Functional Electrical Stimulation – Lower Extremity

Two fair quality RCTs (MacDonell et al., 1994, Kojovic et al., 2009) and one quasi-experimental design study (Solopova et al., 2011) have investigated the effectiveness of FES on ADLs in patients with acute stroke.

The first fair quality RCT (MacDonell et al., 1994) investigated the effectiveness of FES for improving performance in activities of daily living in 38 patients with acute stroke with weakness of dorsiflexion Grade 4 or less measured by the Medical Research Council Scale (MRC). The treatment group received FES five times a week for four weeks producing ankle dorsiflexion of the affected leg in addition to standard physical therapy, while the control group received standard physical therapy alone. Outcomes were measured at four weeks (post-treatment) and at eight weeks. No significant between-group difference was found on performance of activities of daily living at post-treatment (four weeks) or at the eight week follow-up as measured by the Barthel Index.

The second fair quality RCT (Kojovic et al., 2009) investigated the effectiveness of FES for improving ADL in 13 patients with acute stroke who were able to ambulate with a single cane or hand support. The patients were randomly assigned to either the functional electrical therapy (FET) group, or a control group. Both FET and control groups participated in a standard rehabilitation program and walking sessions. During the walking sessions, patients used the tripod cane or were physically assisted by the therapist in addition to instructions given by the therapist on how to improve their walking pattern. The FET group used a sensor-driven electrical stimulator device to stimulate four muscle groups: quadriceps, hamstring, soleus and tibialis anterior of the paretic leg in order to improve knee flexion/extension and ankle flexion/extension during walking. Outcomes were assessed at post-treatment (four weeks). At the end of the four week intervention, the Barthel Index score was significantly higher (reflecting better functioning) in the FET group compared to the control group.

The quasi-experimental design study (Solopova et al., 2011) pseudo-randomized patients with acute stroke to receive FES with assisted passive/active locomotor-like leg movements and progressive limb loading (FES group) or conventional rehabilitation (control group). FES was applied to the soleus, tibialis anterior, hamstring, quadriceps, hip adductors, gluteus maximus and medial gastrocnemius muscles as participants performed leg movements while semi reclined on a tilt table. ADL function was measured by the Barthel Index. At post-treatment (2 weeks) a significant between-group difference in ADL function was seen in favour of the FES group compared to the control group.

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT and 1 quasi-experimental design study that FES is more effective than conventional rehabilitation in improving ADL performance in patients with acute stroke.

Note: However, another fair quality RCT found that FES was not more effective than conventional rehabilitation alone in improving ADL performance among patients with acute stroke. Both fair quality RCTs had small sample sizes, which may in part contribute to different results between studies.

One fair quality RCT (MacDonell et al., 1994) has investigated the effectiveness of FES for improving electrophysiological functioning in patients with acute stroke.

The fair quality RCT (MacDonell et al., 1994) investigated the effectiveness of FES for improving electrophysiological functioning in 38 patients with acute stroke and weakness of dorsiflexion Grade 4 or less as measured by the Medical Research Council Scale (MRC). The treatment group received FES producing ankle dorsiflexion of the affected leg in addition to standard physical therapy and the control group received standard physical therapy alone. Outcomes were measured at post-treatment (four weeks) and at eight weeks. No significant between-group difference was found on any electrophysiological measure at post-treatment or at the eight week follow-up as measured by foot tap frequency, duration and frequency of electromyographic (EMG) burst activity in tibialis anterior, H max/ M max ratio in gastocnemius, degree of vibratory inhibition of the H reflex, and F mean/ M max ratio using responses from the flexor hallucis brevis.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that FES in combination with standard physical therapy is not more effective than standard physical therapy alone for improving electrophysiological functioning in patients with acute stroke.

Two high quality RCTs (Ferrante et al., 2008, Yan et al., 2005) and one quasi-experimental design study (Solopova et al., 2011) investigated the effectiveness of FES for improving maximum isometric voluntary contraction in patients with acute stroke.

The first high quality RCT (Ferrante et al., 2008) investigated the use of FES for improving MIVC in 20 patients with acute stroke. The participants were randomized to receive either (1) standard rehabilitation and FES-cycling (5 minutes of passive cycling, 10 minutes of FES, 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling); or (2) standard rehabilitation alone. Outcomes were measured at post-treatment (four weeks). Following four weeks of treatment, a significant between-group difference in favour of the group receiving FES-cycling was found on the quadriceps isometric maximum force as measured by maximal voluntary contraction.

The second high quality RCT (Yan et al., 2005) investigated the effectiveness of FES on maximum isometric voluntary contraction (MIVC) of ankle dorsiflexors and planter flexors in 46 patients with acute stroke. Patients received either (1) standard rehabilitation and FES-cycling; (2) standard rehabilitation and placebo stimulation (placebo); or (3) standard rehabilitation alone (control) for three weeks. All participants were assessed on MIVC and EMG co-contraction ratio during ankle dorsiflexion and plantarflexion at one, two, three and 8 weeks. During ankle dorsiflexion, percent increases in MIVC torques and integrated EMG of the FES group were significantly larger than those of the control group from week one onward, and significantly larger than the placebo group at week three only. Also, the EMG co-contraction ratio during dorsiflexion of the affected ankle showed a significantly greater reduction in the FES group compared to the other groups from week one or two onward, reflecting greater improvement. For ankle plantarflexion, a significant difference was found in favour of the FES group only at week three only compared to the other groups. No significant between group difference was found on either measure at the 8 week follow-up.

The quasi-experimental design study (Solopova et al., 2011) pseudorandomized patients with acute stroke to receive FES with assisted passive/active locomotor-like leg movements and progressive limb loading (FES group) or conventional rehabilitation (control group). FES was applied to the soleus, tibialis anterior, hamstring, quadriceps, hip adductors, gluteus maximus and medial gastrocnemius muscles as participants performed leg movements while semi reclined on a tilt table. At post-treatment (2 weeks) there were significant between-group differences in maximum isometric voluntary contraction of the paretic flexor and extensor muscles and the nonparetic extensors, in favour of the FES group compared to the control group. There was also a significant between-group difference in EMG patterns of the rectus femoris and biceps femoris muscles during knee flexion of the ipsilateral leg, in favour of the intervention group compared to the control group.

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs and one quasi-experimental study that FES in addition to standard rehabilitation is more effective than standard rehabilitation alone for improving maximal isometric voluntary contraction in patients with acute stroke.

Note:mprovements were not maintained beyond follow-up in one of the high quality RCT.

One high quality RCT (Ferrante et al., 2008) has investigated the effectiveness of FES for improving mobility in patients with acute stroke.

The high quality RCT (Ferrante et al., 2008) examined the effectiveness of FES on mobility in 20 patients with acute stroke. Patients were randomized to receive either (1) standard rehabilitation and FES-cycling (5 minutes of passive cycling, 10 minutes of FES, 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling; or (2) standard rehabilitation alone. Outcomes were measured at post-treatment (four weeks). No significant between-group differences were found for mobility of the ankle, or knee and hip joint during voluntary movement, as measured by the Motricity Index (MI) following 4 weeks of treatment. Due to the small sample size, failure to find group differences may have occurred because of type 2 error.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to standard rehabilitation is not more effective than standard rehabilitation alone for improving mobility in patients with acute stroke.

One high quality RCT (Ferrante et al., 2008) investigated the use of FES for improving motor control in patients with acute stroke.

The high quality RCT (Ferrante et al., 2008), examined the effectiveness of FES on motor control in 20 patients with acute stroke. The participants were randomized to receive either (1) standard rehabilitation and FES-cycling (5 minutes of passive cycling, 10 minutes of FES, 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling); or (2) standard rehabilitation alone. Patients were asked to perform sit to stand trials at 3 different speeds (patient selected, one slower than patient selected, and one faster than patient selected) to test motor control. Outcomes were measured at post-treatment (four weeks). Following four weeks of treatment, a significant between-group difference in favour of the FES-cycling group was found on the percentage ratio between the slow and self selected speeds. The percentage ratio between fast and self selected speed failed to reach significance.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to standard rehabilitation is more effective than standard rehabilitation alone for improving motor control in patients with acute stroke.

Two fair quality RCTs (MacDonell et al., 1994, Kojovic et al., 2009) and one quasi-experimental design study (Solopova et al., 2011) have examined the effectiveness of FES for improving motor function in patients with acute stroke.

The first fair quality RCT (MacDonell et al., 1994) investigated the effectiveness of FES for improving motor function in 38 patients with acute stroke with weakness of dorsiflexion Grade 4 or less measured by the Medical Research Council Scale (MRC). The treatment group received FES five times a week for four weeks producing ankle dorsiflexion of the affected leg in addition to standard physical therapy, while the control group received standard physical therapy alone. Outcomes were measured at post-treatment (four weeks) and at eight weeks. No significant between-group difference was found in motor function at post-treatment or at eight week follow-up measured by the Fugl-Meyer Lower Extremity Motor Assessment Scale.

The second fair quality RCT (Kojovic et al., 2009) investigated the effectiveness of FES for improving motor function in 13 patients with acute stroke who were able to ambulate with a single cane or hand support. The patients were randomly assigned to either the functional electrical therapy (FET) group, or a control group. Both FET and control groups participated in a standard rehabilitation program and walking sessions for four weeks. During the walking sessions, patients used the tripod cane or were physically assisted by the therapist in addition to instructions given by the therapist on how to improve their walking pattern. The FET group used a sensor-driven electrical stimulator device to stimulate four muscle groups: quadriceps, hamstring, soleus and tibialis anterior of the paretic leg in order to improve knee flexion/extension and ankle flexion/extension during walking. At the end of the 4 week intervention, the FET group had significantly higher scores on the Fugl-Meyer Lower Extremity Motor Assessment Scale compared to the control group.

The quasi-experimental design study (Solopova et al., 2011) pseudorandomized patients with acute stroke to receive FES with assisted passive/active locomotor-like leg movements and progressive limb loading (FES group) or conventional rehabilitation (control group). FES was applied to the soleus, tibialis anterior, hamstring, quadriceps, hip adductors, gluteus maximus and medial gastrocnemius muscles as participants performed leg movements while semi reclined on a tilt table. Motor function was measured by the Fugl Meyer Assessment. At post-treatment (2 weeks) a significant between-group difference in motor function was seen in favour of the FES group compared to the control group. Significant between-group differences in motor function were also seen among subgroups of patients with paralysis, severe hemiparesis and pronounced hemiparesis, in favour of the FES group compared to the control group.

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT and 1 quasi-experimental design study that FES is more effective than conventional rehabilitation in improving motor function in patients with acute stroke.

Note: However, another fair quality RCT found that FES was not more effective than conventional rehabilitation alone in patients with acute stroke. Both fair quality RCTs had small sample sizes, which may in part contribute to different results between studies.

One high quality RCT (Ferrante et al., 2008) has investigated the use of FES for improving muscle strength in patients with acute stroke.

The high quality RCT (Ferrante et al., 2008) examined the effectiveness of FES on muscle strength in 20 patients with acute stroke. Patients were randomized to receive either (1) standard rehabilitation and FES cycling FES-cycling (5 minutes of passive cycling, 10 minutes of FES, 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling); or (2) standard rehabilitation alone for four weeks. No significant between group difference was found for muscle strength of the hemiplegic limb, as measured by the Upright Motor Control Test following 4 weeks of treatment. Due to the small sample size, failure to find group differences may have occurred because of type 2 error.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to standard rehabilitation is not more effective than standard rehabilitation alone for improving muscle strength of the hemiplegic limb patients with acute stroke.

One quasi-experimental design study (Solopova et al., 2011) has investigated the effectiveness of FES on severity of stroke in patients with acute stroke.

The quasi-experimental design study (Solopova et al., 2011) pseudorandomized patients with acute stroke to receive FES with assisted passive/active locomotor-like leg movements and progressive limb loading (FES group) or conventional rehabilitation (control group). FES was applied to the soleus, tibialis anterior, hamstring, quadriceps, hip adductors, gluteus maximus and medial gastrocnemius muscles as participants performed leg movements while semi reclined on a tilt table. Stroke severity was measured by the National Institutes of Health Stroke Scale (NIHSS) and European Stroke Scale (ESS). At post-treatment (2 weeks) a significant between-group difference in scores was seen in favour of the FES group compared to the control group.

Conclusion: There is limited evidence (level 2b) from 1 quasi-experimental study that FES is more effective than conventional rehabilitation in improving severity of stroke in patients with acute stroke.

Two high quality RCTs (Yan et al., 2005, Yeh et al., 2010) have investigated the effectiveness of FES for improving spasticity in patients with acute stroke.

The first high quality RCT (Yan et al., 2005) randomized patients to receive (1) standard rehabilitation and FES (FES group); (2) standard rehabilitation and placebo stimulation (placebo group); or standard rehabilitation alone (control group) for 3 weeks. Spasticity was measured by the Composite Spasticity Scale (CSS). At 3 weeks (post-treatment) a significant between-group difference was found in favour of the FES group compared to the placebo group and the control group. No significant between-group difference was found at the 8 week follow-up.

A high quality cross-over design study (Yeh et al., 2010) randomized patients to one 20-minute session of cycling with FES or one 20-minute session of cycling without FES. The groups performed the opposite treatment on the second intervention day. Hypertonicity was measured by the Modified Ashworth Scale (MAS) and the pendulum test (relaxation index and peak velocity). Both treatments were found to improve hypertonia, as seen by significantly reduced MAS scores and significantly improved pendulum test scores from pre- to post-testing. After the second intervention day there were significant between-group differences in all measures of hypertonia, in favour of the FES group compared to the non-FES group.

Conclusion: There is strong evidence (level 1a) from one high quality RCT and one high quality randomized cross-over design study that FES is more effective than standard rehabilitation alone for improving spasticity in patients with acute stroke in the short term.

One high quality RCT (Ferrante et al., 2008) has investigated the effectiveness of FES for improving trunk movement and balance in patients with acute stroke.

The high quality RCT (Ferrante et al., 2008), examined the effectiveness of FES on trunk movements and balance in 20 patients with acute stroke. Patients were randomized to receive either (1) standard rehabilitation and FES-cycling (5 minutes of passive cycling, 10 minutes of FES), 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling; or (2) standard rehabilitation alone for four weeks. No significant between group differences were found for trunk movement and balance, as measured by the Trunk Control Test (TCT) following four weeks of treatment. Due to the small sample size failure to find group differences may have occurred because of type 2 error.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to standard rehabilitation is not more effective than standard rehabilitation alone for improving trunk movement and balance in patients with acute stroke.

Two high quality RCTs (Ferrante et al., 2008, Yan et al., 2005), one fair quality RCT (MacDonell et al., 1994) and one quasi-experimental design study (Solopova et al., 2011) have investigated the effectiveness of FES for improving walking ability in patients with acute stroke.

The first high quality RCT (Ferrante et al., 2008) investigated the use of FES for improving walking ability in 20 patients with acute stroke. The participants were randomized to receive either (1) standard rehabilitation and FES-cycling (5 minutes of passive cycling, 10 minutes of FES, 5 minutes of passive cycling, 10 minutes of FES and 5 minutes of passive cycling); or (2) standard rehabilitation alone for four weeks. Following four weeks of treatment, the group that received FES – cycling showed more improvement in a 50m walk test than the control group, however, the between group difference failed to reach statistical significance.

The second high quality RCT (Yan et al., 2005) investigated the effectiveness of FES for improving walking ability in 46 patients with acute stroke. Patients received either (1) standard rehabilitation or FES; (2) standard rehabilitation and placebo stimulation (placebo); or (3) standard rehabilitation alone (control) for three weeks. Outcomes were measured at one, two, three and 8 weeks. No significant between group difference was found for walking ability, measured by the Timed Up and Go (TUG) test at all periods of assessment.

The fair quality RCT (MacDonell et al., 1994) investigated the effectiveness of FES for improving walking ability in 38 patients with acute stroke and weakness in dorsiflexion Grade 4 or less as measured by the Medical Research Council Scale (MRC). The treatment group received FES producing ankle dorsiflexion of the affected leg in addition to standard physical therapy and the control group received standard physical therapy alone for four weeks. Outcomes were measured at post-treatment and at eight weeks. No significant between-group difference was found in walking ability at post-treatment or at eight week follow-up as measured by the Massachusetts General Hospital Functional Ambulation Categories (MGH FAC) scale.

The quasi-experimental design study (Solopova et al., 2011) pseudorandomized patients with acute stroke to receive FES with assisted passive/active locomotor-like leg movements and progressive limb loading (FES group) or conventional rehabilitation (control group). FES was applied to the soleus, tibialis anterior, hamstring, quadriceps, hip adductors, gluteus maximus and medial gastrocnemius muscles as participants performed leg movements while semi reclined on a tilt table. Walking ability was measured by movement in the knee and ankle joints. At post-treatment (2 weeks) a significant difference in ankle range of movement was found in favour of the FES group compared to the control group, however there were no differences between groups in knee joint motion.

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs, one fair quality RCT and and one quasi-experimental study that FES in addition to standard rehabilitation is not more effective than standard rehabilitation alone for improving walking ability in patients with acute stroke.

One fair quality RCT (Kojovic et al., 2009) has investigated the effectiveness of FES for improving walking efficiency in patients with acute stroke.

In the fair quality RCT, (Kojovic et al., 2009) investigated the effectiveness of FES for improving walking efficiency in 13 patients with acute stroke who were able to ambulate with a single cane or hand support. The patients were randomly assigned to either the functional electrical therapy (FET) group, or a control group. Both FET and control groups participated in a standard rehabilitation program and walking sessions for four weeks. During the walking sessions, patients used the tripod cane or were physically assisted by the therapist in addition to instructions given by the therapist on how to improve their walking pattern. The FET group used a sensor-driven electrical stimulator device to stimulate four muscle groups: quadriceps, hamstring, soleus and tibialis anterior of the paretic leg in order to improve knee flexion/extension and ankle flexion/extension during walking. At the end of the 4 week intervention, a statistically significant decrease in the Physiological Cost Index (PCI) was observed between groups in favour of the FET group, reflecting more efficient walking in the treatment group.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that FES in addition to standard therapy is more effective than standard therapy alone for improving walking efficiency in patients with acute stroke.

One fair quality RCT (Kojovic et al., 2009) has investigated the effectiveness of FES for improving walking speed in patients with acute stroke.

In the fair quality RCT, (Kojovic et al., 2009) investigated the effectiveness of FES for improving walking efficiency in 13 patients with acute stroke who were able to ambulate with a single cane or hand support. The patients were randomly assigned to either the functional electrical therapy (FET) group, or a control group. Both FET and control groups participated in a standard rehabilitation program and walking sessions for four weeks. During the walking sessions, patients used the tripod cane or were physically assisted by the therapist in addition to instructions given by the therapist on how to improve their walking pattern. The FET group used a sensor-driven electrical stimulator device to stimulate four muscle groups: quadriceps, hamstring, soleus and tibialis anterior of the paretic leg in order to improve knee flexion/extension and ankle flexion/extension during walking. At the end of the 4 week intervention, a statistically significant increase in favour of the FES group was found on mean walking velocity over a 6 meter walkway.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that FES in addition to standard therapy is more effective than standard therapy alone for improving walking speed in patients with acute stroke.

One high quality RCT (Tong et al., 2006) and one fair quality RCT (Granat et al., 1996) have investigated the effectiveness of FES on ADL in patients with subacute stroke.

The high quality RCT (Tong et al., 2006) randomized patients with subacute stroke to receive (1) gait training using an electromechanical gait trainer with functional electric stimulation (EGT-FES), (2) gait training using an electromechanical gait trainer (EGT), or (3) conventional gait training (CGT, control) for four weeks. Performance of ADLs was measured using the Barthel Index. At 4 weeks (post-treatment) no significant between-group differences in ADLs were found.

One fair quality crossover RCT (Granat et al., 1996) assigned patients with subacute and chronic stroke to receive four weeks of standard rehabilitation followed by four weeks of FES intervention. ADLs were measured using the Barthel Index. A significant improvement in ADLs was seen following the intervention period but not following the control period.

Conclusion : There is moderate evidence (Level 1b) from one high quality RCT that FES is not more effective than conventional or electromechanical gait training for improving performance of ADLs in patients with subacute stroke.

Note: However, there is limited evidence (level 2a) from one fair quality RCT that FES is more effective than standard rehabilitation in improving performance of ADLs in patients with subacute and chronic stroke.

One high quality RCT (Tong et al., 2006) has investigated the effectiveness of FES on balance in patients with subacute stroke.

In the high quality RCT (Tong et al., 2006) has investigated the effectiveness of gait training using an electromechanical gait trainer with and without functional electric stimulation for improving balance in 46 patients with sub-acute stroke. The patients were randomly assigned to receive either: (1) gait training using an electromechanical gait trainer with functional electric stimulation (EGT-FES); (2) gait training using an electromechanical gait trainer (EGT), or (3) conventional gait training (CGT, control) for four weeks. At the end of the four week intervention, no significant between group difference was found on the Berg Balance Scale.

Conclusion: There is moderate evidence (level 1b) from one high quality RCT that FES not more effective than electromechanical or conventional gait training alone for improving balance in patients with sub-acute stroke.

One high quality cross-over RCT (Bogataj et al., 1995) and 2 fair quality RCTs (Newsam et al., 2004, Winchester et al., 1983) has investigated the effectiveness of FES on maximum isometric voluntary contraction (MIVC) and contraction force in patients with subacute stroke.

The high quality cross-over RCT (Bogataj et al., 1995) randomized patients with subacute stroke and severe hemiplegia to receive either (1) 3 weeks of multi-channel functional electrical stimulation (MFES) followed by 3 weeks of conventional physiotherapy; or (2) 3 weeks of conventional therapy followed by 3 weeks of MFES. Gait stability was measured by the vertical components of ground reaction force and the trajectories of center of pressure (TCP) for each foot in the stance phase. A significant between-group difference in TCP was found, in favour of FES compared to conventional therapy.

The first fair quality RCT (Newsam et al., 2004) assigned patients with subacute stroke to receive physical therapy and an electric stimulation facilitation program of the quadriceps (FES group), or physical therapy alone (control group). A significant between-group difference in supramaximal contraction torque of the quadriceps was found at 3 weeks (post-treatment) in favour of the FES group compared to the control group. There was no significant difference in MIVC of the quadriceps.

The second fair quality RCT (Winchester et al., 1983) randomized patients with subacute stroke to receive either standard physiotherapy and FES (FES group) or standard physiotherapy alone (control group). Contraction force was measured by knee extension torque. A significant between-group difference in contraction force was seen at 3 weeks (during training) and 4 weeks (post-treatment) in favour of the FES group compared to the control group.

Conclusion: There is moderate evidence (level 1b) from 1 high quality cross-over RCT and 2 fair quality RCTs that FES is more effective than standard rehabilitation for increasing contraction force in patients with subacute stroke.

Note: However, there is limited evidence (level 2a) from 1 fair quality RCT that FES is not more effective than standard rehabilitation for improving MIVC in patients with subacute stroke.

Two high quality RCTs (Yavuzer et al., 2006; Yavuzer et al., 2007) and 2 fair quality RCTs (Sabut et al., 2010b, Granat et al., 1996) examined the effectiveness of FES on gait kinematics and parameter in patients with subacute stroke.

The first high quality RCT (Yavuzer et al., 2006) randomly assigned patients with subacute stroke to receive either (1) conventional rehabilitation and neuromuscular electric stimulation applied to the tibialis anterior for 10 minutes per session (NMES group) or (2) conventional rehabilitation alone (control group). Gait kinematics were measured by walking velocity, step length, percentage of stance phase at the paretic side, sagittal plane kinematics of the pelvis, hip, knee and ankle, maximum ankle dorsiflexion at swing and maximum ankle plantar flexion angle at initial contact. No significant between-group differences in gait kinematics were found at 4 weeks (post-intervention).

The second high quality RCT (Yavuzer et al., 2007) randomly assigned patients with subacute stroke to receive (1) conventional rehabilitation and sensory-amplitude electric stimulation (SES) to the paretic leg, or (2) standard rehabilitation alone (control group). Gait kinematics were measured by walking velocity, step length, percentage of stance phase at the paretic side, sagittal plane kinematics of the pelvis, hip, knee and ankle, maximum ankle dorsiflexion at swing and maximum ankle plantar flexion angle at initial contact. No significant between-group differences in gait kinematics were found at 4 weeks (post-treatment).

The first fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control) for 12 weeks. Gait kinematics and parameters were taken as a measure of cadence, step length, step width, and toe-in toe-out. No significant between-group difference in gait kinematics and parameters were seen at 12 weeks (post-treatment).

The second fair quality crossover RCT (Granat et al., 1996) assigned patients with subacute or chronic stroke to receive four weeks of conventional rehabilitation followed by four weeks of FES intervention. Gait kinematics and parameters were taken as a measure of swing symmetry, heel strike and foot inversion. Group results at 11 weeks (post-treatment) revealed a significant improvement in foot inversion on linoleum, carpet and uneven ground, and swing symmetry on linoleum when receiving FES as compared to when not receiving FES.

Conclusion: There is strong evidence (Level 1a) from 2 high quality RCTs and 1 fair quality RCT that FES is not more effective than standard rehabilitation alone for improving gait kinematics and parameters in patients with subacute stroke.

NOTE: However, 1 fair quality crossover RCT found that FES was more effective than conventional rehabilitation in improving some aspects of gait kinematics and parameters in patients with subacute and chronic stroke.

One fair quality RCT (Winchester et al., 1983) has investigated the effect of FES on joint position sense in patients with subacute stroke.

The fair quality RCT (Winchester et al., 1983) investigated the effects of standard physiotherapy and FES (FES group) compared to standard physiotherapy alone (control group) in patients with subacute stroke. No significant between-group difference in joint position sense was found at 4 weeks (post-treatment).

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT that FES and standard rehabilitation is not more effective than standard rehabilitation alone for improving joint position sense in patients with subacute stroke.

One pre-post design study (Sabut et al., 2010a) has investigated the effect of FES on metabolic function in patients with subacute and chronic stroke.

The pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Metabolic function was measured by oxygen consumption, carbon dioxide production, heart rate and energy cost. A significant improvement in metabolic function was seen at 12 weeks (post-treatment).

Conclusion: There is limited evidence (level 2b) from one pre-post design study that FES and standard therapy is effective for improving measures of metabolic response in patients with subacute and chronic stroke.

One high quality RCT (Tong et al., 2006) investigated the effectiveness of FES on mobility in patients with sub-acute stroke.

The high quality RCT (Tong et al., 2006) investigated the effectiveness of gait training using an electromechanical gait trainer with and without functional electric stimulation for improving mobility in 46 patients with sub-acute stroke. The patients were randomly assigned to receive either: (1) gait training using an electromechanical gait trainer with functional electric stimulation (EGT-FES), (2)gait training using an electromechanical gait trainer (EGT), or (3) conventional gait training (CGT, control) for four weeks. At the end of the four week intervention, there was a significant between-group difference on the Elderly Mobility Scale (EMS) and on the Five-meter Walking Speed Test in the EGT and EGT-FES groups compared to controls, however no significant between-group differences were found on the Functional Independence Measure (FIM).

Conclusion: There is moderate evidence (level 1b) from one high quality RCT that FES in addition to gait training is more effective than gait training alone for improving mobility in patients with sub-acute stroke.

Three high quality RCTs (Ambrosini et al., 2011; Bogataj et al., 1995; Tong et al., 2006) and 1 fair quality RCT (Sabut et al., 2010b) have investigated the effectiveness of FES for improving motor function in patients with subacute stroke.

The first high quality RCT (Ambrosini et al., 2011) randomized patients with subacute stroke or traumatic brain injury to either an FES-induced cycling group (FES group) or a placebo FES cycling group (control group). Function of the paretic limb was measured by the Upright Motor Control Test and by pedaling unbalance between the paretic and non-paretic legs, and motor power of the paretic limb was measured by the Motricity Index leg subscale. Significant between-group differences in all measures were seen at 4 weeks (post-treatment), in favour of the FES group compared to the control group. Results remained significant for the Upright Motor Control Test and the Motricity Index at follow-up (3-5 months post-treatment), in favour of the FES group.

The second high quality cross-over RCT (Bogataj et al., 1995) randomized patients with subacute stroke and severe hemiplegia to receive: (1) 3 weeks of multi-channel functional electrical stimulation (MFES) followed by 3 weeks of conventional physiotherapy; or (2) 3 weeks of conventional physiotherapy followed by 3 weeks of MFES. Functional motor status was assessed using the Fugl-Meyer Assessment. A significant improvement in motor function was seen at the middle of intervention (week 3), in favour of FES compared to conventional physiotherapy. However, no differences were seen at week 6.

The third high quality RCT (Tong et al., 2006) investigated the effectiveness of gait training using an electromechanical gait trainer with and without functional electric stimulation for improving motor impairment in 46 patients with sub-acute stroke. The patients were randomly assigned to receive either: (1) gait training using an electromechanical gait trainer with functional electric stimulation (EGT-FES), (2) gait training using an electromechanical gait trainer (EGT), or (3) conventional gait training (CGT, control) for four weeks. At the end of the four week intervention, there was a significant between group difference on the Motricity Index Leg subscale (MI) in the EGT and EGT-FES groups compared to control group.

One fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. Motor function was measured by the Fugl-Meyer Assessment. No significant between-group differences were reported at 12 weeks (post-treatment), although the authors reported “better” improvement was made with FES.

Conclusion: There is strong evidence (level 1a) from 3 high quality RCTs that FES is more effective than control therapies (e.g. placebo intervention, conventional rehabilitation) in improving motor function in patients with subacute stroke. A fair quality RCT also reported improved motor function following FES, although between-group differences were not reported.

Two high quality RCTs (Yavuzer et al., 2007, Yavuzer et al., 2006) have investigated the effect of FES on motor recovery in patients with subacute stroke.

The first high quality RCT (Yavuzer et al., 2007) randomized patients with subacute stroke to receive either sensory amplitude electrical stimulation (SES group) or sham stimulation (control group). Motor recovery was measured using the Brunnstrom Stages of Motor Recovery. No significant between-group difference in motor recovery was seen at 4 weeks (post-treatment).

The second high quality RCT (Yavuzer et al., 2006) randomly assigned patients with subacute stroke to receive either conventional stroke rehabilitation and neuromuscular electric stimulation to the tibialis anterior (NMES group) or conventional therapy alone (control group). Motor function was measured using the Brunnstrom Stages of Motor Recovery. No significant between-group difference in motor recovery was seen at 4 weeks (post-treatment).

Conclusion: There is strong (level 1a) evidence from 2 high quality RCTs that FES is not more effective than standard rehabilitation in improving motor recovery in patients with subacute stroke.

One fair quality RCT (Newsam et al., 2004) has investigated the effectiveness of FES on motor unit recruitment in patients with sub-acute stroke.

In the fair quality RCT (Newsam et al., 2004) investigated the effectiveness of FES on motor unit recruitment in 20 patients with sub-acute stroke. The treatment group received an electric stimulation facilitation program of the quadriceps during gait training and weight-bearing activities as well as standard physical therapy. The control group received standard physical therapy alone for three weeks. To measure the interpolated twitch and determine the motor unit recruitment, a short-duration, supramaximal electric stimulus was delivered at approximately three seconds into each of the maximal tests. After the three week intervention there was a significant between group difference in motor unit recruitment in favour of the patients that received FES.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that FES of the quadriceps in combination with standard physical therapy is more effective than standard physical therapy alone for increasing motor unit recruitment in the quadriceps in patients with sub-acute stroke.

One fair quality RCT (Winchester et al., 1983) has investigated the effects of FES on muscle girth in patients with subacute stroke.

The fair quality RCT (Winchester et al., 1983) investigated the effects of standard physiotherapy and FES (FES group) compared to standard physiotherapy alone in patients with subacute stroke. No significant between-group difference in quadriceps muscle girth was seen at 4 weeks (post-treatment).

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT that FES is not more effective than standard rehabilitation in improving muscle girth of the quadriceps in patients with subacute stroke.

One fair quality RCT (Sabut et al., 2010b) and 1 pre-post design study (Sabut et al., 2010a) have investigated the effect of FES on muscle strength in patients with subacute stroke.

The fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. Muscle output was measured by the maximum value of the root mean square on electromyography analysis. No between-group comparisons in muscle strength were reported, however the authors state that the FES group showed a “better” improvement than the control group at 12 weeks (post-treatment).

The pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Muscle strength was measured by the mean absolute value, root mean square, median frequency and median amplitude of the tibialis anterior muscle EMG signal. A significant effect of FES treatment on muscle strength was found at 12 weeks (post-treatment).

Conclusion: There is limited evidence (level 2b) from 1 fair quality RCT and 1 pre-post design study that FES is effective improving muscle strength in patients with subacute stroke.

Note: Neither study reported between-group differences so it cannot be determined whether FES is more effective than control therapy in improving muscle strength in patients with subacute stroke.

Two fair quality RCTs (Sabut et al., 2010b; Winchester et al., 1983) have investigated the effect of FES on range of motion in patients with subacute stroke.

The first fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. No between-group comparisons on ankle range of motion were made, however the authors state that the FES group showed “better” improvement than the control group at 12 weeks (post-treatment).

The second fair quality RCT (Winchester et al., 1983) randomized patients with subacute stroke to receive either standard physiotherapy and FES (FES group) or standard physiotherapy alone (control group). Range of motion was measured weekly during the 4-week intervention. A significant between-group difference in active range of motion was found at week 2 (during treatment) in favour of the FES group compared to the control group. No significant differences were found at week 1 or 3 of treatment, or at week 4 (post-treatment).

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT that FES is not more effective than standard rehabilitation in improving range of motion in patients with subacute stroke, although temporary gains were seen mid-treatment.

Note: One fair quality RCT found improvement in ankle range of motion following FES, however these results are not used to determine level of evidence as between-group differences were not reported.

Two fair quality RCTs (Sabut et al., 2010b; Winchester et al., 1983) have investigated the effectiveness of FES on spasticity in patients with subacute stroke.

The first fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. Spasticity was measured by the Modified Ashworth Scale. No between-group comparison in spasticity was provided, however, the authors state that the FES group showed a “better” improvement than the control group at 12 weeks (post-treatment).

The second fair quality RCT (Winchester et al., 1983) randomized patients with subacute stroke to receive either standard physiotherapy and FES (FES group) or standard physiotherapy alone (control group). Spasticity was measured on a scale of “none,” “minimal,” “moderate,” or “severe”. No significant between-group difference in spasticity was seen at 4 weeks (post-treatment).

Conclusion: There is limited (level 2a) from 1 fair quality RCT that FES and standard rehabilitation is not more effective than standard rehabilitation alone for decreasing spasticity in patients with subacute stroke.

Note: One fair quality RCT reported reduced spasticity following FES, however these results are not used to determine level of evidence as between-group differences were not reported.

One high quality RCT (Ambrosini et al., 2011) has investigated the effectiveness of FES on trunk control in patients with subacute stroke.

In the high quality RCT (Ambrosini et al., 2011) randomized patients with subacute stroke or traumatic brain injury to either an FES-induced cycling group (FES group) or a placebo FES cycling group (control group). Trunk control was measured by the Trunk Control Test. A significant between-group difference on trunk control was seen at 4 weeks (post-treatment) and follow-up, in favour of the FES group compared to the control group.

Conclusion: There is moderate evidence (level 1b) from 1 high quality RCT that FES is more effective than placebo intervention in improving trunk control in patients with sub-acute stroke.

One high quality RCT (Tong et al., 2006) ), 1 fair quality RCT (Sabut et al., 2010b) and 1 pre-post design study (Sabut et al., 2010a) have investigated the effectiveness of FES on walking efficiency in patients with subacute and chronic stroke.

The high quality RCT (Tong et al., 2006) randomly assigned patients with subacute stroke to receive: (1) gait training using an electromechanical gait trainer with functional electric stimulation (EGT-FES); (2) gait training using an electromechanical gait trainer (EGT); or (3) conventional gait training (CGT, control) for four weeks. Walking efficiency was measured using the Functional Ambulatory Category (FAC). At four weeks (post-intervention), there was a significant between-group difference in walking efficiency in favour of the EGT and EGT-FES groups compared to the control group.

One fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. Walking efficiency was measured using the Physiological Cost Index. No significant between-group difference in walking efficiency was seen at 12 weeks (post-treatment).

One pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Walking efficiency was measured using the Physiological Cost Index. A significant effect of FES treatment on walking efficiency was seen at 12 weeks (post-treatment).

Conclusion: There is moderate evidence (level 1b) from 1 high quality RCT that FES is more effective than conventional gait training or standard rehabilitation in improving walking efficiency in patients with subacute stroke. One pre-post study also reported improved walking efficiency following FES in patients with subacute and chronic stroke.

NOTE: However, 1 fair quality RCT did not find FES to be more effective than standard rehabilitation in improving walking efficiency.

One high quality RCT (Ambrosini et al., 2011), 2 fair quality RCTs (Sabut et al., 2010b; Granat et al., 1996) and 1 pre-post design study (Sabut et al., 2010a) have investigated the effectiveness of FES on walking speed in patients with sub-acute and chronic stroke.

The high quality RCT (Ambrosini et al., 2011) randomized patients with subacute stroke or traumatic brain injury to either an FES-induced cycling group (FES group) or a placebo FES cycling group (control group). Self-selected gait speed was measured by the 50m walking test. No significant between-group difference in gait speed was seen at 4 weeks (post-treatment) or follow-up. However, a subgroup analysis of patients with ischaemic stroke revealed a significant between-group difference in favour of the FES group compared to the control group.

The fair quality RCT (Sabut et al., 2010b) randomized patients with subacute stroke to receive physical therapy and FES (FES group) or physical therapy alone (control group) for 12 weeks. Walking speed was measured using the 10 Meter Walk-Way Test. No significant between-group difference in walking speed was found at 12 weeks (post-treatment).

The fair quality crossover RCT (Granat et al., 1996) assigned patients with subacute or chronic stroke to receive four weeks of standard rehabilitation followed by four weeks of FES intervention. Walking speed was measured by the walk-path length divided by the time to traverse the walk-path. No between-group comparisons in walking speed were made, however the authors reported improved speed following 4 weeks of FES intervention.

One pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Walking speed was measured over 10 metres. A significant effect of FES treatment on walking speed was found at 12 weeks (post-treatment).

Conclusion: There is moderate evidence (level 1b) from 1 high quality RCT that FES is more effective than placebo intervention in improving walking speed in patients with subacute stroke. One pre-post design study also reported improved walking speed following FES.

NOTE: However, one fair quality RCT reported that FES is not more effective than standard rehabilitation alone in improving walking speed. Further, a second fair quality crossover RCT did not find significant improvement in walking speed following FES, although these results are not used to determine level of evidence as between-group differences were not reported.

Two fair quality crossover RCTs (Embrey et al., 2010; Granat et al., 1996) have investigated the effect of FES on Activities of Daily Living (ADLs) and participation in patients with chronic stroke.

The first fair quality crossover RCT (Embrey et al., 2010) randomized patients with chronic stroke and hemiplegia to receive FES or no FES for 3 months (phase I). The groups performed the opposite treatment for the following 3 months (phase II). Function and participation was measured using the Stroke Impact Scale. Patients who received FES in phase I showed significantly improved function and participation compared with patients who received no FES, and retained these improvements following no FES in phase II. Patients who received FES in phase II demonstrated significantly improved function and participation at 6 months (post-treatment) compared to baseline scores.

The second fair quality crossover RCT (Granat et al., 1996) assigned patients with sub-acute or chronic stroke to receive 4 weeks of standard rehabilitation followed by 4 weeks of FES intervention. Performance of ADLs was measured using the Barthel Index. A significant difference in ADLs was seen at 4 weeks (post-treatment) in favour of FES compared to standard rehabilitation.

Conclusion: There is limited evidence (level 2a) from 2 fair quality crossover RCTs that FES is more effective than standard rehabilitation for improving ADLs in patients with chronic stroke.

One fair quality RCT (Kottink et al., 2007) investigated the effectiveness of FES on physical activity in patients with chronic stroke.

The fair quality RCT (Kottink et al., 2007) investigated the effectiveness of using a 2-channel implantable peroneal nerve stimulator for improving physical activity in 29 patients with chronic stroke and foot drop. Participants were randomly assigned to receive either: (1) a 2-channel implantable peroneal nerve stimulator, (stimulation treatment group) or (2) a control group. Participants in the treatment group had the stimulator implanted under the epineurium of the peroneal nerve under general or spinal anesthesia. The control group was able to continue using their regular walking devices (ankle-foot orthosis, orthopedic shoes, or no device). Each participant was assessed on physical activity at baseline and at week 26. After 26 weeks, there was no significant between group difference in physical activity measured by the percentage of time spent stepping or standing, however there was a significant difference between the groups on time spent sitting/lying in favour of the stimulation group.

Conclusion: There is limited evidence (Level 2a) from one fair quality RCT that FES is not more effective than standard therapy for increasing physical activity in patients with chronic stroke.

One high quality RCT (Janssen et al., 2008) has investigated the effectiveness of FES on aerobic capacity in patients with chronic stroke.

The high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on aerobic capacity in 12 patients with chronic stroke. Subjects were randomly assigned to either the treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After six weeks of treatment, there was no significant difference between the groups on VO2 peak scores.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to cycling is not more effective than cycling alone in improving aerobic capacity in patients with chronic stroke.

Two high quality RCTs (Daly et al., 2006, Janssen et al., 2008) have investigated the effectiveness of FES on improving balance in patients with chronic stroke.

The first high quality RCT (Daly et al., 2006) investigated the effectiveness of functional neuromuscular stimulation using intramuscular electrodes (FNS-IM) on balance in 32 patients with chronic stroke. The treatment group received functional neuromuscular stimulation using intramuscular electrodes, body-weight support treadmill training, coordination exercises, over ground walking (OG) and a home exercise program. The control group received the same treatment as the intervention group but without FNS for 12 weeks. No significant between group difference was reported for balance as measured by the Tinetti Balance (TB) after 12 weeks of treatment.

The second high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on balance in 12 patients with chronic stroke. Participants were randomly assigned to either the treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation of the paretic leg evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After the six week intervention, no significant difference between the groups was found on the Berg Balance Scale (BBS).

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs that FES in addition to a gait training or a cycling program compared to a gait training or cycling program alone does not improve balance in patients with chronic stroke.

Two high quality RCTs (Daly et al., 2011, Daly et al., 2006) has investigated the effectiveness of FES for improving coordination in patients with chronic stroke.

The first high quality RCT (Daly et al., 2011) randomized patients with chronic stroke to receive intramuscular FES or no FES. Participants performed strengthening exercises, overground gait training and body-weight supported treadmill training for 1.5 hours 4 times/week for 12 weeks. Lower extremity coordination was measured by the Fugl-Meyer Lower Limb Scale and by the Functional Independence Measure (FIM), Locomotion and Mobility subscales. Both groups demonstrated significant improvements in lower extremity coordination at post-treatment (12 weeks), however between-group differences were not reported for these measures.

The second high quality RCT (Daly et al., 2006) investigated the effectiveness of functional neuromuscular stimulation using intramuscular electrodes (FNS-IM) on coordination in 32 patients with chronic stroke. The treatment group received functional neuromuscular stimulation using intramuscular electrodes, body-weight support treadmill training, coordination exercises, over ground walking (OG) and a home exercise program. The control group received the same treatment as the intervention group except without FNS for 12 weeks. After 12 weeks of treatment there was a significant between-group difference in favour of the treatment group on the Fugl-Meyer knee flexion coordination measure (FMKnFx). No statistically significant between group difference was found on the Fugl-Meyer lower extremity coordination (FMLE).

Conclusion: There is strong evidence (level 1a) from one high quality RCTs that FES is more effective than no FES in improving gait training program alone improves knee coordination but not general lower extremity coordination in patients with chronic stroke. Another high quality RCT also reported significantly improved coordination following FES, however these results are not used to determine level of evidence as between-group differences were not reported.

One fair quality crossover study (Sheffler et al., 2006) has investigated the effectiveness of FES for improving functional ambulation in patients with chronic stroke.

The fair quality crossover study (Sheffler et al., 2006) investigated the effect of a transcutaneous peroneal nerve stimulation device on functional ambulation in 14 patients with chronic stroke and a drop foot. The modified Emory Functional Ambulation Profile was used to measure functional ambulation while the 14 patients underwent the test with the Odstock Dropped-Foot Stimulator (ODFS), with an Ankle-Foot Orthosis (AFO), or with no device. The order of use of the device was randomized, and all evaluations and training were completed over two days. After two days, there was a significant difference in favour of using the ODFS compared to no device as well as in favour of using an AFO compared to no device but there was no significant difference between using an AFO and the ODFS as measured by the modified Emory Functional Ambulation Profile.

Conclusion: There is limited evidence (Level 2a) from one fair quality crossover study that FES (via the ODFS) is more effective than using no device for improving functional ambulation in patients with chronic stroke.

Two high quality RCTs (Daly et al., 2011; Daly et al., 2006) have investigated the effectiveness of FES for improving gait component execution in patients with chronic stroke.

The first high quality RCT (Daly et al., 2011) randomized patients with chronic stroke to receive intramuscular FES or no FES. Participants performed strengthening exercises, overground gait training and body-weight supported treadmill training for 1.5 hours 4 times/week for 12 weeks. Gait component analysis was performed using the Gait Assessment and Intervention Tool (GAIT). A significant between-group difference was found in favour of the FES group compared to the no-FES group at post-treatment (12 weeks) and follow-up (6 months).

The second high quality RCT (Daly et al., 2006) investigated the effectiveness of functional neuromuscular stimulation using intramuscular electrodes (FNS-IM) on gait component execution in 32 patients with chronic stroke. The treatment group received functional neuromuscular stimulation using intramuscular electrodes, body-weight support treadmill training, coordination exercises, over ground walking (OG) and a home exercise program. The control group received the same treatment as the intervention group except without FNS for 12 weeks. There was a significant between-group difference in favour of the treatment group after 12 weeks of treatment on the Tinetti Gait scale (TG) which measured: (1) gait initiation; (2) walking path; (3) trunk alignment; (4) swing phase limb trajectory; (5) step continuity; (6) step symmetry; and (7) swing limb floor clearance and forward swing limb execution.

Conclusion: There is strong evidence (level 1a) from two high quality RCTs that FES is more effective than no FES in improving gait component execution in patients with chronic stroke.

Two fair quality RCTs (Daly et al., 2004, Granat et al., 1996) and 1 quasi-experimental design study (Kesar et al., 2011) have investigated the effectiveness of FES for improving gait kinematics in patients with chronic stroke.

The first fair quality RCT (Daly et al., 2004) assigned patients with chronic stroke to: (1) a treatment group that received functional neuromuscular stimulation (FNS) with intramuscular electrodes during strengthening and coordination exercises, ground gait training, and weight supported treadmill training, or (2) a control group received the same treatment except for the FNS. Data was collected in three sessions: pre-treatment, post-treatment and follow-up (six months after the end of treatment). At each session gait kinematics were measured during volitional over-ground walking at a self-selected speed. Three limb flexion gait components were chosen as indicators of swing phase limb advancement in the sagittal plane: peak swing hip flexion, peak swing knee flexion and mid-swing ankle dorsiflexion. Although the stimulation treatment group showed a significant gain in volitional peak swing knee flexion and volitional mid-swing ankle dorsiflexion after treatment, no group comparisons were reported.

The second fair quality crossover RCT (Granat et al., 1996) assigned patients with sub-acute or chronic stroke to receive four weeks of standard rehabilitation followed by four weeks of FES intervention. Gait parameters were measured by swing symmetry and foot inversion in stance. A significant difference in gait parameters in patients was seen at 4 weeks (post-treatment) in favour of FES compared to standard rehabilitation.

One quasi-experimental design study (Kesar et al., 2011) measured gait kinematics of patients with chronic stroke during four walking conditions: 1) self-selected speed without FES (SS); 2) self-selected speed with FES (SS-FES); 3) faster than self-selected speed without FES (FAST); and 4) faster than self-selected speed with FES (FAST-FES). Gait kinematics were measured by peak anterior ground reaction force during paretic stance (AGRF) and trailing limb angle. Comparison of walking faster than self-selected speed conditions revealed significant differences on the AGRF in favour of FES (condition 4) compared to no FES (condition 3). Comparison of FES conditions at post-evaluation revealed significant differences on the AGRF, trailing limb angle and peak knee flexion in favour of condition 4 (walking faster than self selected speed with FES) compared to condition 2 (walking at a self selected speed with FES). Comparison of non-FES conditions revealed a significant difference in gait kinematics in favour of condition 3 (walking faster than self selected speed) compared to condition 1 (walking at self selected speed). No differences were found when comparing non-FES conditions (condition 3: walking at a faster than self selected speed vs. condition 1: walking at a self selected speed) on peak knee flexion. No differences were found when comparing FES conditions (condition 4: walking faster than self selected speed with FES vs. condition 2: walking at a self selected speed with FES) on trailing limb angle and peak knee flexion. No significant differences were found on any comparison on the percent paretic propulsion.

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT and 1 quasi-experimental designs study that FES is more effective than standard rehabilitation or no FES in improving gait parameters and kinematics in patients with chronic stroke. A second fair quality RCT reported improved gait parameters following FES, however these results are not used to determine level of evidence as between-group differences were not reported.

One high quality RCT (Janssen et al., 2008) has investigated the effectiveness of FES on maximal power output in patients with chronic stroke.

The high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on maximal power output in 12 patients with chronic stroke. Subjects were randomly assigned to either the treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After six weeks of treatment, there was no significant difference between the groups on PO2max scores.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to cycling is not more effective than cycling alone in improving maximal power output in patients with chronic stroke.

One pre-post design study (Sabut et al., 2010a) has investigated the effect of FES on metabolic responses in patients with sub-acute and chronic stroke.

The pre-post design study (Sabut et al., 2010a) assigned patients with sub-acute or chronic stroke to receive FES and standard physical therapy. Metabolic responses were measured according to oxygen consumption, carbon dioxide production, heart rate and energy cost. A significant effect of FES treatment on metabolic responses was seen at 12 weeks (post-treatment) in patients with sub-acute and chronic stroke.

Conclusion: There is limited evidence (level 2b) from one pre-post design study that FES is effective in improving metabolic responses in patients with sub-acute and chronic stroke.

One high quality RCT (Janssen et al., 2008) has investigated the effectiveness of FES on maximum isometric voluntary contraction (MIVC) in patients with chronic stroke.

The high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on maximum isometric voluntary contraction in 12 patients with chronic stroke. Subjects were randomly assigned to either the treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After six weeks of treatment, no significant difference was found between the groups on maximal voluntary knee extension.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to cycling is not more effective than cycling alone in improving maximum isometric voluntary contraction in patients with chronic stroke.

Two high quality RCTs (Daly et al., 2011, Janssen et al., 2008), one fair quality crossover RCT (Embrey et al., 2010), one meta-analysis (Glanz et al., 1996) and one pre-post design study (Sabut et al., 2010a) have investigated the effectiveness of FES on lower limb muscle strength in patients with chronic stroke.

The first high quality RCT (Daly et al., 2011) randomized patients with chronic stroke to receive intramuscular FES or no FES. Participants performed strengthening exercises, overground gait training and body-weight supported treadmill training for 1.5 hours 4 times/week for 12 weeks. Muscle strength was measured by manual muscle testing. Both groups demonstrated significant improvements in muscle strength at post-treatment. Between-group differences were not reported for this outcome.

The second high quality RCT (Janssen et al., 2008) randomized patients to an intervention group that received FES to evoke muscle contractions during cycling exercises, or a control group that received sensible stimulation that did not evoke muscle contractions during cycling exercises. Muscle strength was measured by maximal knee extension torque. No significant difference in muscle strength was seen at 6 weeks (post-intervention).

A fair quality crossover RCT (Embrey et al., 2010) randomized patients with chronic stroke and hemiplegia to receive FES or no FES for 3 months (phase I). The groups performed the opposite treatment for the following 3 months (phase II). Isometric muscle strength was measured by dynamometer. Significant between-group differences in dorsiflexion strength in the paretic leg were seen in favour of the FES group compared to the no-FES group following phase I (3 months). Similar results were found in favour of FES compared to no-FES following phase II (6 months). There were no significant between-group differences in strength of plantar flexor muscles at either time point.

In the meta-analysis (Glanz et al., 1996), four randomized controlled trials were analyzed to assess the effectiveness of functional electrical stimulation (FES) therapy for improving muscular strength in chronic patients post-stroke. The range of mean time since onset of symptoms for the individuals in the four studies included was 1.5 months to 29.2 months. The treatment group received FES for a muscle in their hemiparetic extremity (ankle, knee or wrist), along with standard physical therapy. The control group received physical therapy alone for all studies except one, where these patients received a sham treatment. All four studies generated a positive effect size (0.63), where patients who received FES had significant improvements in muscle strength of the hemiparetic extremity (ankle, knee or wrist) at post-treatment, in comparison to those who received standard physical therapy or sham treatment.

One pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Muscle strength was measured by the mean absolute value, root mean square, median frequency and median amplitude of the tibialis anterior muscle EMG signal. A significant effect of FES treatment on muscle strength was seen at 12 weeks (post-treatment).

Conclusion: There is strong evidence (level 1a) from one meta-analysis and one fair quality RCT that FES is more effective than no FES, standard therapy, or no therapy for improving muscle strength in patients with chronic stroke. One high quality RCT and one pre-post study also found significant improvement in muscle strength following FES, although between-group differences were not reported.

NOTE: One high quality RCT, published after the meta-analysis, found no significant difference between FES treatment and sensible stimulation during cycling.

One high quality RCT (Kottink et al., 2010) has investigated the effectiveness of FES on quality of life in patients with chronic stroke.

The high quality RCT (Kottink et al., 2010) randomized patients with stroke and chronic hemiplegia with foot drop to a group that received FES by an implantable two-channel peroneal nerve stimulator for correction of their foot drop, or to a control group who continued using conventional walking devices (e.g. ankle-foot orthoses, orthopedic shoes or no device). Health status was measured using the Short Form-36 (SF-36), Disability Impact Profile (DIP), and mean preference-based summary indexes of the SF-36 and EuroQoL (EQ-5D). At 26 weeks there was a significant between-group difference in favour of the intervention group compared to the control group on the SF-36 (physical functioning, general health and physical component summary scores), the DIP (mobility, self-care and psychological status scores), and the EQ-5D preference-based summary index. There were no differences in other domains of the DIP (symptoms, social activities, communication), the SF-36 (physical role functioning, bodily pain, social functioning, mental health, emotional role functioning, vitality, mental component summary) or the SF-36 mean preference-based summary index.

Conclusion: There is moderate evidence (level 1b) from one high quality RCT that FES is more effective than control therapies (no FES, orthotic devices) in improving aspects of quality of life in patients with chronic stroke.

One high quality RCT (Cozean et al., 1988) has investigated the effectiveness of FES for improving range of motion in patients with chronic stroke.

The high quality RCT (Cozean et al., 1988) examined the effectiveness of FES on range of motion in 32 patients with chronic stroke who demonstrated a dynamic gait problem of spastic equinus posturing of the affected leg. Patients received either (1) a combination of electromyographic biofeedback (BFB) and functional electrical stimulation (FES), (2) BFB only, (3) FES only, or (4) control therapy (standard physical therapy only). At four weeks post-treatment, statistically significant differences were seen in the improvement indices of knee flexion and ankle dorsiflexion in the combined treatment group compared to the control group who received standard physical therapy only.

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES combined with BFB and physiotherapy compared to physiotherapy alone improves range of motion (knee flexion and ankle dorsiflexion) in patients with chronic stroke.

One high quality RCT (Janssen et al., 2008) has investigated the effectiveness of FES on self-reported functional mobility in patients with chronic stroke.

One high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on self-reported functional mobility in 12 patients with chronic stroke. Subjects were randomly assigned to either the treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After six weeks of treatment, no significant difference was found between the groups on the Rivermead Mobility Index (RMI).

Conclusion: There is moderate evidence (Level 1b) from one high quality RCT that FES in addition to cycling is not more effective than cycling alone in improving self-reported functional mobility in patients with chronic stroke.

One high quality RCT (Chen et al., 2005) and one fair quality RCT (Embrey et al., 2010) have investigated the effect of FES on spasticity in patients with chronic stroke.

The high quality RCT (Chen et al., 2005) randomized patients with chronic stroke to receive electrical stimulation (ES) or placebo ES. Spasticity was measured using the Modified Ashworth Scale and electrophysiological evaluation of lower limb responses (tibial Fmax/Mmax ratio, H-reflex latency and H-reflex recovery curve). Patients who received ES demonstrated a significant decrease in electrophysiology measures of spasticity (Fmax/Mmax ratio, H-reflex latency). More patients from the ES group than the placebo group demonstrated decreased spasticity on the modified Ashworth scale, although between-group differences were not reported.

A fair quality crossover RCT (Embrey et al., 2010) randomized patients with chronic stroke and hemiplegia to receive FES or no FES for 3 months (phase I). The groups performed the opposite treatment for the following 3 months (phase II). Spasticity was measured by the Modified Ashworth Scale. No significant between-group differences in spasticity were seen following phase I or phase II.

Conclusion: There is limited evidence (level 2a) from 1 fair quality crossover RCT that FES is not more effective than no stimulation in reducing spasticity in patients with chronic stroke.

Note: 1 high quality RCT reported significantly reduced spasticity following FES, although between-group differences were not reported.

Two high quality RCTs (Daly et al., 2006, Janssen et al., 2008) have investigated the effectiveness of FES on walking distance in patients with chronic stroke.

The first high quality RCT (Daly et al., 2006) investigated the effectiveness of functional neuromuscular stimulation using intramuscular electrodes (FNS-IM) on walking distance in 32 patients with chronic stroke. The treatment group received functional neuromuscular stimulation using intramuscular electrodes, body-weight support treadmill training, coordination exercises, over ground walking (OG) and a home exercise program. The control group received the same treatment as the intervention group but without FNS for 12 weeks. No significant between group difference was found for walking distance measured by the 6 minute walking test (6MWT) after 12 weeks of treatment.

The second high quality RCT (Janssen et al., 2008) compared the effectiveness of cycling with and without FES on walking distance in 12 patients with chronic stroke. Subjects were randomly assigned to either treatment or control group in which both groups performed a cycling exercise. During cycling, the treatment group received electrical stimulation evoking muscle contractions and the control group received sensible stimulation not evoking muscle contractions for six weeks. After six weeks of treatment, no significant difference was found between the groups on the 6 minute walking test (6MWT).

Conclusion: There is strong evidence (Level 1a) from two high quality RCTs that FES in addition to a gait training or cycling program is not more effective than a gait training or cycling program alone for improving walking distance in patients with chronic stroke.

Two fair quality RCTs (Burridge and McLellan, 2000, Burridge et al., 1997), two pre-post designed studies (Sabut et al., 2010a, Stein et al., 2006) and one systematic review (Kottink et al., 2004) examined the effectiveness of FES on walking efficiency in patients with chronic stroke.

The first fair quality RCT (Burridge and McLellan, 2000) assigned patients with chronic stroke and age-matched controls to receive FES to the common peroneal nerve during swing phase of walking, for a period of 3 months. Walking efficiency was measured by the Physiological Cost Index. Subjects were assessed with and without stimulation. Comparison of scores from baseline to post-intervention revealed a significant improvement in walking efficiency, both with and without stimulation. Note: this study did not provide between-group differences, so is not used to determine level of evidence in the conclusion below.

The second fair quality RCT (Burridge et al., 1997)randomized patients to a group that received FES and physiotherapy (FES group) or a group that received physiotherapy alone (control group). Assessment of the Physiological Cost Index (PCI) was done at the start of the trial, between four and five weeks, and between 12 and 13 weeks. Subjects in the treatment group were assessed with and without stimulation. A statistically significant difference was found between the change in PCI in the control group and change in PCI in the FES group (measured without stimulation at the start of the trial and with stimulation at the end).

The first pre-post design study (Sabut et al., 2010a) assigned patients with subacute or chronic stroke to receive FES and standard physical therapy. Walking efficiency was measured by the Physiological Cost Index. A significant effect of FES treatment on walking efficiency was seen at 12 weeks (post-treatment).

The second pre-post design study (Stein et al., 2006) investigated the use of FES to the peroneal nerve and tibialis anterior for improving walking efficiency in 26 individuals with drop foot as a result of various central nervous system disorders, including 12 patients with stroke.. Physiological Cost Index (PCI) was measured before the subjects took the device home and at approximately monthly intervals for at least three months. At the end of treatment, a trend was seen in the PCI toward lower values (from 1.06 to 1.01) however these results were not significant.

The systematic review (Kottink et al., 2004) examined the literature on the effectiveness of FES for improving walking efficiency in eight studies of patients with stroke. Only one of these studies was an RCT and has been reviewed above (Burridge et al., 1997). A second pre-post designed study by the same authors reported a significant decrease in PCI with and without stimulation after three months in 56 individuals with stroke. Overall this systematic review suggests a positive impact of FES on improving walking efficiency in individuals with chronic stroke.

Conclusion: There is limited evidence (level 2a) from 1 fair quality RCT that FES is more effective than standard rehabilitation alone for improving walking efficiency in patients with sub-acute and chronic stroke. A second fair quality RCT, a pre-post design study and a systematic review also reported significant improvements in walking efficiency following FES, and a second pre-post design study reported a non-significant trend towards improved walking efficiency.

One meta-analysis (Robbins et al., 2006) has investigated the effectiveness of FES for improving walking speed once the FES therapy has ended. Six of the studies included in the meta-analysis (Alon & Ring, 2003; Bogataj et al., 1995; Burridge et al., 1997; Burridge & MacLellan, 2000; Chen et al., 2005; Granat et al., 1996) have been reviewed in this Stroke Engine module.

The meta-analysis (Robbins et al., 2006) which included three controlled trials examined changes in gait speed that had occurred after a series of FES treatments over a period of weeks or months. The main outcome was gait speed. Time since stroke was 3.9 to 59 months with a mean of 34 months. All studies examined differences in gait speed in subjects who had received training with FES for weeks or months compared to those that did not receive FES training. A fixed-effects model of three studies produced a mean difference that was significant, indicating the effectiveness of FES treatment at increasing gait speed in subjects post-stroke.

Conclusion: There is strong evidence (Level 1a) from one meta-analysis that FES is more effective even after the sessions have ended for improving walking speed compared to conventional therapy.

Note:Due to the small number of controlled trials, the authors also included in their discussion the results of 1 pre-post design and 1 crossover study to provide a more comprehensive review. In total three of the five FES studies showed a real improvement in gait speed (defined in this article as a minimal improvement of 7.9% in gait speed). This meta-analysis showed that previous training using FES produces sustained improvements in gait speed even after the FES is turned off (therapeutic effect).

Three high quality RCTs (Cozean et al., 1988; Chen et al., 2005; Daly et al., 2011), four fair quality RCTs (Burridge et al., 1997; Kottink et al., 2007; Alon and Ring, 2003; Burridge and McLellan, 2000), two fair quality crossover RCT (Granat et al., 1996; Embrey et al., 2010), two pre-post designed studies (Sabut et al., 2010a; Stein et al., 2006) and one systematic review (Kottink et al., 2004) investigated the effectiveness of FES on walking speed in patients with chronic stroke.

The first high quality RCT (Cozean et al., 1988) randomized (1) a combination of electromyographic biofeedback (BFB) and functional electrical stimulation (FES), (2) BFB only, (3) FES only, or (4) control therapy (standard physical therapy only). At four weeks post-treatment a statistically significant improvement in the velocity of gait was observed in the combined treatment group, however no group comparisons were made.

The second high quality RCT (Chen et al., 2005) randomized patients with chronic stroke to receive electrical stimulation (ES) for 20 minutes/day, 6 days/week for 1 month, or placebo ES. Walking speed was measured by a 10m walking time assessment. Patients who received ES demonstrated significant improvements in walking time at post-treatment, while patients who received placebo ES did not demonstrate a significant improvement. Between-group differences were not reported.

The third high quality RCT (Daly et al., 2011) randomized patients with chronic stroke to receive intramuscular FES or no FES. Participants performed strengthening exercises, overground gait training and body-weight supported treadmill training for 1.5 hours 4 times/week for 12 weeks. Walking speed was measured using the 6-Minute Walk Test (6MWT). Both groups demonstrated significant improvements in walking speed at post-treatment.

The first fair quality RCT (Burridge et al., 1997) examined the effectiveness of FES and physical therapy on walking speed in patients with chronic stroke whose walking was impaired by a drop foot. The participants were randomized into a treatment group that received FES and physiotherapy or to a control group which received physiotherapy alone. The Odstock Dropped Foot Stimulator provided FES. Walking speed was measured over a 10-meter walkway. Treatment group subjects were asked to perform the walk three times with stimulation and three times without. Control subjects completed the walk three times without stimulation. At the first assessment there was no significant difference between the walking speed of the control and the FES group with stimulation. However, there was a significant difference between the groups when the mean change in walking speed was compared (measured without stimulation at the start of the trial and with stimulation at the end for the FES group).

The second fair quality RCT (Kottink et al., 2007) investigated the effectiveness of using a 2-channel implantable peroneal nerve stimulator for improving walking speed in patients with chronic stroke and foot drop. Participants were randomly assigned to receive either: (1) a 2-channel implantable peroneal nerve stimulator, (stimulation treatment group) or (2) a control group. Participants in the treatment group had the stimulator implanted under the epineurium of the peroneal nerve under general or spinal anesthesia. The implanted device communicates with an external transmitter, and applied continual stimulation to the nerve. The control group was able to continue using their regular walking devices (ankle-foot orthosis, orthopedic shoes, or no device). Each participant was assessed on this measure at baseline, week 12 and week 26. After 12 weeks, there was no significant difference between the groups on the 6MWT or 10-meter walkway test, however, after 26 weeks, there was a significant improvement in walking speed in the stimulation group compared to the control group measured by the 6MWT and the 10-meter walkway test.

The third fair quality RCT (Alon and Ring, 2003) assigned patients with chronic stroke to receive either FES (electrically augmented stimulation training) or a control treatment consisting of the same exercises without stimulation. Walking speed was measured by the 10 Meter Walk Test: time and cadence. A significant between-group difference in walking speed was seen at 2 months (post-treatment) in favour of FES and standard rehabilitation compared to standard rehabilitation only.

The fourth fair quality RCT (Burridge and McLellan, 2000) assigned patients with stroke and age-matched controls to receive 3 months of FES to the common peroneal nerve. Walking speed was measured by the 10 Meter Walk Test. Comparison of scores from baseline to post-intervention revealed a significant improvement in walking speed when the stimulator was turned on.

The first fair quality crossover RCT (Granat et al., 1996) assigned patients with sub-acute or chronic stroke to receive four weeks of standard rehabilitation followed by four weeks of FES intervention. Walking speed was measured by the walk-path length divided by the time to traverse the walk-path. A significant improvement in walking speed was seen after 4 weeks of FES intervention, although between-group comparisons were not reported.

A second fair quality crossover RCT (Embrey et al., 2010) randomized patients with chronic stroke and hemiplegia to receive FES or no FES for 3 months (phase I). The groups performed the opposite treatment for the following 3 months (phase II). Walking speed was measured by the 6-Minute Walk Test (6MWT) and the Emory Functional Ambulatory Profile (EFAP). At 3 months (phase I), participants who had received FES demonstrated a statistically significant improvement in walking speed. At 6 months (phase II) the group that received FES in phase I retained significant improvements from baseline and the group that received FES in phase II demonstrated significantly improved walking speed compared to baseline scores.

The first pre-post design study (Sabut et al., 2010a) assigned patients with sub-acute or chronic stroke to receive FES and standard physical therapy. Walking speed was measured over 10 metres. A significant effect of FES treatment on walking speed was found at 12 weeks (post-treatment).

The second pre-post designed study (Stein et al., 2006) investigated the use of a peroneal nerve and tibialis anterior FES for improving walking efficiency in individuals with drop foot as a result of various central nervous system disorders including 12 patients with stroke. Walking speed was measured over a straight distance of 10 meters as well as by walking around a continuous 10 meter figure of eight. Measurements were taken before the subject took the device home and at approximately monthly intervals for at least three months. Subjects were asked to perform the 10 meter straight walking test twice with FES and twice without FES at each subsequent visit. The average increase in walking speed from the initial measure without FES to the final measure with FES was 12%, which was not a significant difference.

The systematic review (Kottink et al., 2004), investigated the results of eight studies (including only one RCT, Burridge et al., 1997) on the relationship between FES and walking speed. Walking speed was measured in six of the eight studies analyzed for the review, and it was concluded that FES interventions improved walking speed post-stroke.

Conclusion: There is limited evidence (level 2a) from 5 fair quality RCTs (one which using a cross-over design) that FES is more effective than control therapies (physiotherapy, conventional rehabilitation, no FES) in improving walking speed in patients with chronic stroke. Two high quality RCTs, one fair quality cross-over RCT, one pre-post design study and one systematic review found significant improvements in walking speed following FES, although between-group differences were not reported.

Note: One high quality RCT reported significant improvements in walking speed at post-treatment in both groups whereas 1 pre-post design study reported no significant improvement in walking speed following FES.