Sample size
We estimated a total sample size of 2,100, assuming a headache prevalence ≥10% and an absolute margin of error of 1.3% with 95% confidence interval [13].