Twelve RCTs were included in the review (n=1,524 children; range 37 to 576). Of the 12 studies, six reported adequate randomisation generation, and 8 reported adequate allocation concealment. Three trials were open label, four blinded microscopists, and blinding was not possible in the remaining trials due to the trial design. Loss to follow-up ranged from none (seven trials) to 20%. Only three trials provided a power calculation.
Overall, there were no significant differences in mortality between artemisinin and quinine (RR 0.90, 95% CI 0.73 to 1.12; 12 RCTs), but coma resolution was significantly faster with artemisinin (WMD -4.61, 95% CI -7.21 to -2.00; eight RCTs); the significant improvement in coma resolution was no longer observed when only trials with adequate concealment were considered. No significant heterogeneity was observed for these analyses.
There was no significant difference between artemisinin and quinine in terms of parasite (seven RCTs) or fever (eight RCTs) clearance times, neurological sequelae (10 RCTs), cure rates (five RCTs), or adverse events (11 RCTs).
Further results for mortality for each artemisinin drug, and sensitivity and subgroup analyses, were also reported.