Fifteen studies with a total of 1,038 participants were included in the review: 1 controlled trial (partially randomised) (n=66), 3 cohort studies with concurrent control groups (n=79), 3 cohort studies with historical control groups (n=108) and 8 case series (n=785).
Outcome measures for impairment.
One study, which was rated as level IV and weak, reported results favouring CE without statistical evaluation. One study, which was rated as level II and weak, reported statistically significant results in favour of the control group. Four studies (25 data sets), which were rated as level I/strong to level III/weak, reported no statistically significant difference between the groups.
Outcome measures for functional limitation and activity.
Four studies (9 data sets), which were rated level I/strong to level II/weak, reported statistically significant results in favour of CE. Three studies (8 data sets), one rated level II/moderate and two rated level II/weak, reported statistically significant results in favour of the control group. Six studies (81 data sets), which were rated level I/strong to level III/weak, reported no statistically significant difference between the groups.
Outcome measures for societal limitations and contextual factors.
One study, which was rated level II/weak, reported statistically significant results in favour of CE. Four studies (8 data sets), which were rated level I/strong to level III/weak, reported no statistically significant difference between the groups.