Text this: Publication bias in prevalence studies should not be ignored