/clinicaltrials.gov/ct2/show/results/NCT01712009?sect=Xd5601Note the 95% confidence intervals - not very impressive at all, to be honest. It doesn't even seem to show a placebo effect, IMO. But, this type of study should correct for placebo effect.
And, laratadine does have its own list of potential adverse effects.
I agree that no pain is no pain. But, as a mathematician, I constantly remind people that the plural of anecdote is not data.
I'm not saying that people shouldn't take it. I'm just saying that the data do not seem to support the conventional wisdom.