Funnel graph of precision (1/variance) by the natural logarithm of the odds ratio of studies estimating the prevalence of genital human papillomavirus of focussing on those at high-risk oncogenic potential by circumcision status in adult men. Empty triangles represent published studies in which the entire penis was sampled and circumcision status was determined by physical examination. Shaded triangles represent published studies in which only the glans was sampled and circumcision status was determined by physical examination. Inverted shaded triangles represent published studies in which entire penis was sampled and circumcision status was determined by patient report. Shaded circles represent published studies in which only the glans was sampled and circumcision status was determined by patient report. Solid squares represent likely unpublished studies using the “trim and fill” method.