People with transfemoral amputation have been shown to be less efficient walkers than able-bodied individuals and demonstrate a marked increase in metabolic cost during walking.1 Previous studies indicate that people with transfemoral amputation exhibit a 27% to 88% greater metabolic cost compared with able-bodied individuals.2,3
This lack of efficiency is more marked in people with transfemoral amputation compared with transtibial amputation.4 People with transtibial amputation demonstrate a 15% to 55% increase in metabolic cost over able-bodied individuals.5 Increased energy consumption has been attributed to asymmetric limb movements, impaired lower limb sensorimotor control, abnormal muscular contraction patterns, and abnormal energy transfer.6
Researchers have sought ways to minimize the energy costs associated with using knee prostheses. Hydraulic knees provide variable resistance to knee flexion and extension, depending on valve adjustments set manually by the prosthetist, to allow a variety of walking speeds.7 Resistance settings of such hydraulic knees do not account for unexpected obstacles or sudden changes in speed or direction,7 and the individual may have to alter the normal gait pattern to continue walking in some scenarios. Microprocessor knee (MPK) prostheses have multiple sensors that continuously detect joint position, direction, and speed of forces acting upon the knee. The onboard microprocessor uses input from the sensors to determine the real-time phase of gait and provide the appropriate knee resistance to match variations in walking speed and terrain.8 By enabling users to walk at their self-selected walking velocity and instantaneously adapt to changing conditions with reduced abnormal movement compensations, MPK prostheses may minimize energy expenditure during walking.9
The first research on MPK prostheses began in the 1970s and culminated in the 1993 introduction of the first commercially available MPK unit, which controlled swing phase motion: the Endolite Intelligent Prosthesis.10 Since then, new models have been introduced that allow wireless operation and control of both the swing and stance phases of gait.7 Today, common MPK prostheses include the C-Leg® by Otto Bock (Duderstadt, Germany) and the Rheo Knee® by Ossur (Reykjavik, Iceland), as well as more recently introduced models such as the Genium™ by Otto Bock and the Orion by Endolite (Hampshire, UK). Although an MPK has been documented to consistently provide stumble prevention and cost benefits,8 evidence to support reduced energy expenditure with MPK prostheses has been less consistent.6,8 A past systematic review of studies comparing the C-Leg with non-MPK prostheses concluded that the evidence suggested less energy expenditure with C-Leg use but excluded all other MPKs.8 In an investigation of MPKs from three different manufacturers, energy expenditure was similar for most MPKs at most speeds.11 This systematic literature review will include studies of all MPK prostheses used by unilateral transfemoral amputees for the purpose of determining whether using MPK-type prostheses results in reduced energy expenditure, as measured by oxygen consumption, compared with non-MPK prostheses.
A literature search using the PubMed, CINAHL, and MedlinePlus databases; the Google Scholar search engine; and the Otto Bock, Ossur, and Endolite Web sites was conducted in March 2011. Primary search terms included microprocessor knee prostheses, C-Leg prostheses, intelligent prostheses, lower leg prostheses, and lower-limb amputee. Boolean searches were used in all databases, with the terms searched independently and in combination with one of the following secondary search terms: energy expenditure, oxygen consumption, VO2, or gait. References in review articles were examined to ensure that important articles were not overlooked.
References with abstracts written in the English language were compiled. Two reviewers, using the preestablished inclusion and exclusion criteria, independently screened the resulting abstracts. Duplicates were removed. Abstracts were excluded if published more than 15 years ago, pertained to endoprostheses, or were not peer reviewed. The remaining abstracts were reviewed and included if they had an experimental design, a comparative study between non-MPK and MPK prosthesis use, and an outcome measure of oxygen consumption.
QUALITY ASSESSMENT OF INCLUDED STUDIES
All articles meeting inclusion criteria for selection were reviewed and assessed for quality and risk of bias. Each article was assessed by two separate raters using the PEDro rating scale (Figure 1). The PEDro scale is an established scale for comprehensive assessment of methodological quality of clinical trials,12 including prosthetic research8 with moderate reliability for consensus scores13 and good construct validity upon Rasch analysis.14
The PEDro scale has been used to grade the quality of clinical studies based on a rating derived from 10 study criteria that must be explicitly addressed to receive a point (see Appendix). Percentage agreement may overestimate the true reliability of the categorical PEDro scale ratings between raters because a proportion of agreements may occur by chance alone. Thus, κ was calculated to determine the chance-corrected agreement between raters,15 which is typically less than the percentage agreement:16 κ values greater than 0.80 demonstrate excellent agreement, whereas values 0.61 to 0.80 demonstrate substantial agreement.17 If raters differed on the PEDro score of an article, a consensus was reached by discussing differences, and if needed, a third rater was enlisted.
In addition, each article was assessed for possible limitations that may have biased results, including sample selection, assessment process including test order, accommodation to new prostheses, reporting of results, and the sources of sponsorship.
LITERATURE SELECTION AND QUALITY ASSESSMENT
Of the 275 articles compiled, 7 articles were included in this review (Figure 2). Quality assessment of the seven studies demonstrated an average PEDro rating of 3.7 of a possible 10 points, with a rating range of 2 to 5 (Figure 1). Raters agreed on the PEDro score on six of seven articles, with κ = 0.77 demonstrating substantial agreement.17 All seven studies were controlled before-after designs with repeated measures, as classified by the ranking scheme proposed by the American Academy of Orthotists and Prosthetists.18 Because there were no two-group randomized control trials, study quality was low, with PEDro scores limited by the lack of group randomization, concealed allocation, and blinding to different groups of subjects. None of the studies reported effect size assessment. Two studies that examined energy expenditure with oxygen consumption during MPK use were excluded because the comparison groups were able-bodied individuals, rather than people using non-MPK prostheses.19,20
Of the 60 combined subjects from the seven included studies, only 10 (16.7%) were women, only 2 (3.3%) were older than 60 years,21 and only 1 (1.7%) had a vascular amputation (see Figure 1).22 None of the seven studies included subjects at the Medicare K2 functional levels or specified racial ethnicity; only one specified prosthetic alignment.4 In addition, four studies received sole support from the manufacturers of the MPK under investigation. The findings and limitations of the seven studies are briefly summarized below.
DESCRIPTIONS OF STUDIES
BUCKLEY ET AL.23
In the earliest and smallest study, two subjects had reduced mean volume of oxygen consumption per unit of time (VO2; −5.6% to −9.0%) when using the MPK at speeds slower and faster than their self-selected walking speeds; one subject using the non-MPK had increased VO2 (1.7%).23 Only oxygen consumption reductions of more than 5%, the normal variation in submaximal exercise, were considered clinically significant.24 At self-selected speeds, energy consumption between MPK and non-MPK use did not vary by more than 2%.23 Oxygen consumption was reduced for all three subjects when using the MPK versus the non-MPK when walking at variable speeds: oxygen consumption reduction ranged from 2.2% to 7.0%, with a mean of 4.1%. The authors concluded that although differences were small, MPK use may reduce energy most effectively at speeds other than self-selected speed.23 Limitations included 1) fatigue due to the 26-minute minimum testing period, 2) inadequate training because of the unspecified MPK training period, 3) nonrandom test order (MPK before non-MPK), and 4) manufacturer as the sole funding source.
SCHMALZ ET AL.4
A larger study of transfemoral prosthesis users found reductions in net oxygen consumption with MPK use at self-selected and slower speeds (6.0% and 6.2%, respectively; p < 0.05) but not at faster speeds. The authors concluded that energy consumption reductions resulting from MPK use may be speed dependent. One limitation was that non-MPK swing phase knee resistance had been optimized for faster speeds.4 Other limitations include 1) fatigue due to the 25-minute minimum testing period, 2) short 10-minute training period with the MPK, 3) nonrandom test order (non-MPK before MPK), 4) unspecified sex of the subjects, and 5) manufacturer as sole funding source.
DATTA ET AL.25
The largest reductions in oxygen consumption of up to 9.1% were observed when using the MPK at slow walking speeds. Energy cost reductions were observed at fast (8.3%) and medium (7.8%) speeds. Not all subjects were able to walk at the faster speeds, and energy reduction was only statistically reduced with MPK use for speeds less than 0.9 m/s.25 The authors suggested that reductions in energy with MPK use was most likely at speeds other than self-selected speeds, although this could not be confirmed for the faster speeds.25 Limitations include 1) fatigue due to the 18-minute testing period, 2) variable initial training period with the MPK, 3) nonrandom test order (non-MPK before MPK), and 4) comparisons between unequal groups because only 6 of 10 subjects completed testing at all speeds.
JOHANSSON ET AL.26
This study investigated the Rheo Knee concurrently with the C-Leg compared with a non-MPK. Each subject wore all three prostheses but was tested only while walking at a speed averaging between 1.14 and 1.2 m/s.26 Use of the Rheo Knee resulted in a mean reduction in energy cost (5%; p = 0.009) compared with the non-MPK. Six of the eight subjects had reduced energy cost. No significant differences were noted between the Rheo Knee and C-Leg or between the C-Leg and non-MPK (p > 0.05).26 The authors could not address potential energy savings at varying speeds. Limitations include 1) variations in typical self-selected speeds ranging from 0.7 to 1.28 m/s, 2) variations in prosthesis typically used, and 3) manufacturer as sole funding source.
ORENDURFF ET AL.6
This was the only study in the group to find increased group mean energy cost with MPK use at any speed. Whereas there was no significant difference (p > 0.05) in oxygen consumption between non-MPK and MPK use at slow, self-selected normal, and fast speeds, MPK use at 0.8 m/s resulted in an 8.1% increase in energy cost, although at speeds faster than 0.8 m/s, MPK use reduced oxygen costs by 4.5% to 7.2%.6 This study showed that the MPK did not reduce mean energy consumption, although four individuals showed reduced and one showed increased energy consumption. Limitations include 1) more than 50% of the original subject pool dropped out, 2) lack of intent-to-treat analysis, and 3) potential variations in prosthetic alignment and MPK settings.
SEYMOUR ET AL.21
In addition to oxygen consumption, this study also assessed obstacle course performance and quality of life measured by the SF-36v2 normalized survey. Oxygen consumption decreased with MPK use compared with non-MPK use in both the self-selected speed (3.3%, p = 0.04) and fast-paced walk tests (8.0%, p = 0.03) for the 10 subjects.21 The reduction at the self-selected speed, however, was less than necessary for clinical significance.24 Oxygen cost while walking with an MPK was 54% to 55% more and with a non-MPK was 65% to 67% more than that of able-bodied individuals.21 The authors concluded the C-Leg decreased energy consumption compared with the non-MPK, although the MPK user still consumes far more energy than able-bodied counterparts. Subjects using the MPK also performed the obstacle course in faster times, with fewer steps than those using the non-MPK.21 No quality-of-life comparison was possible because the SF-36v2 was obtained only after MPK use. Limitations included 1) comparison of performance using the C-Leg with performance using a mix of six different non-MPKs, 2) wide variation of 2 to 44 months of prosthesis use before testing, 3) mixed amputation level, and 4) the manufacturer as sole funding source.
KAUFMAN ET AL.22
In the largest study to date investigating energy consumption with MPK and non-MPK use, energy consumption in free-living physical activity and patient walking satisfaction were investigated.22 Results showed a statistically insignificant 2.3% reduction in energy consumption with MPK use compared with non-MPK use (p = 0.34) despite the subjects’ perceived greater ease when walking with the MPK (p = 0.02).22 Although the 2.3% energy consumption reduction using the MPK was within the variation typical for general submaximal exertion,24 the authors concluded that the reduction was clinically significant because it paralleled the subjects’ perception of greater ease. In addition, subjects displayed increased physical activity–related energy consumption in their living environments (6%, p = 0.04), as measured by their total daily energy expenditure, suggesting that ease of walking facilitated participation in more physical activity.22 Limitations included 1) comparison of performance using the C-Leg with performance using a mix of four different non-MPKs, 2) wide variation of 10 to 39 weeks of acclimation to the MPK, 3) mixed vascular and nonvascular amputation etiology, and 4) nonrandom test order (non-MPK before MPK).
The current review was limited to studies comparing energy efficiency as measured by oxygen consumption between various MPK and non-MPK users, excluding case-control studies that compared MPK users with nonamputees20 or alternate outcome measures such as heart rate27 that were included in a past review.8 The seven reviewed studies, all published within the past 15 years, included multiple manufacturers to avoid potential bias and examine the effects of the different classes of prosthetic knee units. Results of the current review revealed a pattern of support for the concept that use of an MPK from various manufacturers reduces oxygen consumption during walking, although study results varied in the magnitude of the reduction and the speed of walking, consistent with the weak recommendation28 in favor of using the C-Leg to increase energy efficiency in previous work.8
Gait speed was classified as slow, moderate, and fast based on past reports that gait speed for prosthesis users after transfemoral amputation is typically slower than that for able-bodied individuals,1 which for adults of both sexes can be estimated as approximately 1.35 m/s.29 Moderate walking speed was thus defined as 0.8 to 1.0 m/s, speeds less than 0.8 m/s down to 0.5 m/s were considered slow, and speeds more than 1.0 m/s up to 1.4 m/s were considered fast. Percentage reductions in oxygen consumption for the five studies that reported data for slow, moderate, and fast speeds are presented in Figure 3. Although all five studies reported decreased energy expenditure at selected speeds, revealing a pattern of energy reduction, observed energy reductions did not consistently reach the minimum 5% threshold suggested for clinical significance,24 nor were changes in energy consumption always statistically significant.
Of the five studies that included statistical analysis of the changes in oxygen consumption, two studies found no change6,22 and three found statistically significant reductions in energy consumption at both self-selected and other speeds.4,21,26 Statistical significance is a mathematical indication that the result of an analysis at a specified probability is unlikely to result from chance and differs from clinical significance, which is an indication that a measured result is large enough to make a clinically observable difference.16 In five studies, mean reductions in energy consumption of at least 5%, the level considered clinically significant beyond the normal variation in submaximal exercise,24 were reported at one or more speeds other than the self-selected speed.4,6,21,25,26 In only two studies were 5% reductions in energy consumption observed at the self-selected speed,4,25 and in one study, a greater than 5% increase in oxygen consumption was found at the self-selected speed6 (see Figure 3). Reduction in energy consumption during walking with an MPK was therefore more apparent at speeds other than the self-selected speed.
Another way to assess clinical significance is to calculate Cohen effect size. Three studies reported sufficient data for effect sizes to be calculated, although only at moderate and fast speeds (Figure 3). Effect sizes (d) ranged from 0.44 to 0.92 at moderate speeds4,6,21 and from 0.60 to 0.66 at fast speeds.6,21 The medium to large effect sizes observed at moderate speeds4,6,21 and the d = 0.57 effect size for the self-selected walking pace in one study6 suggest that gains in energy consumption may also be obtained at self-selected speeds for these study populations.
Comparison among the results of the included studies is limited by a number of factors, including the following: differences in walking speeds; different technologies among the different prosthetic manufacturers30; wide variation in time using the prostheses, potentially affecting comfort and skill level of MPK use; the length of the testing duration, which can cause fatigue; and differences in treadmill21–25 and overground4,6,26 oxygen consumption assessment. Application of the findings to a wider population is limited because of small sample size, low representation within the samples of women and older adults, and lack of specific inclusion of vascular amputees. This furthermore limits the ability to make firm inferences based on study results about the larger population of adults with transfemoral amputation using prostheses. In addition, the collection of studies did not specify the race of the subject samples, inconsistently reported years of prosthetic use, and rarely specified prosthetic alignment methods. Furthermore, it remains unexamined whether the largest proportion of people with leg amputations—older adults, people with vascular amputations, and those at the K2 functional level—would derive any reduction in energy consumption from MPK use. Although the results of this review document reduced energy consumption for people younger than 60 years with nonvascular traumatic amputations who use MPK prostheses, overall, results cannot be generalized to a broad population.
Finally, the lack of randomized and controlled comparison groups makes the cause of observed oxygen consumption differences unclear. Effect sizes could be calculated for fewer than half of the studies because all relevant results including means and standard deviations of all pretest and posttest values were not reported; thus, meta-analysis could not be performed. No effort was made to contact the authors to obtain the original data. Essentially, research related to oxygen consumption with MPK use consists of small studies that documented an effect but were not designed to determine clinical effect size and can be considered level 4 evidence,28 in agreement with a previous review.8
It has been suggested that even a small decrease in energy consumption during walking may be felt by the user and lead to greater overall physical activity.8,22 Although oxygen consumption reduction in one study was too small to be statistically or clinically significant, the observed (2.3%) reduction occurred with a concurrent perception of ease in walking and spontaneous out-of-laboratory increase in activity level.22 However, most active prosthesis users engage in activity lasting more than 15 continuous minutes less than once per day, with the average bout lasting less than 2 minutes.31 Many prosthesis users may not walk for long enough durations to obtain the reductions in energy consumption obtained during the test activity periods of 8 to 26 minutes in the studies included in this review. It is possible that the operating features of an MPK such as knee resistance through a greater range of motion,30 regardless of manufacturer, provide eccentric control of stand-to-sit or stair negotiation, improve stability during stumbles, and minimize fall risk,30 thereby giving a sense of safety that contributes to the observed increased physical activity.
This review is limited by the research designs of the study, which precluded meta-analysis. Although completing randomized control trials with distinctly separate control and experimental groups is difficult, such designs would improve the methodological quality of the evidence. A limitation of this review was the lack of perfect agreement on the PEDro scale. PEDro scores within the current study had substantial agreement but differed with those of raters of another review by both plus and minus 1 point on four studies and 2 points on one study.8 Although MPKs of different manufacturers were included, only one study each using the Endolite Intelligent Prosthesis and the Ossur Rheo Knee met the inclusion criteria; no relevant studies were identified involving MPKs from manufacturers other than the three represented in this systematic review. Slow, moderate, and fast walking speeds were defined based on estimated normal walking speed for prosthetic users. However, speeds within each study varied widely, and comparing oxygen consumption for speeds within a speed range rather than an exact speed may not provide a complete picture of the data. Although multiple databases and reference sources were used, this review was limited to articles in the English language.
Some evidence suggests that MPK use may reduce energy consumption for high-functioning individuals with nonvascular amputations. However, there is insufficient evidence to suggest that MPK use decreases energy consumption, in general, for adults with unilateral transfemoral amputation compared with non-MPK use.
1. Czerniecki JM. Rehabilitation in limb deficiency. Gait and motion analysis. Arch Phys Med Rehabil
1996; 77 (3): S3–S8.
2. Gitter A, Czerniecki J, Weaver K. A reassessment of center-of-mass dynamics as a determinate of the metabolic inefficient of above-knee amputee ambulation. Am J Phys Med Rehabil
1995; 74 (5): 332–338.
3. Hoffman MD, Sheldahl LM, Buley KJ, Sandford PR. Physiological comparison of walking among bilateral above-knee amputee and able-bodied subjects, and a model to account for the differences in metabolic cost. Arch Phys Med Rehabil
1997; 78 (4): 385–392.
4. Schmalz T, Blumentritt S, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: the influence of prosthetic alignment and different prosthetic components. Gait Posture
2002; 16: 255–263.
5. Gailey RS, Nash MS, Atchley TA, et al.. The effects of prosthesis mass on metabolic cost of ambulation in non-vascular trans-tibial amputees. Prosthet Orthot Int
1997; 21: 9–16.
6. Orendurff MS, Segal AD, Klute GK, et al.. Gait efficiency using the C-Leg. J Rehabil Res Dev
2006; 43: 239–246.
7. Berry D. Microprocessor prosthetic knees. Phys Med Rehabil Clin N Am
2006; 17: 91–113.
8. Highsmith MJ, Kahle JT, Bongiorni DR, et al.. Safety, energy efficiency, and cost efficacy of the C-leg for transfemoral amputees: a review of the literature. Prosthet Orthot Int
2010; 43: 362–377.
9. Whittle MW. Gait Analysis: An Introduction
. Oxford, UK: Butterworth-Heinemann; 2006.
11. Bellmann M, Schmalz T, Blumentritt S. Comparative biomechanical analysis of current microprocessor controlled prosthetic knee joints. Arch Phys Med Rehabil
2010; 91: 644–652.
12. Bhogal SK, Teasell RW, Foley NC, Speechley MR. The PEDro scale provides a more comprehensive measure of methodological quality than the Jadad Scale in stroke rehabilitation literature. J Clin Epidemiol
2005; 58: 668–673.
13. Maher CG, Sherrington C, Herbert RD, et al.. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther
2003; 83: 713–721.
14. deMorton NA. The PEDro scale is a valid measure of the methodological quality of clinical trials: a demographic study. Aust J Physiother
2009; 55: 129–133.
16. Portney LG, Watkins MP, eds. Foundations of Clinical Research: Applications to Practice
. East Norwalk, CT: Appleton and Lange; 1993.
17. Landis JR, Koch GG. The measurement of observer agreement for categorical data source. Biometrics
. 1977; 33 (1): 159–174.
18. Dillon MP, Fatone S, Hodge MC. Biomechanics of ambulation after partial foot amputation: a systematic literature review. J Prosthet Orthot
2007; 19: 2–61.
19. Chin T, Sawamura S, Shiba R, et al.. Effect of an Intelligent Prosthesis (IP) on the walking ability of young transfemoral amputees: comparison of IP users with able-bodied people. Am J Phys Med Rehabil
2003; 82: 447–451.
20. Chin T, Machida K, Sawamura S, et al.. Comparison of different microprocessor controlled knee joints on the energy consumption during walking in trans-femoral amputees: Intelligent Knee Prosthesis (IP) versus C-Leg. Prosthet Orthot Int
2006; 30: 73–80.
21. Seymour R, Engbretson B, Kott K, et al.. Comparison between the C-Leg microprocessor-controlled prosthetic knee and non-microprocessor control prosthetic knees: a preliminary study of energy expenditure, obstacle course performance, and quality of life survey. Prosthet Orthot Int
2007; 31: 51–61.
22. Kaufman KR, Levine JA, Brey RH, et al.. Energy expenditure and activity of transfemoral amputees using mechanical and microprocessor-controlled prosthetic knees. Arch Phys Med Rehabil
2008; 89: 1380–1385.
23. Buckley JG, Spence WD, Solomonidis SE. Energy cost of walking: comparison of ‘intelligent prosthesis’ with conventional mechanism. Arch Phys Med Rehabil
1997; 78: 330–333.
24. Armstrong LE, Costil DL. Variability of respiration and metabolism: responses to submaximal cycling and running. Res Q Exerc Sport
1985; 56: 93–96.
25. Datta D, Heller B, Howitt J. A comparative evaluation of oxygen consumption and gait pattern in amputees using Intelligent Prostheses and conventionally damped knee swing-phase control. Clin Rehabil
2005; 19: 398 – 403.
26. Johansson JL, Sherrill DM, Riley PO, et al.. A clinical comparison of variable-damping and mechanically passive prosthetic knee devices. Am J Phys Med Rehabil
2005; 84: 563–575.
27. Highsmith MJ, Kahle JT, Fox JL, et al.. Decreased heart rate in a geriatric client after physical therapy intervention and accommodation with the C-leg. J Prosthet Orthot
2009; 21 (1): 43–47.
28. University of Oxford. Centre for Evidence-Based Medicine. Levels of evidence. Available at: http://www.cebm.net/?o=1025
. Accessed October 7, 2011.
29. Bohannon RW, Williams Andrews A. Normal walking speed: a descriptive meta-analysis. Physiotherapy
2011; 97 (3): 182–189.
30. Wong CK, Edelstein JE. Advanced rehabilitation for people with microprocessor knee prostheses. In: Lusardi MM, Nielsen CC, eds. Orthotics and Prosthetics in Rehabilitation
. 3rd Ed. St Louis, MO: Saunders Elsevier; 2013: 735–757.
31. Klute GK, Berge JS, Orendurff MS, et al.. Prosthetic intervention effects on activity of lower-extremity amputees. Arch Phys Med Rehabil
2006; 87: 717–722.
KEY INDEXING TERMS: microprocessor; knee; prosthesis; energy expenditure; gait