2 Clinical need and practice

2.1 The annual incidence of cervical cancer in the UK in 2003 was estimated to be 9.7 per 100,000 population, which corresponds to a mortality rate of 3.9 per 100,000 population (2001). Pre-cancerous cervical cells cause no symptoms and may only be detected by population screening methods. The NHS Cervical Screening Programme (NHSCSP) began national coordination of cervical screening in 1989. The nature of a screening programme is to screen a large subsection of the population (in this case women) to identify a subpopulation that is thought to be at sufficient higher risk of developing a disease such as cervical cancer to warrant further diagnostic investigation and treatment. Diagnostic tests used in screening programmes are not 100% sensitive (some false-negative tests are reported), and there is a possibility that pre-cancerous cells will not be detected in a small number of women. Screening programmes like the NHSCSP and Cervical Screening Wales, which screen women at regular intervals, reduce the likelihood of pre-cancerous cells and invasive cancer being missed on the basis of one false-negative result because they are picked up at subsequent cervical smear tests.

2.2 The NHSCSP and Cervical Screening Wales use the Papanicolaou (Pap) smear test for cytological screening. Women aged 20–64 years are screened at 3–5-yearly intervals (depending on Strategic Health Authority policy) for the early detection and treatment of pre-cancerous cells, with the aim of reducing the incidence and associated mortality of cervical cancer. Approximately 3.9 million women are tested in England each year, equating to coverage of 71.2% for 3-yearly screening and 81.6% for 5-yearly screening in 2001–02.

2.3 The Pap smear is usually carried out by a GP or nurse at a primary care or community clinic. Cervical cells are collected using a disposable spatula device, spread on a glass slide and fixed. The slide is then sent to a hospital laboratory where it is stained and examined by a cytologist.

2.4 Smear tests are evaluated according to morphological features of the cervical cells, which indicate the degree of cellular abnormality (dyskaryosis). In the UK, smears are categorised using the British Society for Clinical Cytologists (BSCC) guidelines as negative, borderline, mild, moderate, severe, '?glandular neoplasia', '?invasive' or inadequate. In the USA, the Bethesda system is used to classify cervical smears as atypical squamous cells of undetermined significance, atypical glandular cells of undetermined significance, low-grade squamous intraepithelial lesions or high-grade squamous intraepithelial lesions. Approximately 90% of cervical cancers are squamous cell carcinomas; the potential precursors of these relate to the borderline, mild, moderate and severe dyskaryosis in the BSCC guidelines or the atypical squamous cells of uncertain significance, low-grade squamous intraepithelial lesions and high-grade squamous intraepithelial lesions in the Bethesda system. Approximately 15% of cervical cancers are adenocarcinomas and are frequently undetected by screening, although potential precursors are recognised (described as cervical glandular intraepithelial neoplasia [CGIN] or adenocarcinoma in situ) and may be detected on cytology as '?glandular neoplasia' in the BSCC classification ('atypical glandular cells of undetermined significance' or adenocarcinoma in situ in the Bethesda system). The BSCC and Bethesda classification systems are similar but not directly comparable.

2.5 Patient management depends on the classification of the smear test. Women with negative tests are invited for re-screening at the standard 3–5-year interval, while those with borderline or mildly dyskaryotic smears are monitored at a reduced screening interval. Women with moderately or severely dyskaryotic smear tests, mildly dyskaryotic smears on a maximum of two tests, or persistent inadequate or borderline tests are referred for additional diagnostic testing, such as visual examination of the cervix with a binocular microscope (colposcopy), when a tissue biopsy may be taken for histological examination.

2.6 The principal criteria used to assess the effectiveness of the LBC method compared with the Pap smear are the sensitivity and specificity of each method, and the rate of 'inadequate' specimens. Sensitivity is the extent to which a test identifies true-positive samples (sensitivity decreases as the number of false-negatives rises), and specificity is the extent to which the test excludes true-negatives (specificity decreases as the number of false-positives increases). Knowledge of the prevalence of pre-cancerous disease is required in order to assess the number of false-negatives, and so surrogates of sensitivity are used, such as detection rates for high-grade and low-grade cytological abnormalities.

2.7 On average, approximately 8% (range of 5.9–11.0%) of Pap smear tests are inadequate; that is they cannot be interpreted because of problems with sample collection or preparation (such as insufficient cervical cells), or the presence of inflammatory cells, blood or mucus, which obscure the sample. Women with inadequate test results are required to attend a repeat test, which is inconvenient and may cause anxiety.