Table 1 Overview of each paper included in the review. Sensor locations: A = anatomical, P = prosthesis, L = liner.

Arch (2017) [30]

Method to Quantify Cadence Variability of Individuals with Lower-Limb Amputation

27 participants

(10 BK and 1AK at K2, 10 BK and 6 AK at K3)

Fitbit One

(P. Ankle)

Activity level – steps, cadence

(7 days)

Develop a method of quantifying real-world cadence variability.

This method of quantifying cadence variability can differentiate between K2 and K3 groups.

Arch (2018) [31]

Step count accuracy of StepWatch and FitBit One among individuals with a unilateral transtibial amputation

50 participants

(All BK)

Fitbit One, StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Compare step count accuracy of Fitbit One and StepWatch monitors during tasks and free-living.

Both monitors were accurate during forward-linear walking, but the StepWatch was more accurate during the Four Square Step Test and the Figure-of-8 Walk Test. The Fitbit counted fewer steps in free-living.

Belter (2014) [32]

Grasp force based taxonomy of split-hook prosthetic terminal devices

1 participant

(BE)

Wide angle camera (Forehead)

Type of grasp and ability to exert force on the environment - prosthesis only

(5 h)

Develop a taxonomy of split-hook grasping and force exertion during real life daily activities at home.

When using a split hook some types of grasp needed the help of the intact hand to pre-position the object. The ability to firmly grasp and hold objects was limited in both voluntary opening and closing hooks.

Chadwell (2018) [33]

Visualisation of upper limb activity using spirals - A new approach to the assessment of daily prosthesis usage

4 participants

(2 BE, 2 Anatomically Intact)

2x ActiGraph

GT3X+

(A. and P. Wrists for prosthesis users, both A. Wrists for anatomically intact controls)

Symmetry of non-specified arm movements - both arms

Self-reported prosthesis wear time

(7 days)

Propose a new method to analyse and visualise upper-limb activity data.

New approaches visualised the distribution of arm movements between the prosthetic and intact limb. A time series data spiral clearly illustrated arm activity over extended periods. The prosthesis users heavily relied on their intact limb.

Coleman (1999) [34]

Step activity monitor: Long-term, continuous recording of ambulatory function

3 participants

(2 BK, 1 Anatomically Intact with heart failure)

Step Activity Monitor – SAM (later ‘StepWatch’) (A. Ankle)

Activity level – steps

(2 weeks)

Provide guidelines and examples of using the Step Activity Monitor (SAM), and results of accuracy and reliability testing.

SAM is accurate, reliable, and can be used to perform long-term step monitoring on a range of subjects. It can quantify differences in ambulatory activity resulting from medical interventions and changes in health status.

Frossard (2008) [35]

Monitoring of the load regime applied on the osseointegrated fixation of a trans-femoral amputee: A tool for evidence-based practice

1 participant

(AK using osseo-integration)

Transducer Model 45E15A

(P. Between residuum and knee)

Loading and activity levels

(5 h)

Describe the continuous recording of the true load regime experienced during daily living by the abutment of a person with a transfemoral amputation fitted with an osseo-integrated fixation.

The participant varied in lengths of time spent active, and averaged 64% of time spent inactive and 36% active, with an average 8 steps/min. Maximum load on the mediolateral, anteroposterior and long axes represented 21, 21 and 120% of the body weight, respectively.

Frossard (2010) [36]

Categorisation of activities of daily living of lower limb amputees during short term use of a portable kinetic recording system: a preliminary study.

1 participant

(AK using osseo-integration)

Transducer Model 45E15A

(P. Between residuum and knee)

Loading and activity levels

(5 h)

Determine the relevance of the categorisation of the load data to assess the functional output and usage of the prosthesis.

Walking, localised movement, and standing occurred 44, 34 and 22% of recording time, respectively. The maximum forces on the mediolateral, anteroposterior and long axes were highest when walking and lowest when standing.

Hornero (2013) [37]

Bioimpedance system for monitoring muscle and cardiovascular activity in the stump of lower-limb amputees

5 participants

(All AK)

Electrical impedance myography, electrical impedance plethys-mography

(A. Residuum)

Breathing, muscle activity on the residual limb, heart-rate

(Not provided)

Continuously monitor as many non-invasive physiological parameters as possible from people with lower-limb amputations using a single bioimpedance electrode configuration system.

The system monitors muscle activity, heart rate and breathing rate. The shape and amplitude of the changes in the electrical impedance myography signal correlated with the type of gait and the force exerted by the muscle.

Jayaraman (2014) [38]

Global Position Sensing and Step Activity as Outcome Measures of Community Mobility and Social Interaction for an Individual With a Transfemoral Amputation Due to Dysvascular Disease

1 participant

(AK)

StepWatch 3.1,

(P.)

QStarz GPS

(A. pocket or purse)

Steps, distance, speed, transport

(1 month)

Objectively quantify community mobility and social interaction in an individual with AK amputation using a step activity monitor and GPS monitoring device over 1 month.

The method sensitively quantified community mobility and social activities, counting steps and recording the reasons the steps were taken, for insight into the participant’s everyday prosthesis use.

Sanders (2018) [39]

A Novel Method for Assessing Prosthesis Use and Accommodation Practices of People with Transtibial Amputation

21 participants

(All BK)

WAFER –inductive sensor to find distance between liner and socket

(P. In socket)

Donning and doffing

(8–13 days)

Present a novel method for assessing prosthesis use and accommodation practices of people with transtibial amputation.

The WAFER showed good ability to detect donning and doffing but it had limited battery life and issues with alignment between the sensor on the socket and the target on the liner.

Shawen (2017) [40]

Fall Detection in Individuals With Lower Limb Amputations Using Mobile Phones: Machine Learning Enhances Robustness for Real-World Applications

17 participants

(7 AK, 10 Anatomically Intact)

Samsung Galaxy S4 accelerometer and gyroscope

(A. Waist, pocket or in hand)

Fall events

(2 days)

Develop a fall-detection classifier that is robust to population, phone location and environmental sources of error, and detects falls for individuals with and without amputations.

The fall-detection classifier, trained using data from healthy anatomically intact individuals, was able to reliably separate falls from daily activities in individuals with AK amputation.

Spiers (2017) [41]

Analyzing at-home prosthesis use in unilateral upper-limb amputees to inform treatment & device design

3 participants (2 BE, 1AE)

GoPro Hero 3+ (forehead)

Type of grasp and grasp count - both arms

(Up to 4 h)

Develop a taxonomy of manipulation suitable for use with unilateral upper-limb prosthesis users and demonstrate its use.

The taxonomy was applicable to the data from all 3 participants. Findings showed that intact hand use dominated across the 3 very different prosthesis users.

Stam (1995) [42]

A device for long-term ambulatory monitoring in trans-tibial amputees

1 participant

(BK)

Continuous

Ambulatory Monitoring of Prosthetic walking (CAMP) accelerometer

(P. Shank)

Steps, walking periods

(5 days)

Develop an actimeter that is lightweight, easily attaches to the prosthesis, and has energy and memory capacity for 5 days for practical clinical use.

The CAMP is lightweight and measures walking time of prosthesis users for up to 5 days. Further work will focus on decreasing size and weight, and increasing memory capacity.

Albert (2013) [43]

Monitoring Functional Capability of Individuals with Lower Limb Amputations Using Mobile Phones

18 participants

(10 AK, 8 Anatomically Intact)

Android phone

(A. Lumbar spine)

Activity level – thresholds of acceleration,

(7 days)

Provide evidence that accelerometry, using mobile phones, can objectively quantify the activity levels of lower-limb prosthesis users.

K1 and K2 subjects are less active than the control subjects. K-levels co-vary with high level activity.

Albert (2014) [44]

Monitoring Daily Function in Persons With Transfemoral Amputations Using a Commercial Activity Monitor: A Feasibility Study

9 participants

(All AK)

Fitbit One

(A. Wrist)

Activity level – steps, distance

(7 days)

Observe the relationship between the measured Fitbit activity levels and K-level classifications.

The percentage of movement time had a predictable relationship to the designated K-level. Activity level, measured outside the clinic, may lend support for K-level classifications.

Balkman (2019) [45]

Prosthetists’ perceptions of information obtained from a lower limb prosthesis monitoring system: a pilot study.

3 participants

(All AK)

WAFER

(P. In socket, distal end)

2x ActiGraph GT3X+

(L. proximal to patella, P. lateral side of ankle)

Donning and doffing, sitting, standing, walking

(2 weeks)

Compare prosthetist-estimated patient activity with the prosthesis use and activity recorded by the sensors.

Prosthetists over- and under-estimated patient activity, relative to sensor data. Prosthetists found features of each presented survey tool to be clinically useful. Prosthesis-mounted monitors may provide prosthetists with improved understanding of their patients’ prosthesis use.

Chadwell (2018) [46]

Upper limb activity in myoelectric prosthesis users is biased towards the intact limb and appears unrelated to goal-directed task performance

40 participants (20 BE, 20 Anatomically Intact)

2x ActiGraphs from the GT3X range

(A. and P. Wrists for prosthesis users, both A. Wrists for anatomically intact controls)

Symmetry of non-specified arm movements - both arms, self-reported and sensor calculate prosthesis wear time

(7 days)

Report the activity of transradial myoelectric prosthesis users, and anatomically intact participants. Assess the extent to which self-report captures everyday patterns of activity, and whether kinematic and gaze-related measures of performance correlate with activity.

Prosthesis users relied heavily on their intact arm during everyday life, while intact adults demonstrated similar reliance on both arms. There was no correlation between the amount of time a prosthesis was worn and reliance on the intact limb, or between these measures and the kinematic and gaze-related measures of performance.

Cuberovic (2019) [47]

Learning of artificial sensation through long-term home use of a sensory-enabled prosthesis

1 participant (BE with pre-implanted 8 channel flat interface nerve electrodes (FINEs))

Aperture sensor,

(P. base of thumb)

3x Flexiforce pressure sensors

(P. thumb, index, and middle finger)

Grasp count – prosthesis only

Hours of sensory feedback

(115 days)

Study the effects of artificial sensory stimulation on perceptions related to the stimulation and functional outcomes, and assess correlation between use time and frequency of grasp use with clinical measures.

During the study, perception of sensation location and quality improved over time. As did prosthesis embodiment, confidence, and other psychosocial measures.

Prosthesis usage did not increase, but was higher than in a previous study without sensory feedback.

Desveaux (2016) [48]

Physical Activity in Adults with Diabetes Following Prosthetic Rehabilitation

15 of 22 participants completed follow-up

(All BK)

StepWatch

(A. Ankle)

Activity level – steps

(5–9 days)

Determine whether individuals with a BK amputation and diabetes meet guidelines for activity intensity and daily step counts, and whether clinical measures of physical function are associated with objective measures of physical activity.

Physical activity levels remained stable after discharge from rehabilitation but fell below recommended guidelines of steps per day and minutes of moderate to vigorous activity per week. Average activity levels correlated with clinical measures.

Godfrey (2018) [49]

The Accuracy and Validity of Modus Trex Activity Monitor in Determining Functional Level in Veterans with Transtibial Amputations

27 participants

(All BK)

StepWatch and GPS

(A. Ankle)

Activity level – steps

(10–16 days)

Investigate the accuracy and reliability of Modus Trex–derived K-level to differentiate between Medicare Functional Classification levels (MCK-levels).

The Modus Trex–derived K-level was reliable and accurate at estimating MCK-levels and can be useful as a component in K-level evaluation.

Halsne (2013) [50]

Long-term activity in and among persons with transfemoral amputation

17 participants

(All AK: 6 at K2, 8 at K3, 3 at K4)

StepWatch

(P. Ankle)

Steps/day

(12 months)

Objectively characterise the mobility dimension of participation in persons with AK amputation using long-term step activity data, and determine how activity varies over extended periods of time.

Activity between K2 and K3 subjects was not significantly different. Relative variation (CoV) was 0.65 across subjects but was lower for those with higher activity levels. Warmer seasons promoted higher activity, but peak temperatures and humidity reduced activity.

Kent (2015) [51]

Step activity and stride-to-stride fluctuations are negatively correlated in individuals with transtibial amputation

22 participants

(All BK at K3)

ActiGraph

(P. Shank)

Steps

(3 weeks)

Determine whether increased stride-stride fluctuations also correspond to a reduced level of activity in daily life in people with BK amputation.

Stride-to-stride fluctuations correlated with decreased step counts, but it is unclear whether high fluctuations promoted decreased activity or less active individuals did not gain the experience to achieve skilled movement.

Lin (2014) [52]

Physical activity, functional capacity, and step variability during walking in people with lower-limb amputation

20 participants

(12 BK, 8 AK)

Pedometer

(A. Waist)

Steps

(7 days)

Explore the associations between physical activity and physical performance measures (self-selected walking speed, 6-min walk test (6MWT), step length variability, and step width variability) in people with BK amputation.

Physical activity correlated with comfortable walking speed, 6MWT, and step width variability, and was inversely correlated with step length variability of the prosthetic leg and of the sound leg.

Mandel (2016) [53]

Balance confidence and activity of community-dwelling patients with transtibial amputation

22 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Examine the relationship between community-based physical activity and balance confidence in people with BK amputation who have low fall-risk.

There was a positive correlation between self-perceived balance confidence and community-based physical activity.

Orendurff (2016) [54]

Comparison of a computerized algorithm and prosthetists’ judgment in rating functional levels based on daily step activity in transtibial amputees

81 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Compare prosthetists’ ratings of K-levels based on a visual inspection of step activity patterns with the ratings calculated by an algorithm based on the same step activity data.

The algorithm produced functional level values that closely matched the average ratings of 14 experienced prosthetists. Linear regression indicated good linearity and concordance across the range of the two scales.

Orendurff (2016) [55]

Functional level assessment of individuals with transtibial limb loss: Evaluation in the clinical setting versus objective community ambulatory activity

12 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Determine the relationship between clinic-based K-level classification and K-level based on everyday ambulatory activity data collected by StepWatch.

There was good agreement between the two methods of determining K-level with R2 = 0.775 (p < 0.001).

Parker (2010) [56]

Ambulation of People With Lower-Limb Amputations: Relationship Between Capacity and Performance Measures

52 participants

(BK and AK, numbers not provided)

StepWatch 3

(P.)

Activity level – steps

(7 days)

Examine the relationship between clinical measures of ambulation capacity (using Locomotor Capabilities Index, 2MWT, and Timed Up and Go Test) and measures of ambulation performance in the community.

Capacity and performance showed moderate correlation (Spearman’s ρ: 0.41–0.78, p < 0.05). The highest correlation was the 2MWT and SAM peak activity index (0.78, p < 0.001). The 2MWT correlated with steps/day (p = 0.026) and TAPES (p = 0.016). Depressive symptoms correlated with decreased performance (p = 0.003, TAPES).

Pepin (2019) [57]

Correlation Between Functional Ability and Physical Activity in Individuals With Transtibial Amputations: A Cross-Sectional Study.

19 participants

(All BK. 5 at K2, 14 at K3/K4)

ActivPAL

(A. Thigh)

Activity – sitting/lying, standing, walking

(7 days)

Examine the association between functional ability and physical activity in individuals with BK amputations.

Participants spent on average 19.7 h per day lying/sitting, 3.5 h standing, and 0.77 h walking. They walked an average of 3145 steps/day, placing them in the sedentary category.

Resnik (2017) [58]

The DEKA hand - A multifunction prosthetic terminal device - patterns of grip usage at home

21 participants (Estimate ~ 40% AE ~ 60% BE, 1 bilateral – level unclear)

On board software - not detailed - logs device on/off state, and motor and joint position data

(P.)

Time prosthesis powered on, position of joint motors, time in each grip pattern.

(Up to 3 months)

To quantify usage of DEKA hand grip patterns during home use and compare patterns of usage at home to test sessions.

The most used grips at home were power, pinch open, and lateral pinch. There were no differences in grip use over time. Power grip was used 55% of the time at home and 23% of the time during lab testing. Fewer grip patterns were used at home than in the lab.

Samuelsen (2017) [59]

The Impact of the Immediate Postoperative Prosthesis on Patient Mobility and Quality of Life after Transtibial Amputation

10 participants

(All BK)

ActiGraph GT3X

(A. Belt buckle)

Activity level – steps, energy expenditure

(10 days – 6 weeks)

Measure activity using accelerometers, assess quality of life with the Medical Outcome Study Short Form-36, and Evaluate expected mobility status using the Amputee Mobility Predictor.

No significant relationships were observed between expected level of function and recorded activity level.

Patients had low physical and emotional Short Form-36 component scores.

Sanders (2018) [60]

Residual limb fluid volume change and volume accommodation: Relationships to activity and self-report outcomes in people with trans-tibial amputation

29 participants

(All BK)

ActiGraph GT3X+

(P.)

Activity - sitting, standing, walking, doffing

(3 h)

Explore whether the morning-to-afternoon limb volume change was associated with time weight-bearing, satisfaction, comfort, or perceived mobility, and whether participants who changed sock thickness to adjust to volume change had better outcomes.

Factors other than time spent weight-bearing (standing and walking) correlated with the rate of morning-to-afternoon limb fluid volume change on BK prosthesis users.

Sions (2018) [61]

Self-Reported Functional Mobility, Balance Confidence, and Prosthetic Use Are Associated With Daily Step Counts Among Individuals With a Unilateral Transtibial Amputation

47 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Determine if self-reported measures, assessing constructs other than physical activity, are associated with accelerometer measurements of physical activity.

Clinical outcome measures may be predictive of daily physical activity as obtained with accelerometers among community ambulating longer-term BK prosthesis users.

Stepien (2007) [62]

Activity Levels Among Lower-Limb Amputees: Self-Report Versus Step Activity Monitor

77 participants

(All BK)

StepWatch 3

(P.)

Activity level – steps

(8 days)

Determine the accuracy of self-reported activity by community-dwelling, individuals with lower-limb amputation.

Participants averaged 3063 ± 1893 steps/day. Self-reported activity in an experimental setting was not accurate, and the measured and self-reported proportion of time spent in various states of activity showed poor agreement in rest, low, medium and high level activity. There was no bias toward either over- or under-reporting.

Agrawal (2010) [63]

A comparison of gait kinetics between prosthetic feet during functional activities - Symmetry in External Work (SEW) approach

11 participants

(All BK)

StepWatch

(Not provided)

Activity level – steps

(4* 10–14 days)

Validate a measure for quantifying gait differences among prosthetic feet. Calculate the reliability of the Symmetry in External Work (SEW) measure and determine its correlation with clinical measures.

The StepWatch results showed no difference in the number of steps or activity level of subjects. There was good correlation between the SEW values for level walking and other clinical outcome measures.

Andrysek (2017) [64]

Long-term clinical evaluation of the automatic stance-phase lock-controlled prosthetic knee joint in young adults with unilateral above-knee amputation

10 participants

(All AK)

Power Walker EX-510

(P. Anterior/ medial of thigh)

Steps/day

(2 weeks)

Compare the ASPL knee to the WBA knee on step count, walking speed and energy expenditure.

Participants did not tend to alter their walking speed or step count with the ASPL knee, but a reduction in energy expenditure was found.

Berge (2005) [65]

Efficacy of shock-absorbing versus rigid pylons for impact reduction in transtibial amputee

15 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Compare a shock-absorbing pylon to a rigid pylon to assess the effect on gait mechanics and functional outcomes using step counts and questionnaires.

No difference was found in number of steps/week. The only difference was that at initial contact, the prosthetic-side knee angle had more flexion with the rigid pylon while walking at a controlled speed (p = 0.004).

Buis (2014) [66]

Measuring the daily stepping activity of people with transtibial amputation using the ActivePAL activity monitor

48 participants

(All BK, 24 TSB sockets,

24 PTB sockets)

ActivPAL

(P. Ankle, anterior)

Steps per minute/hour/day

(7 days)

Compare the level of activity of prostheses users with PTB sockets to prostheses users with TSB sockets.

The differences in socket design did not result in significant differences in activity level.

Christiansen (2018) [67]

Behavior-Change Intervention Targeting Physical Function, Walking, and Disability After Dysvascular Amputation: A Randomized Controlled Pilot Trial

36 participants

(All BK)

ActiGraph GT3X-BTa

(A. Waist)

Activity level – steps

(10 days)

Determine the efficacy of a home-based behaviour change intervention to promote exercise, and disease self-management after dysvascular transtibial amputation.

The home-based behaviour-change intervention improved daily step counts.

Coleman (2004) [68]

Quantification of prosthetic outcomes: Elastomeric gel liner with locking pin suspension versus polyethylene foam liner with neoprene sleeve suspension

13 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(2 weeks)

Compare two socket suspension systems: the Alpha liner with distal locking pin and the Pe-Lite liner with neoprene suspension sleeve, based on ambulatory activity, wear time, comfort and satisfaction.

10 participants preferred the Pe-Lite and 3 the Alpha.

The Pe-Lite was worn for 82% more time and 83% more steps per day. Ambulatory intensity distribution did not differ and no differences were found from questionnaires.

Darter (2007) [69]

The effects of an integrated motor learning based treadmill mobility and aerobic exercise training program in persons with a transfemoral amputation

8 participants

(All AK)

AMP 331 Accelerometer

(P. Ankle)

Activity – steps, distance, speed

(3* 7 days)

Determine daily averages for steps, speed and distance using an activity monitor before, during and after an 8-week mobility and aerobic exercise training intervention.

The mobility and aerobic exercise training was effective in improving gait performance, cardiorespiratory fitness, and locomotion related disability over the length of the intervention.

Gailey (2012) [70]

Application of self-report and performance-based outcome measures to determine functional differences between four categories of prosthetic feet

10 participants

(All BK - 5 with peripheral vascular disease (PVD), 5 without)

Step Activity Monitor – SAM

(later ‘StepWatch’)

(P. Ankle)

Steps/day

(2 weeks)

Determine the ability of self-report and performance-based measurements to detect functional differences between four categories of prosthetic feet, and whether differences exist between cohorts with and without peripheral vascular disease (PVD).

AMPPRO performance-based measure found differences between some feet from baseline (p < 0.05). No other differences were found between feet by performance-based measures (6MWT, SAM) or self-report measures (PEQ-13 and LCI). AMPRO and 6MWT found differences between the PVD and the non-PVD groups (p < 0.05) with the Proprio foot.

Graczyk (2018) [71]

Home use of a neural-connected sensory prosthesis provides the functional and psychosocial experience of having a hand again

2 participants (Both BE with pre-implanted 8 channel flat interface nerve electrodes (FINEs))

Aperture sensor,

(P. base of thumb)

3x Flexiforce pressure sensors

(P. thumb, index, and middle finger)

Grasp count – prosthesis only

Hours of sensory feedback

Self-reported prosthesis wear time

(36–47 days)

Study the effect of electrical stimulation for sensory feedback while using a myoelectric prosthetic on use/non-use of the prosthesis, functional performance and psychosocial experience.

When sensory feedback was provided, participants used the prosthesis more, functional performance improved and psychosocial factors improved (self-efficacy, prosthetic embodiment, self-image, social interaction, and quality of life).

Hafner (2015) [72]

Physical performance and self-report outcomes associated with use of passive, adaptive, and active prosthetic knees in persons with unilateral, transfemoral amputation: Randomized crossover trial

12 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps/day

(2–14 months)

Assess and compare physical performance and self-reported outcomes that may be attributed to use of prosthetic knees with passive, adaptive, and active control in persons with unilateral BK amputation.

Compared with passive control, adaptive control improved comfortable Timed Up and Go (TUG) by 0.91 s (p = 0.001) and reported physical function by a T-score of 1.26 (p = 0.03), and active control increased comfortable TUG, fast TUG, and ramp times by 3.02, 2.66, and 0.96 s, respectively (all p < 0.03), and increased balance confidence by 3.77 (p = 0.003). Steps/day was lower when using the active knee than passive or adaptive knees.

Hafner (2007) [73]

Evaluation of function, performance, and preference as transfemoral amputees transition

17 participants

(All BK)

StepWatch 2

(P. Ankle)

Steps/day, distance

(2 months)

Evaluate differences in function, performance, and preference between mechanical and microprocessor prosthetic knee control technologies.

The microprocessor knee showed improved performance on stairs and hills, reduced frequency of stumbling, and was preferred by participants. No differences were found between knees on step frequency or estimated daily distance travelled.

Highsmith (2016) [74]

Effects of the Genium Knee System on functional level, stair ambulation, perceptive and economic outcomes in transfemoral amputees

20 participants

(All AK)

StepWatch

(P. Ankle)

Activity level – steps

(2 weeks)

Determine if the Genium knee is beneficial, compared to the C-leg, using common clinical assessments.

The Genium knee improved stair walking, multi-directional stepping, functional level, steps/ day and perceived function, compared to the C-Leg.

Highsmith (2012) [75]

Spatiotemporal Parameters and Step Activity of a Specialized Stepping Pattern Used by a Transtibial Amputee During a Denali Mountaineering Expedition

1 participant

(BK)

Sportline ThinQ XA Model 305 Pedometer

(A. Neck)

Steps/day, cadence.

(8 days)

Describe spatiotemporal differences between the French technique and typical walking patterns of the participant. Report the qualitative and quantitative step activity during a climbing expedition.

Stride, step, and double support times were greater in the French technique, but spatially, stride and step lengths were greater in the traditional stepping. Daily step count averaged 10,404 steps on active climbing days.

Hsu (2006) [76]

The Effects of Prosthetic Foot Design on Physiologic Measurements, Self-Selected Walking Velocity, and Physical Activity in People With Transtibial Amputation

8 participants

(All BK)

Yamax Digiwalker Pedometer

(A. Iliac crest)

Steps

(1 month)

Investigate the physiological differences during treadmill walking and the physical activity profiles for the Otto Bock C-Walk foot (C-Walk), Flex-Foot, and solid ankle cushion heel (SACH) foot. Compare feet on step-count during daily physical activity.

The C-Walk trended towards better physiological responses compared with the SACH; however, no differences between feet were statistically significant. The Flex-Foot showed no differences in energy expenditure, gait efficiency, or steps/day, but showed a lower age-predicted maximum heart rate and perceived exertion.

Kaufman (2018) [12]

Functional assessment and satisfaction of transfemoral amputees with low mobility (FASTK2): A clinical trial of microprocessor-controlled vs. nonmicroprocessor- controlled knees

50 participants

(All AK)

ActiGraph GT3X+

(P and A. Waist, thigh, both ankles)

Activity level – steps

(4 days)

Assess if individuals with amputations categorised as K2 would benefit from a microprocessor-controlled knee.

Participants demonstrated a reduction in falls, less time spent sitting, and increased activity when using the microprocessor knee. They also reported better ambulation, improved appearance, and greater utility.

Klute (2006) [77]

Prosthetic Intervention Effects on Activity of Lower-Extremity Amputees

17 participants

(12 BK, 5 AK)

StepWatch

(P. Ankle)

Steps/day

(7 days)

Investigate the effect of prosthetic interventions on the functional mobility of people with lower-limb amputation.

The intervention had no effect on activity level and duration. Individuals with BK amputation had higher activity levels on weekdays than weekends. Prosthetic components should be optimised for 1 to 2 min bouts of activity consisting of a few dozen steps.

Klute (2011) [78]

Vacuum-Assisted Socket Suspension Compared With Pin Suspension for Lower Extremity Amputees: Effect on Fit, Activity, and Limb Volume

5 participants

(All BK)

StepWatch 3

(P. Ankle)

Activity level – steps

(2 weeks)

Compare the fit and function of two socket and suspension systems: 1. A total surface-bearing socket with a vacuum-assisted suspension system (VASS), 2. A modified patellar tendon-bearing socket with a pin lock suspension system.

Activity levels were lower while wearing the vacuum-assisted socket suspension system than the pin suspension. The VASS socket fitted slightly better, as measured by pistoning.

Klute (2016) [79]

Prosthesis management of residual-limb perspiration with subatmospheric vacuum pressure

5 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(2 weeks)

Measure differences in activity levels, residual-limb skin temperatures, perspiration accumulation and expulsion, and subjective experiences between the dynamic air exchange (DAE) prosthesis and a standard-of-care total surface bearing suction socket.

During the 7-day acclimation, no difference in step activity levels was detected (p = 0.22). During the rest-walk-rest protocol, no differences in skin temperatures were observed (p = 0.37). The DAE prosthesis accumulated 1.09 ± 0.90 g and expelled 0.67 ± 0.38 g of perspiration. The suction prosthesis accumulated 0.97 ± 0.75 g. Participants were receptive to both prostheses.

Larson (2014) [80]

Massage therapy effects in a long-term prosthetic user with fibular hemimelia

1 participant

(BK)

Pedometer Yamax SW-200

(A. Hip on prosthetic side)

Steps/day

(50 days)

Evaluate the effectiveness of massage therapy to promote activity level, decrease low-back pain, and improve health-related quality of life (HRQOL) in a long-term prosthetic user with fibular hemimelia.

Pain level decreased, HRQOL increased, and no change occurred in ambulatory activity level.

Littman (2018) [81]

Pilot randomized trial of a telephone-delivered physical activity and weight management intervention for individuals with lower extremity amputation

15 participants

(12 BK, 3 AK)

StepWatch

(Not provided)

Steps

(7 days)

Test the feasibility, acceptability, and safety of a weight management and physical activity intervention and obtain preliminary efficacy estimates for changes in weight, body composition, and physical functioning.

Coached participants had greater decreases in waist circumference and fat mass, but no significant intervention effects were observed for physical functioning or physical activity.

Morgan (2018) [82]

Laboratory- and community-based health outcomes in people with transtibial amputation using crossover and energy storing prosthetic feet: A randomized crossover trial

27 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(1 month)

Assess changes in lab-based endurance, perceived exertion and walking performance, and community-based step activity, self-reported mobility, fatigue, balance confidence, activity restrictions, and satisfaction for crossover and energy storing feet.

Self-reported results showed the users preferred crossover to energy storing devices. Quantitative measures did not show significant differences between prostheses, except for step length.

Raschke (2015) [83]

Biomechanical characteristics, patient preference and activity level with different prosthetic feet: A randomized double blind trial with laboratory and community testing

11 participants

(All BK)

StepWatch

(P. Ankle)

Activity level – steps

(7 days)

Determine if reported preference was related to a biomechanical characteristic of different prosthetic feet, and if this resulted in increased community-based activity.

Each foot was evaluated over a similar number of steps, but no foot increased activity levels.

Sanders (2017) [84]

Effects of Socket Size on Metrics of Socket Fit in Trans-Tibial Prosthesis Users

9 participants

(All BK)

ActiGraph GT3X

(P. Ankle)

Steps

(2 weeks)

Identify metrics of acceptable socket fit in people with BK amputation. Determine if a known change in prosthetic socket size was reflected in objective and subjective measures of fit, comfort, and performance.

Most promising variables for early detection of socket fit deterioration were step time and width asymmetry, anterior and anterior-distal morning-to-afternoon limb fluid volume change, SCS, and subscales of the PEQ.

Segal (2014) [85]

Does a Torsion Adapter Improve Functional Mobility, Pain, and Fatigue in Patients with Transtibial Amputation?

10 participants

(All BK)

StepWatch 3

(P. Shank – lateral side)

Activity level – steps

(7 days)

Determine if a torsion adapter results in improved functional mobility, pain and fatigue, compared to a rigid adapter.

For moderately active individuals with BK amputation the torsion adapter did not show improvements in functional mobility, pain or fatigue. However, small increases in low- and medium intensity activities with less pain were seen. The torsion adapter may benefit individuals who have difficulty navigating everyday environments.

Sherman (2018) [86]

Daily step count of British military males with bilateral lower limb amputations: A comparison of in-patient rehabilitation with the consecutive leave period between admissions

9 participants

(All bilateral, 4 AK/AK, 2 AK/BK, 1AK/BE, 1 AK/AE, 1 AK/BK/BE)

Long-term Activity Monitor - LAM2

(P. Thigh)

Steps/day

(4 weeks)

Characterise daily step counts of military personnel with bilateral lower-limb amputations due to trauma, and compare steps during and between in-patient rehabilitation intervals.

Mean daily step count decreased when away from rehabilitation.

Theeven (2012) [87]

Influence of advanced prosthetic knee joints on perceived performance of life activity

30 participants

(All AK at K2)

ActiGraph GT1M

(A. Waist)

Activity level – steps

(3 weeks)

Investigate changes in perceived performance of K2 persons who transition from a mechanical- to a microprocessor-controlled prosthetic knee joint, and assess whether the transition between components affects the daily activity level.

Participants’ perception regarding ambulation and satisfaction with walking were higher with the microprocessor knee, but the activity level was similar on both knees.

Wurdeman (2017) [88]

Step Activity and 6-Minute Walk Test Outcomes When Wearing Low-Activity or High-Activity Prosthetic Feet

28 participants

(24 unilateral BK, 4 bilateral BK)

Activity monitor

(P. Shank)

Steps

(6 weeks)

Determine changes in average daily step count and 6MWT due to use of low- and high-activity prosthetic feet, and examine the sensitivity of these measures to classify different feet.

Neither daily step count nor 6MWT were responsive to changes in prosthetic feet, so it is not recommended that these measures are used to assess outcomes for different prosthetic feet.

Arch (2016) [89]

Real-world walking performance of individuals with lower-limb amputation classified as Medicare Functional Classification Level 2 and 3

27 participants

(10 BK and 1 AK at K2, and 10 BK and 6 AK at K3)

Fitbit One

(P. Ankle)

Activity level – steps/day

(7 days)

Investigate community walking performance measures (step count, amount of activity, and activity intensity) for individuals with unilateral lower-limb amputation classified as K2 and K3.

The K2 group had a slower self-selected walking speed, walked a shorter distance in 6 min, and had a lower total step count, than those classified as K3. The K2 group tended to spend more time in low-intensity activity and less time in high-intensity activity.

Bussmann (2004) [90]

Daily physical activity and heart rate response with a unilateral transtibial amputation for vascular disease

18 participants

(9 BK participants,

9 anatomically intact controls)

2 uniaxial,

1 biaxial

ADX202

(A. 2 on thigh,

1 on chest)

Lying/sitting/standing/walking/ cycling

(48 h)

Assess whether individuals with unilateral transtibial amputation and vascular disease are less active than persons without known impairments.

Unilateral transtibial amputates with vascular disease were less active than persons without known impairments.

Carmona (2017) [28]

Walking activity in prosthesis-bearing lower-limb amputees

43 participants

(11 AK, 29 BK)

StepWatch3

(P. Ankle. If bilateral – on the side of the longer residuum)

Activity level – steps, cadence

(15 days)

Study walking activity in home-dwelling lower-limb prosthesis users, comparing amount of walking to age, weight, cause of amputation, time since amputation, level of amputation and use of a walking aid.

Individuals with BK amputation without vascular disease or a walking aid walked the most per day. Individuals with AK amputation walked 21% less, and those with walking aids walked 13% less. Body mass index did not correlate with time spent walking but did with walking speed.

Chadwell (2016) [91]

The reality of myoelectric prostheses - understanding what makes these devices difficult for some users to control

3 participants

(2 BE, 1 Anatomically Intact)

2x ActiGraph

GT3X+

(A. and P. Wrists for prosthesis users, both A. Wrists for anatomically intact control)

Symmetry and magnitude of non-specified arm movements - both arms

Self-reported prosthesis wear time

(7 days)

Demonstrate the feasibility of a protocol to understand the factors making myoelectric prostheses difficult to control. The activity monitoring part of the paper aimed to collect and analyse data from wrist worn monitors.

The anatomically intact participant was equally reliant on both arms, whilst the two prosthesis users were more reliant on their intact arm. The prosthesis users reported wearing their prostheses for 10h hours on days when it was worn, but one only wore the prosthesis on 3 out of 7 days.

Chadwell (2019) [92]

Upper limb activity of twenty myoelectric prosthesis users and twenty healthy anatomically intact adults

40 participants (20 BE, 20 Anatomically Intact)

2x ActiGraphs from the GT3X range.

(A. and P. Wrists for prosthesis users, both A. Wrists for anatomically intact controls)

Raw accelerometer data, activity counts in 1 s and 60s epochs, self-reported wear and sleep diaries. Prosthesis non-wear algorithm.

(7 days)

Report the dataset upon which Chadwell et al. [46] was based, making the data publicly available for secondary analysis by other researchers.

The raw accelerometry data, the activity count data, code for estimating wear/non wear and wear diaries from 20 myoelectric prosthesis users are provided.

Chu (2014) [93]

Comparison of prosthetic outcomes between adolescent transtibial and transfemoral amputees after Sichuan earthquake using Step Activity Monitor and Prosthesis Evaluation Questionnaire

21 participants (11 BK, 10 AK)

Stepwatch

(P. Ankle)

Activity level – steps (3 months)

Compare the activity levels of adolescents with a transfemoral amputation to those with a transtibial amputation, and explore differences in prosthesis-related quality of life.

Adolescents with transfemoral amputation were less active than those with transtibial amputation.

No differences were found on Prosthesis Evaluation

Questionnaire outputs.

Hordacre (2015) [94]

Community activity and participation are reduced in transtibial amputee fallers: a wearable technology study

46 participants

(All BK)

StepWatch 3

and GPS

(P. Ankle - lateral side)

Steps

(7 days)

Assess activity and participation characteristics in the home and various community settings for people with transtibial amputation who do and do not have a history of falling.

Participants with a history of falls demonstrated lower levels of community activity and participation. Activity levels were reduced for recreational and commercial roles.

Hordacre (2014) [95]

Use of an Activity Monitor and GPS Device to Assess Community Activity and Participation in Transtibial Amputees

46 participants

(All BK)

StepWatch 3

and GPS

(P. Ankle - lateral side)

Steps

(7 days)

Assess wearable technology (step-counts and GPS) to measure community activity and participation, and whether participants with higher K-levels (as assessed by AMP-PRO and timed mobility measures) were more active in their community.

The study linked accelerometer and GPS data and found that individuals categorised as lower functioning (K1/2) showed lower community activity and participation than K3/4.

No difference was found between K3 and K4 for community activity or participation.

Paxton (2016) [96]

Physical activity, ambulation, and comorbidities in people with diabetes and lower-limb amputation.

46 participants

(22 BK diabetics, 11 Anatomically Intact diabetics, 13 Anatomically Intact healthy)

ActiGraph GT3X

(A. Waist)

Activity level – steps

(10 days)

Characterise physical activity and its relation to physical function and comorbidities in people with/without diabetes and amputation.

Physical activity was related to physical function in the diabetic intact-limb group and in the diabetic reduced-limb group, whereas no such relationship existed in the healthy group.