Neurophysiological and clinical outcome measures of the impact of electrical stimulation on spasticity in spinal cord injury: Systematic review and meta-analysis.

Only three small randomised crossover trials fed the pooled Ashworth-scale result, and the authors name their reliance on data from non-randomised studies as the most important limitation of the review. Twenty-eight of the 29 included studies looked at lower-limb spasticity, with a single paper reporting two upper-limb case studies, so the conclusions describe the leg, not the arm. Studies reported the same outcome measures in different ways and several would not share their original data, so seven studies that met the inclusion criteria could not be pooled at all. One trial comparing three forms of TENS was counted three times in the meta-analysis, which the authors acknowledge is against Cochrane guidance but say was necessary to compare each stimulation type separately. Blinding of participants and of the therapist delivering stimulation was not possible with this intervention, and only some studies blinded the assessor. Only 11 studies reported whether participants were allowed to keep taking anti-spasticity medication, which could have affected the results.

Do not read the headline conclusion as proof that electrical stimulation reduces spasticity after spinal cord injury. The Ashworth-scale benefit rests on three small crossover trials, the two randomised trials that used the pendulum test disagreed with each other, and the authors name their reliance on non-randomised studies as the review's most important limitation. The suggestion that 100 Hz TENS may be better than 50 Hz or below comes from comparing results across studies, not from any trial that tested the frequencies against each other.