South and West Regional Development and Evaluation Committee (DEC) Approval.

In December 1995 the results of the trial were presented to South and West Regional DEC. The report (which follows) also include data from other subjects using the ODFS who were not included in the trial bringing the total number of subjects to 178. The current and proposed services were outlined and a cost benefit study made in terms of QALYs (Quality Added Life Years).

After reviewing the evidence put forward, the DEC recommended the use of the ODFS for use within the NHS.



Swain ID, Taylor PN, Burridge JH, Hagan SA, Wood DE.



1. Statement of the proposed service:

This paper considers the use of Common Peroneal Stimulation for the correction of drop foot.

2. Background

Drop foot is a common problem for patients with Upper Motor Neurone lesions. It is caused partly by poor active motor control of the anterior tibial muscles and partly by spasticity of the calf muscles. Drop foot prevents the patient from effectively swinging the leg when walking. This in consequence leads to an inefficient, unsafe gait characterised by hip hitching and circumduction and causing the patient to stumble. The increased effort involved not only means that walking is slow, tiring and unsafe but also leads to further increase in spasticity.

3. Incidence, prevalence and projected trends

90% of patients using the ODFS have suffered a stroke. There are 100,000 hospital admissions for stroke in Great Britain annually; 20% are fatal and 80% of survivors have a restricted lifestyle. (Stroke Towards better management RCP 1989) There are no statistics to show how many of these patients have a drop foot but 20% (Merletti et al 1979) would be a conservative estimate. 80% of patients referred for treatment are suitable; giving a potential of 12,800 patients per year. 178 patients have been treated in the last three years.

Despite an ageing population there has been a steady decline in the incidence of stroke of 1% per annum since 1970. This is partly accounted for by:

i) Effective treatment of hypertension

ii)Improved lifestyle

iii)Awareness of predisposing factors

The projected trend is for a continued decline of 30% over the next decade.

Incomplete Spinal Cord injuries and MS accounts for fewer than 10% of patients referred for treatment.

4. Outline of current service

Correction of drop foot is conventionally through splinting, either a plastic, semi-flexible ankle-foot orthosis or a caliper. 11% of patients referred for treatment were using a splint. A further 20% had used and rejected this method of correction. Many patients had been discouraged from using a splint because of the problems caused by them such as: increased calf spasticity, discomfort, restricted ankle movement leading to further gait abnormalities.

Few patients receive physiotherapy more than one year post stroke, studies have shown that there is little evidence of continued benefit. There is no evidence that physiotherapy alone can improve ankle dorsiflexion.

5. Outline of proposed service

This proposal is for the Odstock Drop Foot stimulator to be offered as a treatment for drop foot resulting from an Upper Motor Neurone lesion. The service will include supply and maintenance of equipment and consumables, setting up stimulation systems and monitoring of patients progress. Most of the clinical work will be done by a physiotherapist who will often spend some time on gait re-education or liaise with the patients own physiotherapist.

6. Justification for this service

Literature justifying the use of stimulation to correct drop foot rests mainly on case studies, uncontrolled trials and retrospective reviews. Much is anecdotal and outcome measures lack standardisation but oxygen consumption, walking speed and gait analysis predominate. No studies other than the RCT conducted in this department and the Strathclyde study have used patient-centred outcome measures.

A study of a sample of 50 patients using a drop foot stimulator similar to the ODFS, but with less ‘fine tuning,’ in Veruno, Italy 1979 (ref Scan J Rehab ) showed improved gait and reduced oxygen consumption in 75% of patients.

Case studies of 3 hemiplegic subjects (Hesse et al) measured increases of 33% in gait speed, 5% in cadence, 26% stride length and 18% in stair climbing. These patients also increased their score on Rivermead Motor Assessment, 19.5%. Ashworth test showed 11% reduction in spasticity.

A study of 22 patients at the Rehabilitation Institute in Ljubljana, Slovenia, measured a mean increase in walking speed of 0.49 m/s, increased stride speed of 0.49s and increased stride length of 5.49 cms.

A study at Strathclyde University (Granat et al 1995) which used an A-B-A design demonstrated improved gait parameters in 14 out of 15 subjects while using a drop foot stimulator; with no change during the control period. Significant orthotic benefit was seen in 10 out of 17 subjects. The pattern of stimulation used in this study was effective in correcting ankle inversion, the gait analysis results reflect this by improved medio-lateral stability. Flexible electrode positioning and polarity reversal allow for finer control over the stimulated ankle movement with the ODFS than with the stimulator used in the Strathclyde Study.

7. Benefits

80% of patients referred for Functional Electrical Stimulation (FES) for the correction of drop foot have been suitable for treatment.

Drop foot systems have been set up for 178 patients.

25 patients have since discontinued using the drop foot stimulator. Reasons for discontinuing are:

7 - Improved ankle control - orthosis no longer needed

6 - Problems using the system skin irritation, increased spasticity, unable to set up stimulator

3 - Walking deteriorated

5 - Insufficient benefit

1 - Cosmetic

3 - Died

153 patients are continuing to use a drop foot stimulator. The only documented evidence of the effect of long term use of the stimulator is a report by Karsnia et al 1990 reviewing 99 patients who had used drop foot stimulation for 10 years. The first patients to use the ODFS were in 1988, those who are continuing have not demonstrated any problems other than those mentioned above which become apparent in the first few months of use.

The treatment of drop foot by Common peroneal stimulation is supported by the results of a RCT conducted by the Medical Physics and Biomedical Engineering Department Salisbury District Hospital and funded by the Medical Devices Agency at the Department of Health.

The structure and results of the trial are summarised:

The purpose of the trial

The trial assessed the efficacy of the Odstock Dropped Foot Stimulator (ODFS) as part of a physiotherapy programme for stroke patients. It enabled clinical protocols to be developed for its use with patients who are unable to achieve active, functional, ankle dorsiflexion when walking.


Randomised Controlled Trial

Suitable subjects (n = 32, mean age 56.5 years, mean time from stroke 4 years 5 months) referred for treatment were identified following selection criteria (appendix). Subjects gave informed consent and were then allocated to either the treatment (FES) (n = 16) or control group (n = 16). Allocation was made at random following the first assessment. All subjects received the same amount of treatment time. The treatment group used the drop foot stimulator at home and as part of their physiotherapy sessions.

A battery of tests were performed prior to stimulation; after 1 month (during which time subjects received physiotherapy) and at 3 months.


Walking Speed

FES group

At the end of the trial period 13 FES subjects walked faster, 2 no change and 1 walked more slowly. Mean increase in speed (all FES subjects) = 0.22 m/s SD 0.09 mean % increase =18.1%

Increase in speed was significant (paired T test) p <0.0002

Control group

In the control group 6 subjects walked faster, 6 at the same speed and 3 more slowly. Mean increase in speed (all control subjects) = 0.03 m/s SD 0.09. mean % increase = 5.1%

Increase in speed was not significant (paired T test p<0.604)

There was no significant difference in the change in walking speed between the FES and control group (student T test) p<0.0713.

It is interesting to note that:

The mean walking speed of the control group increased from 0.48 m/s to 0.50m/s during the first month when subjects received physiotherapy but fell to 0.49 m/s at the 3 month assessment.

The mean walking speed of the FES group increased from 0.68m/s to 0.75m/s in the first month and this improvement was maintained without further physiotherapy.

There was no significant change in the walking speed of the FES subjects without stimulation.

Physiological Cost Index (PCI)

FES group

PCI is a measure of the effort of walking which relates increased heart rate to walking speed

At the end of the trial period 10 FES subjects had a lower PCI, 4 no change and in 2 subjects it was higher. Mean reduction in PCI (all FES subjects) = 0.21 SD 0.27 mean % reduction = 46.44%

Reduction in PCI was significant p< 0.0067 (paired T test)

Control group

In the control group 5 subjects had a lower PCI, 8 no change and in 3 subjects it was higher. There was a mean reduction (all control subjects) of 0.02 which was not significant p<0.7101 (paired T test).

There is a significant difference in change in PCI between the FES and control group p<0.0466 (student T test)

Although there was a slight improvement in PCI without stimulation in the FES group it was not significant.


Mobility Questionnaire

Maximum score = 16

Mean increase in score during the trial period:

FES group = 2.50

Control group = 0.85

There was a significant difference between the change in the FES group and the change in the control group p< 0.0489 (Fisher exact probability test). No significant difference was seen at the start of the trial.

Hospital Anxiety and Depression Scale (HAD)

Patients taking part in the trial were also asked to complete HAD questionnaires at each assessment. A significant reduction in both anxiety and depression in the FES group over the three month period p<0.0028 (anxiety scores) and p<0.0047 (depression scores) No significant improvement was seen in the control group. The difference between control and FES groups was not significant p<0.0820. It is interesting to note, however, that the improvement over the first month was almost the same in both FES and control groups while both groups received physiotherapy. Improvement continued in the FES group but not in the controls.

8. Disbenefits

12% of subjects showed no benefit, none were made worse.

Skin irritation has been a problem with 3% of subjects 1 is unable to continue using the stimulator. A possible cause of skin irritation is the formation of electro-chemical compounds when stimulation produces a charge imbalance. A recent modification of the stimulator which inverts the wave form on alternate pulses effectively avoids a charge imbalance. 5 patients are using these stimulators. 1 patient continues to have skin irritation.

The stimulation system involves wearing self adhesive electrodes on the leg, a switch in the heel of the shoe, and leads connecting these to the stimulator which is worn either in a pocket or on belt. This is an inconvenience and time is taken setting it up each day. Implanted systems have been used but these continue to have problems. 14 of the 16 subjects in the trial were able to set the stimulator up independently but 3 of these said they usually had help. 2 subjects were unable to set it up without help. 4 subjects said they often had difficulty finding the correct electrode positions and 4 said they found the system an encumbrance. The sensation of stimulation has been described as like a momentary nettle rash, 4 subjects said it was slightly uncomfortable, none found it too uncomfortable to use.

9. Costs

Treatment is charged at £166 per episode. This includes the cost of building and maintaining the stimulator and the supply of consumables.

Careful application, support and training of patients using these devices is important (Karsnia et al 1990). Two treatment episodes are therefore needed to set up the stimulator and to ensure that the patient and/or carer are able to use it effectively. Patients are followed-up at 6 weeks with 2 further appointments during the first year. In subsequent years patients are reviewed at 6 monthly intervals.

Therefore the cost in the first year is £830 and £332 in subsequent years.

If an estimate of a mean of 5 years use is taken the cost per year is £431.60

Assuming that stroke patients effectively using a drop foot stimulator will benefit for a mean of 5 years. (mean age of subjects 54 years). Then an estimated 0.042. QALYs may be gained by each patient and the cost / QALY would be £10,307.

If a patient only uses the stimulator for 1 year the cost would be £830

The cost per QALY would be £19,821

Some patients may benefit for as long as 10 years (Karsnia et al) the cost per year would be £381.80

and the cost per QALY £9118

Benefits to the patient are demonstrated by reduced physical disability through increased walking speed, decreased effort of walking and increased mobility. Improvement in these were measured by the randomised controlled trial.

Scoring for Speed, PCI, Mobility Questionnaire and HAD and correlation to points on the Rosser Matrix.

Walking speed, PCI and Mobility Questionnaire scores are considered to relate to disability. HAD scores are related to Distress. It was estimated that all patients were in the categories E1 to E4 and D2 to D6 at both the start and end of the trial. Any movement was considered to be within these categories. Walking speeds, PCI values, Mobility and HAD scores were ranked in the range 0-10. Speed and Mobility scores were ranked in the range 0-8. Scores for Speed PCI and Mobility were then summed to give a disability value.

Example: subject 8 who was at point D4 on the Rosser Matrix at the start of the trial. Walking speed was 0.37 m/s (40% of normal walking speed) (score 2). PCI 0.98 (3* normal)(score 3). Mobility Questionnaire score 12. (score 6) The total disability score of 11 is equal to the sum of these scores. Subject 8 could stand independently from a wheelchair and walk over 500m but always used a walking stick unless supported, only occasionally walked outside alone and not on uneven ground unless with help. At the end of the trial walking speed was 0.55 m/s (score 3) PCI was 0.33 (score 9). Mobility score was 15 (score 8). Giving a total Disability score of 15. She did not use a stick, walked outside regularly and was able to walk on uneven surfaces which enabled her to walk her dog in the New Forest. Point on the Rosser Matrix was therefore D3.


Walking Speed (m / s)










































PCI beats per (min / metres per min)














































Mobility Questionnaire
























































8- 7


6- 5


4 3




























Total Distress (Emotional) score
































Total Disability score


































Although changes in QALY scores are not great these should be seen in context with alternative treatment available. Stroke is a very common and debilitating disease for which there is little effective medical treatment beyond the initial 3 - 9 month rehabilitation stage.

A clinical controlled trial of the Odstock Dropped Foot Stimulator (ODFS) for correction of dropped foot in chronic stroke.

Taylor PN, Burridge JH, Hagan SA Wood DE, Swain ID.

Department of Medical Physics and Biomedical Engineering, Salisbury District Hospital, Salisbury, Wiltshire, SP2 8BJ, UK

Presented at the first IFESS (International Functional Electrical Stimulation Society) Conference at the Case Western Reserve University, Clevland Ohio, USA. 14 -16 May 1996


There are 100,000 hospital admissions for stroke in the U.K. annually; 20% are fatal and 80% of survivors, despite making some recovery, lead a restricted life-style. (Stroke Towards better management R.C.P. 1989). There are no up-to-date statistics on how many of these patients have a drop-foot but 20% (Merletti et al 1989) would be a conservative estimate. 80% of patients referred for treatment are found to be suitable; giving a potential of 9600 patients per year.

Drop-foot following stroke is therefore a common problem. It is thought to be caused partly by poor active control of the anterior tibialis muscles and partly by increased and inappropriate tone in the extensor muscles of the leg; particularly the calf. Drop-foot prevents the patient from effectively swinging the leg when walking causing an abnormal gait characterised by hip hitching and circumlocution and toe catch . The increased effort required not only means that walking is slow, tiring and sometimes unsafe but may lead to further increase in spasticity.

Drop foot is conventionally corrected by splinting usually a plastic ankle foot orthosis and occasionally a more substantial splint attached to the shoe. Of the 32 subjects who took part in this study 8 were using a splint and 13 had rejected this method of correction and 11 had either been advised not to use it or had not been offered it. All subjects who used electrical stimulation to correct the drop-foot continued with this method in preference to conventional splinting. The disadvantages of splinting are as follows: The splint prevents passive dorsiflexion thus making bringing the centre of gravity forward over the base of support when standing from a chair more difficult. It also prevents active dorsiflexion so that the user cannot 'push-off ' at the end of stance phase. Splinting can lead to further loss of muscle control and reduces the proprioceptive input from the ankle joint resulting in further ankle instability, Many patients find the splint uncomfortable sometimes exacerbating ankle oedema. In some cases patients do not find splinting an effective way of overcoming a drop-foot, it may, in fact lead to further increase in calf tone.

The Odstock Dropped Foot Stimulator produces dorsiflexion by electrically stimulating the common peroneal nerve, timed to the gait cycle by a pressure sensitive switch placed in the shoe. The stimulation elitists the withdrawal reflex which consists of dorsiflexion, knee and hip flexion. The relative amounts joint movement can be adjusted by controlling the stimulation output and by electrode placement. Timing can be controlled by either heel rise, where the foot switch is placed under heel of the effected leg or by heel strike, where the foot switch is placed under the heel of the ipsillateral side. The device, after a few days of practice, can be used all day as an orthotic aid or as a training aid in gait re-education.


1. The correction of dropped foot will lead to a more normal gait pattern, improving the symmetry of temporal and spatial gait parameters.

2. The correction of dropped foot will give a more efficient gait causing a reduction in physiological cost index (PCI).

3. The correction of dropped foot will enable faster gait.

Selection criteria.

1. Chronic Hemiplegia of greater than 6 months.

2. A single dropped foot.

3. Sufficient ankle passive range of movement to allow dorsiflexion.

4. Subjects were not significantly mentally impaired and understood the use of the device.

5. A degree of expressive dysphasia was not a contraindication.

6. Significant medical complications was a contraindication.

7. Subjects could stand unsupported and walk at least 10m with an appropriate walking aid.

8. Subjects were able to stand from sitting independently and walk at least 50 m independently prier to stroke.

9. Subjects were not hypersensitive to the sensation of electrical stimulation.

10. Subjects gave written, informed consent.

11. Dorsiflexion could be produced by electrical stimulation of the common peroneal nerve and/or the antieria tibialis.

12. Subjects were able to attend regularly for assessments and treatment.


All subjects were referred by a GP or Consultant. Some subjects were initially self referred following publicity in a national newspaper and some subjects were initially referred by physiotherapists.

32 subjects were selected and randomly allocated to two groups of 16. The treatment group used the device and received 10 sessions of physiotherapy gait re-education. The control group received the same number of gait re-education sessions but without the stimulator. Thus both groups received equal contact time. The physiotherapy was given by a Bobath trained physiotherapist. While more than one physiotherapist took part, each subject was only treated by one therapist. Most of the gait re-education sessions took place in the fist month of the trial.


All subjects were assessed before the onset of treatment, at one month and at three months. The assessments made considered in this paper are as follows:

1. Walking speed and PCI (Physiological Cost Index) measured over 10 m. A 10 m course was marked out in a corridor. 2m at either end of the course was allowed for acceleration and deceleration. The subject was asked to walk briskly. The time to walk the 10 m was recorded and the final heart rate at the end of the walking. The heart rate was recorded using a Polar Heart Rate Monitor. This device consists of a chest strap containing an ECG detector and radio transmitter. The signal is received by a wrist warn device which displays the mean heart rate over the previous 4 heart beats.

PCI (bt/m) = Walking HR - Resting HR (Bt / min)

Walking speed (m / min)

Three measurements were taken, and the average walking speed and PCI recorded. The resting heart rate was recorded prier to the measurement, after the subject had sat quietly for 3 minutes.

2. Gait parameters measured using a Mainstream Locomotion 2D manual video digitisation system. Reflective markers were placed on the toe, heel, ankle, knee, hip, shoulder, elbow and wrist on each side. Additionally the ears was used as a marker. The subject was videod as they walked over a 6m course. The video was then manually digitised using a Acorn A5000 computer using a frame grabber and "clicking" on each point using a mouse. Each marker is then recorded as a two dimensional co-ordinate. The horizontal co-ordinates of the heel and toe of each foot were then plotted against time, producing a "Helix" diagram (see figures 1,2) By measuring along the X axis, temporal measurements can be made allowing swing time, stance time and double stance time (effected leg leading and non-effected leg leading) to be calculated. Measuring along the Y axis gives spatial measurements allowing step length and the distance one foot is placed in front of the other (both effected non-effected side) to be recorded. Each parameter was recorded twice and the mean value recorded. By comparing effected and non effected sides, a measurement of asymmetry of gait can be obtained.


The results are shown here in tabular form. Only results that were statistically significant are given except where they are used to compare with other data.

Speed (10m walk)




% change between groups

T test

Treatment group

no FES

0.64 m/s

0.63 m/s

1.5% decrease


Treatment group FES

0.68 m/s

0.75 m/s

16% increase


There was a significant increase in walking speed in the treatment group with FES in comparison with the treatment group.


PCI(10m walk)




% change between groups

T test

Treatment group no FES

0.80 Bt.m

0.76 Bt.m



Treatment group FES

0.59 Bt.m

0.54 Bt.m

29% decrease

p< 0.01

There was a significant decrease in PCI in the treatment group when the stimulator was used.

Speed (gait analysis)


1st .


% change

T test

Treatment group no FES

0.44 m/s

0.56 m/s

21.5% increase


treatment group FES

0.46 m/s

0.60 m/s

25% increase


Control group

0.40 m/s

0.46 m/s

13% increase


All groups showed an increase in speed at the 3rd assessment.






% change

T test

Treatment group no FES

34.0 steps/s

37.4 steps/s

9% increase


Treatment group FES

34.1 steps/s

38.2 steps/s

10.8% increase

p= 0.02

The treatment group showed an increase in cadence at the 3rd assessment both with and without the stimulator.


Swing time asymmetry





% change 1st - 2nd

T test 1st - 2nd

Treatment group FES

1.2 s

0.935 s

1.4 s


p = 0.049

There was a reduction in asymmetry at 2nd assessment in the treatment group with FES which was lost at the 3rd assessment.

Stance time




% change

T Test

treatment group FES

1.47 s

1.17 s

20.5% reduction

p =0.043

There was a reduction in the time that the hemi leg is on the ground when using FES at the 3rd assessment compared the first.

Double stance time (effected leg leading)




% change

T - test

Treatment group no FES

0.58 s

0.41 s

29.3% reduction

p =0.04

Treatment group FES

0.50 s

0.37 s

26% reduction

p = 0.02

There was a reduction in double stance time when the hemi leg is leading at 3rd assessment compared with the first when using FES. This is also seen in the treatment group without FES. This is not reflected in measurements of asymmetry.

Step length






Treatment group no FES

0.69 m

0.80 m

16% increase

p= 0.04

Treatment group FES

0.82 m

0.95 m

16% increase

p = 0.04

There was an increase in step length at 3rd assessment compared with first assessment both with and without FES for the treatment group.

One foot in front of the other distance




% change

T -test

treatment FES hemi

0.36 m

0.42 m

14.3% increase


treatment group norm

0.37 m

0.42 m

12 % increase


treatment no FES hemi

0.35 m

0.42 m

16.7 % increase


treatment no FES norm

0.35 m

0.40 m

12.5% increase


Increase in both hemi leading and normal leading distance in both treatment group with and with out FES.


10 m Walk

1. There was a significant increase in walking speed when the stimulator was used.

2. There was a significant decrease in PCI when the stimulator was used.

3. There were no significant effects on speed or PCI on either control or treatment group without stimulator.

Gait analysis

1. There was an increase in walking speed in all groups. This was greatest when the stimulator was used but was maintained when walking unaided.

2. Cadence was increased in the treatment group, both with and without the stimulator.

3. Step length and the distance that one foot either foot is placed in front of the other both increased in the treatment group, both with and without the stimulator.

4. There is a reduction in the double stance time when the effected leg is leading both with and without the stimulator in the treatment group and a reduction in stance time while using FES in the treatment group.

5. There was no significant effect on the symmetry of either temporal or Spatial parameters except for a temporary change in the swing time at the second assessment.


The discrepancy in results between measurements in walking speed over 10 m and measurement by gait analysis must be due to the way the measurements were made. Firstly the gait analysis was performed over 6 m this gave a shorter distance for acceleration and deceleration which would restrict the subjects movement. Secondly the subjects may have felt under more pressure to walk in a manor approved by the physiotherapist when under the scrutiny of the gait analysis video cameras while concentrated more on the simpler task of walking down the corridor on the 10 m walk. Finally the calculation of speed in the gait analysis is derived from the step length divided by the summation of the temporal parameters. This may lead to a greater error in the measurement. This discrepancy also casts doubt on the general significance of the rest of the data from the gait analysis which can only be considered valid in the particular circumstances of this measurement.

The lack of improvement in PCI and speed without the stimulator while walking over 10m may in comparison to the improvements recorded in speed by the gait analysis suggest that while there is some carry-over effect when the subject is walking carefully, this is not having a significant effect on every day life which is probably more accurately represented by the 10 m walk.

No significant effect was found on gait asymmetry. This suggests that while mobility may be improved by correction of the dropped foot, it is only one small component of gait leaving significant other problems. Some of these may be addressed by stimulating additional muscle groups. Increases in step length and reduction in stance and double stance time are consistent with increased walking speed and cadence.

Perhaps the most clinically significant result is the reduction in PCI. This means that using the stimulator saves the user effort while walking. This is a very direct return, enabling the user walk further and longer.


We would like to acknowledge the Department of Health Medical devices agency for funding this work, the physiotherapists who treated the patients and the Wessex Rehabilitation Association for allowing us to inhabit their building. We are particularly indebted to Mrs J Watkins for funding the Mainstream Locomotion Video Gait Analysis System. We would also like to thank James Burridge and Simon Gallaghar for assistance with video digitisation.


  1. Royal College of Physicians, Stroke: Towards better management Dec 1989).
  2. Karsnia, A. Dillner, S. Ebefors, I. Lundmark, P. (1990) Why patients use or reject a peroneal muscle stimulator. Advances in External Control of Human Extremities
  3. Liberson, W.T. Holmquest, H.J. Scott, M.E.D. (1961) Functional Electrotherapy: Stimulation of the Common Peroneal Nerve Synchronised with the swing phase of gait of Hemiplegic subjects. Arch Phys Med and Rehab Fe, 101-105.
  4. Maxwell, D.J. Granat, M. et al. Evaluation of the Peroneal Stimulator. (1995) BES Symposium Strathclyde University.
  5. Merletti, R. Andina, A. Galante, M. Veruno Italy. (1979). Clinical Experience of Electronic Peroneal Stimulators in 50 Hemiparetic patients. Scand J Rehab Med 11:111-121
  6. Waters, R.L., McNeal, D., Perry, J. (1985) Functional Electrical Stimulation of the Peroneal Nerve for Hemiplegia. Journal of Bone and Joint Surgery, vol.67-A No 5 June
  7. Stefancic, M. Rebersek, M. Merletti, R. (1976) The therapeutic effects of the Ljubljana functional electronic brace. Europa Medicophysica. 12,1-9.
  8. Kljajic, M., Acimov, R., Malezic, M., Stanic, U., et al Long Trm follow-up of gait in Hemiparetic patients using subcutaneous Peroneal electrical stimulation. (undated).
  9. Taylor, P.N., Ewins, D. J., Swain, I.D. Development and evaluation of Functional electrical stimulation orthosis for spinal cord injured patients. Report to the Department of Health Oct 1992.
  10. Wade, D. T. Wood, V.A. Heller, A. Maggs, J. Langon Hewer, R. (1987) Walking after stroke. Scand J Rehab Med 19: 25-30.
  11. Bailey, M.J. Ratcliffe, C.M. Reliability of Physiological Cost Index in walking normal subjects using steady-state and non-steady state and post exercise recordings. Physiotherapy (October 1995, vol 81 No 10)
  12. Jennings, S.J., Nene, A.V, Tait, J.H., Major, R.E. (1991) An automated system for measuring Physiological Cost Index during clinical gait assessment. J Biomed Eng
  13. Nene, A. V., ;Jennings, S.J. (1992) Physiological Cost Index of paraplegic locomotion using the ORLAU Para Walker. Paraplegia 30 246-252
  14. J. MacGregor. The objective measurement of physical performance with long term ambulatory surveillance equipment. Proc. 3rd Symposium on Ambulatory Monitoring. 1979, p. 29 -39.