Effect of Ropinirole on Sleep Outcomes in Patients with Restless Legs Syndrome: Meta-Analysis of Pooled Individual Patient Data from Randomized Controlled Trials

Study Objective. To compare the effects of ropinirole with those of placebo on sleep, as evaluated by specific domains of the Medical Outcomes Study (MOS) sleep scale, as well as the Clinical Global Impression‐Improvement (CGI‐I) scale, in patients with restless legs syndrome (RLS). Design. Meta‐analysis of six randomized, double‐blind, placebo‐controlled, parallel‐group trials conducted in the United States and Europe. Patients. A total of 1679 patients aged 18–79 years with primary moderate‐to‐severe RLS who received ropinirole (835 patients) or placebo (844 patients). Measurements and Main Results. A systematic review of MEDLINE (January 1980‐January 2007) and clinical trial registers was performed to identify placebo‐controlled trials of ropinirole that used the 12‐item MOS sleep scale to assess sleep in patients with RLS. Individual patient data from both published and nonpublished trials were pooled for meta‐analysis. In the eligible studies, immediate‐release ropinirole 0.25‐6 mg or placebo had been given for at least 12 weeks. In addition, sleep scale summary scores for the domains of sleep quantity, adequacy, disturbance, and daytime somnolence had to have been assessed at baseline and at 12 weeks. Our meta‐analysis found that at baseline study patients slept an average of 5.8 hours/night. At the end of 12 weeks, ropinirole‐treated patients slept a mean of 2.5 hours/week more and had a 21% greater improvement from baseline in sleep adequacy scores compared with patients receiving placebo. Ropinirole‐treated patients also had 14% less sleep disturbance and 8% less daytime somnolence than patients receiving placebo. Clinicians rated 63% of ropinirole‐treated patients and 47% of patients receiving placebo as responders based on the CGI‐I scale. Mixed effects analysis of covariance was used to estimate treatment effect adjusting for study center as a random effect, as well as the following fixed effects known to affect sleep: baseline sleep characteristics, age, sex, and chronic medical conditions. All differences were statistically significant (p