Current Diagnosis and Treatment in Orthopedics, 4th Edition

Chapter 12. Amputations


Amputations are performed to remove extremities that are severely diseased, injured, or no longer functional. Although medical advances in antibiotics, trauma care, vascular surgery, and the treatment of neoplasms have improved the prospects for limb salvage, in many cases prolonged attempts to save a limb that should be amputated lead to excessive morbidity or even death. To counsel a patient regarding amputation versus limb salvage adequately, the physician must provide sufficient information about the surgical and rehabilitative steps involved with each procedure and must also appraise the probable outcome for function realistically with each alternative. Attempting to salvage a limb is not always in the best interest of the patient.

The decision to amputate is an emotional process for the patient, the patient's family, and the surgeon. The value of taking a positive approach to amputation cannot be overemphasized. It is not a failure and should never be viewed as such. The amputation is a reconstructive procedure designed to help the patient create a new interface with the world and to resume his or her life. The residual limb must be surgically constructed with care to maintain muscle balance, transfer weight loads appropriately, and assume its new role of replacing the original limb.

For patients to achieve maximal function of the residual limb, they also need a clear understanding of what to expect for an early postoperative prosthetic fitting, a rehabilitation program, and for long-term medical and prosthetic needs. For these discussions, the team approach to meeting the patient's needs can be especially rewarding. Nurses, prosthetists, physical and occupational therapists, and amputee support groups can be invaluable in providing the physical, psychologic, emotional, and educational support needed in returning patients to a full and active life. Many new amputees state that a peer visitor program was one of the most helpful events during their hospitalization and rehabilitation. The Amputee Coalition of America, a not-for-profit organization, supports this peer visitor training and can help locate programs that are available throughout the country.


In infants and children, amputations are frequently associated with congenital limb deficiencies, trauma, and tumors. Congenital limb deficiencies are commonly described using the Birch revision of the Frantz and O'Rahilly classification system. Amelia is the complete absence of a limb; hemimelia is the absence of a major portion of a limb; and phocomelia is the attachment of the terminal limb at or near the trunk. Hemimelias can be further classified as terminal or intercalary. A terminal hemimelia is a complete transverse deficit at the end of the limb. An intercalary hemimelia is an internal segmental deficit with variable distal formation. In discussions of limb deficiencies, preaxial refers to the radial or tibial side of a limb, and postaxial refers to the ulnar or fibular side. The International Organization for Standardization (ISO) published a recommended classification for limb deficiencies in 1989 based on standard anatomic and radiologic characteristics and terminology. Although the ISO intentionally avoided the use of the terminology in the Frantz and O'Rahilly system, the older system is widely used, and the definitions and unusual language must still be understood by those caring for children with limb deficiency.

Reamputation of a congenital upper limb deficiency is rarely indicated, and even rudimentary appendages can often be functionally useful. In the lower limb, however, the ability to bear weight and the relative equality of leg lengths are mandatory for maximal function. Reamputation may be indicated in proximal femoral focal deficiency and congenital absence of the fibula or tibia to produce a more functional residual limb and improve prosthetic placement.

In the growing child, proportional change occurs in residual limb length from childhood to adulthood—an important concept to keep in mind when determining the surgical approach. A diaphyseal amputation in an infant or young child removes one of the epiphyseal growth centers, and the involved bone therefore does not keep proportional growth with the rest of the body. What initially appears to be a long transfemoral amputation in a small child can turn out to be a short and troublesome residual limb when the child reaches skeletal maturity. All attempts should be made to save the distal-most epiphysis by disarticulation. If this is not technically possible, the greatest amount of bone length should be saved.

Terminal overgrowth occurs in 8–12% of pediatric patients who had a surgical amputation. The growth of appositional bone at the transected end of a long bone exceeds the growth of the surrounding soft tissues. If left untreated, the appositional bone can penetrate through the skin (Figure 12–1). Terminal overgrowth of the transected bone does not occur as a result of the normal growth from the proximal physis pushing the distal end of the bone through the soft tissues, nor does it occur in limb disarticulations. Terminal overgrowth occurs most commonly in the humerus, fibula, tibia, and femur, in that order. Although numerous surgical procedures are used to manage this problem, the best approach consists of stump revision with adequate bone resection or autogenous osteochondral stump capping as originally described by Marquardt (Figure 12–2). If the stump-capping procedure is done at the time of original amputation, the graft material can be obtained from part of the amputated limb, such as the distal tibia, talus, or calcaneus. If a procedure is done later, the graft material can be obtained from the posterior iliac crest. Although techniques with nonautologous material are used, significant complications are reported. A report of using a modified Ertl osteomyoplasty to prevent terminal overgrowth in childhood limb deficiencies was not successful.

Figure 12–1.


Terminal overgrowth of the transected bone in a pediatric amputee.


Figure 12–2.


Stump-capping procedure. The bone end was split longitudinally and the osteochondral graft fixed temporarily with K-wires.

In a growing child, the fitting of a prosthesis can be challenging and requires frequent adjustments. Specialty pediatric amputee clinics can ease this process, provide family support, and make care more cost effective. The timing of prosthetic fitting should be initiated to coincide closely with normal motor skill development.

Prosthetic fitting for the upper limb should begin near the time the child gains sitting balance, usually around 4–6 months of age. A passive terminal device with blunt rounded edges is used initially. Active cable control and a voluntary opening terminal device are added when the child exhibits initiative in placing objects in the terminal device, usually in the second year of life. Myoelectric devices are usually not prescribed until the child masters traditional body-powered devices. The physical demand placed on prosthetic devices by children can often exceed the durability of current myoelectric designs, so maintenance and repair costs must be considered. The decision to prescribe a myoelectric device for a child is individual and depends on many factors, including the physical characteristics of the residual limb, the desires of the child, the training available, the proximity of prosthetic facilities for fitting and maintenance, and issues about funding.

Prosthetic fitting for the lower limb commonly begins when the child develops the ability to crawl and pull to a standing position, which is usually at 8–12 months. A child with a Syme amputation or a transtibial amputation generally adapts to a prosthesis with surprising ease, and although formal gait training is not required, educational efforts are focused on teaching the parents about the prosthesis. For a child with a transfemoral amputation, control of a knee unit should not be expected immediately. The knee unit should be eliminated or locked in extension until the child is ambulating well and demonstrates proficient use of the prosthesis. The initial gait pattern used by a child with a transfemoral amputation is not a normal heel strike, midstance, toe-off gait pattern but is instead a more circumducted gait pattern with a prolonged foot flat phase. Formal gait training is seldom warranted until the child reaches 5 or 6 years of age. Attempts to force gait training too early can be frustrating for all involved. When pediatric patients are allowed to develop their own gait patterns as they grow and gain improved motor coordination, they are surprisingly adept at discovering the most efficient gait pattern without formal training.

Bernd L et al: The autologous stump plasty: Treatment for bony overgrowth in juvenile amputees. J Bone Joint Surg Br 1991;73:203. [PMID: 2005139] 

Birch JG et al: Syme amputation for the treatment of fibular deficiency. An evaluation of long-term physical and psychological functional status. J Bone Joint Surg Am 1999;81(11):1511. [PMID: 10565642] 

Drvaric DM, Kruger LM: Modified Ertl osteomyoplasty for terminal overgrowth in childhood limb deficiencies. J Pediatr Orthop 2001;21(3):392. [PMID: 1137827] 

Fixsen JA: Major lower limb congenital shortening: A mini review. J Pediatr Orthop B 2003;12:1. [PMID: 124887464] 

Greene WG, Cary JM: Partial foot amputation in children: A comparison of the several types with the Syme's amputation. J Bone Joint Surg Am 1982;64:438. [PMID: 7061561] 

International Organization for Standardization: ISO 8548-1: Prosthetics and orthotics—Limb deficiencies, Part 1: Method of describing limb deficiencies present at birth. International Organization for Standardization, 1989.

Pfeil J et al: The stump-capping procedure to prevent or treat terminal osseous overgrowth. Prosthet Orthot Int 1991;15:96.

Weber M: Neurovascular calcaneo-cutaneus pedicle graft for stump capping in congenital pseudarthrosis of the tibia: Preliminary report of a new technique. J Pediatr Orthop B 2002;11(1):47. [PMID: 11866081] 



Epidemiologic data on the incidence of amputation in the United States from 1993 to 2001 show the number of lower extremity amputations increased 14% from 99,522 to 113,379, and the average hospital charge for this procedure increased 38% from $24,332 to $33,562. Nearly two thirds of amputations are performed in individuals with diabetes, even though people with diabetes represent only 6% of the population.

Preoperative Evaluation & Decision Making

The decision to amputate a limb and the choice of amputation level can be difficult and are often subject to differences in opinion. Advances in the treatment of infection, peripheral vascular disease, replantation, and limb salvage complicate the decision-making process. The goals are to optimize a patient's function and reduce the level of morbidity.


Ischemia resulting from peripheral vascular disease remains the most frequent reason for amputation in the United States. Nearly two thirds of patients with ischemia also have diabetes. The preoperative assessment of these patients includes a physical examination and an evaluation of perfusion, nutrition, and immunocompetence. Preoperative screening tests can be helpful, but no single test is 100% accurate in predicting successful healing. Clinical judgment based on experience in examining and following many patients with vascular disease and diabetes is still the most important factor in preoperative assessment.

Doppler Ultrasound Studies

The most readily available objective measurement of limb blood flow and perfusion is by Doppler ultrasound. Arterial wall calcification increases the pressure needed to compress the vessels of patients with vascular disease, and this often gives an artificially elevated reading. Low-pressure levels are indicative of poor perfusion, but normal and high levels can be confusing because of vessel wall calcification and are not predictive of normal perfusion or of wound healing. Digital vessels are not usually calcified, and blood pressure levels in the toes appear to be more predictive of healing than do those in the ankles.

Transcutaneous Oxygen Tension Measurements

Tests to measure transcutaneous partial pressure of oxygen (PO2) are noninvasive and becoming more readily available in many vascular laboratories. These tests use a special temperature-controlled oxygen electrode to measure the PO2 diffusing through the skin. The ultimate reading is based on several factors: the delivery of oxygen to the tissue, the utilization of oxygen by the tissue, and the diffusion of oxygen through the tissue and skin. Caution in interpreting the transcutaneous PO2 measurements during acute cellulitis or edema is warranted because the presence of either of these disorders can increase oxygen utilization and decrease oxygen diffusion, thereby resulting in lower measurements of PO2. Paradoxical measurements are also reported on the plantar skin of the foot. In spite of these limitations, transcutaneous PO2 and transcutaneous partial pressure of carbon dioxide (PCO2) are both statistically accurate in predicting amputation healing, but this does not rule out false-negative results.

Xenon Studies

Xenon-133 (133Xe) skin clearance studies are used successfully to predict healing of amputations, but the preparation of the mixture containing xenon-133 gas and saline solution and the administration of the test are time consuming, highly technician dependent, and expensive. A small amount of the xenon and saline solution is injected intradermally at various sites, and the rate of washout is monitored by gamma camera. Xenon-133 is almost never used today and is primarily of historical interest.

Fluorescence Studies

Skin fluorescence studies use intravenous injection of fluorescein dye and subjective observation or digital fluorometers to assess skin blood flow and correlate this with the likelihood of successful wound healing. The technique is not commonly used, and studies to assess its accuracy yielded conflicting results.


Arteriography is not helpful in predicting successful healing of amputations, and this invasive test is probably not indicated solely for the purpose of selecting the proper level of amputation. Arteriography is indicated if the patient is truly a candidate for arterial reconstruction or angioplasty.

Nutrition and Immunocompetence Studies

Both nutrition and immunocompetence correlate directly with amputation wound healing. Many laboratory tests are available to assess nutrition and immunocompetence, but some are quite expensive. Screening tests for albumin level and total lymphocyte count are readily available and inexpensive. Several studies show increased healing of amputations in patients who have vascular disorders but have a serum albumin level of at least 3 g/dL and a total lymphocyte count exceeding 1500/mL. Nutritional screening is recommended to allow for nutritional improvement preoperatively and to help determine whether a higher level of amputation is needed.

Other Issues

Activity level, ambulatory potential, cognitive skills, and overall medical condition must be evaluated to determine if the distal-most level of amputation is really appropriate for the patient.

For patients who are likely to remain ambulatory, the goals are to achieve healing at the distal-most level that can be fit with a prosthesis and to make successful rehabilitation possible. Newer studies of patients with vascular insufficiency and diabetes demonstrate that successful wound healing can be achieved in 70–80% of these patients at the transtibial or more distal amputation level. This is in sharp contrast to 25 years ago, when because of a fear of wound failure, surgeons elected to perform 80% of all lower extremity amputations at the transfemoral level.

For nonambulatory patients, the goals are not simply to obtain wound healing but also to minimize complications, improve sitting balance, and facilitate position transfers. Occasionally, a more proximal amputation more successfully meets these goals. For example, a bedridden patient with a knee flexion contracture might be better served with a knee disarticulation than a transtibial amputation, even if the biologic factors are present to heal the more distal amputation. Preoperative assessment of the patient's potential ability to use a prosthesis, the patient's specific needs for maintaining independent transfers, and the best weight distribution for seating can help in making wise decisions concerning the appropriate level of amputation and the most successful type of postoperative rehabilitation program.

Some nonambulatory patients do benefit from a partial foot amputation, or even transtibial amputation with prosthetic fitting, not with the goal of walking but to use that leg as a standing pivot for independent transfers. In these cases, prosthetic fitting is justified.


As vascular reconstruction techniques improved, more attempts to salvage limbs were initially made, often with the result that multiple surgical procedures were subsequently required. In many cases, amputation was ultimately performed after a substantial investment of time, money, and emotional energy. Current studies offer guidelines for immediate or early amputation and show the value of amputation not only in saving lives but also in preventing the emotional, marital, and financial disasters that can follow unwise and desperate limb salvage attempts. Although several scoring systems for mangled limbs are published, none can perfectly predict when an amputation should be performed. These scores can help in the decision-making process, but good clinical experience and judgment are still required.

The absolute indication for amputation in trauma remains an ischemic limb with unreconstructable vascular injury. Massively crushed muscle and ischemic tissue release myoglobin and cell toxins, which can lead to renal failure, adult respiratory distress syndrome, and even death. In two groups of high-risk patients (multiply injured patients and elderly patients with a mangled extremity), limb salvage, even though technically possible, can become life-threatening and generally should be avoided. In all patients, the decision about whether to undertake immediate or early amputation of a mangled limb must also depend on whether it is an upper extremity or lower extremity.

An upper extremity can function with minimal or protective sensation, and even a severely compromised arm can serve as an assistive limb. An assistive upper extremity often functions better than the currently available prosthetic replacements. The decision of salvage versus amputation in the upper limb should be based on the chance of maintaining some useful function, even if that function is limited.

In the lower extremity, weight bearing is mandatory. A lower limb functions poorly without sensation, and an assistive limb is not useful. A salvaged lower limb often functions worse than a modern prosthetic replacement unless the limb can tolerate full weight bearing, is relatively pain free, has enough sensation to provide protective feedback, and has durable skin and soft-tissue coverage that does not break down whenever walking is attempted. The decision to salvage a mangled lower extremity should be based on providing a limb that can tolerate the demands of walking.


Exposure to cold temperatures can directly damage the tissue and cause a related vascular impairment from endothelial vessel injury and increased sympathetic tone. If the foot or hand is wet or directly exposed to the wind, cold injury can result even in temperatures above freezing. The immediate treatment involves restoring the core body temperature and then rewarming the injured body part in a water bath at a temperature of 40–44°C for 20–30 minutes. Rewarming can be painful, and the patient often requires opiate analgesia. After rewarming, the involved part should be kept dry, blisters left intact, and dry gauze dressings used. The goals are to keep the injured extremity clean and dry and to prevent maceration, especially between the digits.

The temptation to perform early amputation should be avoided because the amount of recovery can be dramatic. As the extremity recovers from frostbite, a zone of mummification (dry gangrene) develops distally, and a zone of intermediate tissue injury forms just proximal to this. Even at the time of clear demarcation, the tissue just proximal to the zone of mummification continues to heal from the cold insult, and although the outward appearance is often pink and healthy, this tissue is not totally normal. Delaying amputation can improve the chance of primary wound healing. It is not unusual to wait 2–6 months for definitive surgery. In spite of having mummified tissue, infection is rare if the tissue is kept clean and dry.


Patients with musculoskeletal neoplasms face new choices in treatment with the development of limb salvage techniques and adjuvant chemotherapy and radiation therapy. If an amputation is chosen, the amputation incisions must be carefully planned to achieve the appropriate surgical margin.

Surgical margins (Figure 12–3) are characterized by the relationship of the surgical incision to the lesion, to the inflammatory zone surrounding the lesion, and to the anatomic compartment in which the lesion is located. The four types of margins are the intralesional margin, in which the surgical incision enters the lesion; the marginal margin, in which the incision enters the inflammatory zone but not the lesion; the wide margin, in which the incision enters the same anatomic compartment as the lesion but is outside of the inflammatory zone; and the radical margin, in which the incision remains outside of the involved anatomic compartment. Biopsy incisions and amputation incisions must be planned with careful consideration as to the tumor margin required.

Figure 12–3.


Surgical margins in tumors of the extremity.

Newer studies continue to evaluate the complex issues and outcomes of amputation versus limb-sparing procedures for patients with extremity sarcomas. Studies still suggest that functional outcomes in terms of kinesiologic parameters are comparable with either limb salvage or amputation. Both treatment groups report quality of life problems involving employment, health insurance, social isolation, and poor self-esteem. Overall survival remains comparable with either treatment. In some tumors, amputation may achieve better local disease control. These results confirm that the decision about treatment must be made on an individual basis, according to the specific lifestyle and needs of the patient.

Surgical Definitions & Techniques

Terminology for amputation level now uses an accepted international nomenclature. Transtibial should be used instead of below knee, and transfemoral instead of above knee. In the upper extremity, the terms transradial and transhumeral replace the older terms below elbow and above elbow.

Careful surgical techniques, especially in soft-tissue handling, are more critical to wound healing and functional outcome in amputation procedures than in many other surgical procedures. The tissues are often traumatized or poorly vascularized, and the risk of wound failure is high, particularly if close attention is not paid to soft-tissue technique. Flaps should be kept thick, avoiding unnecessary dissection between the skin and subcutaneous, fascial, and muscle planes. In adults, periosteum should not be stripped proximal to the level of transection. In children, however, removing 0.5 cm of the distal periosteum may help prevent terminal overgrowth. The rounding of all bone edges and the beveling of prominences are necessary for optimal prosthetic use.

Muscle loses its contractile function when the skeletal attachments are divided during amputation. Stabilizing the distal insertion of muscle can improve residual limb function by preventing muscle atrophy, providing counterbalance to the deforming forces resulting from amputation, and providing stable padding over the end of the bone. Myodesis is the direct suturing of muscle or tendon to the bone or the periosteum. Myodesis techniques are most effective in stabilizing strong muscles needed to counteract strong antagonistic muscle forces, such as in cases involving transfemoral or transhumeral amputation and in cases involving knee or elbow disarticulation. Myoplasty involves the suturing of muscle to muscle over the end of the bone. The distal stabilization of the muscle is more secure with myodesis than with myoplasty. Care must be taken to prevent a mobile sling of muscle over the distal end of the bone, which usually results in a painful bursa.

The transection of nerves always results in neuroma formation, but all neuromas are not symptomatic. Historical attempts to diminish symptomatic neuromas include clean transection, ligation, crushing, cauterization, capping, perineural closure, and end-loop anastomosis. No technique is more effective than careful and meticulous isolation, retraction, and clean transection of the nerve. This allows the cut end to retract into the soft tissues, away from the scar, pulsating vessels, and prosthetic pressure points. Ligation of a nerve is still indicated to control bleeding from the blood vessels contained within larger nerves, such as the sciatic.

Split-thickness skin grafts are generally discouraged except as a means to save a knee or elbow joint that has a stable bone and good muscle coverage. Skin grafts do best with adequate soft-tissue support and are least durable when closely adherent to bone. New prosthetic interfaces, such as silicone-based liners, can help reduce the shear at the interface and improve durability in skin-grafted residual limbs.

An open amputation is occasionally necessary to control a severe ascending infection. The term guillotine amputation should be avoided because it gives the impression that the limb is transected at one level through skin, muscle, and bone. Open amputations need to be performed with careful planning and forethought as to how the amputation will eventually be closed. The surgical plan must obviously consider adequate debridement of tissue necrosis and drainage of infection but must also consider the surgical flaps and tissue needed for a functional closure of the amputation to allow prosthetic fitting.

The problem of ascending infection is seen, for example, in a diabetic patient with a severe infection of the foot and cellulitis extending upward to the calf. The open amputation removes the source of infection, provides adequate drainage, and allows the acute cellulitis to resolve. After resolution, a definitive amputation and closure can be done safely. In the case of a diabetic foot infection, an open ankle disarticulation is simple, relatively bloodless, and preserves the posterior calf flap for a definitive transtibial amputation. Occasionally, it is necessary to make a longitudinal incision to drain the posterior tibial, anterior tibial, or peroneal tendon sheaths, in which case care should be taken not to violate the posterior flap of the definitive amputation. This approach often prevents having open, transected muscle bellies that can retract and become edematous—a problem that commonly occurs if an open calf-level amputation was initially performed and one that can make the definitive amputation difficult. In more severe infections or in cases in which the level of the definitive amputation will clearly be transfemoral, an open knee disarticulation has the same advantages as the open ankle disarticulation.

Postoperative Care


The terminal amputation allows the unique opportunity to manipulate the physical environment of the wound during healing. A variety of methods are described, including rigid dressings, soft dressings, controlled environment chambers, air splints, and skin traction. The use of a rigid dressing controls edema, protects the limb from trauma, decreases postoperative pain, and allows early mobilization and rehabilitation.

The use of an immediate postoperative prosthesis, or IPOP (Figure 12–4), is effective in decreasing the time to limb maturation and the time to definitive prosthetic fitting. In most cases involving a lower limb amputation, the surgeon has the patient start with partial weight bearing if the wound appears stable after the first cast change, which usually takes place between the fifth and tenth day after surgery. Immediate postoperative weight bearing can be initiated safely in selected patients, usually young patients in whom an amputation was performed following a traumatic injury and above the zone of injury. Rigid dressings and the IPOP need to be applied carefully, but their application is easily learned and well within the scope of interested physicians. For upper extremity amputations, an IPOP can be applied immediately. Early training with an IPOP is believed to increase the long-term acceptance and use of a prosthesis. Chapter 13 offers a detailed discussion of rehabilitation.

Figure 12–4.


Immediate postoperative prosthetic cast for transtibial amputation.

To counsel patients adequately , some insight into the typical surgical and postoperative course can be helpful. Many patients require inpatient hospital care for 5–8 days after a transtibial amputation. Epidural or patient-controlled analgesia is usually required for pain control. Assistance with basic mobility and emotional support are also necessary. Antibiotics can minimize the risk of infection. The cast applied at the end of the surgical procedure is changed about postoperative day 5. If the wound healing is adequate, a new cast with a foot attachment is applied, and the patient can begin ambulating with approximately 30 lb of weight on the amputated extremity. Transtibial amputees are discharged to home or a nursing facility typically 5 or 8 days after surgery. Outpatient visits are scheduled weekly to change the cast, which frequently becomes loose as edema lessens, and to monitor wound healing and allow suture removal. Active and active-assisted knee range of motion (ROM) is performed between each cast. On average, approximately six casts are applied on a weekly basis until the wound heals, edema resolves, wrinkles return to the skin, and the patient is ready for prosthetic fitting. The cast and the prosthetic foot attachment are applied and aligned by either the surgeon or the prosthetist. New prefabricated, removable postoperative prosthetic systems are alternatives to the traditional casting techniques. Unfortunately, comparison trails versus traditional techniques have not been done.

Close interaction between the patient, the physical therapist, and the prosthetist is required in the first 12–18 months. The socket made for the first prosthesis must allow modifications as the residual limb continues to change shape during this time. Volume changes and mismatch between the shape of the socket and the evolving shape of the residual limb are treated with amputation socks and by adding pads to the socket or socket liner. Pads are usually needed in the region that contacts the anteromedial and anterolateral tibial flares, and posteriorly, in the popliteal region. Even with careful modifications, the prosthetic socket must be changed two or three times in the first 18 months. Because of these frequent prosthetic modifications, encouraging the patient to work with a prosthetic provider who is located close to the patient's residence can help tremendously in this rehabilitation phase. Many patients have an immediate desire to have the most advanced and high-tech components in their first prosthesis. But often these components are designed for higher activity levels than are typically achieved in the rehabilitation phase and are too rigid. Discussing how the prosthesis will evolve and be upgraded as the patient's activity increases can ease this process. A new prosthesis is typically required around month 18; the old components often can be turned into a shower leg.


Failure of the Wound to Heal Properly

Problems with wound healing, especially in diabetic and ischemic limbs, occur as the result of insufficient blood supply, infection, or errors in surgical technique. Healing failure rates are difficult to interpret because they depend so much on the level of amputation selected. Low failure rates can be achieved by doing amputations at an extremely proximal level in the majority of cases, but this sacrifices the rehabilitation potential of many patients because the ability to ambulate decreases dramatically with a transfemoral amputation. Wound healing failure that necessitates reamputation at a more proximal level occurs in approximately 5–10% of cases at centers specializing in amputee treatment.

Most surgeons prefer open wound care if the wound gap is less than 1 cm wide and prefer revision surgery if the gap is wider. If the surgical edema has resolved and some atrophy has already occurred, a wedge excision of all nonviable tissue can be performed and still allow primary closure without any tension at the original level. If it is not possible to oppose the viable tissue gently without tension, bone shortening or reamputation at a more proximal level should be performed.

In patients with small local areas of wound-healing failure, successful treatment with rigid dressings and an IPOP is reported. The wounds are debrided weekly and packed open, and the IPOP is applied to allow some weight bearing. The stimulation of weight bearing can increase local circulation, decrease edema, and promote wound healing.


Infection without widespread tissue necrosis or flap failure may be seen after surgery, especially if active distal infection was present at the time of the definitive amputation or if the amputation was done near the zone of a traumatic injury. Hematomas can also predispose a wound to infection. In cases involving infection or hematomas, the wound must be opened, drained, and debrided. If the wound is allowed to remain open for an extended time, the flaps retract and become edematous, which makes delayed closure difficult or impossible without shortening the bone. One solution, which can be instituted after thorough debridement and irrigation, is to close only the central one third to one half of the amputation wound and to use open packing for the medial and lateral corners (Figure 12–5). This method provides coverage of the bone but also allows adequate drainage and open wound management for the edges. If the original problem was truly infection and not tissue failure, the open portions of the wound heal secondarily, and the result is still a residual limb suitable for prosthetic fitting.

Figure 12–5.


Partial closure of the infected transtibial amputation.

Phantom Sensation

Phantom sensation is the feeling that all or a part of the amputated limb is still present. This sensation is felt by nearly everyone who undergoes surgical amputation, but it is not always bothersome. Phantom sensation usually diminishes over time, and telescoping (the sensation that the phantom foot or hand has moved proximally toward the stump) commonly occurs.

Pain and Phantom Pain

Phantom pain is defined as a bothersome, painful, or burning sensation in the part of the limb that is missing. Although from 80% to 90% of patients with acquired amputation experience some episodes of phantom pain, the episodes are often infrequent and brief. The dreaded problem of unrelenting phantom pain fortunately occurs only in a much smaller minority of patients. Surgical intervention for this problem is not very successful.

Local physical measures, including massage, cold packs, exercise, neuromuscular stimulation by external electrical currents, acupuncture, and regional sympathectomy, may under given circumstances have a place in therapy when the pain is intractable. A technique that has gained some acceptance and success is the use of transcutaneous electrical nerve stimulation (TENS), incorporated either into a prosthesis or used as an isolated unit. The TENS system can be worn by the amputee at night and even during the day with the battery pack attached to the belt or inside a pocket. We use this TENS system with moderate short-term success, but it is rare to see a patient who continues to use a TENS system for more than a year.

Pharmacologic treatment shows some success with several oral agents including gabapentin, amitriptyline, carbamazepine, phenytoin, and mexiletine. Medications can decrease the frequency of phantom pain episodes and decrease the intensity of these episodes. The appropriate use of an intravenous lidocaine challenge is predictive of a favorable response to oral mexiletine. Unfortunately, no indicators are good at predicting who will respond to treatment with gabapentin, amitriptyline, carbamazepine, or phenytoin. Psychological support can be beneficial, particularly when personality problems seem to accentuate the occurrence of pain. The individual needs patience and reassurance that the discomfort will improve with time, especially when a supportive social environment is present.

The sensations described by patients with phantom pain may be similar to the symptoms of reflex sympathetic dystrophy after an injury. Reflex sympathetic dystrophy can occur in amputated limbs and should be treated aggressively if present. Although rare, pain unrelated to the amputation can easily be overlooked. The differential diagnosis includes radicular nerve pain from proximal entrapment or disk herniation, arthritis of proximal joints, ischemic pain, and referred visceral pain.

Research has progressed in the prevention of phantom limb pain. Several authors document that the use of perioperative epidural anesthesia or intraneural anesthesia can block the acute pain associated with amputation surgery and decrease the opiate requirements in the immediate postoperative period. They also suggest that perioperative analgesia can prevent or decrease the later incidence of phantom pain, although this is difficult to document. The literature unfortunately is not conclusive on whether preemptive measures can truly reduce the frequency or severity of phantom limb pain. Some reports dispute the claims that preemptive analgesia reduces the frequency of phantom limb problems. A randomized trial by Lambert and colleagues found that perioperative epidural block started 24 hours before amputation is not superior to infusion of local anesthetic via a perineural catheter in preventing phantom pain but does give better relief in the immediate postoperative period.


Postoperative edema is common in patients who have undergone amputation. Rigid dressings can help reduce this problem. If soft dressings are used, they should be combined with stump wrapping to control edema, especially if the patient is a prosthetic candidate. The ideal shape of a residual limb is cylindrical, not conical. One common mistake is wrapping the stump too tightly at the proximal end, which can lead to congestion and worsening edema and also cause the residual limb to become shaped like a dumbbell. Another common mistake is not wrapping transfemoral amputations in a waist-high soft spica cast that includes the groin. If wrapped incorrectly, the limb has a narrow, conical shape, and a large adductor roll develops. Because of the difficulty in wrapping the transfemoral amputation with elastic bandages, shrinker socks with a waist belt are frequently used as a safer alternative for the transfemoral level.

Stump edema syndrome is a condition commonly caused by proximal constriction and characterized by edema, pain, blood in the skin, and increased pigmentation. The syndrome usually responds to temporary removal of the prosthesis, elevation of the residual limb, and compression.

Joint Contractures

Joint contractures usually develop between the time of amputation and prosthetic fitting. Contractures that exist preoperatively can seldom be corrected postoperatively.

In transfemoral amputees, the deforming forces are flexion and abduction. Adductor and hamstring stabilization can oppose the deforming forces. During the postoperative period, patients should avoid propping up the residual limb on a pillow and should begin active and passive motion exercises early, including lying prone to stretch the hip.

In transtibial amputees, knee flexion contractures greater than 15 degrees can cause major prosthetic problems and failure. Long leg rigid dressings, early postoperative prosthetic fitting, quadriceps-strengthening exercises, and hamstring stretching can prevent this complication. Because contractures in below-knee amputees can seldom be corrected, their prevention is paramount.

In the upper extremity amputee, shoulder and elbow flexion contractures often follow amputation, especially with short residual limbs. Efforts should be directed at prevention, with aggressive physical therapy beginning soon after surgery.

Dermatologic Problems

Good general hygiene includes keeping the residual limb and prosthetic socket clean, rinsed well to remove all soap residual, and thoroughly dry. Patients should avoid the application of foreign materials and be encouraged not to shave a residual lower limb. Shaving seems to increase the problems with ingrown hairs and folliculitis.

Reactive hyperemia is the early onset of redness and tenderness after amputation. It is usually related to pressure and resolves spontaneously.

Epidermoid cysts commonly occur at the prosthetic socket brim, especially posteriorly. These cysts are difficult to treat and commonly recur, even after excision. The best initial approach is to modify the socket and relieve pressure over the cyst. Warm heat, often with a warm tea bag; topical agents; and oral antibiotics can be required as local treatment.

Verrucous hyperplasia is a wartlike overgrowth of skin that can occur on the distal end of the residual limb. It is caused by a lack of distal contact and failure to remove normal keratin. The disorder is characterized by a thick mass of keratin, sometimes accompanied by fissuring, oozing, and infection. The infection should be addressed first, and then the limb should be soaked and treated with salicylic acid paste to soften the keratin. Topical hydrocortisone is occasionally helpful in resistant cases. Prosthetic modifications to improve distal contact must be made to prevent recurrences. Because the distal limb is often tender and prosthetic modifications are uncomfortable, an aggressive preventive approach is warranted.

Contact dermatitis sometimes occurs in amputees and can be confused with infection. The primary irritation type of dermatitis is caused by contact with acids, bases, or caustics and frequently results from failure to rinse detergents and soaps from prosthetic socks. Patients should be instructed to use mild soap and to rinse extremely well. Allergic contact dermatitis is commonly caused by the nickel and chrome in metal, antioxidants in rubber, carbon in neoprene, chromium salts used to treat leather, and unpolymerized epoxy and polyester resins in plastic laminated sockets. After infection is ruled out and contact dermatitis is confirmed, treatment begins and consists of removal of the irritant and use of soaks, corticosteroid creams, and compression with elastic wraps or shrinkers.

Superficial skin infections are common in amputees. Folliculitis occurs in hairy areas, often soon after the patient starts to wear a prosthesis. Pustules develop in the eccrine sweat glands surrounding the hair follicles, and this problem is often worse if the patient shaves. Hidradenitis, which occurs in apocrine glands in the groin and axilla, tends to be chronic and responds poorly to treatment. Socket modification to relieve any pressure in these areas can be helpful. Candidiasis and other dermatophytoses present with scaly, itchy skin, often with vesicles at the border and clearing centrally. Dermatophytoses are diagnosed with a potassium hydroxide preparation and treated with topical antifungal agents.


The ultimate function of an amputation depends on both the length of the bone and the quality of the soft-tissue envelope for the residual limb. Ilizarov techniques of distraction osteogenesis are applied to lengthen the tibia or ulna in amputees. Bone lengthening can be successful, but often issues of soft-tissue coverage remain. Although great success is described in a small series of congenital short transradial amputations, another author describes the pending necrosis of the skin over the tip of the lengthened ulna. Nonadherent, mobile soft tissue that can pad the distal end of the bone is vitally important to successful prosthetic fitting. Microsurgical techniques are also applied to use free tissue transfer to supply this type of coverage over bone in select patients, most often in trauma or tumor surgery. By using these techniques, the gracilis or latissimus dorsi muscle can be transferred to the end of the residual limb and covered with skin graft. The transposed tissues do not have sensation, and the bulk of the flap can lead to tremendous volume changes over the first 2 years. Lack of sensation and volume issues do complicate prosthetic fitting and function. These extraordinary techniques are probably best reserved for very select and unique circumstances.


For lower limb prostheses, the major advances include the development of new lightweight structural materials (see Chapter 1), the incorporation of elastic response ("energy-storing") designs, the use of computer-assisted design and computer-assisted manufacturing technology in sockets, and microprocessor control of the prosthetic knee joint. For upper limb use, new electronic technology has increased the success and durability of myoelectric prostheses. The surgeon who prescribes prosthetic limbs should have a basic understanding of the general features available to match optimally the components with the patient's specific needs.

A good prosthetic prescription specifies the socket type, suspension, shank construction, specific joints, and terminal device. The socket can be a hard socket with no or minimal interface, or it can incorporate a liner. For the transfemoral amputee, a wide variety of socket shapes are available and range from the traditional quadrilateral design to the newer narrow mediolateral design. The prosthesis is suspended from the body by straps, belts, socket contour, liners that roll on the limb and then lock to the socket, suction, friction, or physiologic muscle control.

Shank construction can be either exoskeletal or endoskeletal. The older exoskeletal type has a rigid outer shell that is hollow in the center. The endoskeletal type has a central pylon or pipe surrounded by a soft and lightweight cosmetic foam cover. In the past, exoskeletal systems were more durable; however, as materials technology has improved, so has the durability and cosmetic appearance of endoskeletal systems. The endoskeletal systems also allow more adjustment and fine tuning of alignment and are now considered structurally as durable as the older exoskeletal designs. However, the cosmetic and foam covers for the endoskeletal systems are not as durable as an exoskeletal shell. Exoskeletal systems are rarely prescribed, except for very active patients without easy access to prosthetic services or for those involved in activities that would stain, tear, or destroy the endoskeletal cover. As the public's impression of disability has evolved, many active patients now decide not to cover the prosthesis and often take pride in the high-tech look of the titanium or carbon fiber components incorporated in an endoskeletal prosthesis.

A large variety of elbow, wrist, knee, and ankle joints are now available, as well as numerous terminal devices, including hands, hooks, feet, and special adaptive equipment for sports and work. The choice of an appropriate terminal device is extremely important. For an upper extremity amputee, there is no sensation in the prosthesis, and the critical feedback of touch and proprioception is missing. Initially, a hook may be a better choice than a prosthetic hand because vision must substitute for upper extremity proprioception, and a prosthetic hand blocks vision and makes dexterous use of the terminal device difficult and clumsy. In each case, the prosthetic prescription must be individualized to ensure the most efficient system for a particular patient.

Nearly all prosthetic sockets are fabricated by forming a thermoplastic or laminate socket over a plaster mold. An exact mold of the residual limb does not make a good socket for a prosthesis. The original mold must be modified to relieve the socket over areas that cannot tolerate pressure and to indent the socket over areas that can. Test sockets of clear plastic are commonly made to visualize the blanching of the skin at troublesome areas. Automated fabrication of mobility aids (AFMA) technology uses computer-assisted design and manufacturing to aid the prosthetist by digitizing the residual limb, adding the standard modifications usually applied to a mold, and allowing additional fine manipulation of the shape on the computer screen. The computer can direct the carving of the mold or fabrication of the socket. AFMA technology can decrease the time needed for the fabrication of prostheses and increase the time available for the evaluation and training of patients. The best use of AFMA is to allow fabrication of multiple sockets for one patient during the fitting process. By using computer modifications, refinements are added in each iteration, ultimately to optimize the fit and comfort of the final socket. Before AFMA, this technique was not cost effective.

Myoelectric components are exciting but should generally not be prescribed for patients until they master traditional body-powered devices and their residual limb volume is stable. Myoelectric devices are used most successfully by patients with a midlength transradial amputation. Although a long below-elbow limb has better rotation, it is less able to contain the electronics. The need for myoelectric devices is greater in patients with a more proximal upper extremity amputation, but the weight and slow speed of the myoelectric components is a deterrent for their use. Hybrid devices utilizing body power and myoelectric components can be effective. Muscles stabilized by myodesis or myoplasty techniques seem to generate a better signal for myoelectric use.

Microprocessor control systems are applied to the knee units for transfemoral amputees. The microprocessor control alters the resistance of the knee unit to flexion or extension appropriately by sensing the position and velocity of the shank relative to the thigh. The current microprocessor-controlled knee units still do not provide power for active knee extension that would assist in rising from the sitting position or in providing power to the amputee's gait and rise up stairs. The new microprocessor-controlled so-called intelligent knee units do offer superior control when walking at varied speeds, descending ramps and stairs, and walking on uneven surfaces. Patients report improved confidence and a decrease in the tendency for the knee unit to buckle. One transfemoral amputee credits this new technology for his survival by allowing him to descend 70 stories in the World Trade Center at a normal pace during the terrorist attacks.

Attal N, Rouaud J, Brasseur L et al: Systemic lidocaine in pain due to peripheral nerve injury and predictors of response. Neurology 2004;62(2):218. [PMID: 14745057] 

Bone M, Critchley P, Buggy DJ: Gabapentin in postamputation phantom limb pain: A randomized, double-blind, placebo-controlled, cross-over study. Reg Anesth Pain Med 2002;27(5):481. [PMID: 12373695] 

Bosse MJ et al: A prospective evaluation of the clinical utility of the lower-extremity injury-severity scores. J Bone Joint Surg Am 2001;83-A(1):3. [PMID: 11205855] 

Boyko EJ et al: A prospective study of risk factors for diabetic foot ulcer. The Seattle Diabetic Foot Study. Diabetes Care 1999;22(7):1036. [PMID: 10388963] 

Brooks B et al: TBI or not TBI: That is the question. Is it better to measure toe pressure than ankle pressure in diabetic patients? Diabet Med 2001;18(7):528. [PMID: 11553180] 

Burgess EM et al: The Management of Lower-Extremity Amputations. Publication TR 10-6. U.S. Government Printing Office, 1969.

Carter SA, Tate RB: The value of toe pulse waves in determination of risks for limb amputation and death in patients with peripheral arterial disease and skin ulcers or gangrene. J Vasc Surg 2001;33(4):708. [PMID: 11296321] 

Ehde DM et al: Chronic phantom sensations, phantom pain, residual limb pain, and other regional pain after lower limb amputation. Arch Phys Med Rehabil 2000;81(8):1039. [PMID: 10943752] 

Lambert AW et al: Randomized prospective study comparing preoperative epidural and intraoperative perineural analgesia for the prevention of postoperative stump and phantom limb pain following major amputation. Reg Anesth Pain Med 2001;26(4):316. [PMID: 11464349] 

Lane JM et al: Rehabilitation for limb salvage patients: Kinesiological parameters and psychological assessment. Cancer 2001;92(Suppl 4):1013. [PMID: 11519028] 

Marks LJ, Michael JW: Science, medicine, and the future: Artificial limbs. BMJ 2001;323(7315):732. [PMID: 11576982] 

Mayfield JA et al: Trends in lower limb amputation in the Veterans Health Administration, 1989–1998. J Rehabil Res Dev 2000;37(1):23. [PMID: 10847569] 

Melzack R: Phantom limbs. Sci Am 1992;266:120. [PMID: 1566028] 

Mertens P, Lammens J: Short amputation stump lengthening with the Ilizarov method: Risks versus benefits. Acta Orthop Belg 2001;67(3):274. [PMID: 11486691] 

Misuri A et al: Predictive value of transcutaneous oximetry for selection of the amputation level. J Cardiovasc Surg (Torino) 2000;41(1):83. [PMID: 10836229] 

Nikolajsen L et al: Phantom limb pain. Curr Rev Pain 2000;4(2):166. [PMID: 10998730] 

Nikolajsen L et al: Randomized trial of epidural bupivacaine and morphine in prevention of stump and phantom pain in lower limb amputation. Lancet 1997;350:1353.

Peabody TD et al: Evaluation and staging of musculoskeletal neoplasms. J Bone Joint Surg Am 1998;80(8):1204. [PMID: 9730132] 

Reiber GE, Ledoux W: Epidemiology of foot ulcers and amputations in the diabetic foot: Evidence for prevention. In Williams R et al (eds): The Evidence Base for Diabetes Care. Wiley, 2002.

Siddle L: The challenge and management of phantom limb pain after amputation. Br J Nurs 2004;13(11):664. [PMID: 15218432] 

Smith DG, Burgess EM: The use of CAD/CAM technology in prosthetics and orthotics—Current clinical models and a view to the future. J Rehabil Res Dev 2001;38(3):327. [PMID: 11440264] 

Smith DG, McFarland LV, Sangeorzan BJ et al: Postoperative dressing and management strategies for transtibial amputations: A critical review. J Rehabil Res Dev 2003;40(3):213. Review. [PMID: 14582525] 

Soucacos PN: Indications and selection for digital amputation and replantation. J Hand Surg [Br] 2001;26(6):572. Review. [PMID: 11884116] 

Stojadinovic A et al: Amputation for recurrent soft tissue sarcoma of the extremity: Indications and outcome. Ann Surg Oncol 2001;8(6):509. [PMID: 11456050] 

Waters RL et al: The energy cost of walking of amputees: Influence of level of amputation. J Bone Joint Surg AM 1976;58:42. [PMID: 1249111] 


Hand Amputation

Although microsurgical replantation techniques have reduced the incidence of hand amputations, for many patients replantation is still not feasible or results in failure. There is considerable controversy about the best treatment for any given hand injury, and the optimal treatment takes into consideration the injured patient's occupation, hobbies, skills, and hand of dominance. The hand is a highly visible and important part of body image. Many patients with partial hand amputations can benefit tremendously from using a cosmetic partial hand prosthesis.


Fingertip injuries occur frequently, and fingertip amputation is the most common type of amputation. The treatment of choice usually depends on the geometry of the defect and whether or not bone is exposed. Although a large variety of local flap procedures are used to cover defects of different shapes and sizes, there is also a growing understanding that allowing secondary healing of fingertip injuries is the treatment least prone to complications in adults as well as in children. Even if bone is exposed, simply rongeuring back the exposed bone proximal to the soft-tissue defect and allowing secondary healing can give excellent results. The amount of the bone that can be removed is limited because at least a third of the distal phalanx must be left intact to prevent a hook deformity of the nail.

Two problems frequently result from fingertip amputations: cold intolerance and hypersensitivity. Overall, regardless of which treatment is chosen, approximately 30–50% of patients experience these problems. One criticism of the many local flap procedures used to obtain coverage and primary wound healing is that all of them involve incising and advancing uninjured tissue, which extends the area of scarring and damages the fine branches of the digital nerves. Newer studies suggest that the incidence of cold intolerance and hypersensitivity may be lower with secondary healing than with skin grafts or local flaps.


The thumb, with its unique range of motion, plays the major role in all three prehensile activities of the hand: palmar grip, side-to-side pinch, and tip-to-tip pinch. Amputation of the thumb can result in the loss of virtually all hand function. Thumb amputations can involve (1) the distal third of the thumb (ie, distal to the interphalangeal joint), (2) the middle third of the thumb (ie, from the metacarpophalangeal joint to the interphalangeal joint), or (3) the proximal third of the thumb.

Thumb amputation of the distal third allows the patient to retain a tremendous amount of thumb function. Cold intolerance and hypersensitivity are frequent problems, as noted in the previous discussion of fingertip amputations. Treatment of distal third injuries should allow secondary healing of the thumb or should use relatively uncomplicated techniques for coverage.

Thumb amputation in the middle third is more complicated. The issues here are length, stability, and sensate skin coverage. More aggressive procedures may well be warranted and may consist of cross-finger flaps, volar advancement flaps, neurovascular island flaps from the dorsal index finger (radial nerve) or volar middle finger (median nerve), bone lengthening, or web space deepening.

Thumb amputation in the proximal third has a devastating impact on hand function. Local reconstruction for this degree of loss is not generally successful. Pollicization of another digit, a toe-to-hand transfer, or other complicated surgical techniques may be indicated to restore function.


Isolated amputation of a lesser digit can cause a variety of functional and cosmetic problems. Digit amputations distal to the insertion of the sublimis flexor tendon retain active flexor tendon activity and maintain useful metacarpophalangeal joint flexion. The long flexor tendon should not be sewn to the extensor tendon because it limits the excursion of both tendons and definitely limits the function of the remaining digits.

Amputations proximal to the sublimis tendon insertion retain approximately 45 degrees of proximal phalanx flexion at the metacarpophalangeal joint through the action of the intrinsic muscles. This is usually enough to keep small objects from falling through the defect and to allow the residual finger to participate to some degree in grip. If the patient uses a cosmetic finger prosthesis and wears a ring to cover the proximal edge of the prosthesis, the amputation is almost unnoticeable.

The index finger participates principally in side-to-side and tip-to-tip pinch with the thumb. After an amputation of the index finger at the metacarpophalangeal joint, the middle finger assumes this important role. The residual second metacarpal can interfere with side-to-side pinch between the thumb and the middle finger, however. Converting this amputation to a ray amputation often can improve function and cosmesis, but the drawback is that it also narrows the width of the palm and can decrease grip and torque strength significantly. Surgical decisions must be individualized, but the second metacarpal should probably be retained if the patient uses hand tools extensively, as does a carpenter or machinist.

Amputation of the middle or ring finger at the metacarpophalangeal joint can make it difficult for the patient to hold small objects because they tend to fall through the defect. Full ray resection can narrow the central defect and occasionally improve function, but narrowing the palm can decrease grip and torque strength.

Amputation of the small finger at the metacarpophalangeal joint is often cosmetically unacceptable because of the abrupt and noticeable change in contour of the hand. Although converting a fifth digital amputation to a ray amputation by including the metacarpal can improve cosmesis, it also narrows the width of the palm and can decrease grip and torque strength. Surgical decisions must be based on individual factors and concerns.


Amputations through the carpus are generally discouraged. Most surgeons believe the result to have no real advantages over a wrist disarticulation or transradial amputation. There are isolated reports of patients valuing the little bit of wrist flexion and extension that allows them to hold objects against their body and to stabilize objects for two-handed grasp. The flexor and extensor carpi radialis and ulnaris tendons must be reattached to provide this limited motion. The prosthetic options are less standard and generally considered to be less functional than the traditional transradial designs.

Carpus amputations should probably be considered in bilateral cases. Although rare, more patients sustaining tissue loss from ischemia are seen in the intensive care unit after prolonged resuscitations and the use of vasopressors. Without the vasopressors, these patients would die. Unfortunately, part of the body's response to these lifesaving medications can be to shunt blood flow from the distal extremities, resulting in demarcation and dry gangrene in the hands and feet. Just as in frostbite, if infection is not present, it is worthwhile to delay any surgical intervention and allow adequate time for tissue demarcation and recovery. Partial hand amputation is occasionally necessary, and if required, the carpus level should be considered.

Wrist Disarticulation

Wrist disarticulation continues to be controversial. Proponents frequently argue that it has two advantages over the shorter transradial amputation: It retains the distal radioulnar joint, which preserves more forearm rotation, and it retains the distal radial flare, which dramatically improves prosthetic suspension. Volar and dorsal fish-mouth incisions are usually best, and removal of the radial and ulnar styloids can prevent painful pressure points. Tenodesis of the major forearm motors stabilizes the muscle units and thereby improves physiologic and myoelectric performance.

Opponents of wrist disarticulation argue that prosthetic substitution after this procedure is slightly more complicated than it is after a standard transradial amputation. The prosthetic socket is more difficult to fabricate because of the bone contours. Conventional wrist units add too much length to the prosthetic arm after wrist disarticulation and therefore cannot be used. The terminal device for a wrist disarticulation also needs to be modified because of length. Myoelectric prostheses are difficult to fit because there is less space to conceal the electronics and power supply.

In spite of these prosthetic concerns, wrist disarticulation patients are often excellent upper extremity prosthetic users. Some patients with an unsatisfactory hand can gain improved function by undergoing a wrist disarticulation and using a standard prosthesis. This decision must be individualized and based on contributory factors such as severity of tissue loss, pain, functional requirements, and the patient's body image.

Transradial Amputation

The transradial amputation is extremely functional, and successful prosthetic rehabilitation and sustained use are achieved in 70–80% of patients who undergo amputation at this level. Forearm rotation and strength are proportional to the length retained. Surgical incisions are best with equal volar and dorsal flaps. A myodesis should be performed to prevent a painful bursa, facilitate physiologic muscular suspension, and allow for myoelectric prosthetic use. An extremely short transradial residual limb requires the use of a Muenster-type socket, which molds up around the humeral condyles for added suspension. Occasionally, side hinges and a humeral cuff are required to achieve suspension of the prosthesis. Both of these types of suspension preserve elbow flexion and extension but limit rotation.

The value of preserving the elbow joint cannot be overemphasized. Skin grafts and even composite grafts should be considered to retain the tremendous functional benefit of an elbow with some active motion. Even a limited range of elbow motion can be useful, and an ingeniously designed, geared step-up elbow hinge can convert a limited active range of elbow motion to an improved prosthetic ROM. Although body-powered prostheses are extremely functional at the transradial level of amputation, this level is also the most successful level at which to use myoelectric devices.

Krukenberg Amputation

The Krukenberg kineplastic operation transforms the transradial amputation stump into radial and ulnar digits that are capable of strong prehension and have excellent manipulative ability because of retained sensation on the "fingers" of the forearm. The operation should not be performed as a primary amputation.

The Krukenberg amputation can be performed as a secondary procedure in a transradial amputee who has a residual limb of at least 10 cm from the tip of the olecranon, an elbow flexion contracture of less than 70 degrees, and good psychological preparation and acceptance. In this case, the amputee can become completely independent in daily activities because of the retained sensory ability of the pincers as well as the quality of the grasping mechanism (Figure 12–6). The Krukenberg amputation traditionally was indicated for blind patients with bilateral below-elbow amputations, but it also may be indicated at least unilaterally in bilateral below-elbow amputees who are able to see and in those who have limited access to prosthetic facilities.

Figure 12–6.


A patient with bilateral Krukenberg hands demonstrates bimanual dexterity in sharpening a pencil.

(Reproduced, with permission, from Garst RJ: The Krukenberg hand. J Bone Joint Surg Br 1991;73:385.)

A conventional prosthesis can be worn over the Krukenberg forearm, and myoelectric devices can be adapted to use the forearm motion. The major disadvantage is the appearance of the arm, which many people consider grotesque and do not accept. As society continues to become more understanding and accepting of disabled individuals, concerns about appearance may diminish. Intensive preoperative preparation and counseling are mandatory.

Elbow Disarticulation

Elbow disarticulation can be a satisfactory amputation level and has the advantage of retaining the condylar flare to improve prosthetic suspension and allow for the transfer of humeral rotation to the prosthesis. The longer lever arm improves strength. The disadvantage is in the design of the prosthetic elbow hinge. An outside hinge is bulky and hard on clothing, whereas the conventional elbow unit provides a disproportionately long upper arm and short forearm. Whether the advantages of the elbow disarticulation outweigh the disadvantages remains controversial. Surgically, volar and dorsal flaps work best, and myodesis of the biceps and triceps tendons are needed to preserve the distal muscle attachments.

Transhumeral Amputation

When transhumeral amputation is performed, efforts should be made to retain as much as possible of the bone length that has suitable soft-tissue coverage. Even if only the humeral head remains and no functional length is salvageable, an improved shoulder contour and cosmetic appearance results. Myodesis helps preserve biceps and triceps strength, prosthetic control, and myoelectric signals. In most cases of transhumeral amputation, an immediate postoperative prosthesis and rigid dressings can be used successfully. Physical therapy should focus on proximal joint and muscle function. Because the terminal prosthetic device is usually controlled by active shoulder girdle motion, early prosthetic use and therapy can prevent contracture and maintain strength.

Prosthetic suspension traditionally was incorporated in the body-powered harness, which can be somewhat uncomfortable. Among the alternative techniques are humeral angulation osteotomy (rarely used), socket-suction suspension, and the newer elastomeric roll on locking liners. Many prosthetic options are available for the transhumeral amputee. One option is a prosthesis that is totally body powered. Another is a hybrid prosthesis that uses myoelectric control of one component (either the terminal device or the elbow device) and body-powered control of the other. The transhumeral prosthesis is heavy, often considered slow, and requires much mental concentration to use effectively. These issues lead many unilateral transhumeral amputees to choose not to wear a prosthesis at all or to wear only a lightweight cosmetic prosthesis for special occasions.

Transhumeral amputation is sometimes elected to manage a dysfunctional arm following a severe brachial plexus injury. The advantages of amputation are that it unloads the weight from the shoulder and scapulothoracic joints and eliminates the problem of having a paralyzed arm that gets in the way and hinders body function. The decision to undertake shoulder arthrodesis in combination with transhumeral amputation is controversial and should be made on an individualized basis. Investigators who compared two groups of patients with transhumeral amputation because of brachial plexus injury—one group without shoulder arthrodesis and one group with it—found a somewhat better return-to-work rate in the group without shoulder arthrodesis. Prosthetic expectations in these patients should be limited because prosthetic fitting adds weight to a dysfunctional shoulder girdle, often defeating one of the original goals of the amputation.

Shoulder Disarticulation & Scapulothoracic (Forequarter) Amputation

The performance of shoulder disarticulation (Figure 12–7) or scapulothoracic amputation (Figure 12–8) is rare. When either operation is performed, it is usually in cases of cancer or severe trauma. Either operation results in a loss of the normal shoulder contour and causes the patient difficulty because clothing does not fit well. Saving the humeral head, if possible, can improve the contour of a shoulder disarticulation tremendously. The scapulothoracic amputation, usually performed for proximal tumors, removes the arm, scapula, and clavicle. Dissection often extends into the neck and into the thorax.

Figure 12–7.


Shoulder disarticulation.


Figure 12–8.


Forequarter amputation.

Elaborate myoelectric prostheses are available for patients but are expensive, heavy, and require intensive maintenance. Body-powered prostheses are also heavy, hard to suspend comfortably, and difficult to use. Most patients request prosthetic help for improved cosmesis and fitting of clothes. Often a simple soft mold to fill out the shoulder meets these expectations and is an alternative to a full-arm cosmetic prosthesis.

Postural Abnormalities after High Upper Extremity Amputation

Normally, the weight of the arm and the muscle activity associated with shoulder and arm function keep the shoulders appropriately level. Unilateral hypertrophy of an upper limb, including the shoulder girdle, occurs in certain occupations and is also seen in some sports. Some people are born with a degree of asymmetry of their shoulders, which is a relatively minor postural abnormality and does not require special clothing.

When the arm is removed and the clavicle and scapula remain, the muscles elevating the shoulder girdle are unopposed by both the weight of the arm and those muscles that pass across the shoulder and tend to depress the shoulder and arm. The consequence of this imbalance is an upward elevation described as "hiking" of the shoulder girdle. This high shoulder tends to accentuate the cosmetic loss, even when the individual is wearing a cosmetic shoulder filler or a cosmetic limb. Abnormal shoulder elevation can be countered by corrective exercises beginning as soon as they can be tolerated after the amputation. The wearing of a prosthesis with its dependent weight also diminishes shoulder hike. In most circumstances, the shoulder girdle elevation is inevitable; however, its degree can be minimized by appropriate physical measures.

Removal of the entire upper limb in the growing skeleton can result in a scoliosis of the spine. Muscular imbalance is considered to be the cause of the deformity. It may be seen to a slighter degree in the adult but is primarily confined to the growing skeleton. The combined postural deformity of upper dorsal spine scoliosis and elevation of the shoulder girdle produces asymmetry of the head and neck on the trunk, with the head appearing to be placed asymmetrically as the person stands.

In general, no corrective splinting or orthotic device can successfully counteract the postural changes associated with shoulder-level amputation. Neck and shoulder-girdle exercises offer the most effective prophylaxis and treatment. The postural deficits are particularly evident with forequarter amputation. Soft, light polyurethane cosmetic restoration, either as part of a cosmetic prosthesis or separately used with the empty sleeve, counters to some degree the unsightly upper body contour.

Hand Transplantation

Hand transplantation and the suppression of rejections is now technically possible. Approximately 25 documented cases of hand transplantation have been performed with varying degrees of success. The potential benefits for the amputee are certainly many, but they must be balanced against the real risks. In general, skin, muscle, and bone marrow appear to reject earlier and more aggressively than bone, cartilage, or tendon. Preventing this rejection is an ongoing and lasting issue, with real consequences for the individual's health and life expectancy. The current immunosuppressive drugs needed to prevent rejection of a composite hand transplant include toxic side effects, opportunistic infections, and increase in malignancies.

Also, the real psychological impact following hand and other organ transplantation should not be underestimated. One study examining the issues 5 years following heart transplant showed a significant increase in emotional issues such as irritability, depression, and low self-esteem. Even for a patient with no preexisting psychological issues, living with a hand transplantation, which remains constantly in view, may not be easy.

Baumeister S, Kleist C, Dohler B et al: Risks of allogeneic hand transplantation. Microsurgery 2004;24(2):98. Review. [PMID: 15038013] 

Crandall RC, Tomhave W: Pediatric unilateral below-elbow amputees: Retrospective analysis of 34 patients given multiple prosthetic options. J Pediatr Orthop 2002;22(3):380. [PMID: 11961460] 

Goel A et al: Replantation and amputation of digits: User analysis. Am J Phys Med Rehabil 1995;74(2):134. [PMID: 7710728] 

Hatrick NC, Tonkin MA: Hand transplantation: A current perspective. ANZ J Surg 2001;71(4):245. [PMID: 11354126] 

Martin C, Gonzalez del Pino J: Controversies in the treatment of fingertip amputations. Conservative versus surgical reconstruction. Clin Orthop 1998;(353):63. [PMID: 9728160] 

Neusel E et al: Results of humeral stump angulation osteotomy. Arch Orthop Trauma Surg 1997;116:263. [PMID: 9177800] 

Peimer CA et al: Hand function following single ray amputation. J Hand Surg [Am] 1999;24(6):1245. [PMID: 10584948] 

Schatz RL, Rosenwasser MP: Krukenberg kineplasty: A case study. J Hand Ther 2002;15(3):260. [PMID: 12206329] 

Waikakul S, Sakkarnkosol S, Vanadurongwan V et al: Results of 1018 digital replantations in 552 patients. Injury 2000;31(1):33. [PMID: 10716048] 

Wilkinson MC et al: Brachial plexus injury: When to amputate? Injury 1993;24(9):603. [PMID: 8288380] 


Foot Amputation


Toe amputations can be performed with side-to-side or plantar-to-dorsal flaps to use the best available soft tissue. The bone should be shortened to a level that allows adequate soft-tissue closure without tension.

In great toe amputations, if the entire proximal phalanx is removed, the sesamoids can retract and expose the keel-shaped plantar surface of the first metatarsal to weight bearing. This often leads to high local pressure, callous formation, or ulceration. The sesamoids can be stabilized in position for weight bearing by leaving the base of the proximal phalanx intact or by performing tenodesis of the flexor hallucis brevis tendon.

An isolated amputation of the second toe commonly results in severe hallux valgus deformity of first toe (Figure 12–9). This situation may be prevented by amputation of the second ray or by fusion of the first metatarsal and phalanx. In the shorter toe amputations at the metatarsophalangeal joint level, transferring the extensor tendon to the capsule may help elevate the metatarsal head and maintain an even distribution for weight bearing. Prosthetic replacement is not required after toe amputations.

Figure 12–9.


Severe hallux valgus deformity occurring after isolated second toe amputation.


A ray amputation removes the toe and all or some of the corresponding metatarsal. Isolated ray amputations can be durable. Multiple ray amputations, however, especially in patients with vascular disease, can narrow the foot excessively. This increases the amount of weight that must be borne by the remaining metatarsal heads and can lead to new areas of increased pressure, callous formation, and ulceration. Surgically, it is often difficult to achieve primary closure of ray amputation wounds because more skin is usually required than is readily apparent. Instead of closing these wounds under tension, it is usually advisable to leave them open and allow for secondary healing.

The fifth ray amputation is the most useful of all the ray amputations. Plantar and lateral ulcers around the fifth metatarsal head often lead to exposed bone and osteomyelitis. A fifth ray amputation allows the entire ulcer to be excised and the wound to be closed primarily (Figure 12–10). In general, for more extensive involvement of the foot, a transverse amputation at the transmetatarsal level is more durable. Prosthetic requirements after ray amputations include extra-depth shoes with custom-molded insoles. The insole should include a metatarsal pad that loads the shafts of the metatarsal and unloads some of the pressure at the metatarsal heads.

Figure 12–10.


Fifth ray amputation for fifth metatarsal head ulcer.


The transmetatarsal and Lisfranc amputations are reliable and durable. The Lisfranc amputation is actually a disarticulation just proximal to the metatarsals where the cuneiform and cuboid bones are retained. Surgically, a healthy, durable soft-tissue envelope is more important than a specific anatomic level of amputation, so the length of bone to be removed should be based on the ability to perform soft-tissue closure without tension. A long plantar flap is preferable, but equal dorsal and plantar flaps work well, especially for transmetatarsal amputation in the treatment of metatarsal head ulcers (Figure 12–11).

Figure 12–11.


Transmetatarsal amputation with Achilles tendon lengthening.

Muscle balance around the foot should be carefully evaluated preoperatively, with specific attention to tightness of the heel cord and strength of the anterior tibial, posterior tibial, and peroneal muscles. Midfoot amputations significantly shorten the lever arm of the foot, so lengthening of the Achilles tendon should be done if necessary. Tibial or peroneal muscle insertions should be reattached if they are released during bone resection. For example, if the base of the fifth metatarsal is resected, the peroneus brevis tendon should be reinserted into the cuboid bone. In patients with vascular disease, this can be performed with a minimal amount of dissection to prevent further compromise of the tissues.

Postoperative casting prevents deformities, controls edema, and speeds rehabilitation. Prosthetic requirements can vary widely. During the first year following amputation, many patients benefit from the use of an ankle-foot orthosis (AFO) with a long footplate and a toe filler. To prevent an equinus deformity from developing, patients should be advised to wear the orthosis except when taking a bath or shower. Later, the use of a simple toe filler combined with a stiff-soled shoe may be adequate. Cosmetic partial foot prostheses are also available.


A Chopart amputation removes the forefoot and midfoot and saves only the talus and calcaneus. Rebalancing procedures are required to prevent equinus and varus deformities. Achilles tenotomy, transfer of the anterior tibial or extensor digitorum tendons, and postoperative casting are all usually necessary. Although tendon transfer to the talus was previously recommended, transfer to the calcaneus is now done to minimize varus positioning. Beveling the inferior, anterior surface of the calcaneus can remove a potential bone pressure point.

Two other types of hindfoot amputations are the Boyd and the Pirogoff amputations. The Boyd procedure consists of a talectomy and calcaneal-tibial arthrodesis after forward translation of the calcaneus. The Pirogoff procedure consists of a talectomy with calcaneal-tibial arthrodesis after the vertical transection of the calcaneus through the midbody and a forward rotation of the posterior process of the calcaneus under the tibia. These two types of hindfoot amputations are done mostly in children to preserve length and growth centers, prevent heel pad migration, and improve socket suspension.

Studies in which various procedures in children are compared showed that a hindfoot amputation results in better function than a Syme amputation (see section on Syme amputation) in cases in which the hindfoot is balanced and no equinus deformity has developed.

The hindfoot prosthesis requires more secure stabilization than a midfoot prosthesis to keep the heel from pistoning during gait. An anterior shell can be added to an ankle-foot prosthesis, or a posterior opening socket prosthesis can be used.


Partial calcanectomy, which consists of excising the posterior process of the calcaneus (Figure 12–12), should be considered an amputation of the back of the foot. In selected patients with large heel ulcerations or calcaneal osteomyelitis, partial calcanectomy can be a functional alternative to transtibial amputation. The removal of the entire posterior process of the calcaneus allows for fairly large soft-tissue defects to be closed primarily. Patients must have adequate vascular perfusion and nutritional competence for wound healing to occur. As with other amputations, partial calcanectomy creates a functional and cosmetic deformity. Use of an ankle-foot prosthesis with a cushion heel is usually required to replace the missing heel and prevent further skin ulceration.

Figure 12–12.


Partial calcanectomy.

Syme Amputation

In the Syme amputation, the surgeon removes the calcaneus and talus while carefully dissecting on bone to preserve the heel skin and fat pad to cover the distal tibia (Figure 12–13). The surgeon must also remove and contour the malleoli, but whether this should be done during the initial operation or 6–8 weeks later remains controversial. Proponents of the two-stage procedure argue that it can improve healing in patients with vascular disease. Opponents point out that it delays rehabilitation because the patient cannot bear weight until after the second stage of the operation. One series supports the use of the one-stage procedure, even in the presence of vascular disease or diabetes. A late complication of the Syme amputation is the posterior and medial migration of the fat pad. One of these surgical procedures can be done to stabilize the fat pad: tenodesis of the Achilles tendon to the posterior margin of the tibia through drill holes; transfer of the anterior tibial and extensor digitorum tendons to the anterior aspect of the fat pad; or removal of the cartilage and subchondral bone to allow scarring of the fat pad to bone, with or without pin fixation. Careful postoperative casting can also help keep the fat pad centered under the tibia during healing. The Syme amputation is one of the most difficult amputations to perform in terms of surgical technique and achievement of primary healing and heel pad stability.

Figure 12–13.


Syme amputation with tenodesis of the Achilles tendon to the distal tibia.

Syme amputation should be designed to allow end bearing. Retaining the smooth, broad surface of the distal tibia and the heel pad allows direct transfer of weight from the end of the residual limb to the prosthesis. A transtibial or transfemoral amputation does not allow this direct transfer of weight. Because of the ability to end-bear, the amputee can occasionally ambulate without a prosthesis in emergency situations or for bathroom activities.

The Syme prosthesis is wider at the ankle level than is a transtibial prosthesis, and this cosmetic problem can be bothersome to some patients. The surgical narrowing of the malleolar flare and the use of new materials in the prosthesis, however, can improve the appearance of the final prosthesis. Moreover, patients can now benefit from energy-storing technology provided by the newly designed lower profile elastic response feet. Sockets do not need the high contour of a patellar-tendon bearing design because of the end-bearing quality of the residual limb. The socket can be windowed either posteriorly or medially if the limb is bulbous, or a flexible socket within a rigid frame design can be used if the limb is less bulbous. Because of the tibial flare, the socket used following Syme amputation is usually self-suspending.

Transtibial Amputation

Transtibial amputation is the most commonly performed major limb amputation. The long posterior flap technique (Figure 12–14) is now standard, and good results can be expected even in the majority of patients with vascular disease. Anterior and posterior flaps, sagittal flaps, and skewed flaps can be helpful in specific patients.

Figure 12–14.


Transtibial amputation with long posterior flap technique.


Efforts should be made to preserve as much bone length as possible between the tibial tubercle and the junction of the middle and distal thirds of the tibia, based on the available healthy soft tissues. Amputations in the distal third of the tibia should be avoided because they result in poor soft-tissue padding and are more difficult to fit comfortably with a prosthesis. The goal is a cylindrically shaped residual limb with muscle stabilization, distal tibial padding, and a nontender and nonadherent scar (Figure 12–15). The transtibial amputation is especially well suited to rigid dressings and immediate postoperative prosthetic management.

Figure 12–15.


Bilateral transtibial amputations that emphasize the benefits of the long posterior flap technique. The right limb, amputated by using equal anterior and posterior flaps, is conically shaped and atrophic. The left limb, amputated by using the long posterior flap technique, is cylindrical and well padded.

(Reproduced, with permission, from Smith DG, Burgess EM, Zettl JH: Fitting and Training the Bilateral Lower-Limb Amputee, in Bowker JH, Michael JW (eds): Atlas of Limb Prosthetics Surgical, Prosthetic, and Rehabilitation Principles. Rosemont, IL, American Academy of Orthopaedic Surgeons, 2002, pp 599–622.)

Distal tibiofibular synostosis (Ertl procedure) should be considered for the treatment of a wide trauma-induced diastasis to improve stabilization of the bone and soft tissue. The procedure is less often indicated in the treatment of patients with vascular disease. The synostosis is developed to create a broad bone mass terminally to improve the distal end-bearing property of the limb and minimize motion between the tibia and fibula. Although there is renewed interest in these techniques, true comparison of patients with osteomyoplastic techniques versus standard techniques has not been done.

A wide variety of prosthetic designs are available for the transtibial amputee. Sockets can be designed to incorporate a liner, which offers the advantages of increased comfort and accommodation of minor changes in residual limb volume. Disadvantages include increased perspiration and a less sanitary, less comfortable feeling in hot humid weather. Hard sockets are designed to have cotton or wool stump socks of an appropriate ply or thickness as the interface between the leg and the socket. Hard sockets are easier to clean and more durable than the liners are.

The Icelandic-Swedish-New York (ISNY) socket refers to the use of a more flexible socket material that is supported by a rigid frame. The flexible socket changes shape to accommodate underlying muscle contraction. This socket style can also be useful for limbs that are scarred or difficult to fit. Open-ended sockets with side joints and a thigh corset are not used much today except by patients who wore them successfully in the past and by patients with limited access to prosthetic care. The patellar tendon-bearing shape is most commonly used for the transtibial amputee. In spite of its name, the majority of the weight is borne on the medial tibial flare and laterally on the interosseus space, whereas the rest of the weight is borne on the patellar tendon area. Even the new so-called total-contact transtibial socket, which is designed to have increased contact on all areas of the residual limb, preferentially loads the tibial flare and patellar tendon regions.

Numerous types of suspension devices are available for the transtibial prosthesis. The simplest and most common is a suprapatellar strap, which wraps above the femoral condyles and patella. Sockets can be designed to incorporate a supracondylar mold or wedge to grip above the femoral condyles, but this higher profile is bulkier and less cosmetic when the patient is sitting. A waist belt and fork strap are helpful for the patient who has a very short transtibial residual limb because these devices decrease pistoning in the socket; they are also helpful for the patient whose activities require extremely secure suspension. If the patient has a limb with poor soft tissue or has intrinsic knee pain, side hinges and a thigh corset can help unload the lower leg and transfer some of the weight to the thigh.

External suspension sleeves made of latex or neoprene are still used quite frequently. Latex is more cosmetic but less durable and can be constricting. Neoprene is more durable and not as constricting but sometimes causes contact dermatitis. The newest suspension uses an elastomeric or silicone-based liner that is rolled on over the residual leg and offers an intimate friction fit. A small metal post on the distal end of the liner then locks into a catch in the prosthetic socket to suspend the socket securely to the liner. Many patients who use these elastomeric locking liners like the secure suspension and feeling of improved control of the prosthesis. The liners have the disadvantages of being less durable and requiring frequent replacement. These elastomeric locking liners can be expensive. Although elastomeric locking liners were originally touted as preventing skin problems; rashes, skin irritation, and skin breakdown remain a frequent complaint even with this new technology, however. Approximately a third of amputees cannot tolerate the forces generated at the distal part of the amputation with liners using the metal post or pin lock system. New techniques were designed to attach the elastomeric liner to the socket with vacuum pumps, clips on the side of the liner, or sealing liners and one-way socket valves to maintain a suction between the liner and the socket. Suspension must be individualized, and no system is yet proven acceptable to all amputees.

Many different designs for prosthetic feet are now available, ranging from the original solid ankle cushion heel (SACH) foot to the newer elastic response technology with a variety of keel, ankle, and pylon designs. Cost and function can vary widely, and care should be used in prescribing an appropriate prosthetic foot for an individual patient. A common error is to prescribe a foot that is either too stiff or does not get to feel flat quickly enough for an individual patient, especially in the first 12–18 months after an amputation.

Knee Disarticulation

Disarticulation through the knee joint (Figure 12–16) is indicated in ambulatory patients when a below-knee amputation is not possible but suitable soft tissue is present for a knee disarticulation. These circumstances are most commonly found in cases involving traumatic injuries. In patients with vascular disease, the blood supply is such that if a knee disarticulation would heal, a short transtibial amputation would usually heal as well. The knee disarticulation is indicated in patients who have vascular problems and are nonambulatory, especially if knee flexion contractures or spasticity are present. Although sagittal flaps or the traditional long posterior flap can be used to take advantage of the best available soft-tissue coverage, newer literature supports use of the posterior flap technique when possible. The patella is retained and the patellar tendon sutured to the cruciate stumps to stabilize the quadriceps muscle complex. The biceps tendons can also be stabilized to the patellar tendon. A short section of gastrocnemius muscle can be sutured to the anterior capsule to pad the distal end. Although many techniques are described to trim the condyles of the femur, trimming is rarely necessary, and radical trimming can decrease some of the advantages of the knee disarticulation.

Figure 12–16.


Knee disarticulation.

For ambulatory patients, the advantages of a knee disarticulation over a transfemoral amputation include improved socket suspension by contouring above the femoral condyles, the added strength of a longer lever arm, the retained muscle balance of the thigh, and, most important, the end-bearing potential to transfer weight directly to the prosthesis. In the past, the objections to a bulky prosthesis and asymmetric knee-joint level led many surgeons to abandon the practice of performing knee disarticulations. New materials allow a less bulky prosthesis to be fabricated, and the four-bar linkage knee unit, which can fold under the socket, improves the appearance of the prosthesis when the patient is sitting. The four-bar linkage knee remains the prosthetic knee of choice for a knee disarticulation. It is low profile, has excellent stability, and can incorporate a hydraulic unit for control during the swing phase of gait in patients who can walk at different cadences.

For nonambulatory patients, a knee disarticulation eliminates the problem of knee flexion contractures, provides a balanced thigh to decrease hip contractures, and provides a long lever arm for good sitting support and transfers.

The Gritti-Stokes amputation is not recommended. In this operation, the patella is advanced distally and fused by arthrodesis to the distal femur, theoretically to allow direct weight bearing. The concept behind this operation is flawed because even in normal kneeling, the weight is borne on the pretibial and patellar tendon areas and not on the patella. The added length and the asymmetry of the knee joints complicate prosthetic fitting.

Transcondylar amputation can be performed, but the end-bearing comfort and improved suspension of a transcondylar amputation appear to be diminished when compared with the true knee disarticulation.

Transfemoral Amputation

Transfemoral amputation is usually performed with equal anterior and posterior fish-mouth flaps. Atypical flaps can and should be used to save all possible femoral length in cases of trauma because the amount of function is directly proportional to the length of the residual limb.

Muscle stabilization is more important in the transfemoral amputation than in any other major limb amputation. The major deforming force is into abduction and flexion. Myodesis of the adductor muscles through drill holes in the femur can counteract the abductors, prevent a difficult adductor tissue roll in the groin, and improve prosthetic control (Figure 12–17). Without muscle stabilization, the femur commonly migrates laterally through the soft-tissue envelope to a subcutaneous location. Newer transfemoral socket designs attempt to better control the position of the femur, but they are not as effective as muscle stabilization. Even in nonambulatory patients, muscle stabilization is helpful in creating a more durable, padded residual limb by preventing migration of the femur.

Figure 12–17.


Transfemoral amputation with adductor myodesis.


An IPOP and rigid dressings are more difficult to apply and keep positioned after a transfemoral amputation than after more distal amputations. IPOP techniques do offer the advantages of early rehabilitation and control of edema and pain, and these techniques are preferred if the expertise to use them is available. The major complaints of patients with the transfemoral IPOP are the weight of the cast and the discomfort when sitting. In many cases, only a soft compressive dressing alone is used, and in these patients, the dressing should be carried proximally around the waist as a spica to better suspend the dressing and to include the medial thigh and prevent the development of an adductor roll of tissue. Proper postoperative positioning and therapy are essential to prevent hip flexion contractures. The limb should be positioned flat on the bed, rather than elevated on a pillow, and hip extension exercises and prone positioning should be started early.

Suspension of the prosthesis is more complicated in transfemoral amputations than in more distal amputations because of the short residual limb, the lack of bony contours, and the increased weight of the prosthesis. The transfemoral amputation prosthesis can be suspended by suction, Silesian bandage, hip-joint and pelvic band, or by the newer elastomeric locking liners.

Traditional socket-suction suspension works when the skin forms an airtight seal against the socket. Air is forced distally through a small one-way valve when the prosthesis is donned and with each step during gait, thus maintaining negative pressure distally in the socket. No prosthetic sock or other liner is used between the hard socket and the limb because air leaks out around the sock and prevents suction from developing. Donning a socket-suction prosthesis requires skill and exertion, and patients must have good coordination, upper extremity function, and balance to perform this task. Socket-suction systems work well for average-to-long transfemoral residual limbs that have adequate soft tissues and stable shape and volume. It is usually comfortable and the most cosmetically acceptable method of socket suspension.

A Silesian bandage is a flexible strap that attaches laterally to the prosthesis, wraps back around the waist and over the contralateral iliac crest, and then comes forward to attach to the anterior proximal socket (Figure 12–18). It provides good suspension and added rotational control of the prosthesis. A Silesian bandage is commonly used to augment suction suspension for patients who have shorter-length limbs or for patients whose activities require more secure suspension than suction alone can offer.

Figure 12–18.


Silesian band suspension of a transfemoral prosthesis.

As with the transtibial prosthesis, the newer elastomeric locking liners can provide excellent suspension and control. An elastomeric or silicone-based liner is rolled onto the leg similar to the way a condom is applied. This liner has an intimate fit with the residual limb and avoids pistoning and rotational forces. A small metal post at the distal end of the liner locks down into a catch at the bottom of the prosthetic socket to create a secure mechanical suspension. A small button must be pushed to disengage the lock and release the prosthesis. Many amputees express an improved sense of security and improved proprioception with these systems. The disadvantages continue to be the added cost, the need to replace the liners as they tear, and, rarely, developing a contact dermatitis. As discussed with transtibial amputees, approximately a third of amputees cannot tolerate the forces generated at the distal part of the amputation with liners using the metal post or pin lock system. For these patients, new methods to attach the liner to the socket must be explored.

The hip joint and pelvic band provides extremely secure suspension and control, but the band is bulky, the least cosmetically acceptable method of suspension, and the least comfortable, especially when the patient is sitting. The pelvic band, made of metal or plastic, is thicker than a Silesian bandage. The pelvic band runs from the hip hinge, around the waist, between the contralateral iliac crest and trochanter, and back to the hip hinge. The hinge is located laterally, just anterior to the trochanter, over the anatomic axis of the hip joint. Hip joint and pelvic band suspension is indicated for very short transfemoral limbs, geriatric patients who cannot don a suction suspension, and obese patients who cannot get adequate control with suction, silicone suspension sleeves, or Silesian band suspension.

Socket design for the transfemoral amputation has changed. The traditional quadrilateral socket has a narrow anteroposterior diameter to keep the ischium positioned back and up on top of the posterior brim of the socket for weight bearing. The anterior wall of the socket is 5–7 cm higher than the posterior wall to hold the leg back on the ischial seat. Anterior pain is a frequent complaint and should be addressed by modification of the prosthetic socket in a small local area such as over the anterior superior iliac spine. If the entire anterior wall is lowered or relieved, the ischium slips inside the socket and totally alters the load transfer and pressure areas. Even though the lateral wall is contoured to hold the femur in adduction, the overall dimensions of the quadrilateral socket are not anatomic and provide poor femoral stability in the coronal plane.

Narrow mediolateral transfemoral socket designs attempt to solve the problems of a traditional quadrilateral socket by contouring the posterior wall to set the ischium down inside the socket, not up on the brim. Weight is transferred through the gluteal muscle mass and lateral thigh instead of the ischium, which eliminates the need for anterior pressure from a high anterior wall. Attention is then focused on a narrow mediolateral contour to better hold the femur in adduction and minimize the relative motion between the limb and the socket. The normal shape and normal alignment (NSNA) socket and the contoured adducted trochanteric-controlled alignment method (CAT-CAM) socket are two of the narrow mediolateral designs available.

A socket made of flexible material with a rigid frame can also be used. The flexible material allows socket wall expansion with underlying muscle contraction. A flexible socket can be made in either the traditional quadrilateral or narrow mediolateral shapes. Advantages of this type of socket include improved comfort in walking and sitting and possibly improved muscular efficiency. One drawback is that the flexible material is less durable, and cracks can result in the loss of suction suspension and skin irritation.

Prosthetic knee joints are available in many designs to address specific patient needs. The traditional standard was the single-axis constant-friction knee. The constant-friction knee is simple, durable, lightweight, and inexpensive. The friction can be set at only one level to optimize function at one cadence, and patients have difficulty when walking at different speeds.

Outside hinges were the old standard for the knee disarticulation patient, to better approximate the center of motion of the knee. Outside hinges are cosmetically poor but still available for patients who used them successfully in the past and remain satisfied with them. For new patients, other types of knee units are used.

The term stance control knee has replaced the term safety knee. It refers to a knee unit that has weight-activated friction to increase stability and resistance to buckling as more of the amputee's body weight is applied. This unit is particularly useful for patients who are older, feel less secure, and have a very short residual limb, weak hip extensors, or hip flexion contractures.

A polycentric knee provides a changing center of rotation that is located more posteriorly than other knee joints. The posterior center of rotation offers more stability during stance and the first few degrees of flexion than other knee units do. The four-bar knee is one of many polycentric knee units available.

A hydraulic or pneumatic unit can be added to most knee joints to provide superior control of the prosthesis in swing phase by using fluid hydraulics to vary the resistance according to the speed of gait. This option is useful in active amputees who walk and run at different speeds.

The variable-friction knee unit can be a less expensive way to accommodate patients who walk at different speeds. This knee changes the friction according to the degree of flexion in the knee unit and leads to an improvement in the swing phase of walking. Although a variable-friction knee is less costly and requires less maintenance than a hydraulic unit, it is not as effective in allowing the amputee to walk at different cadences.

A manual locking option can also be added to most knee units to lock the knee in full extension. Locking is helpful if the patient is blind, feels less secure, has a very short residual limb, or is a bilateral amputee.

As mentioned previously, microprocessor-controlled so-called intelligent knee units incorporate the latest technology to provide superior control of the swing and stance characteristics or the knee and respond to the amputee's speed, cadence, and accelerations. Technology has not yet advanced enough for knee units to replace the tremendous motor power lost when an amputation is done above the knee.

Specifically designed prostheses known as stubbies are initially recommended for bilateral knee disarticulation or transfemoral amputees, regardless of age, who have lost both legs simultaneously but are candidates for ambulation. Stubbies consist of prosthetic sockets mounted directly over rocker-bottom platforms that serve as feet. The rocker-bottom platforms have a long posterior extension to prevent the patient from falling backward, and they have a shortened anterior process that allows smooth rollover into the push-off phase of gait. These prostheses look as if the foot were positioned backward. The use of stubbies results in a lowering of the center of gravity, and the rocker bottom provides a broad base of support that teaches trunk balance, provides stability, and allows the patient to build confidence during standing and ambulation. As the patient's confidence and skills improve, periodic lengthening of the stubbies is permitted until the height becomes nearly compatible with full-length prostheses, at which time the transition is attempted. Many patients reject full-length prostheses and prefer the stability and balance afforded by the stubbies.

Hip Disarticulation

Hip disarticulation (Figure 12–19) is rarely performed. Surgically, the traditional racket-shaped incision with an anterior apex is used in patients with vascular problems and in trauma-injured patients when possible. In tumor surgery, creative flaps based on the uninvolved anatomic compartments must be designed.

Figure 12–19.


Hip disarticulation.

Prosthetic replacement can be successful in healthy young patients who required hip disarticulation because of trauma or cancer but is generally not indicated for patients with vascular disease. The standard prosthesis is the Canadian hip disarticulation prosthesis. The socket contains the involved hemipelvis and suspends over the iliac crests. Although the hip joint and other endoskeletal components are made of lightweight materials in an effort to keep the weight to a minimum, the prosthesis is still heavy and difficult to manipulate. Ambulation with the prosthesis usually requires more energy than it would take to ambulate with crutches and a swing-through gait. For this reason, many ambulatory patients use crutches and no prosthesis. The advantage of the prosthesis is that it does allow freer use of the upper extremities.


Although a hemipelvectomy (Figure 12–20) is even less frequently required than a hip disarticulation, it is sometimes indicated for trauma injuries or cancer involving the pelvis. Use of a prosthesis after this procedure is extremely rare because the body weight must be transferred onto the sacrum and thorax. Special considerations for seating are usually required after hemipelvectomy.

Figure 12–20.



Prosthetic Prescription Following Amputation at or above the Knee

To be considered a candidate for a high anatomic level prosthesis (knee disarticulation and higher), a patient must be able to transfer independently, rise from sitting to standing independently, and ambulate using one leg and a swing-through gait over a distance of 100 feet on the parallel bars or with a walker. Although these requirements seem extreme, they are necessary for the successful use of this heavy and complicated prosthesis. The use of a transtibial prosthesis can make it easier to transfer and to ambulate. But a current transfemoral prosthesis can make it much more difficult to rise from sitting to standing because the powerful motor force required to extend the knee is not present. High-level prosthetic devices can actually increase the energy required for walking compared with one-leg swing-through gait. Unfortunately, without the ability to meet the activity demands unassisted, a prosthesis acts as an anchor to decrease overall independence. We use these same activity requirements as a functional test before prescribing a prosthesis for all transfemoral, hip disarticulation, and hemipelvectomy amputees.

Percutaneous Direct Skeletal Attachment of Artificial Limbs

The benefits of attaching prosthetic limbs through the skin, directly to the skeleton, was envisioned for nearly 100 years. Documentation of temporary external fixation for fractures dates to Malgaigne in 1845. During and just after World War II, independent attempts were made in Germany and the United States to attach a transtibial prosthesis directly to the tibia. Four humans were fit in May 1946 by Drummer, a general surgeon in Pinneberg, Germany. The two major hurdles continue to be the bone–implant interface, and the skin–implant interface. Breakthrough work by Branemark in Gothenburg, Sweden, advanced the use of titanium and improved design implants that led to over 30 years of successful dental and maxillofacial reconstruction with prosthetic devices directly connected to the bone of the mouth and face.

The skin of the extremities posed a larger challenge to the cutaneous–implant interface. Improvements in implant design and surgical technique, however, made it possible to implant and fit thumb, forearm, and transfemoral amputees successfully. Approximately 60 amputees have undergone surgical implantation and prosthetic fitting in Sweden, the United Kingdom, and Australia.

The early results confirm the potential promise of major improvements in attachment, proprioception, and function of osseointegrated prosthetic limbs compared with socket-style prostheses. Much work remains to be accomplished, however, especially in the skin–implant interface. A tremendous improvement in the bone–implant interface led to results that far outdistance historical attempts at directly attaching artificial limbs to the skeleton. Without true cutaneous–implant integration that provides a durable and biologic barrier, however, the risk of bacterial migration causing infection and loosening continues. It is fantastic to see this dream continue and advance.

Smith DG, Michael JW, Bowker JH: Atlas of Amputations and Limb Deficiencies: Surgical, Prosthetic, and Rehabilitation Principles, 3rd ed. American Academy of Orthopaedic Surgeons, 2004.

Bowker JH et al: North American experience with knee disarticulation with use of a posterior myofasciocutaneous flap. Healing rate and functional results in seventy-seven patients. J Bone Joint Surg Am 2000;82-A(11):1571. [PMID: 11097446] 

Branemark R et al: Osseointegration in skeletal reconstruction and rehabilitation: A review. J Rehabil Res Dev 2001;38(2):175. Review. No abstract available. [PMID: 11392650] 

Gaine WJ, McCreath SW: Syme's amputation revisited: A review of 46 cases. J Bone Joint Surg Br 1996;78:461.

Gottschalk F et al: Does socket configuration influence the position of the femur in above-knee amputation? J Prosthet Orthot 1989;2:94.

Kock HJ, Friederichs J, Ouchmaev A et al: Long-term results of through-knee amputation with dorsal musculocutaneous flap in patients with end-stage arterial occlusive disease. World J Surg 2004;28(8):801. Epub accessed August 3, 2004. [PMID: 15457362] 

Pinzur MS, Bowker JH, Smith DG et al: Amputation surgery in peripheral vascular disease. Instr Course Lect 1999;48:687. [PMID: 10098097] 

Unruh T et al: Hip disarticulation: An eleven-year experience. Arch Surg 1990;125:791.