Fixed-effect models have not been run as sensitivity
analyses in this study, as simulations studies suggested
that fixed-effect model tended to underestimate standard
errors of pooled estimates in indirect comparison.24 We
therefore only used the random-effects model, which has
been validated in Bucher method,24 for the quantitative
pooling in both the direct and the adjusted indirect comparisons.
Dexlansoprazole is different from any other PPI drugs;
it is formulated with an improved technology – dual
delayed-release system, which has been confirmed in
studies that dexlansoprazole was well tolerated, effective
in the healing and maintenance of EO, and in the treatment
of NERD,13 that is consistent with the findings in
our systematic review and meta-analyses. Our study further
made it possible to compare dexlansoprazole with
another widely used PPI drug esomeprazole to evaluate
the comparative treatment effect in the absence of the
head-to-head RCTs between dexlansoprazole and
esomeprazole.