Gravar-mail: Should adjustment for covariates be used in prevalence estimations?