Seven RCTs (n=714) were included.
Overall, the methodological quality of the primary studies was deemed to be poor. One of the included studies was considered to be of high quality; however, while the randomisation method was satisfactory, concealment of treatment allocation was not deemed satisfactory in this trial. There were a high proportion of patient drop-outs in the included studies (39%). Agreement between reviewers for initial decisions was fair (kappa 0.58).
Pain (3 studies, n=152).
One high-quality trial demonstrated a statistically significant effect in favour of the supervised rehabilitation group at short-term follow-up. No significant differences were observed in the immediate, intermediate or long term.
Swelling (2 studies, n=128).
One study demonstrated a statistically significant improvement in the supervised rehabilitation group at short-term follow-up. Compared with standard treatment, supervised rehabilitation did not show a statistically significant effect on swelling reduction, immediately or at intermediate follow-up, in either study.
Functional instability (subjective feeling of giving way) (4 studies, n=224).
The authors of one low-quality primary study reported a significant difference in the incidence of functional instability after a mean follow-up of 230 days in favour of the exercise group (effect size not reported). Three other studies did not find any statistically significant differences between the treatment groups.
Re-injury (4 studies, n=267).
One study demonstrated a statistically significant beneficial effect of supervised rehabilitation (early exercise instruction combined with balance training) on the incidence of re-injury at long-term follow-up. Three additional studies did not report any statistically significant differences between supervised rehabilitation and standard treatment on the incidence of re-injury.
Return to work (4 studies, n=278).
One study reported a statistically significantly shorter time to return to work in the supervised exercise group at immediate follow-up, while no statistically significant between group differences were shown in another (high-quality) trial at short-term or intermediate follow-up. Effect sizes were not calculated in two of the studies reporting on time taken to return to work after injury.
Return to sport (2 studies, n=96).
Neither study (one of which was deemed to be of high quality) showed a statistically significant effect between treatment groups on time taken to return to sport after injury.
Patient satisfaction (2 studies, n=123).
Greater patient satisfaction was reported with the supervised rehabilitation group at long-term but not intermediate follow-up; this was based on one study. Another study reported no benefit of supervised rehabilitation at long-term follow-up.
Range of motion (1 study, n=80).
No statistically significant difference between the treatment groups was shown on passive dorsal and plantar flexion at immediate or intermediate follow-up.