Does temporary socket removal affect residual limb fluid volume of trans-tibial amputees?

Participants

Human subject approval from an institutional review board at the University of Washington was obtained before test procedures were initiated. Trans-tibial amputee subject volunteers were recruited to participate in this study. Inclusion criteria were a trans-tibial amputation performed at least 1 year prior, use of a definitive prosthesis, capability to walk at least 5 min continuously on a treadmill, capability to stand 2 min continuously with equal weight-bearing, and a residual limb length that allowed at least 5.5 cm between voltage-sensing electrodes (necessary for proper bioimpedance measurement ⁹ ). Exclusion criteria were skin breakdown or presence of metal implants within the limb which can distort bioimpedance results. The research practitioner interviewed the subject and inspected the residual limb at the outset of each test session to ensure the subject met all inclusion criteria.

Instrumentation

A bioimpedance analyzer and custom electrodes constructed specifically for testing amputee residual limb fluid volume were used in this study. The system is described in detail elsewhere. ¹⁰ In brief, the instrument generated short packets of constant alternating electrical current (~300 µA). Each set of packets included 24 frequencies between 5 kHz and 1 MHz. Approximately 20 packets/s were transmitted via cable to electrodes positioned proximally on the thigh and distally on the inferior surface of the residual limb (Figure 1). Voltage was sensed via two channels, one from the anterior surface and another from the posterior surface. On each surface (anterior and posterior), one voltage-sensing electrode was positioned at the level of the patellar tendon and the other at the level of the distal tibia. The instrument collected and processed the current and voltage data and then stored results to disk. No filtering was performed on the signal.

OPEN IN VIEWER

Figure 1.

Electrode layout. Current injecting electrodes were placed on the thigh (one anterior, one posterior) and on the distal inferior residual limb. Two voltage-sensing electrodes were placed on the anterior lateral side for the anterior channel (ant) and two on the posterior side for the posterior channel (post).

The electrodes were constructed using a conductive polymer (ARCare 8881; Adhesives Research Inc., Glen Rock, PA, USA (thickness 0.09 mm), hydrogel (KM10B, Katecho; Des Moines, IA, USA (thickness 0.81 mm), and a very thin layer of coupling agent between the hydrogel and skin (ultrasonic coupling gel; GE Panametrics, West Chester, OH, USA). Multi-stranded silver-platedcopper wire with an aramid core and poly vinyl chloride insulation (32 AWG; New England Wire, Lisbon, NH, USA) (thickness: 0.76 mm) and a flattened gold crimp at the wire-insulation boundary was used for the electrode leads. The proximal current injecting electrodes were 1.5 × 15.0 cm, the distal current injecting electrode was of 3.5 cm diameter, and the voltage-sensing electro-des were 1.5 × 5.0 cm. Electrodes were covered with 0.03-mm-thickness Tegaderm Transparent Film Dressing ( 3M, St. Paul, MN, USA).

Test procedure

On each of three test days, the participant sat for 10 min after arrival to achieve a homeostatic condition. While sitting, participants answered questions about their health history and recent changes to their prosthesis. Temperature in the room was controlled via the building thermostat to 22°C. Medical history was recorded and medications that might affect residual limb volume identified. The subject’s body mass was measured. Participants then doffed their prosthesis, liner, and socks (if worn). The participant’s skin was cleaned (Red Dot Trace Prep; 3M) and electrodes were placed on the residual limb. If a participant had bilateral trans-tibial amputation, the limb expected to produce the strongest bioimpedance signal (i.e. longest length, least skin scarring near electrode sites) was selected for testing. Wires from the electrodes were extended proximally along the lateral aspect of the thigh. Care was taken to ensure wires were laid flat and covered with Tegaderm at the brim to avoid air gaps and ensure no loss of prosthetic suspension.

After the electrodes were placed on the limb, bioimpedance data collection was initiated. The participant donned the prosthesis and sat in a chair for 90 s. We reminded the participant to keep a proper sitting posture by avoiding excessive knee flexion that might occlude blood flow and by avoiding too much extension that might cause a slouching posture. Bioimpedance data were observed by the researchers on a computer monitor in approximately real time (~2-s delay) to ensure the instrumentation was performing properly. The participant stood for 90 s on a platform with an electronic scale embedded in the surface to monitor weight bearing on the prosthetic limb. If the prosthetic-side weight deviated by more than 10% of half the body weight, the participant was instructed to shift his or her weight to restore equal weight bearing. The 10% criteria was based on prior investigation where we found that this range of weight-bearing change did not typically induce changes in bioimpedance results and thus did not require continual instruction to the subject for weight shifting. ¹¹ Next, the participant walked for 5 min on a treadmill at the self-selected walking speed established during subject inclusion/exclusion testing. This 5 min of treadmill walking was followed by standing briefly (5–10 s) on the platform with the embedded electronic scale. The participant repeated the sit (90 s)/stand (90 s)/walk (5 min)/brief stand (5–10 s) cycle two more times for a total of three cycles. The participant then sat down for 30 min with the prosthesis in one of three conditions (one condition was performed on each of the three test days): (1) ON: prosthesis and liner left donned; (2) LINER: prosthesis doffed but liner left donned; (3) OFF: prosthesis and liner doffed. After the 30-min period, the subject donned the prosthesis (if it was removed) and conducted three more sit (90 s)/stand (90 s)/walk (5 min)/brief stand (5–10 s) cycles as performed earlier in the test session. At the conclusion of the cycles, the subject doffed the prosthesis, bioimpedance data collection was terminated, and the electrodes were removed. The three test conditions (ON, LINER, OFF) were conducted on three different test days in a randomized order. We attempted to schedule test sessions at the same time of day for each subject and asked subjects to perform similar physical activities each day before coming to the lab.

Analysis

Collected bioimpedance data were post-processed using custom-written code in MATLAB (MathWorks, Natick, MA, USA) that performed de Lorenzo’s version of the Cole modeling strategy. ¹² Data from the model were converted to extracellular fluid volume using a limb segment model. ¹³ No filtering of the signal was conducted.

Percentages limb fluid volume change during the 30-min rest period, short-term fluid volume response, and long-term fluid volume response for each test condition were calculated (Figure 2)

% a g e V 30 min r e s t p e r i o d = [ ( V e n d r e s t − V b e g r e s t ) V p o s t w a l k 3 ] x 100 % % a g e V s h o r t − t e r m r e s p o n s e = [ ( V p o s t w a l k 4 − V p o s t w a l k 3 ) V p o s t w a l k 3 ] x 100 % % a g e V l o n g − t e r m r e s p o n s e = [ ( V p o s t w a l k 6 − V p o s t w a l k 3 ) V p o s t w a l k 3 ] x 100 %

(1)

OPEN IN VIEWER

Figure 2.

Data points used in analysis. Labeled points correspond to variables in equation (1).

We used the V_{post walk3} as the reference (denominator) for all three calculations because the experimental protocol was identical up to that point for all protocol conditions (ON, LINER, OFF). We were interested in effects of our intervention strategy (condition) relative to this point. Results are presented as percentage changes since subjects had different residual limb sizes.

We conducted a repeated-measures analysis of variance (ANOVA) with Greenhouse–Geisser adjustment for lack of sphericity, and Bonferroni adjustment for post hoc tests to assess whether there were differ-ences in % a g e V 30 min r e s t p e r i o d , % a g e V s h o r t − t e r m r e s p o n s e , and % a g e V l o n g − t e r m r e s p o n s e between the three experimental conditions (ON, LINER, OFF). Data from anterior and posterior channels were considered separately. Additionally, we analyzed for differences between anterior and posterior results running a repeated-measures ANOVA with two factors: side (anterior vs posterior) and condition (ON, LINER, OFF).

Anterior versus posterior regions

Anterior and posterior percentage fluid volume changes were similar for all comparisons for ON and OFF conditions and for the 30-min rest for the LINER condition (Figures 3(a) and (b), 4(a) and (b), and 5(a) and (b)); subjects are ordered the same in all figures, from low to high anterior 30-min rest fluid volume change).

For the LINER condition, measures of mean percentage fluid volume changes for both short and long term on the anterior side tended to be larger than those on the posterior side. However, repeated-measures ANOVA including side (anterior vs posterior) as a second factor showed that only condition was statistically significant (p < 0.001 for both short and long term); there was no statistically significant effect of side (p = 0.39), or interaction between condition and side (p = 0.24).

Formal subgroup analysis for this small sample size would not have meaningful interpretation. We considered exploratory analysis, but refrained from performing it because certain subgroups had so few observations that we could not obtain meaningful estimates of the mean and standard deviation for each subgroup (e.g. only two bilateral amputations, only three females, most of the suspension types observed in one or two individuals).

Clinical application

Socket release may be an effective strategy to reduce residual limb fluid volume loss over the day. A person with limb amputation may benefit by removing the prosthesis during periods of inactivity (e.g. a lunch break). Our previous results illustrate that the degree of fluid volume recovery is higher if the prosthesis is doffed right after activity compared with doffing after rest. ¹⁴ The optimal doffing duration to achieve daily limb fluid volume stabilization is unknown and may need to be customized for the individual patient.

The greater short-term and long-term fluid volume recovery for the OFF condition compared to the LINER condition supports the conclusion that the lower the interface pressure during rest, the longer the recovered fluid is retained during subsequent activity. Wearing a tight liner and keeping the liner on after doffing the prosthesis will likely not gain and retain as much fluid from socket release compared with wearing a less restrictive liner or removing the liner altogether during socket release.

One limitation of our study was that, due to the small sample size, we could not assess whether certain factors, such as sex, weight, unilateral or bilateral amputation, type of suspension, and comorbidities, influenced the change in limb fluid volume. Furthermore, test sessions were conducted over a relatively short time period (less than 2 h). Bioimpedance monitoring for longer durations will be needed to quantify duration recovered fluid volume is maintained within the residual limb and if/when a second socket release is needed.

Future research

We suspect that socket release intervals of less than 30 min will accomplish effective limb fluid volume recovery, although it is unclear at this point to what degree the magnitude of recovery and retention during subsequent walking is dependent upon the duration of doffing. A sensitivity analysis of the duration of socket release may serve to establish clinical guidelines for appropriate doffing intervals.

Although results from the present investigation suggest that even the relatively low interface pressure from a donned elastomeric liner can affect the duration of fluid retained from socket release, the mode of action of the liner on limb physiology is still not clear. Three possible modes of action include (1) the prosthetic liner restricts the volume of blood entering the residual limb, (2) the liner restricts transport of fluid across the arterial walls into the interstitium, and (3) the liner enhances transport of fluid from the interstitium into the veins and out of the residual limb. Identifying the dominant mode of action would help toward the design of effective socket release strategies and devices.

Automated socket release systems that adjust socket size at appropriate times rather than by user intervention should also be considered. Actuators that relieve the socket wall based on posture and/or activity (e.g. using accelerometers, inclinometers, and/or force sensors) may be effective, although such systems may need to accommodate for day-to-day differences in physiological fluid transport rates of the prosthesis user. Furthermore, it will be important for such systems to carefully re-tighten the socket upon subsequent donning so that the prosthesis user is stable during weight-bearing. Finally, a liner that is not excessively tight may be an important criterion for such systems to achieve effective socket release.