Seventeen studies were included in the review (4,508 patients received treatment and 1,880 were controls). Six studies were open-label.
Eleven RCTs were included in meta-analyses (6,388 patients). These included four RCTs that evaluated duloxetine (n=1,411 randomised), three RCTs that evaluated milnacipran (n=2,213 randomised) and four RCTs that evaluated pregabalin (n=2,768 randomised). Sample size ranged from 125 to 1,200. Median duration of the randomised phase was 24 weeks (range six to 28 weeks). The median percentage of patients randomised was 55%. All studies were initiated by pharmaceutical companies. All studies were high quality (van Tulder scores of 8 or 9).
The median percentage of patients who completed the studies was 64% (range 55% to 78%) for active drugs and 68% (range 50% to 77%) for placebo.
Compared with placebo: All three drugs were more effective than placebo, except duloxetine for fatigue, milnacipran for sleep disturbance and pregabalin for depressed mood. Results were reported in the review.
Comparative efficacy and harms: There was no significant difference between duloxetine, milnacipran and pregabalin in the proportion of patients with more than 30% pain relief and for withdrawals due to adverse events.
Compared to milnacipran, duloxetine and pregabalin significantly reduced pain (SMD 1.74, 1.68 to 1.80 for duloxetine versus milnacipran and SMD 0.70, 95% CI 0.58 to 0.82 for milnacipran versus pregabalin) and sleep disturbance (SMD 6.20, 95% CI 6.05 to 6.35 for duloxetine versus milnacipran and SMD 0.84, 95% CI 0.69 to 0.99 for milnacipran versus pregabalin).
Duloxetine significantly improved depressed mood compared to milnacipran (SMD 2.45, 95% CI 2.32 to 2.58) and pregabalin (SMD 27.0, 95% CI 26.83 to 27.17).
Milnacipran and pregabalin significantly improved fatigue compared to duloxetine (SMD 0.77, 95% CI 0.67 to 0.87 for duloxetine versus milnacipran and SMD 0.62, 95% CI 0.52 to 0.72 for duloxetine versus pregabalin).
Duloxetine and milnacipran significantly increased headache and nausea compared to pregabalin (RR of headache 2.24, 95% CI 1.83 to 2.65 for duloxetine versus pregabalin and RR of headache 1.81, 95% CI 1.48 to 2.14 for milnacipran versus pregabalin) and (RR of nausea 3.25, 95% CI 2.13 to 4.37 for duloxetine versus pregabalin and RR of nausea 1.90, 95% CI 1.37 to 2.43 for milnacipran versus pregabalin).
Duloxetine significantly increased diarrhoea compared to milnacipran (RR 2.21, 95% CI 1.64 to 2.78) and pregabalin (RR 2.01, 95% CI 1.23 to 2.99).
Duloxetine significantly reduced health-related quality of life (HRQoL) compared to milnacipran (SMD1.47, 95% CI 1.29 to 1.65) and milnacipran significantly reduced HRQoL compared to pregabalin (SMD 0.44, 95% CI 0.28 to 0.60).
Other results were presented in tables. Results of the open uncontrolled trials were reported.
The small number of studies precluded investigation of potential publication bias.