Which are more correctly diagnosed: conventional Papanicolaou smears or Thinprep samples? A comparative study of 9 years of external quality-assurance testing

• Published in: Cancer cytopathology Vol. 123; no. 2; p. 108
• Main Authors: Cummings, Margaret C, Marquart, Louise, Pelecanos, Anita M, Perkins, Gail, Papadimos, David, O'Rourke, Peter, Ross, Jennifer A
• Format: Journal Article
• Language: English
• Published: United States 01.02.2015
• Subjects: Australia; Female; Humans; New Zealand; Papanicolaou Test; Quality Assurance, Health Care; Uterine Cervical Neoplasms - diagnosis; New Zealand; Australia; cytology; proficiency testing; Papanicolaou smears; cervical smears
• ISSN: 1934-6638
• DOI: 10.1002/cncy.21498

The Royal College of Pathologists of Australasia Cytopathology Quality Assurance Program offers external testing in gynecologic cytology to Australasian and international laboratories. Laboratory interpretation of conventional Papanicolaou (Pap) smears is compared with interpretation of liquid-based cervical cytologic (ThinPrep) samples. Conventional Pap smears and ThinPrep samples were distributed to participating laboratories annually over 9 years (from 2004 to 2012), and a range of entities was tested. Target responses and major error rates and diagnostic trends over time were explored. In total, 23,373 conventional Pap smears and 14,104 ThinPrep samples were reported. Both Australasian (P = .003) and international (P < .001) laboratories achieved a higher percentage of the target diagnosis of squamous dysplasia with ThinPrep samples. Australasian laboratories more accurately diagnosed endocervical adenocarcinoma in situ with conventional smears (P = .036), whereas international laboratories performed better with ThinPrep samples (P = .006). Sampling of the lower uterine segment was more accurately diagnosed by both Australasian (P < .001) and international (P = .001) laboratories using conventional Pap smears. Significant improvements in achieving the target diagnosis over time for squamous dysplasias using both modalities were observed for Australasian and international laboratories (P < .001 for both). There was improvement in diagnosing high sampling using ThinPrep for both groups (P = .001 and P = .015, respectively). Australasian performance declined over time in reaching the target of normal (no infections) for both conventional (P = .001) and ThinPrep (P < .001) techniques and for international laboratories with the ThinPrep technique (P < .001). Participation in external proficiency testing in cervical cytology allows an analysis of performance, the identification of areas of diagnostic difficulty, a review of trends over time, and the highlighting of topics for ongoing education.