Measurement characteristics of anal cytology, histopathology, and high-resolution anoscopic visual impression in an anal dysplasia screening program

Background: The study aims were (1) to estimate agreement between consecutive anal cytologic examinations, between concurrent cytologic examination and histopathology, and between high-resolution anoscopy (HRA) visual impression and histopathology and (2) to estimate the prevalence of severe dysplasia by concurrent cytologic category. Methods: Prospective study of HIV-infected patients receiving anal dysplasia screening as part of routine care. Agreement between measures was estimated by weighted kappa-statistics. Results: Between July 2000 and September 2003, 1864 patients underwent 2947 anal cytology tests. Excluding unsatisfactory tests (6%), 642 patients had repeat cytologic evaluation and 154 had concurrent cytology tests and biopsy. Using 4-category cytology grading, kappa-agreement between the first 2 cytologic measurements was 0.36. Comparing concurrent cytology tests and biopsy, kappa-agreement was 0.36. Comparing the most severe HRA visual impression and biopsy, kappa-agreement was 0.32. The prevalence of anal intraepithelial neoplasia 3 at biopsy by concurrent cytology category was 0 (cytology normal), 21% (atypical squamous cells of uncertain significance), 27% (low-grade squamous intraepithelial lesion), and 54% (high-grade squamous intraepithelial lesion). Conclusions: These data suggest that the reproducibility of key screening measures is moderate at best but of similar magnitude to that of other studies of anal and cervical dysplasia screening. As candidate interventions to treat or prevent precursor lesions enter clinical development, standardization and improvement of measurement methods are essential.