction compared with sleeve (p = 0.049 suction vs. sleeve). There was no difference in TUG between VASS and suction (p = 0.541).
The maximum walking speed attained with VASS during the 10MWT was 0.07 m/s faster than that with suction and over 0.10 m/s faster than that with sleeve (p = 0.027 VASS vs. suction; p = 0.011 VASS vs. sleeve). Maximum walking speed was not different between suction and sleeve (p = 0.17 suction vs. sleeve). There was no effect of suspension on SSS (p = 0.40). There was no effect of suspension on SI for SL (p = 0.35), step time (p = 0.93), or ST (p = 0.07).
The use of VASS resulted in an average within-subject increase of 1.5 and 3 points on SFCS compared to suction and sleeve, respectively (p = 0.001 VASS vs. suction and VASS vs. sleeve; p = 0.014 suction vs. sleeve; Table 2).
There was no effect of group on any self-reported measures including ABC scores (p = 0.73; Table 3). Table 3 summarizes group average values and corresponding p values for overall F test for all self-reported outcome measures.
The purpose of this study was to quantify the extent to which VASS improved energetic costs of gait, performance-based outcomes measures, and self-reported outcome measures related to functional mobility and well-being. The first hypothesis, that in current users of VASS, a loss of vacuum would increase the cost of transport at a SSS, was not supported. The second hypothesis, that in current users of VASS, a loss of vacuum would worsen performance-based outcomes measures related to function and mobility, was partly supported. When participants used VASS, they obtained faster maximum walking speeds relative to the two other conditions and performed TUG more quickly, although only relative to sleeve condition. However, SSS was unaffected by acute changes in suspension. The third hypothesis, that in current users of VASS, a loss of vacuum would reduce socket comfort, was supported. The fourth hypothesis, that long-term use of VASS would improve self-reported outcomes measures related to prosthetic wear and use, locomotor capabilities, and well-being, was not supported; there were no significant differences in Houghton scale, LCI5, or well-being domain of the PEQ in current users of VASS versus non-VASS users.
The fact that VASS did not affect the cost of transport, although contrary to our expectations, seems to be consistent with the absence of improved symmetry. Previous studies have demonstrated that energetics of gait increase with asymmetry. For example, when healthy young subjects received visual feedback to induce asymmetries in stride time, a 23% increase in SI resulted in a 17% increase in metabolic power.20 Mattes et al.27 demonstrated a similar association between asymmetry and energetic costs in persons with amputation, although in that study, asymmetry was induced by altering inertial properties of the prosthesis, which may differently affect energetics than manipulating SI with visual feedback. Although there is agreement between cost of transport and symmetry measures, given the work of Board et al.,3 it is somewhat surprising that VASS did not reduce asymmetries in SL or ST. An important difference between that study and our study is that Board et al. observed an immediate effect of VASS after providing vacuum to nonvacuum users, which is the reverse process studied here. Moreover, even if their results were replicated, it is unclear if this would have an effect on cost of transport as Board et al. observed only a 1% to 2% change in SI.
The systematic across-condition increase in maximum walking speed observed with VASS seems to be clinically meaningful. For example, a study evaluating the minimum clinically important difference (MCID) in the 10MWT speed, at least for older adults performing the test at a preferred walking speed, concluded that changes of 0.05 and 0.1 m/s represent small and substantially meaningful differences, respectively.28 The speed differences between VASS and suction and between VASS and sleeve are within this range. To our knowledge, similar MCID for the 10MWT for prosthetic users is absent from the literature.
Although we did not quantify stability during 10MWT or SSS, walking at speeds above SSS has been shown to produce less stable locomotor patterns (as measured by variables derived from dynamic system theory).29,30 Nonetheless, VASS allows users to attain faster maximum walking speeds, perhaps because it reduces instability at these speeds. Although we would expect any stabilizing effects of VASS to generalize across speeds and activities, and to contribute to the increased balance confidence reported by Ferraro, for our higher-functioning participants, the impact on outcome measures may be task specific. To observe an effect on outcomes, it may be necessary to present these individuals with a challenge to stability that exceeds that provided either by walking at SSS or by performing basic activities of daily living. The nonsystematic effects of VASS on TUG times are consistent with this notion. Only a small fraction of the TUG involves the dynamically challenging tasks of turning or gait initiation/termination; much of the total TUG time involves sit-to-stand and stand-to-sit transfers and walking at or near SSS. This is, in part, the rationale for using the L Test of Functional Mobility (L Test) to assess functional mobility in high-function individuals. Like TUG, the L Test includes transferring and turning but has participants walking longer distances and performing more turns including at least one on the each side.31
To our knowledge, this is the first study to explicitly report immediate reductions in socket comfort after removal of vacuum. To some extent, this acute effect may also manifest as long-term improvements in comfort after prescription of VASS, which could be one reason why the case subject in the study by Sutton et al.7 reported walking and wearing his prosthesis more after receiving VASS. Accordingly, it is somewhat surprising that in study 2, we observed no differences in any self-reported outcomes between current VASS and non-VASS users. However, the Houghton Scale,16,17 which was used to measure prosthetic use in our study, may have ceiling effects16,17 and may not be sensitive enough to detect small differences in physical activity between high-functioning VASS and non-VASS groups. For example, the scale includes questions such as “When going outside wearing your prosthesis do you use: a wheelchair, two crutches/cane/walker, one cane, nothing?” Physical activity monitoring may provide a more sensitive means to quantify the effects of VASS on prosthetic use in these individuals. A recent crossover study of current users of pin suspension used activity monitors to compare step count between pin and VASS suspension and, surprisingly, reported that participants took nearly twice as many steps with pin suspension.32 However, that study was limited to five participants because of a large number of dropouts after fitting with VASS. In contrast, the study of Ferraro demonstrated a trend toward more activity with VASS as indicated by six of nine subjects reporting an “increase [in] walking time with vacuum (in comparison with pin).” Further work is needed to evaluate the effects of suspension, particularly VASS, on physical activity.
Ceiling effects are also likely to explain, in part, the absence of an effect of VASS on locomotor capabilities as assessed in this study. Referring again to the case study of Sutton, the participant in that study demonstrated immediate improvements in LCI5, achieving a perfect score after only 1 week of VASS. Thus, LCI5 was unable to show further improvements. Despite the fact that LCI5 includes five (rather than four) response choices to avoid earlier observed ceiling effects,17,18 the additional level may be best suited to differentiate persons who do and do not require an aid to independently perform locomotor tasks,18 and the survey is likely to show ceiling effects for higher-functioning subjects. In the current study, 56% of control subjects and 47% of VASS subjects reported perfect scores on the scale. Similarly, the previously reported ceiling effects of the ABC scale33 may explain differences in the current study and that of Ferraro.
In addition to ceiling effects, the study design may have affected our ability to detect between-group differences in self-reported outcome measures. For example, whereas the study of Ferraro used a crossover design to compare individual increases in ABC scores with changes in suspension, the current study compared ABC scores between groups of prosthetic users. Given that the VASS and non-VASS groups were not matched for factors such as level of amputation or time since amputation (see Table 1), the between-group comparison used herein may be inadequate to detect an effect of suspension in absence of a large sample size to account for individual differences. In summary, despite a logical a priori rationale for the choice of surveys used herein, a posteriori knowledge suggests that these surveys are not well suited for identifying difference in high-functioning populations and, in absence of large sample sizes, may be better suited for within-subject comparison.
There are several limitations of the current study. First, only pen-and-paper versions of the surveys from study 2 have been validated (see review by Condie et al.23). Although there is increasing evidence that digital versions of self-reported scales may provide valid assessments of outcome measures,34,35 their use for the measures included herein has yet to be determined. Second, by evaluating current VASS users, it was not possible to blind the participant to alterations in suspension in the laboratory; participants could easily perceive when vacuum was removed.
A third limitation of the study is the lack of accommodation period after altering suspension; the effects of VASS may not have been washed out during the other conditions. Randomization of suspension conditions may have partially countered these effects; for example, within a given task, only one condition was performed directly after removing VASS. In addition, primary outcomes (task 2 and 3) were performed only after participants gained familiarity with all conditions in task 1. Whereas the lack of accommodation may limit our ability to draw conclusions about the effects of suspension on functional mobility, it does provide unique information about an individuals' ability to immediately cope with loss of active vacuum. Many times, patients may have a problem with their vacuum system (leaks or malfunctioning pump), which to some extent was simulated here, but they can manage for several days before visiting a prosthetist. The fact that most subjects completed the protocol with altered suspension corroborates that many patients may be able to perform daily tasks (e.g., walking at SSS or short burst of fast walking) without a properly functioning prosthesis but with potentially compromised comfort. However, there appear to be individuals who rely heavily on the benefits of VASS and may be limited in their ability to perform daily activities in absence of a fully functional vacuum system, for example, subjects who did not complete tasks 2 or 3 because of pain (Figure 1). One possibility, albeit speculative, is that these individuals have residua that are highly sensitive to discomforting (painful) stimuli; one of the two subjects unable to complete the protocol suffered a mangled limb during the trauma that led to her amputation. It would be informative to identify common characteristics in persons incapable of performing activities without VASS to better identify ideal candidates for VASS.
Although accommodation may be important, to our knowledge, no study indicates the amount of time needed to ensure accommodation to VASS. In a previous study of non-VASS users, 30 minutes of continuous walking was sufficient to observe an effect of VASS on limb volume.3,32 Thus, a relatively short accommodation period that allows for changes in limb volume may be sufficient to assess the immediate effects of suspension alteration. In the current study, the fact that VASS did not improve symmetry, SSS, or cost of transport may reflect the fact that sufficient time was not provided during testing to allow for substantial volume loss. However, as participants were current VASS users, it could be contraindicative to provide a long accommodation period or long walking period with other suspension conditions.
A final limitation of this study was that the non-VASS suspension conditions were tested with components and socket design specific for VASS. Although this allows us to consider only the effects of suspension on outcome variables, it creates conditions that are not directly translatable to the clinical setting. For example, the weight of the vacuum was not removed during non-VASS conditions and the added mass could alter kinematics and economy of gait. Indeed, Mattes et al.27 reported an effect of added mass on gait symmetry and even energetic costs of gait. However, vacuum pumps (and batteries) currently available on the market range in total mass from ~130 to 600 g (based on manufacturers' information available online). These values are on the very low end of the mass perturbations in the study of Mattes. Thus, any effects of added mass would likely be small relative to effects reported in their study. An earlier review by Selles et al.36 also points to limited effects of adding up to 0.6 kg of mass on gait economy, stride length or SSS. Regarding the socket, during the sleeve condition, participants wore a total surface-bearing socket, designed for VASS, which may have provided more stability and comfort than afforded by a true sleeve suspension and weakened the observed effects of VASS.
In absence of active vacuum, current users of VASS experience an immediate reduction in comfort, presumably reflecting worse fit, and this may limit their ability to attain faster walking speeds. There is no immediate negative effect of loss of vacuum on the cost of transport or any clear difference in self-reported measures with and without VASS. Therefore, acute loss of vacuum may not negatively impact patients' ability to perform daily tasks before visiting their prosthetist. At the same time, providing VASS could benefit prosthetic users who feel limited by the ability to achieve faster speeds, although a specific study is warranted to test this idea. The self-reported measures used to quantify the effects of suspension on prosthetic use, locomotor capabilities, and domains reflecting prosthesis-related quality of life may not be well suited for identifying difference in the high-functioning, heterogeneous population considered in the current study.
1. Ziegler-Graham K, MacKenzie EJ, Ephraim PL, et al. Estimating the prevalence of limb loss in the united states: 2005 to 2050. Arch Phys Med Rehabil
2. Legro MW, Reiber G, del Aguila M, et al. Issues of importance reported by persons with lower limb amputations and prostheses. J Rehabil Res Dev
3. Board WJ, Street GM, Caspers C. A comparison of trans-tibial amputee suction and vacuum socket conditions. Prosthet Orthot Int
4. Goswami J, Lynn R, Street G, Harlander M. Walking in a vacuum-assisted socket shifts the stump fluid balance. Prosthet Orthot Int
5. Beil TL, Street GM, Covey SJ. Interface pressures during ambulation using suction and vacuum-assisted prosthetic sockets. J Rehabil Res Dev
6. Ferraro C. Outcomes study of transtibial amputees using elevated vacuum suspension in comparison with pin suspension. J Prosthet Orthot
7. Sutton E, Hoskins R, Fosnight T. Using elevated vacuum to improve functional outcomes: a case report. J Prosthet Orthot
8. Arndt B, Caldwell R, Fatone S. Use of a partial foot prosthesis with vacuum-assisted suspension: a case study. J Prosthet Orthot
9. Waters RL, Perry J, Chambers R. Energy expenditure in amputee gait. In: Moore W, ed. Lower Extremity Amputation
. 1st Ed. Phillidelphia: WB Saunders Co; 1989:250.
10. Waters RL, Mulroy S. The energy expenditure of normal and pathologic gait. Gait Posture
11. Schrack JA, Simonsick EM, Ferrucci L. The energetic pathway to mobility loss: an emerging new framework for longitudinal studies on aging. J Am Geriatr Soc
12. Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. JAMA
13. Guralnik JM, Ferrucci L, Pieper CF, et al. Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance battery. J Gerontol A Biol Sci Med Sci
14. Wentink EC, Prinsen EC, Rietman JS, Veltink PH. Comparison of muscle activity patterns of transfemoral amputees and control subjects during walking. J Neuroeng Rehabil
15. Ali S, Abu Osman NA, Naqshbandi MM, et al. Qualitative study of prosthetic suspension systems on transtibial amputees' satisfaction and perceived problems with their prosthetic devices. Arch Phys Med Rehabil
16. Devlin M, Pauley T, Head K, Garfinkel S. Houghton scale of prosthetic use in people with lower-extremity amputations: Reliability, validity, and responsiveness to change. Arch Phys Med Rehabil
17. Miller WC, Deathe AB, Speechley M. Lower extremity prosthetic mobility: a comparison of 3 self-report scales. Arch Phys Med Rehabil
18. Franchignoni F, Orlandini D, Ferriero G, Moscato TA. Reliability, validity, and responsiveness of the Locomotor Capabilities Index in adults with lower-limb amputation undergoing prosthetic training. Arch Phys Med Rehabil
19. Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res
20. Ellis RG, Howard KC, Kram R. The metabolic and mechanical costs of step time asymmetry in walking. Proc Biol Sci
21. Herzog W, Nigg BM, Read LJ, Olsson E. Asymmetries in ground reaction force patterns in normal human gait. Med Sci Sports Exerc
22. Schoppen T, Boonstra A, Groothoff JW, et al. The timed “up and go” test: reliability and validity in persons with unilateral lower limb amputation. Arch Phys Med Rehabil
23. Condie E, Scott H, Treweek S. Lower limb prosthetic outcome measures: a review of the literature 1995 to 2005. J Prosthet Orthot
24. Hanspal RS, Fisher K, Nieveen R. Prosthetic socket fit comfort score. Disabil Rehabil
25. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform
26. Legro MW, Reiber GD, Smith DG, et al. Prosthesis evaluation questionnaire for persons with lower limb amputations: assessing prosthesis-related quality of life. Arch Phys Med Rehabil
27. Mattes SJ, Martin PE, Royer TD. Walking symmetry and energy cost in persons with unilateral transtibial amputations: Matching prosthetic and intact limb inertial properties. Arch Phys Med Rehabil
28. Perera S, Mody SH, Woodman RC, Studenski SA. Meaningful change and responsiveness in common physical performance measures in older adults. J Am Geriatr Soc
29. Kang HG, Dingwell JB. Effects of walking speed
, strength and range of motion on gait stability in healthy older adults. J Biomech
30. Granata KP, Lockhart TE. Dynamic stability differences in fall-prone and healthy adults. J Electromyogr Kinesiol
31. Deathe AB, Miller WC. The L test of functional mobility: measurement properties of a modified version of the timed “up & go” test designed for people with lower-limb amputations. Phys Ther
32. Klute GK, Berge JS, Biggs W, et al. Vacuum-assisted socket suspension compared with pin suspension for lower extremity amputees: effect on fit, activity, and limb volume. Arch Phys Med Rehabil
33. Miller WC, Deathe AB, Speechley M. Psychometric properties of the activities-specific balance confidence scale among individuals with a lower-limb amputation. Arch Phys Med Rehabil
34. Bird M, Callisaya ML, Cannell J, et al. Accuracy, validity, and reliability of an electronic visual analog scale for pain on a touch screen tablet in healthy older adults: a clinical trial. Interact J Med Res
35. Gwaltney CJ, Shields AL, Shiffman S. Equivalence of electronic and paper‐and‐pencil administration of patient‐reported outcome measures: a meta‐analytic review. Value Health
36. Selles RW, Bussmann JB, Wagenaar RC, Stam HJ. Effects of prosthetic mass and mass distribution on kinematics and energetics of prosthetic gait: A systematic review. Arch Phys Med Rehabil