Review of cytology and histopathology as part of the NHS Cervical Screening Programme audit of invasive cervical cancers

Cytopathology. 2012 Feb;23(1):13-22. doi: 10.1111/j.1365-2303.2011.00948.x.

Abstract

Objective: To audit pathology slide reporting in the Cervical Screening Programme in England by reviewing cytology and histology slides from women subsequently diagnosed with invasive cervical cancer.

Methods: Between April 2007 and March 2010, 6113 women diagnosed with cervical cancer were identified. Cervical cytology and histology slides taken within 10 years of diagnosis were identified and where possible reviewed after a nationally agreed protocol. Reviewers were not blinded to the original reading of each sample. Most cytology samples before 2005 were conventional, most after 2007 liquid based.

Results: Of 13,745 cytology results from women developing cervical cancer, 55% were reviewed. The review result was identical for 55% of slides. Of 3759 originally normal slides, only 45% were normal on review: 11% were inadequate, 21% low grade (borderline or mild dyskaryosis) and 23% high grade (moderate dyskaryosis or worse). Of tests originally normal taken over 5.5 years before diagnosis, 14% were upgraded to high grade compared with 37% within 3.5 years of diagnosis. Of 5159 histology specimens recorded within 10 years of diagnosis of a cancer, 3895 were reviewed. Overall, 94% of samples reviewed retained the original diagnosis. One per cent (33/3012) of cancers were downgraded and 5% (6/112) of negative samples were upgraded to cancer upon review (four of which were taken within 2 months of diagnosis). In comparison, 15% (14/91) of cervical glandular intraepithelial neoplasia (CGIN) and 12% (38/314) of cervical intraepithelial neoplasia grade 3 (CIN3) were upgraded to cancer.

Conclusions: In spite of the excellent quality of cytology in England, a high proportion of negative cytology taken up to three and a half years before diagnosis were considered to contain abnormal cells by reviewers informed of the subsequent cancer. Continuing these reviews, with a strong focus on education, will ensure a clear understanding of these slides and further reduce the risk of developing cervical cancer.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Cytodiagnosis / statistics & numerical data*
  • Diagnostic Errors*
  • England
  • Female
  • Humans
  • Mass Screening / methods
  • Medical Audit*
  • Neoplasm Invasiveness
  • State Medicine*
  • Uterine Cervical Dysplasia / pathology*
  • Uterine Cervical Neoplasms / pathology*
  • Vaginal Smears / methods