Does the efficacy of neurodynamic treatments depend on the presence and type of criteria used to define neural mechanosensitivity in spinally-referred leg pain? A systematic review and meta-analysis.

Eighteen of the 21 included trials were rated at high risk of bias, most often because participants, staff or outcome assessors were not blinded, or because outcome data were incomplete. Most trials were small, with 11 to 56 people per group, and the longest follow-up was two months, so nothing is known about whether any benefit lasts. Only three trials, covering 111 people, met the recommended criteria for confirming nerve mechanosensitivity, so the comparison the review was built around rests on very little data. The 16 trials in the unclear group varied widely in their results, which makes the pooled estimate for that group uncertain. Neural mobilisation was combined with an exercise programme in 15 of the 21 trials, so how much of the benefit came from the mobilisation itself cannot be worked out. No trial recruited patients with negative neurodynamic tests, which is the question the review set out to answer. Study selection for the new search was done by a single reviewer rather than two, and the review protocol was not registered in advance. Reporting of the criteria used to interpret neurodynamic tests was often too vague to classify, and the authors accept that some trials may have been placed in the wrong group.

The text provided carries no funding statement and no declaration of competing interests.

Do not read this as proof that neural mobilisation works. In 15 of the 21 trials it was given alongside an exercise programme, so its own contribution cannot be separated, and 18 of the 21 trials were at high risk of bias with small samples and follow-up of two months at most. The review also could not test anyone with a negative neurodynamic test, which was the question it set out to answer.