Prosthetic sockets form the interface between the residual limb and prosthesis, acting to support the body and transfer forces.1 To do this comfortably and efficiently, the socket must be both well conformed and well coupled to the residual limb. Coupling has been achieved using various types of suspension mechanisms, including belts, lanyards, passive suction, locking liners, and most recently active suction using vacuum pumps.2,3 Poor suspension leads to problems such as “pistoning” (i.e., relative vertical motion between the socket and residual limb). It has been proposed that vacuum-assisted suspension (VAS) results in the least pistoning of current suspension systems.2,4,5
Use of vacuum pumps has increased dramatically since their introduction in the late 1990s, with a reported increase in VAS device-specific Medicare billing codes from $1.1 million in 2003 to $7.1 million in 2013.6 Although initially used primarily by persons with transtibial amputation,7–13 vacuum pumps are now also being used by persons with transfemoral amputation.14–16 As more pumps become commercially available, there is need for better understanding of pump function including comparative effectiveness using standardized methods for both bench and in vivo evaluation.16–18 Previous work by Komolafe et al.17 described a method for bench-top performance evaluation of commercially available mechanical and electrical pumps. Gerschutz et al.18 proposed that real-time vacuum pressure monitoring was necessary to understand how vacuum varies with time and usage by patients and illustrated use of a tool for doing so in persons with transtibial amputation. Major et al.16 used both bench-top and in vivo methods to evaluate a newly designed hybrid mechanical-electrical vacuum pump in a single subject with transfemoral amputation. The purpose of this study was to conduct in vivo tests to evaluate the effectiveness of two commonly used commercially available electric pumps, the Ohio Willow Wood LimbLogic (Sterling, OH, USA) and Otto Bock Harmony e-pulse (Duderstadt, Germany), on participants with transfemoral amputation.
This study was approved by the university's institutional review board, and participants provided written informed consent before data collection. A convenience sample of individuals with a unilateral transfemoral amputation who routinely used VAS was recruited to participate. Participants were tested at two sites, a research laboratory and a prosthesis clinical facility, using each pump during two testing protocols:
- Quiet standing—participants were instructed to stand quietly as the pressure within the socket-liner interface was brought to baseline atmospheric pressure, and a pump was then used to decrease pressure until 17 in-Hg below atmospheric pressure was achieved. This test was repeated five times.
- Walking—participants were instructed to first stand quietly until the socket-liner interface pressure was brought to 17 in-Hg below atmospheric pressure using a pump, and then to walk at a comfortable, self-selected speed on a level treadmill (T170; Cosmed, Rome, Italy) for 10 minutes. Both pumps were programmed to allow a minimum vacuum pressure of 13 in-Hg below atmospheric pressure before reactivating to reestablish 17 in-Hg below atmospheric pressure.
The order of pump testing was randomized for each participant, and socket-liner conditions remained the same for both pumps. To ensure the same socket attachment for both pump systems, the LimbLogic pump was connected to the socket volume via a barbed fitting similar to how the e-pulse is typically connected (Figure 1).
For both test protocols, instantaneous pressure in the socket-liner interface was measured with a digital vacuum pressure gauge (model 2 L760, DigiVac, Matawan, NJ, USA) and recorded using custom Labview software (National Instruments Corporation, Austin, TX, USA). For each test protocol, the following outcome metrics were estimated:
- Quiet standing—the rate of evacuation, estimated as the slope of a best-fit linear approximation applied to the linear portion of the pressure temporal profile, and the total evacuation time from pump activation until pressure of 17 in-Hg below atmospheric pressure was achieved. These data were analyzed using Excel (Microsoft Corporation, Redmond, WA, USA) and averaged across the five standing trials for each participant.
- Walking—the number of times the pump reactivated to reestablish a pressure of 17 in-Hg below atmospheric pressure.
These outcome metrics were selected as they represent clinical information that may assist with device recommendations. For example, a clinician must consider if the time required for a pump to achieve a desired level of pressure is important for a given patient based on their activity demands and need for rapidly generated suspension. In addition, as pump reactivation for reestablishing pressure levels would consume additional battery power beyond that of pressure monitoring, the number of reactivations over a specific time would suggest relative frequency of battery recharging during operation. Although pump reactivations are a necessary response due to leakage resulting from features across the entire prosthetic system, some portion of leakage may be due to pump interfacing with the prosthesis and socket-liner interface.
The Shapiro-Wilk test was used to assess data normality with the results suggesting that the data sets were of a nonnormal distribution. Consequently, the Wilcoxon signed-rank test for paired samples was used to statistically assess differences in evacuation rate, evacuation time, and number of reactivations between each pump. The critical alpha was set at 0.05, but applying a Bonferroni correction to account for the familywise type I error rate lowered this threshold to 0.02.
Data for the quiet standing analysis were collected on 18 individuals (13 male, 5 female, 53 ± 14 years, 177 ± 7 cm, 82 ± 8 kg), 9 of whom participated in the walking analysis (8 male, 1 female, 51 ± 13 years, 179 ± 6 cm, 84 ± 10 kg). Fewer subjects participated in the walking analysis because data were collected at two sites, and only one site was equipped with a treadmill.
A representative set of data for the temporal profile of instantaneous pressure for both pumps during the standing protocol is presented in Figure 2, and this behavior was observed for all subjects. Each pump demonstrated a characteristic S-shape profile during evacuation, with three distinct periods of 1) accelerated, 2) constant, and 3) decelerated vacuum pressure rate. The estimated evacuation rate and time across participants are displayed in Figures 3 and 4. These indicated that the LimbLogic required less time to achieve the desired pressure level of 17 in-Hg below atmospheric pressure, and this difference was significant (Table 1). A representative set of data for the temporal profile of instantaneous pressure for both pumps during the walking protocol is presented in Figure 5, and this behavior was observed for all subjects. Both pumps exhibited similar profiles of pressure leakage, but the rate of leakage for the prosthetic system when using the LimbLogic was more rapid, resulting in a significantly greater median reactivation number of one, whereas the e-pulse required no reactivations (Figure 6, Table 1).
Overall, the differences in evacuation rate (23 in-Hg/min) and time (5 seconds) between both pump systems were statistically significant. Although these differences were small, they are likely clinically important. The reduced time needed by the LimbLogic system to achieve full evacuation shortens the period of noise emission and may improve patient compliance and satisfaction with pump use. In addition, less time to achieve the desired pressure level may facilitate longer periods of ideal socket fit during use.
The number of reactivations during walking is interesting as this activity increases the rate of battery power drainage compared with periods of pressure monitoring. A more rapid depletion of battery power would require more frequent charging to maintain socket suspension during operation. The pump-specific reasons for this difference in leakage rate despite no change in the socket is unknown and warrants further investigation as minimizing leakage would maximize battery life. Using a connection (i.e., the barbed fitting) that is not commonly implemented clinically with the LimbLogic may have partially contributed to this increased rate in vacuum leakage.
In addition to these types of installation issues, future research may consider patient-specific factors that likely affect pump evacuation time and vacuum leakage (e.g., residuum dimensions, residuum tissue properties, and liner type) to develop a more nuanced understanding of pump function. Regardless, both pumps were successful in identifying a loss in pressure that necessitated pump reactivation for maintaining appropriate pressure levels for suspension. In this study, pumps were evaluated for the walking trials immediately after evacuation during standing. Clinical experience suggests that pump reactivations occur less frequently with increasing wear time, potentially due to a continuous reduction in air pockets and better seating in the socket. Future studies could explore such issues by monitoring vacuum activity outside the laboratory during prolonged usage.
Although outcome differences between pumps were statistically significant and may be clinically relevant, results from these in vivo tests suggest that both electric pump systems are equally effective in creating and maintaining socket-liner interface pressure for VAS in transfemoral sockets. Importantly, this study demonstrates the use of a simple set of experiments and limited equipment to capture clinically relevant outcome metrics for assessing the effectiveness of pumps for VAS. As use of this suspension method increases, this study addresses the need for standard methods to assess comparative effectiveness of commercial pumps. A combination of in vivo and bench testing methods are required to fully characterize the properties of these devices and their relationship with overall prosthesis function to inform clinical recommendations of appropriate prosthetic technology and development of future designs.16,17,19 Clinical selection of electric pump designs is likely a function of patient activity demands, device evacuation rate, and device-related vacuum leakage.
Based on the time required for achieving clinically recommended levels of pressure for VAS and the number of pump reactivations to maintain that level during walking, the results from this study suggest that the LimbLogic and Harmony e-pulse are equally effective electric pumps despite the observed differences in outcome metrics. Importantly, this study aids in developing standard evaluation methods of commercial pump systems for generating clinically relevant information. Future research should consider investigations on patient- and device-specific factors related to vacuum leakage rate and methods for minimizing this leakage to maximize battery life.
The authors would like to thank Sean Wood for writing the program used for data collection, as well as Oluseeni Komolafe and William Johnson for their assistance with data collection. The authors acknowledge the Jesse Brown VA Medical Center Motion Analysis Research Laboratory for use of their facilities for data collection.
1. Kapp SL. Transfemoral socket
design and suspension
options. Phys Med Rehabil Clin N Am
2000; 11: 569–583 vi.
2. Kahle JT, Orriola JJ, Johnston W, et al. The effects of vacuum
on residual limb physiology, wound healing, and function: a systematic review. Technology & Innovation
2014; 15: 333–341.
3. Gholizadeh H, Abu Osman NA, Eshraghi A, et al. Transfemoral prosthesis suspension
systems: a systematic review of the literature. Am J Phys Med Rehabil
2014; 93: 809–823.
4. Darter BJ, Sinitski K, Wilken JM. The effect of elevated vacuum suspension
on axial bone-socket
displacement in persons with a traumatic transtibial amputation. 35th Annual Meeting of the American Society of Biomechanics
. Long Beach, CA; 2011.
5. Papaioannou G, Mitrogiannis C, Nianios G, et al. Assessment of amputee socket
-stump-residual bone kinematics during strenuous activities using Dynamic Roentgen Stereogrammetric Analysis. J Biomech
2010; 43: 871–878.
6. Centers for Medicare & Medicaid Services. The Part B national summary data file: prosthetic procedures (L5000-L9999). In: United States Department of Health and Human Services (ed.). 2013.
7. Beil TL, Street GM, Covey SJ. Interface pressures during ambulation using suction and vacuum
-assisted prosthetic sockets. J Rehabil Res Dev
2002; 39: 693–700.
8. Board WJ, Street GM, Caspers C. A comparison of trans-tibial amputee suction and vacuum socket
conditions. Prosthet Orthot Int
2001; 25: 202–209.
9. Gerschutz MJ, Denune JA, Colvin JM, et al. Elevated vacuum suspension
influence on lower limb amputee's residual limb volume at different vacuum
pressure settings. J Prosthet Orthot
2010; 22: 252–256.
10. Goswami J, Lynn R, Street G, et al. Walking in a vacuum
shifts the stump fluid balance. Prosthet Orthot Int
2003; 27: 107–113.
11. 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
2011; 92: 1570–1575.
12. Ferraro C. Outcomes study of transtibial amputees using elevated vacuum suspension
in comparison with pin suspension
. J Prosthet Orthot
2011; 23: 78–81.
13. Brunelli S, Averna T, Delusso S, et al. Vacuum
system in transtibial amputees: clinical report. The Orthopadie Technik
14. Kahle JT, Highsmith MJ. Transfemoral sockets with vacuum
comparison of hip kinematics, socket
position, contact pressure, and preference: ischial containment versus brimless. J Rehabil Res Dev
2013; 50: 1241–1252.
15. Kahle JT, Highsmith MJ. Transfemoral interfaces with vacuum
comparison of gait, balance, and subjective analysis: ischial containment versus brimless. Gait Posture
2014; 40: 315–320.
16. Major MJ, Caldwell R, Fatone S. Evaluation of a hybrid vacuum
pump to provide vacuum
for above-knee prostheses. J Med Devices
17. Komolafe O, Wood S, Caldwell R, et al. Methods for characterization of mechanical and electrical prosthetic vacuum
pumps. J Rehabil Res Dev
2013; 50: 1069–1078.
18. Gerschutz M, Haynes M, Colvin J, et al. A vacuum suspension
measurement tool for use in prosthetic research and clinical outcomes: validation and analysis of vacuum
pressure in a prosthetic socket
. J Prosthet Orthot
2010; 22: 172–176.
19. Major MJ, Kenney LP, Twiste M, et al. Stance phase mechanical characterization of transtibial prostheses distal to the socket
: a review. J Rehabil Res Dev
2012; 49: 815–829.