Interventions for motor rehabilitation in people with transtibial amputation due to peripheral arterial disease or diabetes

Abstract

Background

Amputation is described as the removal of an external part of the body by trauma, medical illness or surgery. Amputations caused by vascular diseases (dysvascular amputations) are increasingly frequent, commonly due to peripheral arterial disease (PAD), associated with an ageing population, and increased incidence of diabetes and atherosclerotic disease. Interventions for motor rehabilitation might work as a precursor to enhance the rehabilitation process and prosthetic use. Effective rehabilitation can improve mobility, allow people to take up activities again with minimum functional loss and may enhance the quality of life (QoL). Strength training is a commonly used technique for motor rehabilitation following transtibial (below‐knee) amputation, aiming to increase muscular strength. Other interventions such as motor imaging (MI), virtual environments (VEs) and proprioceptive neuromuscular facilitation (PNF) may improve the rehabilitation process and, if these interventions can be performed at home, the overall expense of the rehabilitation process may decrease. Due to the increased prevalence, economic impact and long‐term rehabilitation process in people with dysvascular amputations, a review investigating the effectiveness of motor rehabilitation interventions in people with dysvascular transtibial amputations is warranted.

Objectives

To evaluate the benefits and harms of interventions for motor rehabilitation in people with transtibial (below‐knee) amputations resulting from peripheral arterial disease or diabetes (dysvascular causes).

Search methods

We used standard, extensive Cochrane search methods. The latest search date was 9 January 2023.

Selection criteria

We included randomised controlled trials (RCT) in people with transtibial amputations resulting from PAD or diabetes (dysvascular causes) comparing interventions for motor rehabilitation such as strength training (including gait training), MI, VEs and PNF against each other.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were 1. prosthesis use, and 2. adverse events. Our secondary outcomes were 3. mortality, 4. QoL, 5. mobility assessment and 6. phantom limb pain. We use GRADE to assess certainty of evidence for each outcome.

Main results

We included two RCTs with a combined total of 30 participants. One study evaluated MI combined with physical practice of walking versus physical practice of walking alone. One study compared two different gait training protocols. The two studies recruited people who already used prosthesis; therefore, we could not assess prosthesis use. The studies did not report mortality, QoL or phantom limb pain. There was a lack of blinding of participants and imprecision as a result of the small number of participants, which downgraded the certainty of the evidence.

We identified no studies that compared VE or PNF with usual care or with each other.

MI combined with physical practice of walking versus physical practice of walking (one RCT, eight participants) showed very low‐certainty evidence of no difference in mobility assessment assessed using walking speed, step length, asymmetry of step length, asymmetry of the mean amount of support on the prosthetic side and on the non‐amputee side and Timed Up‐and‐Go test. The study did not assess adverse events.

One study compared two different gait training protocols (one RCT, 22 participants). The study used change scores to evaluate if the different gait training strategies led to a difference in improvement between baseline (day three) and post‐intervention (day 10). There were no clear differences using velocity, Berg Balance Scale (BBS) or Amputee Mobility Predictor with PROsthesis (AMPPRO) in training approaches in functional outcome (very low‐certainty evidence). There was very low‐certainty evidence of little or no difference in adverse events comparing the two different gait training protocols.

Authors' conclusions

Overall, there is a paucity of research in the field of motor rehabilitation in dysvascular amputation. We identified very low‐certainty evidence that gait training protocols showed little or no difference between the groups in mobility assessments and adverse events. MI combined with physical practice of walking versus physical practice of walking alone showed no clear difference in mobility assessment (very low‐certainty evidence). The included studies did not report mortality, QoL, and phantom limb pain, and evaluated participants already using prosthesis, precluding the evaluation of prosthesis use.

Due to the very low‐certainty evidence available based on only two small trials, it remains unclear whether these interventions have an effect on the prosthesis use, adverse events, mobility assessment, mortality, QoL and phantom limb pain. Further well‐designed studies that address interventions for motor rehabilitation in dysvascular transtibial amputation may be important to clarify this uncertainty.

Plain language summary

Interventions for motor rehabilitation in people with below‐knee amputation due to peripheral arterial disease or diabetes

Key messages

– There is little research in the field of motor rehabilitation in people with amputation due to vascular causes.

– The studies reported some positive results, but we have little confidence in the results due to small numbers of participants and limited data.

What is amputation and what can happen after amputation?

Amputation is the removal of an external part of the body. Amputations due to vascular causes are commonly associated with diabetes and peripheral arterial disease. Peripheral arterial disease is caused by fatty deposits on the walls of the arteries (also called atherosclerosis) that leads to narrowing of the arteries (also called stenosis) and obstructions in the major blood vessels supplying the lower legs.

After lower‐limb amputation the rehabilitation process requires physical adaptation. Motor rehabilitation aims to enhance the rehabilitation process and prosthetic use improving mobility, allowing the return to usual activities, with minimum functional loss and better quality of life. Strength training (elaborated to build strength in one muscle group at a time) is a commonly used technique for motor rehabilitation following lower‐limb amputation. Strength training includes exercises for the surrounding hip muscles and muscles of the residual limb aiming to increase muscular strength. Recent interventions which might enhance the rehabilitation process can be performed at home. These include motor imaging (simulated movement or mentally rehearsing the action without really performing the movement), virtual environments (computer‐generated simulations) and proprioceptive neuromuscular facilitation (stretching the muscles, aiming to achieve maximal static flexibility).

What did we want to find out?

We wanted to find out if these interventions were effective for motor rehabilitation as the convenience of home training and lower costs may be attractive for healthcare practitioners and patients.

What did we do?

We searched medical databases for well‐designed studies in people with below‐knee amputations resulting from peripheral arterial disease or diabetes comparing different interventions for motor rehabilitation against each other. The interventions could have been given alone or combined with usual care (for example, wound dressing and stump care).

What did we find?

We found two randomised controlled trials (studies in which the participants were divided between treatment groups through a random method) with 30 participants (most recent search 9 January 2023). One study with eight participants evaluated motor imaging combined with walking versus walking alone. One study with 22 participants compared two different gait training protocols (one focused on impairment level versus one focused on task level). We found no studies that used other interventions such as virtual environments.

Main results

There was no clear difference in mobility assessment between motor imaging combined with walking and walking alone.

Both gait training protocols may slightly improve from baseline to after treatment for mobility assessment. There was little or no difference in side effects comparing the two different gait training protocols.

What are the limitations of the evidence?

We are not confident in the evidence because it was based on only two trials with a small number of participants. People knew which treatment they received, which could have affected the study's results.

How up to date is this evidence?

The most recent search was 9 January 2023.

Summary of findings

Summary of findings 1. Motor imaging compared to usual care for motor rehabilitation in people with transtibial amputation due to peripheral arterial disease or diabetes.

Motor imaging compared to usual care for motor rehabilitation in people with transtibial amputation due to peripheral arterial disease or diabetes
Patient or population: people with transtibial amputation due to peripheral arterial disease or diabetes Setting: no information Intervention: motor imaging Comparison: usual care
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with usual care	Risk with motor imaging
Prosthesis use Follow‐up: mean 14 days	—	—	—	8 (1 RCT)	—	All participants in this study used a prosthesis at the start of the study. Therefore, we were unable to assess differences in prosthesis use for the motor imaging and usual care groups.
Adverse events	—	—	—	—	—	The single study in this comparison did not assess this outcome.
Mortality – all causes	—	—	—	—	—	The single study in this comparison did not assess this outcome.
Quality of life	—	—	—	—	—	The single study in this comparison did not assess this outcome.
Mobility assessment assessed with: Timed Up‐and‐Go test, walking speed, step length, asymmetry of step length Follow‐up: mean 14 days	See comment		See comment	8 (1 RCT)	⊕⊝⊝⊝ Very lowa,b	The study reported no clear differences between baseline and follow‐up for any the mobility assessments for the motor imaging and usual care groups.
Phantom limb pain frequency	—	—	—	—	—	The single study in this comparison did not assess this outcome.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RCT: randomised controlled trial.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded one level as no blinding of participants.
^bDowngraded two levels as small number of participants (number of participants fewer than 400).

Summary of findings 2. Gait training protocol focused on impairment level compared to gait training protocol focused on task level for motor rehabilitation in people with transtibial amputation due to peripheral arterial disease or diabetes.

Gait training protocol focused on impairment level compared to gait training protocol focused on task level for motor rehabilitation in people with transtibial amputation due to peripheral arterial disease or diabetes
Patient or population: people with transtibial amputation due to peripheral arterial disease or diabetes Setting: inpatient + outpatient Intervention: gait training protocol focused on impairment levela Comparison: gait training protocol focused on task levelb
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with gait training protocol focused on task level	Risk with gait training protocol focused on impairment level
Prosthesis use Follow‐up: mean 10 days	—	—	—	22 (1 RCT)	—	All participants in this study used a prosthesis at the start of the study. Therefore, we were unable to assess differences in prosthesis use.
Adverse events – pain assessed with: NRS VAS Follow‐up: mean 10 days	See comment		—	22 (1 RCT)	⊕⊝⊝⊝ Very lowc,d	The study author reported that in the Impairment‐oriented group mean pain using VAS was 0.9 (SD 1.1) and in the task‐oriented group was 1.1 (SD 1.6) (P = 0.77).
Mortality – all cause	—	—	—	—	—	The single study in this comparison did not assess this outcome.
Quality of life	—	—	—	—	—	The single study in this comparison did not assess this outcome.
Mobility assessment assessed with: Locomotor Capabilities Index, BBS, Amputee Mobility Predictor, mean normalised velocity, Cadence, Double Limb Support time, Symmetry Index for Single Limb Support, Symmetry Index for Step Length, Percentage of Gait Cycle in Stance and Swing Phase Follow‐up: mean 10 days	See comment	—	—	22 (1 RCT)	⊕⊝⊝⊝⊝ Very lowc,d	To evaluate if the training strategy (impairment‐oriented or task‐oriented) led to difference in improvement between baseline (day 3) and post‐intervention (day 10) the change scores were used but those scores did not show any clear differences in the velocity, BBS or AMPPRO denoting no clear difference in training approaches in functional outcomes
Phantom limb pain frequency	—	—	—	—	—	The single study in this comparison did not assess this outcome.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). AMPPRO: Amputee Mobility Predictor with PROsthesis; BBS: Berg Balance Scale; CI: confidence interval; NRS: Numeric Rating Scale; RCT: randomised controlled trial; SD: standard deviation; VAS: Visual Analogue Scale.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

^aTraining protocol consisted of breaking down the functional activities into parts. Pregait training activities were practised standing (balance training, weight shifting in all planes, dynamic reaching, stepping up and down from a low step and single stepping). Practised ≥ 50% of their overall upright physiotherapy time. Remainder of treatment time with therapist was continuous corrective walking. For corrective walking, therapists gave verbal and manual cues while participants were walking to promote a symmetrical gait pattern using an appropriate assistive device as determined by training therapists.
^bPerformed functional tasks as a whole. Continuous corrective walking used for ≥ 90% of overall treatment time with the physiotherapist. For corrective walking, therapists gave verbal and manual cues while participants were walking to promote a symmetrical gait pattern using an appropriate assistive device as determined by training therapists.
^cDowngraded one level due to no blinding of participants.
^dDowngraded two levels due to small number of participants (number of participants fewer than 400).

Background

Description of the condition

Amputation is the removal of an external part of the body by trauma, medical illness or surgery. The most common factors leading to limb amputation are circulation diseases, infections, cancer, accidents and congenital malformation. When the diseased/injured limb cannot heal and the person's life is endangered, amputation may be necessary (Murdoch 1996). Transtibial amputations, also known as lower‐extremity amputations, have been performed since the prehistoric period (Azar 2017). This procedure causes a dramatic impairment to the individual's function and activity level, interfering with their participation in society (Üstün 2003). Amputations caused by vascular diseases (dysvascular amputations) are increasingly frequent, most commonly due to peripheral arterial disease (PAD), associated with an ageing population, and increased incidence of diabetes and atherosclerotic disease (Dillingham 2002; Lin 2022; Renzi 2006).

Lower limb PAD refers to the obstruction or narrowing of the large arteries of the lower limbs, most commonly caused by an atheromatous plaque or thrombosis of the residual lumen of an atheromatous plaque (Dormandy 1999; Fowkes 2008; Hooi 2004; Pande 2011). Although in many cases of less severe disease people can be asymptomatic, the major clinical manifestations of PAD are intermittent claudication (IC) and critical limb ischaemia (CLI). IC typically presents as exercise‐induced ischaemic pain in the leg muscles, from which relief is generally gained by rest. If underlying arterial stenosis continues to progress, CLI may develop. An individual with CLI will experience (often extreme) pain in the foot at rest, while the skin and other tissues of the affected limb may become more susceptible to ulceration and poor wound healing, including the development of gangrene. This progression from asymptomatic disease is often categorised using the Fontaine classification criteria (Fontaine 1954; see Table 3). The prevalence of lower limb PAD increases with age and is more common in men than women (Dormandy 2000). Revascularisation procedures (including angioplasty with or without stenting (Fakhry 2018), stenting (Bachoo 2010), and bypass grafting (Antoniou 2017)), may be required in those in whom disease is severe or does not improve with non‐surgical interventions. Individuals with chronic limb ischaemia are more urgently referred for possible revascularisation after thorough assessment by a vascular multidisciplinary team and adequate pain management (NICE 2012). Finally, in those with disease that is not amenable to revascularisation procedures, or with established gangrene, amputation may be needed. The second Trans‐Atlantic Inter‐Society Consensus Working Group (TASC II) documented an incidence of major amputations due to PAD ranging from 12 to 50 per 100,000 individuals per year (Norgren 2007). The ageing population was expected, in 2008, to increase this number by 50% in the following five years (Ziegler‐Graham 2008). In Scotland in 2010, the prevalence of IC ranged from 0.7% to 1.7% in people aged 16 to 54 years and was 7.4% in people aged over 74 years (Bromley 2011). Globally, an estimated 202 million people had PAD in 2010 (Fowkes 2013), an increase of 28.7% in low‐ and middle‐income countries and of 13.1% in higher‐income countries compared with the preceding 10 years (Fowkes 2013). The absolute numbers of PAD prevalent and incident cases  increased significantly between 1990 and 2019 (Lin 2022).

1. Fontaine classification of peripheral arterial disease.

Stage	Description
I	Asymptomatic
II	Mild claudication pain
IIa	Claudication distance > 200 m
IIb	Claudication distance < 200 m
III	Rest pain (especially at night)
IV	Ulceration or gangrene (or both) of the limb

Fontaine 1954.

People with diabetes with high glucose (blood sugar) levels over an extended period may develop defects in the walls of large arteries, commonly in the legs and feet, leading to the formation of blood clots or haemodynamically significant stenosis (or both), that can cause blockage (Madonna 2011). As a result, the blood flow throughout the arteries is slowed down, compromising the circulation to the lower limbs (Holt 2016). People with diabetes are more likely to develop PAD (one in four people), and have a 10‐fold increased risk for lower‐limb amputation, usually involving younger individuals and lower amputation levels compared to people without diabetes (Newton 1999).

The term 'amputation level' is used to describe the location at which the body part is amputated. Amputations below the knee (e.g. transtibial (28.2%), foot (10.6%) and toe (33.2%)) are much more common than those above the knee (transfemoral (26.1%); through‐knee, hip disarticulation and hemi‐pelvectomy amputations combined represent 1.5% of all amputations) (Dillingham 2002; Esquenazi 2016; Renzi 2006). According to a Cochrane Review encompassing unilateral transfemoral amputation there is limited evidence to inform the choice of prosthetic rehabilitation for these patients (Barr 2018).

Operative mortality from amputation for dysvascular diseases remains high (Stern 2017). The rehabilitation process requires physical adaptation as well as psycho‐social adjustment. Impaired mobility is the main factor responsible for the amputees' reported social isolation and emotional disturbance (Pell 1993).

Strength training (ST) is a commonly used technique for motor rehabilitation of lower‐extremity amputation (Wilhoite 2020). More recent interventions such as motor imaging (MI), virtual environments (VEs) and proprioceptive neuromuscular facilitation (PNF) might work as a precursor to enhance the rehabilitation process and prosthetic use in people with dysvascular amputations below the knee. Once these interventions can be performed at home, the overall expense of the rehabilitation process may decrease. Thus, if these interventions prove to be effective, feasible and have a good cost–benefit ratio it may direct healthcare practitioners towards these treatments and encourage researchers to investigate ways to improve them.

Description of the intervention

Mobilising and rehabilitating people after amputation at any level is a great challenge, and more so in dysvascular cases, because these people have the difficulties of the amputation itself, but also the main disease that led to this condition, frequently seen with cardiac dysfunction and other comorbidities that impact the rehabilitation process. Factors including old age, emotional distress, pain and low muscle strength can make rehabilitation for these amputees more difficult (Madsen 2018). In addition, prior to the amputation, most of these patients have been restricted in mobility due to pain and months of treatment directed to the salvage of the limb (Goodridge 2005).

ST aims to increase muscular strength and the regimen should include exercises for the surrounding hip muscles (particularly the hip abductor and hip extensor groups for pelvic stabilisation) and quadriceps/hamstring of the transtibial residual limb (crucial role in knee stability, which will be needed when a prosthetic device is used) (Czerniecki 1992; Powers 1996). ST may be used alongside usual care, which generally involves the treatment of comorbidities, wound healing and stump care. Additional rehabilitation techniques may include MI, VEs and PNF.

MI consists of simulated movement or mentally rehearsing the action without really performing the movement; the individuals feel themselves accomplishing the movement, however. It has been reported that MI improves motor skills (Halsband 2006; Lotze 2006; Mizuguchi 2012; Pascual‐Leone 1995). The combination of MI and ST has shown more benefits than ST alone (Allami 2008).

VEs consist of computer‐generated simulations. These simulations are interactive and immersible, allowing amputees to practice many daily tasks in addition to the ones that, for safety reasons, are difficult to actually practice (Lehman 2015). VEs may be used at home via a personal computer, increasing availability and opportunity to practice, while removing the need to travel to the health service (Lehman 2015). Treatments can be personalised and provide a standardised exercise record and assessment while providing motivation, functionality and a purposeful context to the therapy (Sveistrup 2003; Sveistrup 2004; Weiss 2003). The settings in VEs can vary from simple game‐like applications (played on computer screens) to three‐dimensional cave systems. Initially, the settings used only visual and auditory input. More recently, haptic (relating to the sense of touch) input devices such as pens, gloves, exoskeletons and joysticks enable the users with the sense of touch and changes in texture. Haptic information has been proved effective in increasing joint range of motion and force, which are the treatment objectives (Jack 2001). VEs offer the advantage of variable task difficulty which can be easily matched with the person's capabilities (Rizzo 1997). Users of VE programs report more enthusiasm and enjoyment upon exercise training, with improved confidence (Bisson 2004).

PNF consists of stretching the muscles, aiming to achieve maximal static flexibility, and is usually performed with a trainer or partner. This technique uses a series of contractions/relaxations with enforced stretching during the relaxation phase. It is a technique of passive stretching combined with isometric stretching (the joint angle and muscle length do not change during the contraction). PNF was designed to relax muscles, increase tone or activity, and has been effective in improving muscle elasticity and active/passive range of motion (Funk 2003; Lucas 1984; Wallin 1985).

How the intervention might work

Interventions for rehabilitation after amputation aim to allow return to function, and work by increasing muscle strength (Ranganathan 2004; Yue 1992), improving motor skills (Halsband 2006; Lotze 2006; Mizuguchi 2012; Pascual‐Leone 1995), improving muscle elasticity/range of motion (Funk 2003; Lucas 1984; Sharman 2006; Wallin 1985), retraining of balance/gait, pain management, limb rehabilitation, management of multitasking skills and executive functioning skills (Deutsch 2007; Mirelman 2011).

ST consists of performing exercises to increase muscle strength. The amputee performs those exercises under the guidance of a physiotherapist, usually at the rehabilitation centre. In transtibial amputees, the most important muscle groups to be strengthened include the hamstrings and the quadriceps, as well as all hip muscles associated with the residual limb (hip flexors, hip extensors, hip abductors, hip adductors). Exercises directed at increasing the strength of those muscles will help with gait, co‐ordination, balance and overall function. Preserved thigh muscle strength is related to a better walking capacity (Renström 1983).

MI is performed with the amputee imagining doing the motion in his/her mind. In other words, the amputee mentally rehearses the movement of the body part affected without ever actually attempting to perform the movement. It can be performed at the amputee's home as well as at the rehabilitation centre. MI leads to the activation of the same brain areas as actual movement, and is evidenced by activity on the front‐parietal motor regions while performing MI (Guillot 2008). MI activates some motor cortex areas, parietal lobule areas, basal ganglia and cerebellum (Decety 1994; Guillot 2009; Hétu 2013; Lotze 1999; Mizuguchi 2013a; Mizuguchi 2013b; Mizuguchi 2014a; Mizuguchi 2014b; Naito 2002). These brain areas are similar to the ones activated during actual motor execution (Hanakawa 2003; Hanakawa 2008; Zabicki 2017). During MI, the enhancement of corticospinal excitability is muscle specific (Fadiga 1999), and associated with force level and imagined movement phase (Hashimoto 1999; Mizuguchi 2013b), indicating that MI's neural processing is similar to the processing of an actual execution. With MI training it is possible to obtain muscle strength, and improved motor skills (Halsband 2006; Lotze 2006; Mizuguchi 2012; Pascual‐Leone 1995), but without muscle hypertrophy (Ranganathan 2004; Yue 1992).

VEs allow the participant the sensation of being inside the VE and having command over it, despite not being there physically (Witmer 1998); and can lead to a behaviour compatible with the experience in the VE (Slater 2003). With VE use, the individual may have activation of brain regions related to motor skills. This has been shown with the use of functional magnetic resonance imaging (fMRI), where an individual practising a task in the VEs had activation of brain areas related to secondary motor systems (August 2006). VE settings permit retraining of balance/gait, pain management, limb rehabilitation, management of multitasking skills and executive functioning skills (Deutsch 2007; Mirelman 2011). The most important feature in all VEs is the interaction. The users are allowed to interact with the VEs and virtual objects by using their own bodies within the environment, leading to a better interaction with the objects. It appears that the rehabilitation process depends on the action related to skill acquisition, not simply the affected limb being repeatedly moved (Plautz 2000).

Overall, the amputee performs PNF with a trainer or partner at home or at the rehabilitation centre. PNF works by reflexes arising from stimuli (stretching/contraction) on the Golgi tendon organs (located in the tendons of the target muscles or in the antagonist muscle to the target muscle) increasing the range of motion (Sharman 2006).

Amputation rehabilitation usually requires many visits to the health provider due to the need for treatment planning, detailed instruction, practice and feedback (Allami 2008). Besides the treatment session time, patients face the costs of moving to and from the health provider's clinic and face difficulties related to these logistics. Frequently, therapy performed in person is impractical, and may not be effective (Lo Priore 2002). PNF and MI may be easily performed at home after initial training, while VEs may be used at home via a personal computer, removing the barrier of travel (Levy 2011). The incorporation of these techniques into the rehabilitation process may provide more interesting and engaging tasks than formal repetitive therapy, as well as reduced demand on staff time and therefore cost. It has been reported that people using VEs have the perception of more benefits compared to the people on the traditional rehabilitation programme (Sveistrup 2004).

Why it is important to do this review

This Cochrane Review is particularly topical at this time, and of interest to professionals who study vascular diseases and to people diagnosed with these conditions. Dysvascular amputation is a growing public health problem largely due to the ageing population with elevated prevalence of dysvascular diseases (Dillingham 2002; Lin 2022; Renzi 2006). Permanent disability and impaired mobility can cause social isolation and emotional disturbance. Effective rehabilitation can improve mobility, allow people the opportunity to take up activities again with minimum functional loss and may enhance the quality of life (QoL) of this population (Pell 1993). A quality amputation rehabilitation service should be a priority due to the prevalence, economic impact and long‐term rehabilitation process (Bedotto 2006; Dillingham 2001; Pezzin 2004). This review will investigate the effectiveness of motor rehabilitation interventions such as ST, MI, VE and PNF techniques as a precursor to prosthetic rehabilitation in people with dysvascular below‐knee amputations. It will also benefit from a risk of bias and GRADE assessment. This review will help consumers, policy makers and healthcare providers make a healthcare decision in addition to bring information with implication for future research projects.

Objectives

To evaluate the benefits and harms of interventions for motor rehabilitation in people with transtibial (below‐knee) amputations resulting from peripheral arterial disease or diabetes (dysvascular causes).

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) and quasi‐randomised controlled trials (qRCTs) comparing interventions for motor rehabilitation. Any method of randomisation was eligible.

Types of participants

We included participants of both sexes, who were at least 18 years of age, with transtibial amputations due to PAD, diabetes, or both (dysvascular causes). We excluded people who had a transtibial amputation due to trauma or malignancy. We excluded elderly people with transfemoral amputation as this was investigated in a separate Cochrane Review (Barr 2018).

Types of interventions

We included all interventions for motor rehabilitation following transtibial amputations. We aimed to include studies that compared interventions for motor rehabilitation (such as ST (including gait training), MI, VEs, PNF) against each other. Motor rehabilitation interventions may have been administered alone or combined with usual care.

We defined 'usual care' as treatment of comorbidities, wound healing and stump care.

Types of outcome measures

The expected time points of evaluation of the outcomes were three to six months after amputation (to evaluate the early phase of postamputation recovery) and over six months after amputation (to evaluate the late phase of postamputation). Most amputees describe a six‐month recovery period as the necessary time to regain functional independence, defined as the level of mobility that allows them to perform daily activities with minimal assistance (Columbo 2018).

Primary outcomes

• Prosthesis use: use of prosthesis achieved after rehabilitation including walking freely or need of a support (walking stick, walker or crutches).

• Adverse events: such as falls (an unintentional loss of balance resulting in the individual coming to rest on the ground) and stump pain.

Secondary outcomes

• Mortality – all causes.

• Quality of life (QoL): we considered any validated questionnaires including the 36‐item Short Form (SF‐36), Physical Activity Scale for the Elderly, TAPES‐R (Trinity Amputation and Prosthesis Experience Scales – Revised), ABIS (Amputation Body Image Scale), ABIS‐R (shortened version of ABIS), Prosthetic Profile of the Amputee (PPA), Prosthesis Evaluation Questionnaire – Mobility Section (PEQ‐MS).

• Mobility assessment: we considered any reliable and valid measures of gait and mobility, such as Locomotor Capability Index (LCI), Timed Up‐and‐Go test, Amputee Mobility Predictor, Berg Balance Scale (BBS), mean normalised velocity (MNV), Cadence, Double Limb Support (DLS) time, Symmetry Index for Single Limb Support, Symmetry Index for Step Length, Percentage of Gait Cycle in Stance and Swing Phase, step length, walking speed, step count, walking capacity test (assessed with the two‐minute walk test (participants are instructed to walk as far as possible in the designated amount of time, on an established, standardised pathway and this distance is recorded; it is important that the participants' self‐selected walking speed is not altered by obstacles or traffic in the pathway, others walking alongside them or the therapist guarding/assisting)) (Brooks 2001; Brooks 2002).

• Phantom limb pain frequency: presence of pain that feels like it originates from a body part that is no longer there. After an amputation, areas of the spinal cord and brain lose input from the missing limb and adjust to this detachment in unpredictable ways which can trigger pain (Subedi 2011).

Search methods for identification of studies

We did not apply any language or publication restrictions.

Electronic searches

The Cochrane Vascular Information Specialist conducted systematic searches of the following databases for RCTs, qRCTs and controlled clinical trials without language, publication year or publication status restrictions:

• Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web);

• Cochrane Central Register of Controlled Trials (CENTRAL; Issue 12, 2022) via the Cochrane Register of Studies Online (CRSO);

• MEDLINE (Ovid MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE);

• Embase Ovid;

• CINAHL EBSCO.

The most recent searches were carried out on 9 January 2023.

The Information Specialist modelled search strategies for other databases on the search strategy designed for MEDLINE. Where appropriate, they were combined with adaptations of the highly sensitive search strategy designed by Cochrane for identifying RCTs and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Chapter 6; Lefebvre 2022). Search strategies for major databases are provided in Appendix 1.

The Information Specialist searched the following trials registries on 9 January 2023:

• World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch);

• ClinicalTrials.gov (clinicaltrials.gov).

Searching other resources

We checked the bibliographies of included trials for further references to relevant trials. We contacted specialists in the field and authors of the included trials for any possible unpublished data, as well as other sources, such as internal reports and conference proceedings.

Data collection and analysis

Selection of studies

We examined titles and abstracts and selected the relevant reports after merging the search results and removing duplicate records. Two review authors (LBA and CF), independently, screened the trials identified by the literature search. We retrieved and examined the full text of the selected trials for compliance with the eligibility criteria. We resolved conflicts through discussion, and if necessary, by involving a third review author, who had the final vote (FMJ or VT) in the case of any disagreement. We illustrated the study selection process in a PRISMA diagram (Liberati 2009). We included studies only published as an abstract if there were sufficient data to determine study eligibility. We attempted to contact the authors of the abstract for further information. We listed all articles excluded after full‐text assessment in the characteristics of excluded studies table and provided the reasons for their exclusion.

Data extraction and management

Two review authors (LBA and CF) extracted data independently and recorded them on a paper data extraction form. We resolved disagreements by discussion within the review team. We collected the following information.

• Study features: publication details (e.g. year, country, authors); study design; population data (e.g. age, comorbidities, sex, time since amputation; number of participants randomised into each intervention group; number of participants lost to follow‐up; duration of follow‐up).

• Outcomes: types of outcomes measured; timing of outcomes; adverse events.

We extracted data of estimates of variability presented as graphic information using Plot Digitizer (Plot Digitizer).

Assessment of risk of bias in included studies

Two review authors (LBA and CF) independently evaluated the studies included in the review and assess risk of bias using the RoB 1 tool according to the domains and criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We evaluated the following domains and characterised them as low, high or unclear risk of bias.

• Random sequence generation

• Allocation concealment

• Blinding of personnel and participants

• Blinding of outcome assessors

• Incomplete outcome data

• Selective outcome reporting

• Any other potential threat to validity

We reported these assessments for each study individually in the risk of bias table in the Characteristics of included studies table. We contacted the study author(s) to seek clarification in cases of uncertainty over methodology or data.

Measures of treatment effect

We planned to calculate the risk ratio (RR) and 95% confidence intervals (CIs) for the dichotomous variables (prosthesis use, adverse events, mortality and phantom limb pain frequency) using Review Manager 5 (Review Manager 2014). We planned to calculate the mean difference (MD) and 95% CI for continuous variables (QoL and mobility assessment) that used similar scales. We planned to calculate the standardised mean difference (SMD) and 95% CIs for continuous variables that used different scales.

Unit of analysis issues

We considered the individual participant as our unit of analysis (unit to be distributed randomly for interventions to be compared), that is, the number of observations in the analysis should have matched the number of randomised participants. We grouped studies evaluating various treatment groups relevant to the review to create a single pairwise comparison.

Dealing with missing data

We noted partial and incomplete data on the data collection form, and took this into account when assessing the overall risk of bias of the study. We contacted the study authors for additional information where data were missing or unavailable. In the case of non‐response, irrespective of the type of data, we reported dropout rates in the Characteristics of included studies table of the review.

Assessment of heterogeneity

We planned to quantify inconsistency among the pooled estimates using the I² statistic (where I² = ([Q − df]/Q) × 100% where Q is the Chi² statistic, and 'df' represents the degree of freedom). This illustrates the percentage of variability in effect estimates resulting from heterogeneity rather than sampling error (Higgins 2011). We intended to interpret the thresholds for the I² statistic as follows: 0% to 25% indicates low heterogeneity; 25% to 75% indicates moderate heterogeneity; and more than 75% indicates substantial heterogeneity (Higgins 2011). If studies differ methodologically and clinically, it may be preferable not to pool the results.

Assessment of reporting biases

If, for futures updates, we are able to include more than 10 studies in the meta‐analysis, we intend to assess the presence of publication bias and other reporting bias using funnel plots. If asymmetry is present, we intend to explore possible causes, including publication bias, poor methodological quality and true heterogeneity (Higgins 2011).

Data synthesis

We planned to carry out statistical analysis using Review Manager 5 (Review Manager 2014). We planned to use a random‐effects model to synthesise the data because of the complexity and differences in the interventions for motor rehabilitation. We intended to use RR if the data were dichotomous, or MDs if the data were continuous. We reported data narratively if it was not appropriate to combine data in a meta‐analysis.

Subgroup analysis and investigation of heterogeneity

We planned to perform subgroup analyses to consider the following.

• Age

• Gender

• Comorbidities (presence of diabetes, hypertension)

• Types of utilised devices for rehabilitation

Sensitivity analysis

If sufficient studies were identified, we planned to conduct sensitivity analysis based on the separation of studies according to the method of randomisation, allocation concealment and blinding of outcome assessment by excluding trials assessed in the risk of bias table as being at 'high' or 'unclear' risk. We planned to present these results and compare them with the overall findings. However, due to the limited data available this was not possible.

Summary of findings and assessment of the certainty of the evidence

Using GRADEpro GDT we developed summary of findings tables to provide the key information presented in the review comparing the interventions for motor rehabilitation (GRADEpro GDT). We created a table for each comparison reported in the review. We included the outcomes described in the Types of outcome measures.

• Prosthesis use

• Adverse events

• Mortality – all causes

• QoL

• Mobility assessment

• Phantom limb pain frequency

We assessed the certainty of the evidence for each outcome as high, moderate, low or very low, based on the criteria of risk of bias of the included studies, the directness of the evidence, the inconsistency of the results, the precision of the estimates and the risk of publication bias (GRADE 2017). We based the tables on methods described in Chapters 11 and 12 of the Cochrane Handbook for Systematic Reviews of Interventions, and planned to justify any departures from the standard methods (GRADE 2004; Higgins 2011).

Results

Description of studies

Results of the search

The search identified 2991 records after duplicates were removed. After reading the titles and abstracts, we excluded 2826 irrelevant records and considered 165 full‐text articles for eligibility. We excluded 145 full‐text articles and provided a selection of 36 records (32 studies) with reasons (Characteristics of excluded studies table). We included two studies (two records) that met the review inclusion criteria (Characteristics of included studies table). We identified 10 studies (10 records) that are awaiting classification (Characteristics of studies awaiting classification table) and eight ongoing studies (eight records; Characteristics of ongoing studies table) (see Figure 1 for PRISMA flow diagram).

1.

Study flow diagram. RCT: randomised controlled trial.

Included studies

We included two studies (Hyland 2009; Vanmairis 2018). See Characteristics of included studies table.

Type of study

The results of the RCTs were published in 2009 and 2018. The included studies evaluated people with transtibial amputation due to PAD or diabetes. Only Hyland 2009 calculated the sample size.

Setting

One study was conducted in France (Vanmairis 2018), and the other in the US (Hyland 2009).

Unit of analysis

The two studies used participants as a unit of analysis.

Study participants

The two studies provided data for 30 participants. The number of participants was eight (Vanmairis 2018), and 22 (Hyland 2009). The participants had transtibial amputation due to PAD or diabetes. Inclusion and exclusion criteria of the included studies varied.

The mean age of the participants ranged from 64.4 to 65 years.

Both studies evaluated women and men, but the majority were men.

Vanmairis 2018 did not evaluate comorbidities. Hyland 2009 did not specify the participant's comorbidities. 

Interventions

The two studies assessed different interventions for motor rehabilitation.

Vanmairis 2018 evaluated MI combined with physical practice of walking versus physical practice of walking.

Hyland 2009 evaluated two gait training protocols with one focused at an impairment level, and the other at a task level. Hyland 2009 described the gait training protocols as follows: "Impairment‐oriented group: the training protocol consisted of breaking down the functional activities into parts. Pre‐gait training activities were practised in standing and included balance training, weight shifting in all planes, dynamic reaching, stepping up and down from a low step, and single stepping. Participants practised these activities for no less than 50% of their overall upright physical therapy time. The remainder of the treatment time with a therapist was made up of continuous corrective walking. For corrective walking, therapists gave verbal and manual cues to the participants while they were walking to promote symmetrical gait pattern. The participants used the appropriate assistive device as determined by the training therapists during walking. Task‐oriented group: this group performed functional tasks as a whole. Continuous corrective walking was utilised for at least 90% of their overall treatment time with the physical therapist. For corrective walking, therapists gave verbal and manual cues to the participants while they were walking to promote a symmetrical gait pattern. The participants utilised the appropriate assistive device during walking."

The duration of follow‐up and time since amputation varied. Hyland 2009 evaluated participants with a mean length of time since amputation ranging from 2.7 (standard deviation (SD) 3.3) to 4.2 (SD 1.4) months for a 10‐day protocol (not including non‐rehabilitation days on which the participant did not receive therapy). Vanmairis 2018 evaluated participants after a mean of 271 days postamputation (experimental group) and 201 days (control group) for two weeks.

We found no eligible studies of interventions for motor rehabilitation in dysvascular transtibial amputation using interventions such as VEs and PNF.

Outcomes

The two studies evaluated people already using prostheses. Hyland 2009 reported the adverse event "stump pain intensity" using a Numeric Rating Scale (NRS) Visual Analogue Scale (VAS). Vanmairis 2018 did not describe any adverse events; however, this study excluded people with phantom sensations or pains. Neither study described mortality or evaluated QoL. The two studies measured mobility assessment using valid measures such as LCI, walking speed, step length, and Timed Up‐and‐Go test. Hyland 2009 did not report phantom limb pain and Vanmairis 2018 excluded people with phantom limb pain.

Funding

Neither study obtained funding or support.

Excluded studies

In total, we excluded 145 studies. In this review, we present 32 of the 145 studies with reasons. Reasons for exclusion were: not RCT (Anderson 2021; Chin 2016; Christiansen 2020; Johannesson 2010; MacLean 1994; NCT04120038; NCT04431817; Schack 2021; Scott 2017; Silva 2021; Torbjörnsson 2020); included people with amputation due to non‐dysvascular causes (Barnett 2009; Erbahceci 2001; Gailey 2020; Igor 2013; Iman 2017; Mazari 2010; NCT03930199; Schafer 2021; Tousignant 2015); did not evaluate motor rehabilitation (Durand 2020; Lass 2013; Liedberg 1983; NCT03149432; Saimpont 2021; Silva‐Filho 2019; Traballesi 2012; Woodburn 2004; Younesian 2020); levels of amputation were transfemoral and transtibial (Godlwana 2020); one study was a study protocol (Bourque 2019); and one study was cancelled by the authors (NCT04086069). See Characteristics of excluded studies table.

Studies awaiting classification

Ten studies were classified as awaiting classification (Anjum 2016; Chapman 2011; Gauthier‐Gagnon 1986; Kirdi 2017; Lamberg 2012; NCT02328859; NCT03872193; Shu 2014; Tao 2019; Topuz 2013).

See Characteristics of studies awaiting classification table for details.

Ongoing studies

We identified eight ongoing studies (NCT02761447; NCT03995238; NCT04083456; NCT04750876; NCT04968691; NCT05095805; NCT05656924; RBR‐4s5nkh). See Characteristics of ongoing studies table.

Risk of bias in included studies

Figure 2 and Figure 3 provide an overall summary of bias present within each of the included studies (see also Characteristics of included studies table for further details).

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation

The studies were both RCTs, but did not provide details about the generation of random sequences in addition to a statement of 'randomised'. Vanmairis 2018 reported that the age of the participants was used as a stratification variable, a criterion that could represent a possible confounding factor. The studies were considered at unclear risk of bias due to lack of information.

Allocation concealment

The studies were at unclear risk of bias due to lack of information.

Blinding

Because the nature of exercise‐based studies involves performing an activity versus standard care, it is impossible to blind the participants and difficult to blind the professionals who perform the therapeutic approach. Neither study blinded the participants. However, the healthcare professionals performing the outcomes assessments were blinded. Therefore, both studies were at high risk of performance bias.

Both studies blinded the evaluators and, therefore, were at low risk of detection bias.

Incomplete outcome data

Neither study excluded participants and, therefore, were at low risk of attrition bias.

Selective reporting

Both studies reported all their planned outcomes and were at low risk of reporting bias.

Other potential sources of bias

Both studies were at unclear risk of bias as we identified no other potential sources of bias.

Effects of interventions

See: Table 1; Table 2

The data presented in the study papers, even after communication with authors, were too scarce to allow analysis 'between groups' in a global way. In addition, the studies used a wide variety of outcome measures; therefore, interventions and outcome measures were considered heterogeneous. We present the results narratively.

Motor imaging plus usual care versus usual care alone

One study evaluated MI combined with physical practice of walking (intervention group) versus physical practice of walking alone (control group) (Vanmairis 2018; Table 1). Although the participants could not be blinded to the interventions, the investigators were.

Primary outcomes

Prosthesis use

The study evaluated people already using a prosthesis, who were able to walk 10 m with or without a walking aid, in order to perform the initial assessment. In the intervention group, the mean time since amputation was 271.5 (SD 275.4) days and the mean time since the placement of the first prosthesis was 92.75 (SD 16.7) days. In the control group, the mean time since amputation was 130.5 (SD 10.3) days and mean time since the placement of the first prosthesis was 96.12 (SD 17.6) days. As the study recruited people who already used a prosthesis, the outcome prosthesis use could not be meaningfully assessed.

Adverse events

The study did not describe any adverse events; however they excluded people with presence of phantom sensations or pains during task imagination greater than 2/10 on the digital scale, since it may cause interference during MI.

Secondary outcomes

Mortality – all causes

The study did not report mortality.

Quality of life

The study did not report QoL.

Mobility assessment

The study assessed mobility using gait parameters assessed by laboratory kinetic analysis.

The study author reported that walking speed improved by +0.21 m/second in the intervention group and +0.20 m/second in the control group with no difference between groups (P value not reported; Table 4).

2. Mobility assessment (walking speed, step length Symmetry Index, ASL, AAAS, TUG).

Mobility assessment	Motor imaging group (mean ± SD)	Control group (mean ± SD)
Walking speed (m/s)
Baseline	0.27 ± 0.17	0.40 ± 0.20
Post‐treatment	0.48 ± 0.38	0.60 ± 0.24
Step length Symmetry Index (%)
Baseline	38.94 ± 36.0	24.08 ± 32.92
Post‐treatment	13.07 ± 6.93	19.5 ± 20.0
Asymmetry of the average and maximum amount of support (%)
Baseline	−18.85 ± 11.65	−17.50 ± 19.15
Post‐treatment	−10.92 ± 12.08	−10.23 ± 8.02
Initial and maximum amount of support (%)
Baseline	−11.69 ± 15.31	−9.81 ± 21.19
Post‐treatment	−7.46 ± 3.54	−13.54 ± 24.04
TUG test (s)
Baseline	35.07 ± 23.07	32.66 ± 8.66
Post‐treatment	19.23 ± 8.73	24.16 ± 7.86

ASL: asymmetry of step length; AAAS: asymmetry of average amount of support; SD: standard deviation; TUG: Timed Up and Go test.

Vanmairis 2018.

The asymmetry of step length improved in the intervention group, while the asymmetry of step length worsened in the control group. The study author also reported that step length Symmetry Index improved in the prosthetic and non‐amputee sides in both the intervention and control groups, with no difference between groups (Table 4).

The asymmetry of the average and maximum amount of support on the prosthetic side and on the non‐amputee side improved in intervention and control groups, with no difference between groups. The asymmetry of the maximum amount of support improved in the intervention group while it worsened in the control group (Table 4).

The Timed Up‐and‐Go test improved in the intervention and control groups with no difference between the groups (Table 4).

Phantom limb pain frequency

The study excluded people with phantom limb pain.

Gait training protocol focused on impairment level versus gait training protocol focused on task level

One study compared two different gait training protocols (Table 2; Hyland 2009).

Primary outcomes

Prosthesis use

The study evaluated people already using a prosthesis who tolerated wearing the prosthesis for at least 15 minutes. The participants were able to use their prosthesis similarly, with no difference between groups. Therefore, the outcome prosthesis use could not be meaningfully assessed.

Adverse events

The study reported no serious adverse events. Hyland 2009 evaluated pain in both gait training groups using an NRS VAS. There was little to no difference between groups (VAS mean 0.9 (SD 1.1) in impairment‐oriented group versus 1.1 (SD 1.6) in task‐oriented group).

Secondary outcomes

Mortality – all causes

The study did not report mortality.

Quality of life

The study did not report QoL.

Mobility assessment

Hyland 2009 compared two different gait training protocols (focused on impairment level versus focused on task level). For LCI, BBS and AMPPRO there was improvement within each study group from baseline to post‐treatment. However, there were no clear differences between the study groups for those parameters (Table 5).

3. Mobility assessment table (LCI, BBS, AMPPRO).

Mobility assessment	Impairment‐oriented group (mean ± SD)	Task‐oriented group (mean ± SD)	Mean difference	95% confidence interval
LCI (score)
Baseline	21.11 ± 8.6	21.50 ± 7.4	—	—
Post‐treatment	24.78 ± 9.4	28.00 ± 11.8	—	—
Change	3.67 ± 1.94	6.88 ± 6.38	−3.32	−8.80 to 2.19
Berg Balance Scale (score)
Baseline	15.45 ± 8.6	22.45 ± 11.7	—	—
Post‐treatment	27.09 ± 9.2	32.91 ± 12.96	—	—
Change	11.64 ± 5.8	10.46 ± 5.1	1.18	−3.7 to 6.06
AMPPRO (score)
Baseline	14.36 ± 7.1	19.73 ± 9.5	—	—
Post‐treatment	25.27 ± 8.1	29.64 ± 9.50	—	—
Change	10.91 ± 3.86	9.82 ± 53.49	1.09	−2.18 to 4.36

AMPPRO: Amputee Mobility Predictor with PROsthesis; BBS: Berg Balance Scale; LCI: Locomotor Capability Index; SD: standard deviation.

Hyland 2009.

There were differences in the impairment‐oriented group from baseline to post‐treatment for some parameters (cadence, MNV and DLS time; Table 6). There was a difference within the task‐oriented group only in MNV from baseline to post‐treatment (Table 6). For Symmetry Index for step length there was a difference from baseline to post‐treatment in both groups (Table 6). The impairment‐oriented and task‐oriented groups demonstrated no clear improvement for Symmetry Index for Single Limb Support (SI Index) or single limb stance time from baseline to post‐treatment (see Table 6).

4. Mobility assessment table (mean normalised ratio, cadence, double limb support and SI).

Mobility assessment	Impairment‐oriented group (mean ± SD)	Task‐oriented group (mean ± SD)	Mean difference	95% confidence interval
Mean normalised velocity (LL/s)
Baseline	0.13 ± 0.07	0.34 ± 0.10	—	—
Post‐treatment	0.22 ± 0.10	0.49 ± 0.20	—	—
Change	0.09 ± 0.07	0.14 ± 0.16	−0.52	−0.17 to 0.07
Cadence (steps/minute)
Baseline	31.57 ± 8.2	59.30 ± 13.8	—	—
Post‐treatment	42.74 ± 7.3	65.84 ± 16.9	—	—
Change	11.17 ± 7.23	16.32 ± 12.59	−5.15	−15.5 to 5.22
Double limb support (% gait cycle)
Baseline	70.74 ± 11.80	52.34 ± 9.8	—	—
Post treatment	62.62 ± 13.5	48.89 ± 12.9	—	—
Change	8.68 ± 8.86	6.70 ± 7.42	1.98	−5.98 to 9.94
SI for Single Limb Support (SI Index)
Baseline	−25.47 ± 37.2	11.07 ± 59.4	—	—
Post‐treatment	−14.03 ± 33.8	2.44 ± 29.6	—	—
Change	18.55 ± 16.29	25.68 ± 25.11	−7.13	−25.96 to 11.69
SI of Step Length (Index)
Baseline	63.89 ± 76.5	−13.95 ± 69.6	—	—
Post‐treatment	67.17 ± 65.4	−0.32 ± 48.0	—	—
Change	44.20 ± 35.48	22.25 ± 25.16	21.96	−5.40 to 49.31

SD: standard deviation; SI: Symmetry Index.

Hyland 2009.

Percentage of Gait Cycle in Stance and Swing Phase showed differences within the impairment‐oriented group from baseline to post‐treatment in both phases (stance and swing) for both limbs. In the task‐oriented group, there was no clear improvement post‐treatment in intact limb and prosthetic limb in both stance and swing phases (Table 7).

5. Mobility assessment table (percentage of gait cycle in stance and swing phase).

Percentage of gait cycle in stance and swing phase (%)	Normala	Impairment‐oriented group (mean ± SD)		Task‐oriented group (mean ± SD)
		Intact limb	Prosthetic limb	Intact limb	Prosthetic limb
Baseline
Stance phase	60	88.61 ± 5.0	83.38 ± 8.9	76.91 ± 9.7	72.50 ± 8.7
Swing phase	40	11.84 ± 6.2	16.61 ± 8.9	24.92 ± 14.8	27.48 ± 8.7
Post‐treatment
Stance phase	60	83.45 ± 5.7	79.16 ± 8.2	76.71 ± 5.4	72.56 ± 8.0
Swing phase	40	16.54 ± 5.7	20.84 ± 8.2	23.30 ± 5.4	27.51 ± 7.9
Change score
Stance phase		5.16 ± 3.5	5.12 ± 14.5	6.98 ± 6.4	3.79 ± 3.7
Swing phase		4.70 ± 3.8	5.12 ± 14.5	8.87 ± 11.1	3.74 ± 3.7

^a The stance phase makes up approximately 60% of the gait cycle and the swing phase makes up the other 40%. The stance/swing ratio in healthy adults is 60/40.
SD: standard deviation.
Hyland 2009.

We noted differences between the groups for DLS time, stance/swing ratio and cadence (Table 6; Table 7).

Hyland 2009 used change scores to evaluate if there was difference in improvement from baseline (day three) to post‐intervention (day 10) according to the training strategy used (impairment‐oriented or task‐oriented). There was no evidence of any clear differences in scores for velocity, BBS or AMPPRO.

Phantom limb pain frequency

The study did not report phantom limb pain.

Subgroup and sensitivity analysis

We were unable to perform subgroup analysis due to lack of information in the trials and the small number of trials identified.

Discussion

Summary of main results

We found two trials with 30 participants that studied interventions for motor rehabilitation in people with transtibial amputation due to PAD or diabetes. The studies involved different interventions for motor rehabilitation namely MI (one study) and gait training protocols (one study). We did not find any studies investigating VE or PNF.

Motor imaging

One study compared MI combined with physical practice of walking versus physical practice of walking alone (Vanmairis 2018). Very low‐certainty evidence showed that there was no evidence of a difference in mobility assessment (walking speed, step length symmetry, symmetry of the amount of support, Timed Up‐and‐Go test) between the groups. This study evaluated people already using a prosthesis. The study did not assess the other outcomes of interest such as adverse events, mortality, QoL and phantom limb pain.

Gait training protocols

Hyland 2009 compared two different gait training protocols (focused on impairment level versus focused on task level) identifying, with very low‐certainty evidence, improvement within each study group from baseline to post‐treatment for LCI, AMPPRO and BBS. There were no clear differences between the  groups for those parameters. The impairment‐oriented group presented differences from baseline to post‐treatment for some parameters: cadence, MNV, double‐limb support time and percentage of the gait cycle in stance and swing for both lower extremities. There was a difference within the task‐oriented group only in MNV from baseline to post‐treatment. For Symmetry Index for Step Length there was a difference from baseline to post‐treatment in both groups. Both groups demonstrated no clear improvement for Symmetry Index for Single Limb Support (SI Index) or single limb stance time from baseline to post‐treatment. We noted differences between the groups for double‐limb support time, stance/swing ratio and cadence with significant improvement in the impairment‐oriented group on all three measures, whereas the task‐oriented treatment group did not.

There was no clear difference in adverse events comparing the two different gait training protocols. Hyland 2009 evaluated people already using a prosthesis. There were no data for the other outcomes of interest such as mortality, QoL and phantom limb pain.

Overall completeness and applicability of evidence

We included two studies that reported only one or two outcomes of interest for this review. The studies ranged from 10 days to two weeks, with outcomes reported at different time points. In addition, studies used different scales to assess mobility. Together with the small number of included studies, the small number of participants, the impossibility of blinding of participants and difficulty in blinding personnel conducting the treatments, the certainty of evidence was affected, the applicability of the evidence was weakened and this should be considered when interpreting the results.

We expected to find studies using VE or PNF for motor rehabilitation of people with dysvascular amputation but we did not identify such studies.

We included only RCTs and qRCTs but other non‐randomised evidence is available. A selection is reported under Agreements and disagreements with other studies or reviews.

Some studies are awaiting classification due to the lack of information about the cause and level of amputation in these studies. To date we have received no additional information received from the contacted study authors. There are eight ongoing studies that may contribute to the evidence at review update.

Quality of the evidence

We judged the overall certainty of the evidence to be very low according to the GRADE approach, as described in the Cochrane Handbook for Systematic Reviews of Interventions (GRADE 2004; Higgins 2011).

The risk of bias summary showed that selection bias was unclear in both studies (Figure 3). The two studies were at high risk for performance bias but at low risk for detection bias. Risk of attrition and reporting bias was low in the two studies. Both studies had an unclear risk of other potential bias. Due to the different interventions, we could not meta‐analyse the studies.

Motor imaging versus usual care

We could not assess the certainty of the evidence for the outcome prosthesis use since the participants recruited were already using prosthesis. We downgraded the certainty of evidence three levels to very low for mobility assessment because of lack of blinding of participants and imprecision. The study did not report the other outcomes of interest (i.e. adverse events, mortality, QoL and phantom limb pain). This study had some methodological limitations such as the small number of participants (only eight participants), study length (two weeks) and time since amputations was longer in the intervention group, the change in prosthesis of participants during the study was not evaluated, and failure to perform other tests (such as Time‐Dependent Motor Imagery to better evaluate the MI capabilities of the amputees), decreasing the study reliability. The study also stratified randomisation of the study population using the age of the participants as a stratification variable, a criterion that could represent a possible confounding factor. See Table 1.

Gait training protocol focused on impairment level versus gait training protocol focused on task level

We could not assess the certainty of the evidence for prosthesis use since the participants recruited were already using prosthesis. For the outcomes adverse events and mobility assessment, we downgraded the certainty of the evidence three levels to very low because of lack of blinding of participants and imprecision. The study did not report other outcomes of interest, mortality, QoL and phantom limb pain. The study used three centres and nine therapists, with the subacute rehabilitation setting of training performing 30 to 60 minutes of therapy (mean 45 minutes) and the inpatient rehabilitation setting performing 60 to 90 minutes (mean 70 minutes); however, on day 10 of training there was no clear difference. This study had methodological limitations such as the small number of participants (only 22 participants) and the study length (10 days). See Table 2.

Potential biases in the review process

We attempted to limit biases at every stage in our review process, following guidance from the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

The risk of bias in locating studies was low. There were no arbitrary search limiters such as geographical location of study or year of publication.

The risk of bias in synthesising studies was low. All expected outcomes were reported on, irrespective of statistical significance.

We requested relevant data from study authors when necessary. The inclusion and exclusion criteria described in the protocol were observed in order to limit any type of non‐compliance with the protocol.

We did not perform a funnel plot assessment because of the insufficient number of included studies (fewer than 10) in each comparison.

Agreements and disagreements with other studies or reviews

We are not aware of previous reviews of interventions for motor rehabilitation in people with dysvascular transtibial amputations.

Saruco 2019 reported in their observational and longitudinal study of MI ability of nine people with lower‐limb amputation a significant improvement in walking task, Timed Up‐and‐Go test and stairs (up and down) during the rehabilitation programme. However, this was no longer evident six weeks after the end of the programme.

Matalon 2021 evaluated one participant with a transfemoral amputation who underwent three sessions of MI per week for four weeks, with a retention testing of subjective and functional assessments at baseline, post‐test and one week. Matalon 2021 reported important and sustained improvement on the Short Form BBS and Tinetti Performance Oriented Mobility Assessment scores, decreased phantom limb pain and the participant could walk a short distance independently for the first time.

Highsmith 2016 reported in their systematic review of gait training interventions for lower extremity amputees from 18 studies (three RCTs, 15 non‐RCTs) that the overground training combined with awareness intervention (auditory, manual and psychological) was effective in improving gait. Treadmill‐based training was effective in improving gait in the following situations: as a supplement to overground training, independently (when augmented with bodyweight support or visual feedback, or both) or as part of a home‐exercise plan.

Wong 2016, in their systematic review of exercise programmes to improve gait performance in people with lower‐limb amputation, from eight studies (three RCTs, five non‐RCTs) and different exercise programmes (specific muscle strengthening, supervised walking, part‐to‐whole gait training, balance training and functional gait/activity training) reported improvements in gait performance (self‐selected gait speed).

Űlger 2018 in their systematic review of physiotherapy and rehabilitation approaches to lower‐limb amputation from nine studies (one RCT, eight non‐RCTs) evaluated different interventions such as strengthening exercises, balance training using video games, balance exercises with Phyaction software, strengthening and range of motion exercises, weight bearing and walking, home‐based treadmill training. The authors demonstrated positive results with a physiotherapy programme in terms of walking and balance ability, weight‐lifting capacity with prosthesis, functional status and acute care process. Űlger 2018 concluded that software‐based programmes and virtual reality are getting more support as part of a rehabilitation program; however, conventional methods remain important.

Christiansen 2018, in their RCT pilot of 38 participants with dysvascular amputations performed telephone sessions (with a physiotherapist) of a behaviour‐change intervention, based on social cognitive and control theories of behaviour change, encompassing self‐monitoring, action planning, tailored feedback, barrier and facilitator identification, problem‐solving promotion and encouragement. In this trial, participants in the intervention group increased their mean daily step count compared to control group, but there were no between‐group differences in change scores for any other physical function measure over time.

Authors' conclusions

Implications for practice.

This review is based on two small randomised controlled trials (RCTs) with a low number of participants. There is very low‐certainty evidence (one trial, 22 participants) that gait training protocols may improve mobility assessment and very low‐certainty evidence (one RCT, eight participants) there is no evidence of a difference in mobility assessments when using motor imaging in people with transtibial amputation due to diabetes or peripheral arterial disease. The effects were demonstrated after one week of the interventions, although there was no standardisation of exercises and interventions adopted. The studies reported some positive results, but provided insufficient evidence to support data, with small numbers of participants and limited data. Certainty of the evidence is very low due to the high risk of bias in the studies and the small number of participants.

Implications for research.

Additional high‐quality studies involving interventions such as virtual environments, proprioceptive neuromuscular facilitation and motor imaging for people with dysvascular amputations are needed. Future trials need to be sufficiently large to detect significant clinical outcomes, include all main clinical outcomes (prosthesis use, adverse events, mobility assessment, quality of life) and have uniform continuous data, using similar scales/scores (specially for mobility assessment and quality of life).

History

Protocol first published: Issue 8, 2020

Notes

Parts of the Methods section of the protocol were based on a standard template established by Cochrane Vascular.

Appendices

Appendix 1. Sources searched and search strategies

Source	Search strategy	Hits retrieved
Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web) (date of most recent search: 9 January 2023)	#1 Amput* OR "residua* limb*" OR (below adj3 knee*) OR disarticulat* OR exarticulat* OR postamputation OR BKA AND INREGISTER #2 "motor rehabilitation" OR Exercise OR Video OR "Physical Therapy " OR rehabilit* OR "computer generated simulation*" OR "virtual environment*" OR "motor imagery" AND INREGISTER #3 #1 AND #2	Sep 2020: 74 Oct 2021: 3 Jan 2023: 4
CENTRAL via CRSO (date of most recent search: 9 January 2023)	#1           MESH DESCRIPTOR Amputation EXPLODE ALL TREES                406 #2           MESH DESCRIPTOR Amputation Stumps EXPLODE ALL TREES    53 #3           MESH DESCRIPTOR Amputees EXPLODE ALL TREES         107 #4           ("residua* limb*"):TI,AB,KY                96 #5           (phantom adj6 limb*):TI,AB,KY                251 #6           amput*:TI,AB,KY              2839 #7           (below adj3 knee*):TI,AB,KY                697 #8           BKA:TI,AB,KY     9 #9           disarticulat*:TI,AB,KY     32 #10         exarticulat*:TI,AB,KY      0 #11         postamputation*:TI,AB,KY                34 #12         post‐amputation*:TI,AB,KY                48 #13         stump*:TI,AB,KY              617 #14         #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13          3835 #15         MESH DESCRIPTOR Exercise Therapy EXPLODE ALL TREES       13193 #16         MESH DESCRIPTOR Image Interpretation, Computer‐Assisted EXPLODE ALL TREES         7425 #17         MESH DESCRIPTOR Video Games EXPLODE ALL TREES          650 #18         MESH DESCRIPTOR Physical Therapy Modalities EXPLODE ALL TREES    24506 #19         MESH DESCRIPTOR Telerehabilitation EXPLODE ALL TREES                101 #20         ("strength train*" or "resistance train*" or "resistance exercise*" or "bodyweight train*"):TI,AB,KY              11718 #21         ("Proprioceptive Neuromuscular Facilitation"):TI,AB,KY                278 #22         ("motor imagery"):TI,AB,KY                432 #23         ("virtual environment*"):TI,AB,KY              400 #24         ("computer generated simulation*"):TI,AB,KY  3 #25         ("exercise therap*"):TI,AB,KY                11861 #26         (personal* adj4 "exercise program*"):TI,AB,KY      66 #27         rehabilit*:TI,AB,KY          47807 #28         #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27          80758 #29         #14 AND #28      404	Sep 2020: 404 Oct 2021: 113 Jan 2023: 39
ClinicalTrials.gov (date of most recent search: 9 January 2023)	Amputation OR amputee OR "residual limb" OR "below knee" OR disarticulat* OR exarticulat* OR postamputation OR BKA | "motor rehabilitation" OR Exercise OR Video OR "Physical Therapy " OR rehabilitate OR rehabilitation OR "computer generated simulation" OR "virtual environment" OR "motor imagery"	Sep 2020: 148 Oct 2021: 28  Jan 2023: 16
ICTRP Search Portal (date of most recent search: 9 January 2023)	Amputation OR amputee OR "residual limb" OR "below knee" OR disarticulat* OR exarticulat* OR postamputation OR BKA | "motor rehabilitation" OR Exercise OR Video OR "Physical Therapy " OR rehabilitate OR rehabilitation OR "computer generated simulation" OR "virtual environment" OR "motor imagery"	Oct 2021: 30 Jan 2023: 7
MEDLINE (Ovid MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE) 1946 to present (date of most recent search: 9 January 2023)	1 exp Amputation/ 2 exp Amputation Stumps/ 3 exp Amputees/ 4 "residua* limb*".ti,ab. 5 (phantom adj6 limb*).ti,ab. 6 amput*.ti,ab. 7 (below adj3 knee*).ti,ab. 8 BKA.ti,ab. 9 disarticulat*.ti,ab. 10 exarticulat*.ti,ab. 11 postamputation*.ti,ab. 12 post‐amputation*.ti,ab. 13 stump*.ti,ab. 14 or/1‐13 15 exp Exercise Therapy/ 16 exp Image Interpretation, Computer‐Assisted/ 17 exp Video Games/ 18 exp Physical Therapy Modalities/ 19 exp Telerehabilitation/ 20 ("strength train*" or "resistance train*" or "resistance exercise*" or "bodyweight train*").ti,ab. 21 "Proprioceptive Neuromuscular Facilitation".ti,ab. 22 "motor imagery".ti,ab. 23 "virtual environment*".ti,ab. 24 "computer generated simulation*".ti,ab. 25 "exercise therap*".ti,ab. 26 (personal* adj4 "exercise program*").ti,ab. 27 rehabilit*.ti,ab. 28 or/15‐27 29 14 and 28 30 randomized controlled trial.pt. 31 controlled clinical trial.pt. 32 randomized.ab. 33 placebo.ab. 34 drug therapy.fs. 35 randomly.ab. 36 trial.ab. 37 groups.ab. 38 or/30‐37 39 exp animals/ not humans.sh. 40 38 not 39 41 29 and 40	Sep 2020: 843 Oct 2021: 99 Jan 2023: 124
Embase via Ovid (Date of most recent search: 9 January 2023)	1 exp amputation/ 2 exp amputation stump/ 3 exp amputee/ 4 "residua* limb*".ti,ab. 5 (phantom adj6 limb*).ti,ab. 6 amput*.ti,ab. 7 (below adj3 knee*).ti,ab. 8 BKA.ti,ab. 9 disarticulat*.ti,ab. 10 exarticulat*.ti,ab. 11 postamputation*.ti,ab. 12 post‐amputation*.ti,ab. 13 stump*.ti,ab. 14 or/1‐13 15 exp kinesiotherapy/ 16 exp video game/ 17 exp physiotherapy/ 18 exp telerehabilitation/ 19 ("strength train*" or "resistance train*" or "resistance exercise*" or "bodyweight train*").ti,ab. 20 "Proprioceptive Neuromuscular Facilitation".ti,ab. 21 "motor imagery".ti,ab. 22 "virtual environment*".ti,ab. 23 "computer generated simulation*".ti,ab. 24 "exercise therap*".ti,ab. 25 (personal* adj4 "exercise program*").ti,ab. 26 rehabilit*.ti,ab. 27 or/15‐26 28 14 and 27 29 randomized controlled trial/ 30 controlled clinical trial/ 31 random$.ti,ab. 32 randomization/ 33 intermethod comparison/ 34 placebo.ti,ab. 35 (compare or compared or comparison).ti. 36 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 37 (open adj label).ti,ab. 38 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 39 double blind procedure/ 40 parallel group$1.ti,ab. 41 (crossover or cross over).ti,ab. 42 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 43 (assigned or allocated).ti,ab. 44 (controlled adj7 (study or design or trial)).ti,ab. 45 (volunteer or volunteers).ti,ab. 46 trial.ti. 47 or/29‐46 48 28 and 47	Sep 2020:  858 Oct 2021: 140 Jan 2023: 173
CINAHL via EBSCO (date of most recent search: 9 January 2023)	S45 S29 AND S44  S44 S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 OR S40 OR S41 OR S42 OR S43  S43 MH "Random Assignment"  S42 MH "Triple‐Blind Studies"  S41 MH "Double‐Blind Studies"  S40 MH "Single‐Blind Studies"  S39 MH "Crossover Design"  S38 MH "Factorial Design"  S37 MH "Placebos"  S36 MH "Clinical Trials"  S35 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study"  S34 TX crossover OR "cross‐over"  S33 AB placebo*  S32 TX random*  S31 TX trial*  S30 TX "latin square"  S29 S14 AND S28  S28 S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27  S27 TX rehabilit*  S26 TX personal* N4 "exercise program*"  S25 TX "exercise therap*"  S24 TX "computer generated simulation*"  S23 TX "virtual environment*"  S22 TX "motor imagery"  S21 TX "Proprioceptive Neuromuscular Facilitation"  S20 TX "strength train*" or "resistance train*" or "resistance exercise*" or "bodyweight train*"  S19 (MH "Telerehabilitation")  S18 (MH "Physical Therapy+")  S17 (MH "Video Games+")  S16 (MH "Image Interpretation, Computer Assisted+")  S15 (MH "Therapeutic Exercise+")  S14 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13  S13 TX stump*  S12 TX post‐amputation*  S11 TX postamputation*  S10 TX exarticulat*  S9 TX disarticulat*  S8 TX BKA  S7 TX below N3 knee*  S6 TX amput*  S5 TX phantom N6 limb*  S4 TX "residua* limb*"  S3 (MH "Amputees")  S2 (MH "Amputation Stumps")  S1 (MH "Amputation+")	Sep 2020: 1029 Oct 2021: 63 Jan 2023: 67
TOTAL before deduplication		Sep 2020: 3356 Oct 2021: 476 Jan 2023: 430
TOTAL after deduplication		Sep 2020: 2532 Oct 2021: 385 Jan 2023: 327

 

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Hyland 2009.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group Number of participants: 22 (impairment‐oriented group: 11; task‐oriented group: 11) Country: USA (3 centres) Setting: inpatients Start date: January 2005 Duration of participation: 10 days End date: May 2008 Method of randomisation: not described Exclusions after randomisation: 27 people were identified, but 2 qualified people did not consent, and 3 were discharged early (2 for medical reasons and 1 for insurance reasons) leaving 22 participants completing the study. Losses to follow‐up: 0 Intention‐to‐treat analysis: no Blinding: the first author was blinded to group allocation. Participants were not. Power calculation: to achieve 80% power for all outcome variables the sample size calculated was 54 (27 per group). Due to changes in healthcare and the limited number of people who met the inclusion criteria of the study, data collection was stopped after 22 participants in a period of 3.5 years. Power was recalculated for these data and was as follows: MNV and cadence 94%, symmetry of step length 76%, DLS 61%, stance and swing time 43%, AMPPRO and BBS 20%, and symmetry of single limb support 14%. Unit of allocation: participants Source of funding: none
Participants	Mean age: impairment‐oriented group: 65.2 years; task‐oriented group: 63.6 years Gender: impairment‐oriented group: 6 men, 5 women; task‐oriented group: 9 men, 2 women Inclusion criteria: unilateral transtibial amputation secondary to dysvascular causes, adult aged > 30 years, able to follow verbal directions, first‐time prosthetic user, suture line closure, inpatient rehabilitation, medically stable, able to tolerate wearing the prosthesis for 15 minutes Exclusion criteria: neurological co‐diagnosis, ≥ 10° loss of knee extension, absent sensation in the intact limb, lower extremity musculoskeletal surgery within the last 6 months
Interventions	Task‐oriented approach to gait training involved walking practiced as a whole task and not broken down into different gait components. However, if participant was unable to perform the functional task as a whole, they were evaluated at the impairment level to identify the neuromuscular component that was inhibiting successful completion of the task. Once functional task of walking on level surfaces was successful, both approaches add variety to the training sessions. Impairment‐oriented group Training protocol consisted of breaking down the functional activities into parts. Pregait training activities were practised standing (balance training, weight shifting in all planes, dynamic reaching, stepping up and down from a low step and single stepping). Practised ≥ 50% of their overall upright physiotherapy time. Remainder of treatment time with therapist was continuous corrective walking. For corrective walking, therapists gave verbal and manual cues while participants were walking to promote a symmetrical gait pattern using an appropriate assistive device as determined by training therapists. Task‐oriented group Performed functional tasks as a whole. Continuous corrective walking used for ≥ 90% of overall treatment time with the physiotherapist. For corrective walking, therapists gave verbal and manual cues while participants were walking to promote a symmetrical gait pattern using an appropriate assistive device as determined by training therapists. Duration of follow‐up: 10 days (not including non‐rehabilitation days on which the participant did not receive therapy). Cointerventions: physiotherapy programme for both protocol groups (individual time with the physiotherapist, a group session consisting of leg exercises, transfer training or ambulation) and occupational/recreational therapy.
Outcomes	Outcome measures relevant for this review (collected) BBS GAITRite Locomotor Capabilities Index Modified Locomotor Capability Index Reported outcomes NRS/VAS for pain All outcomes measured at days 3 and 10.
Notes	Declaration of interest: no potential conflicts of interest reported Submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Health Sciences Seton Hall University.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information about sequence generation.
Allocation concealment (selection bias)	Unclear risk	Insufficient information about allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Single blinded, participants not blinded. First study author blinded to group allocation.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The primary researcher, a physiotherapist blinded to group assignment, performed all testing.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No participants were excluded during the study.
Selective reporting (reporting bias)	Low risk	All outcomes planned were described to the 2 gait training approaches.
Other bias	Unclear risk	3 centres were utilised in this study with 9 therapists. The subacute setting delivered 30–60 minutes (mean 45 minutes) of therapy and the inpatient rehabilitation setting 60–90 minutes (mean 70 minutes) of therapy. There were no differences between the 2 settings for day 10.

Vanmairis 2018.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group Number of participants: 8 participants (intervention group: 4; control group: 4) Country: France Setting: not reported Start date: not reported Duration of participation: 2 weeks End date: not reported Method of randomisation: stratified randomisation of study population performed using age of participants as a stratification variable Exclusions after randomisation: not reported Losses to follow‐up: 0 Intention‐to‐treat analysis: not reported Blinding: examiners blinded; participants not blinded Power calculation: not reported Unit of allocation: participants Source of funding: not reported
Participants	Mean age: interventional group: 68.8 years; control group: 62.3 years Gender: interventional group: 3 men, 1 women; control group: 3 men; 1 women Inclusion criteria: unilateral transtibial amputees of arterial aetiology; aged > 18 years; given written consent for prestudy; not having severe cognitive impairments assessed by the MMSE 24/30; able to walk 10 m with or without walking aid in order to carry out the initial assessment. Exclusion criteria: bilateral amputation or transfemoral amputation; aetiology other than vascular (traumatic, congenital or cancerous); presence of phantom sensations or pain during the imagination of tasks, > 2/10 on the numerical scale; residual limb pain or sores; neurological (cognitive or motor) disorders; uncorrected visual disorders; vestibular disorders; orthopaedic disorders; alcohol‐addiction or addiction confirmed
Interventions	MI + physical practice of walking (interventional) 20‐minute MI sessions conducted 5 times a week for 2 weeks, individually. The practice sessions of MI were conducted, indiscriminately, before or after the physical practice of walking because of the constraint of participant availability. Participant was seated in a quiet room, had to keep his eyes closed throughout the session and the instructions were prerecorded on a mobile phone. The participant had to wear their prosthesis during MI and had to imagine themselves walking with it. During the session, a metronome was used as an auditory stimulus to improve performance. The metronome gave the walking rate and the participant had to imagine taking a step at each sound in order to ensure the isochrony, that is, the adequacy between the duration of the mental visualisation and the physical execution time of an action. Then the metronome was switched off and participants were asked to continue the imagination of walking at the same rate. Physical practice of walking (control group) The physical practice of walking (5 × 90‐minute sessions per week for 2 weeks) was carried out in a gymnasium and was separated into 2 × 45‐minute sessions: 1 session in the morning and 1 in the afternoon. The session involved the participant walking alone between the parallel bars for 15 minutes. Then the participant repeated the next sequence outside the parallel bars under supervision of a physiotherapist: they would get up from a chair, walk 25 m (gymnasium length) with their walking aid, turn around and come back to sit. This sequence was repeated as many times as possible by the participant for 30 minutes, pausing between each exercise if needed. The physiotherapists gave verbal indications to the participant as they walked to promote a more symmetrical walking pattern. Participants used the appropriate assistance device (walker, canes) as determined by the physiotherapists during the walk. Duration of follow‐up: 2 weeks Cointerventions: 8 participants in the study responded to the MIQ‐RS, which assessed their MI abilities. Questionnaire was supplemented by 12 healthy volunteers of similar age to the study population to assess the feasibility of the study. It was applied in a quiet room, in the presence of an examiner, under standardised condition.
Outcomes	Outcomes relevant for this review (collected) Gait parameters (assessed by laboratory kinetic analysis). Participant walked at comfortable speed on a blue strip 10 m long and 60 cm wide. Test was repeated 3 times without setting a recovery time; each participant had to feel ready to repeat the test. Best of 3 times was used for the study (walking speed, step length symmetry, symmetry of the amount of support). Risk of falling (assessed using Timed Up‐and‐Go test), which consisted of getting up from a chair (40 cm high seating with 2 armrests), travelling 3 m with usual walking aids, making a U‐turn and returning to sit. The test was repeated 3 times without setting a time. Reported outcomes Feasibility of using MI in amputee patients (MIQ‐RS). Clinical data collected throughout the 2‐week rehabilitation programme.
Notes	Declaration of interest: study authors stated that they have no ties of interest.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Age of participants was used as a stratification variable, a criterion that could represent a possible confounding factor.
Allocation concealment (selection bias)	Unclear risk	Not reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Single blinded, participants not blinded. First study author blinded to group allocation.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Examiner blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No participants were excluded during the study.
Selective reporting (reporting bias)	Low risk	All outcomes planned were described.
Other bias	Unclear risk	No evidence of other bias.

AMPPRO: Amputee Mobility Predictor with PROsthesis; BBS: Berg Balance Scale; DLS: double limb support; MI: motor imaging; MIQ‐RS: Movement Imagery Questionnaire – Revised Second version; MMSE: Mini‐Mental State Examination; MNV: mean normalised velocity; NRS: Numeric Rating Scale; VAS: Visual Analogue Scale.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Anderson 2021	Secondary analysis of data from 2 RCTs examining health behaviour interventions for dysvascular amputations.
Barnett 2009	Causes of the amputations were vascular and non‐vascular.
Bourque 2019	Study protocol for a randomised sham‐control clinical trial. Trial designed to test the effects of CBPT intervention on balance confidence and functional mobility in lower limb prosthesis users.
Chin 2016	Retrospective study.
Christiansen 2020	Not an RCT. Cross‐over study.
Durand 2020	Did not evaluate an intervention for motor rehabilitation (evaluated the value of quantifying the gait quality indices during the 2‐minute walking test of people with transtibial amputation).
Erbahceci 2001	All participants were amputated because of trauma.
Gailey 2020	Causes of amputations were dysvascular and trauma.
Godlwana 2020	Levels of amputation were transfemoral and transtibial.
Igor 2013	Causes of amputation were diabetic gangrene, vascular aetiology and trauma. The levels of amputation were transtibial and transfemoral.
Iman 2017	Causes of amputation were vascular, trauma and cancer. Levels of amputation were transfemoral and transtibial.
Johannesson 2010	Not an RCT. Prospective cohort study.
Lass 2013	Did not evaluate an intervention for motor rehabilitation (evaluated a flexible device prosthesis).
Liedberg 1983	Did not evaluate an intervention for motor rehabilitation (evaluated a type of provisory prosthesis).
MacLean 1994	Not an RCT. Participants were alternatively assigned. 3 participants were included in the study twice due to amputation on the other leg (our unit of analysis was the individual participant).
Mazari 2010	Treatment group with amputations due to neoplasia, venous ulcers, neuropathic painful foot, ischaemia, infected diabetic foot.
NCT03149432	Evaluated the effect of mirror therapy (not an intervention for motor rehabilitation) on phantom pain following amputation of lower limb. Intervention: rehabilitation + gabapentin + mirror therapy during the first 4 weeks. Control: rehabilitation + gabapentin.
NCT03930199	Participants were young, combat‐injured service members, not an RCT. Amputation at transtibial and transfemoral level.
NCT04086069	Study withdrawn (main clinician cancelled collaboration for personal reasons).
NCT04120038	Not an RCT. Single group assignment.
NCT04431817	Not an RCT. Prospective case‐control study.
Saimpont 2021	Did not evaluate an intervention for motor rehabilitation. Investigated the effects of body position and actual execution on the temporal characteristics of motor imaging in amputees, comparing with healthy control participants.
Schack 2021	Not an RCT. Cross‐sectional and part of an observational study examining cortical activity during prosthetic ambulation.
Schafer 2021	Causes of amputation were vascular, trauma and other. Levels of amputation were transfemoral and transtibial.
Scott 2017	Not an RCT or pseudo‐RCT. Mixed methods, retrospective cross‐sectional analysis of data from a national audit conducted by the Scottish Physiotherapy Amputee Research Group.
Silva 2021	Not an RCT. Quasi‐experimental study. 1 group was composed of elderly people with transtibial amputation and the other group by elderly people with transfemoral amputation.
Silva‐Filho 2019	Did not evaluate an intervention for motor rehabilitation (evaluated only analgesic effects of auriculotherapy).
Torbjörnsson 2020	Not an RCT. Prospective cohort study.
Tousignant 2015	The sample consisted of 3 lower limb amputees of which 1 amputation was due to diabetes/trauma (not specified).
Traballesi 2012	Did not evaluate an intervention for motor rehabilitation (evaluated a type of provisory prosthesis).
Woodburn 2004	Did not evaluate an intervention for motor rehabilitation (evaluated a type of stump dressing).
Younesian 2020	Did not evaluate an intervention for motor rehabilitation (evaluated the spatiotemporal characteristics of the 6‐minute walking test in amputees).

CBPT: cognitive behavioural‐physical therapy; RCT: randomised controlled trial.

Characteristics of studies awaiting classification [ordered by study ID]

Anjum 2016.

Methods	Randomised controlled trial
Participants	Unilateral transtibial amputation. PNF group: 31 participants; traditional strength training group: 32 participants
Interventions	Traditional strength training group: 30‐minute session of weight bearing, weight shifting, balance exercise, single limb loading, stepping and strength training using sandbag. PNF group: 30‐minute session of weight bearing, weight shifting, balance exercise, single limb loading and stepping. The PNF principals (e.g. manual contact; verbal command; vision; and timing for emphasis, resistance, approximation, stretch, slow reversal and rhythmic stabilisation) were applied.
Outcomes	Gait parameters including step width, step length, cadence, stride length and the Locomotor Capabilities Index (LCI)
Notes	Causes of amputations not described. Additional information requested from study authors, but received no reply.

Chapman 2011.

Methods	Multiple baseline single subject design randomly assigned
Participants	3 participants with new unilateral transtibial or transfemoral amputations
Interventions	Participants randomly assigned to a variable baseline period followed by an intervention period, with 24 data points collected for each participant. Intervention consisted of 30‐minute daily training on the NWF balance platform in addition to the standard lower extremity amputee rehabilitation programme. Used a battery of standardised outcome measures designed to assess strength, balance and ambulation. NWF activities graded from simple to complex as participants' abilities improved.
Outcomes	Standardised outcome measures to assess strength, balance and ambulation such as 6‐minute walk test and L‐Test.
Notes	Amputation cause not described. Additional information requested from study authors, but received no reply.

Gauthier‐Gagnon 1986.

Methods	Randomised controlled trial
Participants	Elderly people with below‐knee amputations. Intervention group: 5 participants; control group: 6 participants
Interventions	Intervention group used audio feedback from a Limb Load Monitor during the standing balance exercises. Control group used mirrors and verbal feedback
Outcomes	Sway and weight‐bearing patterns evaluated using a force plate, prior to and following 8 non‐consecutive days of treatment.
Notes	Cause of amputations not described. Additional information requested from study authors, but received no reply.

Kirdi 2017.

Methods	Randomised controlled trial
Participants	10 participants randomised
Interventions	Intervention: mirror visual feedback + classical prosthetic rehabilitation programme Control: classical prosthetic rehabilitation program consisting of weight transfer and balance exercises
Outcomes	Amputee weight bearing, 1‐leg stance test and functional reach test before and after treatment
Notes	Causes of amputations not described. Additional information requested from study authors, but received no reply.

Lamberg 2012.

Methods	Pseudo‐randomised, controlled study
Participants	8 participants: intervention group: 4; control group: 4
Interventions	Intervention: 12 × 30‐minute training sessions over 4 weeks, starting training with 30% of bodyweight removed. The amount of weight unloaded was progressively reduced and the speed increased over the duration of the training. Control: participants wore the harness, without weight removed, to mimic the environment of the intervention group and for safety.
Outcomes	6‐minute walk test; Timed Up‐and‐Go test; walking speed and step length variability
Notes	Causes and levels of amputations not described. Additional information requested from study authors, but received no reply.

NCT02328859.

Methods	Randomised controlled trial
Participants	Actual enrolment: 4 participants Total expected: 20 participants
Interventions	Intervention: participants with transtibial amputation on 6 weeks of training with the VGRS Control: participants with transtibial amputation on 6 weeks of training using a conventional treadmill protocol
Outcomes	Rate of improvement over the course of each training programme; gait analysis, postural sway and standard balance and mobility instruments including the 6‐Minute Walk Test, Four Square Step Test, the Berg Balance Scale and the Dynamic Gait Index to assess improvements in gait and mobility.
Notes	Causes of amputations not described. Additional information requested from study authors, but received no reply.

NCT03872193.

Methods	Randomised controlled trial
Participants	20 participants with transtibial amputation
Interventions	Intervention: virtual reality applications plus standard physiotherapy Control: standard physiotherapy alone
Outcomes	Quality of life: SF‐36 scale; balance (Single Leg Balance Test); performance (6‐Minute Walk Test); prosthesis satisfaction (Trinity Amputation Prothesis Experience Scale); gait parameters (Wearable Biometric Device)
Notes	Causes of amputations not described. Additional information requested from study authors, but received no reply.

Shu 2014.

Methods	Randomised controlled trial
Participants	40 participants with unilateral amputation randomly divided into 2 groups
Interventions	Intervention: functional feedback therapy (using Smartstep feedback training method) Control: trained only by the verbal guidance method
Outcomes	Gait temporal‐spatial parameters, weight‐bearing capacity
Notes	Causes and levels of amputations not described. Additional information requested from study authors, but received no reply.

Tao 2019.

Methods	Multisite, parallel, evaluator‐blind randomised controlled trial
Participants	Community‐dwelling adults aged > 50 years with unilateral transtibial or transfemoral amputation
Interventions	Intervention group (38 participants) used modified Wii Fit Balance Board activities Control group (33 participants) used Big Brain Academy: Wii Degree
Outcomes	Primary outcome measure: 2‐Minute Walking Test Secondary outcome measure: activities‐specific balance confidence scale
Notes	Causes of amputations not described; participants with transtibial or transfemoral amputation. Additional information requested from study authors, but received no reply.

Topuz 2013.

Methods	Randomised trial
Participants	20 participants with unilateral below‐knee amputations
Interventions	Intervention: complex decongestive physiotherapy (10 participants) Control: classical elastic bandage application (10 participants)
Outcomes	Stump circumferential measurement; transition period to permanent prosthesis
Notes	Causes of amputations not described. Additional information requested from study authors, but received no reply.

LCI: Locomotor Capability Index; NWF: Nintendo Wii Fit; PNF: proprioceptive neuromuscular facilitation; SF‐36: 36‐item Short‐Form; VGRS: virtual gait retraining system.

Characteristics of ongoing studies [ordered by study ID]

NCT02761447.

Study name	Effectivity of motor imagery and mirror therapy in amputees
Methods	RCT
Participants	Amputees with pain in the stump area for ≥ 3 months, no intermittent claudication
Interventions	Intervention group: traditional and imaginary motor programme Control group: traditional and mirror therapy
Outcomes	Primary outcome: changes from baseline NRS at 1 and 3 months Secondary outcome: MPQ, Vividness of Visual Imagery Questionnaire, Beck Depression Inventory, Neuropathic Pain Questionnaire (DN4), pressure algometer (kg/cm2), West‐Haven‐Yale Multidimensional Pain Inventory, Barthel Index of ADLs, Instrumental Activities of Daily Living Scale, Functional Independence Measure Scale, Quality of Life Questionnaire scoring (SF‐36), Multidimensional Body Self Relations Questionnaire, Tampa Scale for Kinesiophobia, Pain Catastrophizing Scale, Pain Management Inventory
Starting date	4 May 2016
Contact information	Adelaida María Castro‐Sánchez, Universidad de Almeria, Almería, Spain
Notes

NCT03995238.

Study name	Optimizing gait rehabilitation for veterans with non‐traumatic lower limb amputation (GEM)
Methods	RCT
Participants	Estimated enrolment: 54 participants with transtibial amputation due to PAD or diabetes
Interventions	3 groups: Error‐augmentation gait training Error‐correction gait training Supervised walking
Outcomes	Primary outcome: step length symmetry Secondary outcome measures: 6‐Minute Walk Test, free‐living daily step count, World Health Organization Disability Assessment Schedule 2.0 Other outcomes: transcutaneous oximetry, Socket Comfort Score
Starting date	9 October 2019
Contact information	Cory L Christiansen, PhD; Cory.Christiansen@va.gov Noel L So, MD
Notes

NCT04083456.

Study name	Dysvascular amputation self‐management of health (DASH)
Methods	RCT
Participants	Estimated enrolment: 64 participants
Interventions	Intervention group: walking biobehavioural intervention including self‐monitoring (use of the FitBit sensor and software reviewed with the interventionist at each session), tailored feedback (from the FitBit sensor as well as feedback from the interventionist), identification of barriers and facilitators, problem solving, action planning and encouragement + conventional outpatient training. Control group: receive the same computer tablets with tele‐health software as the intervention group + conventional outpatient training
Outcomes	Primary outcome: physical activity (step count with ActivPAL activity monitor) Secondary outcome measures: participant‐reported outcomes measurement information system; PROMIS: Self‐Efficacy for Managing Symptoms; PROMIS: Self‐Efficacy for Managing Daily Activities; PROMIS: Ability to Participate in Social Roles and Activities; Prosthesis Evaluation Questionnaire – Mobility Score; Timed Up‐and‐Go test; World Health Organization Disability Assessment Schedule 2.0; Self‐Efficacy for Managing Chronic Disease. Other outcome measures: Activities‐Specific Balance Confidence Scale; 2‐Minute Walk Test; 5‐Metre Walk Test; Study Intervention Reach; Outpatient Practitioner Intervention Fidelity; Biobehavioral Practitioner Intervention Fidelity; FitBit Use Adherence
Starting date	10 September 2019
Contact information	Emily R Hager, MS; emily.hager@ucdenver.edu John Eun, MD; john.eun@cuanschutz.edu
Notes	Included transtibial, knee disarticulation or transfemoral lower‐limb amputation

NCT04750876.

Study name	Early evaluation of the introduction of pre‐ and post‐operative psychological and physiotherapeutic follow‐up in vascular surgery in amputee patients (EPKIVA)
Methods	RCT
Participants	Estimated enrolment 40 participants with transtibial or transfemoral amputation related to obliterative arteriopathy of the lower limbs
Interventions	Intervention group: consultations with a psychologist; participants will also be cared for on a daily basis by a physiotherapist; usual care of amputee patients Conventional support group: usual care of amputee patients in the vascular surgery department. They will be cared for by the surgeon, nurses and nursing auxiliaries, who will provide individualised care (nursing care, mobilisation, pain management, technical care, etc.), listening and support.
Outcomes	Primary outcome: mobility and emotional reactions scores using Nottingham Health Profile scale at 10 days postoperatively
Starting date	11 February 2021
Contact information	Chloé Moreau: 0251446327; chloe.moreau@chd‐vendee.fr
Notes

NCT04968691.

Study name	Telerehabilitation in individuals with unilateral transtibial amputation
Methods	RCT
Participants	Estimated enrolment: 40 participants with unilateral transtibial amputation, using active vacuum system prosthesis and carbon foot for ≥ 1 year, being at the level of mobility level K2‐K3
Interventions	Intervention group: structured exercise programme supported by telerehabilitation applied 3 days a week + home exercise programme on remaining days of the week. Participants shown how to complete the exercises online by the researcher before each session, 3 sessions a week for 6 weeks, with mobile telecommunication applications, and it will be ensured that the participants perform the exercises correctly and are followed up. At beginning of each week, participants will receive videos with the exercise content of that week. Control group: home exercise programme alone. Exercise cards given and participants asked to complete the exercise treatment at home as 3 sessions a week for 6 weeks.
Outcomes	Primary outcome measures: Timed Up‐and‐Go test; 30‐Second Sit to Stand Test Secondary outcome measures: Trinity Amputation and Prosthesis Experiences Scales; Activities‐specific Balance Confidence Scale Other outcome measures: Amputee Body Image Scale (ABIS); Nottingham Health Profile
Starting date	1 August 2021
Contact information	Mehmet Kurtaran, Istanbul University‐Cerrahpasa, Istanbul, Turkey
Notes

NCT05095805.

Study name	Weight‐bearing exercise for trans‐tibial amputees (Prototest)
Methods	Interventional randomised clinical trial
Participants	20 participants
Interventions	Intervention: additional weight‐bearing exercise with Nintendo Wii Balance Board, real‐time visual feedback, weight‐bearing assessment by centre of pressure Control: standard weight‐bearing exercise conventional training for fitting of initial prosthesis
Outcomes	Primary outcome measures: change in centre of pressure at baseline and 2 weeks. Calculation of centre of pressure (transversal plane, Symmetry Index, lateral, sagittal directions) from 4 pressure sensors built into the balance board; change in weight distribution in quiet standing at baseline and 2 weeks. centre of pressure assessment for 20 seconds of quiet standing Secondary outcome measures: change in 10‐m Walk Test at baseline and 2 weeks; change in L‐test at baseline and 2 weeks
Starting date	27 October 2021
Contact information	Helena Burger, MD, PhD; University Rehabilitation Institute, Republic of Slovenia
Notes

NCT05656924.

Study name	Deploying digital prosthetic interface technology and exercise in dysvascular amputees
Methods	Interventional randomised clinical trial
Participants	38 participants
Interventions	Intervention: digital prosthetic interface technology group. Participants will use the digital prosthetic interface technology developed by Bionic Skins, a socket and liner that are manufactured using a novel technique meant to achieve a better distribution of pressure on the residuum. Study based on the hypothesis that an improvement in pressure distribution on the residuum will enable high adherence to an exercise programme meant to improve health status in people with dysvascular amputations. No intervention: traditional socket and liner technology group. Participants will use a traditional socket‐liner technology (i.e. their own liner and socket system).
Outcomes	Primary outcome measures: difference between groups in intrasocket interface pressure at baseline (pressure sensors inserted in the socket will be used to measure the pressure on the residuum when study participants are weight bearing). The pressure is measured in kilopascals (kPa). 1 kPa is approximately the pressure exerted by a 10‐g mass resting on a 1 cm2 area; difference between groups in step counts at 12 months. Mean number of steps over 2 weeks will be provided by a wearable sensor (Oura ring). Data will be expressed in number of steps per day. Secondary outcome measures: difference in thermal imaging measures of the temperature of the residuum in the 2 groups at baseline (thermal camera will be used to measure the peak temperature of the residuum. The residuum temperature will be measured in degrees Celsius); difference in Socket Evaluation Questionnaire scores in the 2 groups at baseline (questionnaire consists of a battery of questions inquiring about participant's comfort using a socket and uses a Visual Analogue Scale format ranging from 0 (worst) to 100 (best)); difference in Prosthesis Evaluation Questionnaire scores in the 2 groups at baseline (questionnaire consists of a battery of questions inquiring about participant's perception of his/her prosthesis and uses a Visual Analogue Scale format ranging from 0 (worst) to 100 (best)).
Starting date	19 December 2022
Contact information	Paolo Bonato, PhD; Spaulding Rehabilitation Hospital Boston, Massachusetts, US; 02129617‐952‐6319; pbonato@mgh.harvard.edu
Notes

RBR‐4s5nkh.

Study name	Assessment of gait biomechanics and quality of life of individuals with unilateral lower limb amputation
Methods	RCT
Participants	Intervention group: 8 participants; control group: 8 participants; health group: 10 participants
Interventions	Preprosthetic phase (24 sessions, 2 times a week, with 1‐hour sessions): scar mobilisation (1 minute); cicatricial desensitisation with soft sponge (1 minute); desensitisation of stump with rough sponge (1 minute); stump desensitisation with cryotherapy (ice) (1 minute); stretching of the flexor, extensor, adductor and abductor muscle chains (2 sets of 10 repetitions each); muscle strengthening of flexors, extensors, adductors and abductors, abdomen and buttocks (mini‐band exercises, ball (2 sets of 10 repetitions) and Russian electrical current stimulation) (20 minutes); discharge of weight in amputated limb (individuals standing in parallel bar for 5 minutes). Phase postprosthesis (8 sessions, 2 times a week with 1‐hour sessions): weight loss training under prosthesis (weight transfer work from 1 limb to the other); parallel bar gait training (previous gait, back, side and obstacle gait); idle training (outside the parallel bar); trampoline balance training; fast circuit training of changes of direction with cones; stairs training (up and down stairs); outboard training (out of hospital predetermined course).
Outcomes	Primary outcome: gait biomechanics (kinetics, kinematics and force platform) Secondary outcomes: fall risk assessment method, test execution time, neuropathic pain, quality of life (WHOQOL).
Starting date	10 June 2018
Contact information	Leticia Vargas Almeida; (55)1935217000; email: levargasdealmeida@gmail.com
Notes

ADL: activities of daily living; DN4: Douleur Neuropathique en 4 Questions; MPQ: McGiII Pain Questionnaire; NRS: Numerical Rating Scale; PAD: peripheral arterial disease; PROMIS: Patient Reported Outcomes Measurement Information System; RCT: randomised controlled trial; SF‐36: 36‐item Short Form; WHOQOL: World Health Organization Quality of Life.

Differences between protocol and review

In Types of participants, we clarified that participants should be aged 18 years or over.

We have simplified and clarified the interventions we wish to study in the Types of interventions section.

In Types of outcome measures, we added an additional example of possible adverse events (i.e. stump pain) and listed additional validated measures of gait and mobility to those listed in the protocol.

Instead of using a fixed‐effect model in analyses with no substantial heterogeneity (I² of 75% or less), we planned to use a random‐effects model to synthesise the data because of the complexity and differences in the interventions for motor rehabilitation.

Contributions of authors

LBA: guarantor of the review, protocol drafting, acquiring trials reports, trial selection, data extraction, data analysis, data interpretation, review drafting and future review updates.

CF: protocol drafting, acquiring trials reports, trial selection, data extraction, data analysis, data interpretation, review drafting and future review updates.

VT: data analysis, data interpretation, review drafting.

FMJ: data analysis, data interpretation, review drafting.

Sources of support

Internal sources

• CAPES, Brazil

  Foundation linked to the Education Ministry of Brazil 

External sources

• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK

  The Cochrane Vascular editorial base is supported by the Chief Scientist Office.

Declarations of interest

LBA: none.

CF: none. CF reports working as a Professor of Vascular Surgery, São Camilo University Centre, São Paulo.

VT: none.

FMJ: none.

New