Six RCTs (n=1,703) were included in the review. Sample sizes ranged from 104 to 576. Randomisation methods were adequate in three RCTs. Allocation concealment was adequate in only one RCT. None of the studies were double-blind. Patient withdrawals were described in all studies.
A significant difference in mean intention-to-treat eradication rate (RR 1.09, 95% CI 1.01 to 1.17; six RCTs) was reported in favour of high-dose PPI (cure rate 82%) in comparison with standard-dose PPI (cure rate 74%). This analysis was associated with evidence of significant heterogeneity (I2=55%) and publication bias.
Sensitivity analyses showed that both publication bias and heterogeneity (I2=14.4%) were reduced by removal of one study (RR 1.06, 95% CI 1.00 to 1.12, I2=not reported; five RCTs). Similar effects were observed for per protocol eradication rates with no evidence of publication bias.
Subgroup analyses showed significant differences in eradication rates in favour of first generation standard-dose PPIs (20mg omeprazole) versus second generation high-dose PPIs (40mg pantoprazole) (RR 1.12, 95% CI 1.04, 1.20; four RCTs). There was no significant difference between standard- and high-dose PPIs of the same generation (three RCTs, I2=0%). A planned analysis to compare effect sizes in ulcer and non-ulcer patients was not possible due to a lack of studies.
No severe adverse events were reported. There was no statistically significant difference in the mild to moderate adverse event profiles between high-dose and standard-dose treatment groups.