Until recently, balancing the risks and harms in cervical cancer prevention programmes was arbitrary and opaque. Annual cytology was recommended to conform to the periodicity of the Earth's orbit rather than to epidemiological evidence. Post-colposcopy management was based on expert opinion alone.
Recently, however, the assessment and reduction of risk whilst minimising patient burden has become more transparent and scientific. In 2012, the American Society for Colposcopy and Cervical Pathology with staff from the US National Cancer Institute developed guidelines for the management of women with screening abnormalities on the basis of explicit risk measures using ‘big data’. From a database with 8 years of follow-up results from women screened through the Kaiser Permanente Northern California health maintenance organisation, analysts measured 5-year risks for cervical intraepithelial neoplasia grade 3+ (CIN3+) (Katki et al., J Lower Genit Tract Dis 2013;17:S28–35). Benchmarks for follow-up were established on the basis of prior US consensus: 3-year screening intervals for women with negative Papanicolaou (Pap) tests, and colposcopy for women with Pap tests read as low-grade squamous intraepithelial lesion (LSIL). Similar management was targeted to women with similar risks for CIN3+. Testing after cytology modified risk: a positive test for high-risk human papillomavirus (HPV) increased risk, whereas negative colposcopy decreased risk. Management followed from the level of risk measured after combinations of factors. For example, women with a negative Pap test but a test positive for HPV16 had a risk for CIN3+ greater than women with LSIL, and so were tracked to colposcopy. Women with a Pap test read as atypical squamous cells of undetermined significance (ASCUS) with a negative HPV test had CIN3+ risk near that of a negative Pap test, and so were tracked to 3-year retesting. Other results, such as colposcopy and biopsy findings, also modified risk and led to more nuanced management.
US management guidelines cannot be translated to other societies with different thresholds that result from differences in cytology diagnostics, use of HPV testing, malpractice risk, central organisation of screening and skills certification. Using a population from the UK, the TOMBOLA (Trial of Management of Borderline and Other Low-Grade Abnormal Smears) Group used explicit risk assessment to recommend 2-year follow-up for women with CIN1 after a previous low-grade smear, given their 5% risk of CIN3 (Gurumurthy et al., J Lower Genit Tract Dis 2014;18:203–9). The accompanying paper extends TOMBOLA data to show that, after low-grade cytology and negative colposcopy, the 3-year risk of CIN3+ is <1%, allowing routine recall with cytology at 3-year intervals (Cruikshank et al., Intern J Obstet Gynaecol doi. 10.1111/1471-0528.12932). This risk falls well within the risk band used by Katki et al. in the USA to assign women to 3-year follow-up. The risk of CIN3+ after moderate or severe dyskaryosis is probably higher, and more aggressive management, either large loop excision of the transformation zone (LLETZ) or more careful observation, is probably indicated for these women despite negative colposcopy.
As clinicians continue to amass outcomes data on large patient populations, it will become increasingly simple to query databases to determine the downstream risk after various concatenations of risk markers: cytology, HPV tests, HPV genotype results, colposcopy impression, biopsy and endocervical curettage histology, and findings with p16ink4a and other biomarkers. From this, country-specific or even system-specific management guidelines may be deduced. The disadvantage to greater knowledge will be greater complexity. Soon clinicians managing women will be unable to remember all the data combination and risk results. Instead, they will need to rely on computer programs and smartphone applications.
Disclosure of interests
I have no financial conflicts of interest relevant to this paper. I have published articles on the accuracy of colposcopy and have one under consideration at BJOG.