Interlaboratory reproducibility of liquid-based equivocal cervical cytology within a randomized controlled trial framework

Within a multicentre controlled trial framework, an external quality control (EQC) was scheduled to evaluate the interlaboratory reproducibility of liquid-based cytology. In particular, this EQC intended to evaluate the reproducibility of the ASCUS diagnosis. A selected set of 30 slides (4 within normal litnit cases, 16 atypical squamous cells of undetermined significance; 4 lowgrade squamous intraepithelial lesions and 6 high-grade squamous intraepithelial lesions) circulated among the 13 laboratories involved in the trial. Kappa values were obtained fi-oin the comparison between individual laboratory diagnoses and majority diagnoses with target diagnoses. Specific kappa v,alues resulted moderate to high for HSIL and low to moderate for LSIL and WNL. Meanwhile, the specific kappa for ASCUS was below 0.4 in 12 of 13 participating laboratories. The lack of reproducibility for ASCUS was not a result of the introduction of this new technology but rather to the low repro-ducibility of the ASCUS category itself stemming from intrinsic at icerta in ties in the reporting criteria.