Selection of the optimal prosthetic knee mechanism is one critical element in prosthetic rehabilitation of the person with a limb amputated at a higher level. As is the case in so many other aspects of prosthetic clinical practice, the available technology has improved remarkably during the past decade. This paper will review the types of prosthetic knees presently available and their basic prescription criteria, with an emphasis on the more recent advances.
At this time, prosthetic prescription is an art based largely on clinical experience rather than scientific certainty.28 Available scientific studies and objective data simply are insufficient at this time to guarantee an individual's success with one specific prosthetic component without verification by a clinical trial. However, those options likely to be of little value to the person with an amputated limb in question can be identified easily and eliminated from consideration. The selection of one particular knee prosthesis from the many plausible alternatives remains largely influenced by local experience, philosophy, and custom.
FUNCTIONAL CLASSIFICATION OF PROSTHETIC KNEES
There are numerous classification schemes that have been proposed by various authors to introduce order to the universe of prosthetic components.22 One simple way to classify the available alternatives is to divide them into two broad functional groups: those with exclusively mechanical control properties, and those that have the added versatility of microprocessor control. This latter group only recently has become clinically available, and will be discussed later in this article.3 Prosthetic knees can be subdivided additionally according to the complexity of the stance stability and swing phase control provided.21
The simplest articulated knee available has a single axis hinge that allows the knee to bend freely but offers absolutely no swing phase control. With the exception of the developing world, this device now rarely is used. In more developed countries, the most basic knee prosthesis available adds an adjustable friction cell that presses with a fixed force against the knee axle and provides some damping of swing phase motion. Most commonly, a spring loaded or elastic extension assist also is included to help limit heel rise and to propel the shin into full extension just before heel strike.
The virtue of this single axis constant friction design is that it is mechanically simple and therefore rarely needs servicing, and it is the most inexpensive articulated knee prosthesis. The primary indication for this type of knee prosthesis is if followup care will be difficult or impossible, such as when the person with the amputated limb lives in a remote geographic area.
There are two major limitations to the single axis constant friction knee prosthesis. First, such devices are stable only when the net ground reaction force passes anterior to the knee center; otherwise the knee will buckle abruptly.15 Indoors or on level surfaces, this requires a perfectly consistent gait from the person with the amputated limb. Out of doors or on irregular surfaces, the person with an amputated limb consistently must compensate for changes in the ground reaction force or the knee will collapse. Many elderly people with amputations lack the ability to respond to changing conditions this quickly. This knee prosthesis is much more likely to be prescribed for the pediatric patient than for the geriatric patient.13
The second limitation is that, similar to all knee prostheses with constant friction swing controls, even when optimally adjusted this device only will swing properly at one fixed cadence.27 When the patient attempts to walk faster, the added momentum causes the shin to swing into more flexion, forcing him or her to wait for the leg to swing through.20 Although use of an extension aid spring or application of mechanical friction to the knee axis may limit heel rise for brisk walking, the prosthesis then becomes too difficult to flex during slow walking. Therefore, prostheses with mechanical swing phase controls are optimal only for those individuals who are not capable of varying their cadence.10
The development of Fluid Controlled knee prostheses during the past 50 years has overcome many of the limitations of constant friction designs.26 Originally, such hydraulic or pneumatic knee prostheses were heavy, costly, and required annual or more frequent servicing. Modern versions add a few ounces and as little a 15% to the overall cost of the prosthesis, and may be warranted for 2 years or more. In general, the more complex hydraulic designs with numerous stance and swing control features have a higher cost and weight than those more basic units that simply provide cadence responsive swing control.
Pneumatic control cylinders usually are filled with air, which is compressible like all gases.24 Because very vigorous ambulators may be able to out walk such units, pneumatic swing controls often have been recommended for slow to moderate cadence walkers. Recent design enhancements have narrowed the distinction between pneumatic and hydraulic controls, as will be discussed in the segment on microprocessor controlled knee prostheses. The primary functional advantage of pneumatic dampers is that they are unaffected by changes in ambient temperature, so the knee prosthesis resistance is the same in a warm room as it will be after several hours of subzero outdoor winter activities.
Hydraulic dampers are the most common variable cadence control used clinically, in part because restricting the flow of the incompressible liquid they contain (usually silicone oil) creates a very high damping force to control the shin.18 Some of the more basic designs respond over a fairly narrow band of walking speeds, particularly when the flow of the hydraulic fluid is smooth or laminar. The more advanced hydraulic controls are engineered so the fluid flow becomes turbulent at higher cadences, markedly increasing the damping force applied to the shin.27 When properly aligned and adjusted, such hydraulic swing phase controls allow the patient with an amputated limb to walk at any speed from very slow to a race walking pace, and the knee resistance compensates automatically.
Fluid swing phase control, whether pneumatic or hydraulic, also has been shown to offer the smoothest, most nearly normal swing phase movement possible.10 In general, fluid swing controls should be offered to all amputees capable of varying their walking pace. Less active individuals, whose walking speed varies from slow to moderate, will do well with lower cost, simpler hydraulic dampers or pneumatic units. More vigorous ambulators usually need the degree of control offered by the more advanced hydraulic knee prostheses that incorporate turbulent fluid flow.
The strong damping force that can be achieved with hydraulic knee controls also may be used to augment stance stability. Hans Mauch11 developed the first clinically effective stance and swing control hydraulic knee in the 1950s, after decades of intensive, federally funded research. The Mauch S-N-S cylinder (Mauch industries, Dayton, OH) still is used widely today because of its excellent clinical functions.30 With practice, patients with an amputated limb with good strength and reflexes can learn to walk down ramps and declines confidently with an S-N-S hydraulic knee prosthesis. Some even master stair descent, step over step, allowing the stance resistance to gradually lower them from one riser to the next. The patient also has the option to lock the knee prosthesis against flexion for selected activities.
A more recent development uses a mechanically simpler mechanism to achieve similar hydraulic stance and swing control, although it lacks the locking feature. Stance control in the 3R80 rotary hydraulic knee prosthesis (Otto Bock, Duderstadt, Germany) is engaged by normal weightbearing and disengaged automatically as weight is transferred to the leading limb during normal walking. In addition, the geometry of the 3R80 permits several degrees of controlled knee flexion during the weight acceptance phase of gait. Such a stance flexion feature absorbs some of the shock impact of walking, reduces the rise in the center of gravity, and more closely simulates the kinematics of normal gait.1
Knee prostheses with mechanical swing phase controls also may offer stance stability features. Manually locked knee prostheses normally are reserved for only the most feeble or unsteady amputee, because the stiff legged gait that results is not only abnormal but also requires extra effort from the patient.16 If such a knee mechanism is used in the initial prosthesis, every effort should be made to replace it with a free swinging knee prosthesis at the earliest possible juncture. Sometimes more active patients with amputated limbs use the locking feature selectively, for example, by walking with a free knee prosthesis indoors but locking it for use on irregular ground or in jostling crowds.
Some mechanical knee prostheses also offer a simplified stance control based on a weight activated friction brake. Such mechanical stance control knee prostheses are best suited for very limited ambulators who walk rather slowly. Unfortunately, as the patient with an amputated limb tries to walk at a more normal pace the brake stability is likely to interfere with initiating knee flexion during preswing. As was the case with the manually locked knee prosthesis, it is best to use friction brake knee prostheses only when absolutely necessary, and to provide the patient with a more biomechanically sophisticated option as soon as they are capable of using a different prosthetic knee.14
Polycentric prosthetic knees have many clinical and biomechanical advantages over the more basic designs, which accounts for their increasing use worldwide. Most have four points of rotation, each connected by a linkage bar, and also may be referred to as four-bar knee prostheses. The added mechanical complexity of such devices allows the engineering designers to optimize selected stance and swing phase features. For example, one group of four-bar polycentric knee prostheses is designed primarily to minimize the protrusion of the mechanism when flexed to 90°, specifically to improve the sitting appearance of prostheses for knee disarticulation and similar length residual limbs.12
Another group of polycentric knee prostheses is designed to offer significantly increased stance stability combined with ease of flexion during preswing.23 This apparent paradox is accomplished because of the changing locus of rotation that characterizes these complex mechanisms. The point at which a polycentric knee prosthesis appears to be bending at a given moment is referred to as its instant center of rotation. Many polycentric knee prostheses have an instant center of rotation that is located in a very proximal and posterior location compared with the anatomic knee center. A detailed discussion of the biomechanical advantages that result is beyond the scope of this paper. However, the more posterior the instant center of rotation is located with reference to the ground reaction force, the greater the knee extension moment developed in early stance and the more stable the prosthesis becomes.
The position of the instant center of rotation changes as the polycentric knee is flexed, typically moving along a curved pathway in an anterior and distal direction. The biomechanical advantage of this mechanical characteristic is that once the patient voluntarily has flexed the polycentric knee a few degrees, the instant center of rotation has moved anterior to the ground reaction force and the resulting flexion moment makes continued knee flexion nearly effortless32 (Fig 1). This characteristic ease of voluntary flexion combined with inherent mechanical stance stability makes this type of polycentric knee the device of choice for that broad group of patients with an amputated limb who would benefit from added stability yet are capable of walking at a moderate or higher pace.
Polycentric knee prostheses with friction swing phase controls are reserved for those individuals who walk at one speed only, for the reasons previously discussed. But, compared with a single axis knee, Polycentric knees offer an additional swing phase advantage: greater toe clearance at midswing. The geometry of four-bar knees causes the prosthetic shin to move in a trajectory such that as knee flexion increases, the overall length of the prosthesis slightly decreases and the foot is dorsiflexed simultaneously with reference to the floor. The net result is increased toe clearance of as much as 1 to 2 cm, which the person with an amputation perceives as reducing the risk of stubbing the toe and stumbling.4
Individuals who can vary their walking pace benefit greatly from the addition of a fluid controlled swing phase system to the basic polycentric framework. Many such hybrid knee prostheses are now available, offering a range of hydraulic and pneumatic cadence responsive options in addition to the stance and swing advantages noted above. Some also offer a manual locking feature, for use under special conditions such as descending steep, muddy banks while fishing or when climbing ladders. This combination of desirable features is one reason that successful clinical application of hybrid prosthetic knees is increasing rapidly.
MICROPROCESSOR CONTROLLED PROSTHETIC KNEES
Although proposed by Kobe Steel researchers decades ago, the clinical application of microprocessors controls to prosthetic knee mechanisms is a recent development.5 Two microprocessor controlled pneumatic knee prostheses using the Kobe technology are available: the Endolite Intelligent Prosthesis Plus (Chas. Blatchford & Sons, London, England) and the Seattle Limb Systems Power Knee (Seattle Limb Systems, Seattle WA). Although there are some differences between these devices, both use a single onboard sensor to detect when the knee is in full extension and adjust a pneumatic swing control cylinder accordingly.31
Without microprocessor control, the prosthetist must adjust the knee resistances to a setting that represents the best compromise between very fast and very slow walking. The cadence responsiveness then is caused by the fluid flow characteristics within the damper, as discussed previously. Because of the compressibility of gases such as air, mechanical knees with pneumatic swing control have been characterized as having a narrower cadence band than most hydraulic devices.
However, the addition of microprocessor control allows designers to overcome many of the limitations previously associated with mechanical devices. An onboard computer adjusts the pneumatic resistance of these prosthetic knees, optimizing the adjustment for a broad range of gait speeds from very slow to very fast. Clinically, the prosthetist is able to specify several different optimal adjustments during dynamic alignment that the computer later selects and applies according to the pace of ambulation the person with an amputated limb chooses.
In addition to the obvious advantages of improved pneumatic swing phase responsiveness and better gait symmetry,9 there seems to be a secondary clinical benefit from the versatility added by microprocessor control: the patient with an amputated limb perceives the knee as behaving more consistently and therefore develops more confidence in the prosthesis. In fact, case studies are now emerging that raise the possibility that such devices may decrease the energy the patient expends during ambulation, perhaps because they no longer need to exert as much effort trying to control the swing phase timing of the artificial leg.2,25,29
A more advanced type of computer controlled prosthetic knee and shin system has been released recently, termed the C-Leg (Otto Bock), that uses a hydraulic cylinder to provide not only superior swing phase responsiveness but also to provide variable hydraulic stance phase control. This design also is unique in that it uses multiple sensors that are integrated into the prosthetic shin structure to gather and calculate biomechanical data such as the amount of vertical loading, the sagittal plane ankle moment, and the position, direction, and angular acceleration of the knee joint.7,8 These data are sampled 50 times per second, allowing the computer to readjust the knee accordingly. In essence, the C-Leg uses a software gait analysis algorithm to optimize hydraulic stance and swing control resistances up to 60 times within a typical 1.2-second gait cycle (Fig 2).
The predominantly positive subjective feedback from patients who have used these newest prosthetic knees suggests that microprocessor control can provide a clinically significant improvement in prosthetic design, particularly for the more active individual and for those who have mastered mechanical devices yet wish to be even more mobile. Because of their recent availability and the lack of widespread experience with computer controlled prosthetic knees, objective data about their efficacy are very limited, making it difficult to draw more detailed inferences at this time.
The prescription of prosthetic components is not yet completely scientific, although the amount of objective data available increases with each new published study. But it now is possible for clinicians to rule out specific components as being unlikely to benefit a patient with an amputated limb based on a biomechanical analysis of that person's goals and capabilities. Rehabilitation team discussion then can be limited to those plausible alternatives most likely to be clinically useful to the person for whom the prescription is being written.
To illustrate, the clinic team can determine each individual's stance and swing control needs by answering four questions. Can this person be expected to use their remaining neuromuscular capacity to:(1) control prosthetic knee stability under all circumstances; (2) flex the prosthetic knee in a controlled manner during preswing; (3) walk with a prosthesis at differing speeds; and (4) walk with a prosthesis at a moderate or faster pace?
Those patients who can control voluntarily the stability of the prosthesis under all conditions may use one of the basic single axis designs. The remaining majority should be offered one of the more stable knee mechanisms according to their functional needs.
Those few patients who are totally unable to control the knee, perhaps because of a cerebral vascular accident affecting the amputated limb, may require a manual locking knee despite the gait deviations such a device causes. Whenever possible, however, patients should be offered the opportunity to master a prosthetic knee that will bend during swing phase.
Selected patients who had a recent amputation and feeble patients whose overall physical condition will restrict them to very slow walking speeds may find a friction brake stance control mechanical knee sufficient. However, because this design also may interfere with knee flexion under weightbearing during preswing, it is not recommended for more active users. Many of the Polycentric designs offer excellent inherent stability in combinations with ease of knee flexion for swing phase clearance, making Polycentric knees the preferred choice for most patients needing some degree of stance stability who are otherwise in reasonably good health. Individuals with an unusually long residual limb usually will prefer the special class of four-bar knee prostheses designed to improve sitting cosmesis.
If the patient with an amputated limb is expected to vary their walking speed, then mechanical swing control devices will be insufficient and fluid control should be considered.6 Conversely, one-speed ambulators will derive limited benefit from advanced swing phase controls so more basic mechanical swing control knee prostheses often may be adequate.
As noted before, pneumatic and hydraulic swing phase controls are available. The smaller and lighter hydraulic units and pneumatic cylinders usually are sufficient for slow to moderate cadence walking, and often are prescribed for smaller stature individuals, including women and children. Larger, more powerful hydraulic swing phase controls are particularly appropriate for the taller individual, the more active ambulator, and for those who participate in competitive sports events. Figure 3 summarizes this prescription algorithm.
The added stability offered by those units offering integrated hydraulic stance and swing control make them particularly well suited for community ambulators and for those with an amputation who wish to participate in outdoor activities, because of vocational or avocational interests. Patients with bilateral limb loss often benefit significantly from the yielding stance resistance that is unique to such stance and swing control hydraulic knee components, particularly when sitting down or arising from a seated position.14 Even if the ability to walk faster is somewhat limited, prescription of hydraulic stance and swing control knee components often is very helpful to patients with bilateral amputations as these prosthetic components also offer the smoothest possible gait.19
Specific indications for presently available microprocessor controlled knee prostheses have not yet been established. Because their long term durability is not yet fully documented, the ability of the user to return for followup adjustments or servicing is certainly an important consideration at this time. In general, it seems that those amputees who would be reasonably good ambulators with conventional technology will derive the greatest benefits from such sophisticated knee controls. Although microprocessor controlled prosthetic knees are a logical progression for the individual who previously has mastered conventional fluid controlled components, it is more economic to prescribe such advanced technology initially for appropriate candidates.
Today's clinician must consider a plethora of prosthetic knee mechanisms when formulating the prescription for an artificial limb. A basic understanding of the limitations of the different types of prosthetic knee control is useful to rule out inappropriate alternatives for a given individual. Careful consideration of the functional capabilities of each person with an amputated limb allows the rehabilitation team to identify those prosthetic knee designs that offer the required amount of stance and swing phase control. Recent technologic advances, including the addition of microprocessor controls, have resulted in prosthetic knees which offer excellent stance stability and enhanced swing phase responsiveness. Proper application of today's advanced prosthetic technology allows the active patient with an amputated limb to participate in a broad range of vocational and avocational activities, including recreational and competitive sports.17
1. Blumentritt S, Scherer HW, Wellershaus U, Michael JW: Gait analysis of transfemoral amputees walking on a rotary hydraulic prosthetic knee mechanism: A preliminary report. J Prosthet Orthot 10: 61-70, 1998.
2. Buckley JG, Spence WD, Solomonidis S: Energy cost of walking: Comparison of intelligent prosthesis with conventional mechanism. Arch Phys Med Rehabil 78:330-333, 1997.
3. Dietl H, Bargehr H: Der einsatz von elektronik bei prothesen zur versorgung der unteren extremitat (the application of electronics in prosthetics for lower extremities). Med Orthop Tech 117: 31-35, 1997.
4. Gard SA, Childress DS, Uellendahl JE: The influence of four-bar linkage knees on prosthetic swingphase floor clearance. J Prosthet Orthot 8: 34-40, 1996.
5. Hadfield P, Clery D: Electronic leg helps walkers get in the swing. New Scientist 12: 21, 1992.
6. Hicks R, Tashman S, Cary JM, Altman RF, Gage JR: Swing phase control with knee friction in juvenile amputees. J Orthop Res 3: 198-201, 1985.
7. James K: Swing and stance computer control of an above knee prosthesis. Abstracts of the 8th World Congress, International Society for Prosthetics and Orthotics Melbourne, Australia 6, 1995.
8. James K, Stein RB, Rolf R, Tepavic D: Active suspension above knee prosthesis. Proceeding of the World Congress of the International Society for Prosthetics and Orthotics 7:346, 1992.
9. Kirker S, Keymers S, Talbot J, et al: An assessment of the intelligent knee prosthesis. Clin Rehabil 10: 267-273, 1996.
10. Lewis EA: Fluid-control knee mechanisms: Clinical considerations. Bull Prosthet Res 10: 24-56, 1965.
11. Mauch HA: Stance control for above-knee artificial legs - design considerations in the S-N-S knee. Bull Prosthet Res 10: 61-72, 1968.
12. Michael JW: Component selection criteria: Lower limb disarticulations. Clin Prosthet Orthot 12: 99-108, 1988.
13. Michael JW: Pediatric Prosthetics and Orthotics. In Hedman G (ed). Physical and Occupational Therapy in Pediatrics. Glen Hedman New York, Haworth Press 123-146, 1990.
14. Michael JW: Prosthetic knee mechanisms. Phys Med Rehabil: State Art Rev 8: 147-164, 1994.
15. Michael JW: Prosthetic Management and Biomechanics for Transfemoral Amputation. In Murdoch G, Wilson AB (eds). Amputation Surgery: Surgical Practice and Patient Management 1996.
16. Michael JW: Management of High Level and Bilateral Lower Limb Amputations. In Nielsen C (ed). Orthotics & Prosthetics in Rehabilitation 1998. In press.
17. Michael JW, Gailey RS, Bowker JH: New developments in recreational prostheses and adaptive devices for the amputee. Clin Orthop 256: 64-75, 1990.
18. Murphy EF: The swing phase of walking with above-knee prostheses. Bull Prosthet Res 10: 5-39, 1964.
19. Murray MP, Mollinger LA, Sepic SB, Gardner GM: Gait patterns in above-knee amputee patients: Hydraulic swing control vs. constant-friction knee components. Arch Phys Med Rehabil 64: 339-345, 1983.
20. Murray MP, Sepic SB: Gait patterns of above-knee amputees using constant-friction knee components. Bull Prosthet Res 10: 35-45, 1980.
21. Radcliffe CW: Biomechanical design of a lower extremity prosthesis. Am Soc Mech Eng 60-WA-305: 1-15, 1961.
22. Radcliffe CW: Above knee prosthetics. Prosthet Orthot Int 1:146-160, 1977.
23. Radcliffe CW: Four-bar linkage prosthetic knee mechanisms: Kinematics, alignment and prescription criteria. Prosthet Orthot Int 18: 159-173, 1994.
24. Radcliffe CW, Lamoreux L: UC-BL Pneumatic swing-control unit for above-knee prostheses: Design, adjustment and installation. Bull Prosthet Res 10: 73-89, 1968.
25. Schmalz T, Blumentritt S, Tsukishiro K, Kocher L, Dietl H: Energy efficiency of trans-femoral amputees walking on computer-controlled prosthetic knee joint "C-Leg". Abstracts of the Ninth World Congress, International Society for Prosthetics and Orthotics Amsterdam, The Netherlands 459-460, 1998.
26. Staros A: The principles of swing-phase control: The advantages of fluid mechanisms. Prostheses Braces Tech Aids 13: 11-16, 1964.
27. Staros A, Murphy EF: Properties of fluid flow applied to above-knee prostheses. Bull Prosthet Res 10: 40-65, 1964.
28. Staros A, Rubin G: Prescription considerations in modern above-knee prosthetics. Phys Med Rehabil Clin North Am 2: 311-324, 1991.
29. Taylor MB, Clark E, Offord EA: A comparison of energy expenditure by a high level transfemoral amputee using the intelligent prosthesis and conventionally damped prosthetic limbs. Prosthet Orthot Int 20: 116-121, 1996.
30. Volatile TB, Roberson JR, Whitesides TE: The Mauch hydraulic knee unit for above knee amputation. Orthopedics 8: 229-230, 1985.
31. Zahedi S: Bewertung und biomechanik der intelligenten prothes-eine zwei-jahres-studie (evaluation and biomechanics of the intelligent prosthesis-a two year study). Orthop Tech 46: 32-40, 1995.
32. Zarrugh MY, Radcliffe CW: Simulation of swing phase dynamics in above-knee prostheses. J Biomech 9: 283-292, 1976.