The area of management in abnormal cervical cytology that has always caused the greatest debate is that relating to minor abnormalities. To a large extent this is represented by the dilemma between immediate referral to colposcopy and follow-up in the form of repeat cytology. In the UK, in common with other European countries, minor cytological abnormalities are responsible for around 50% of colposcopy referrals.1 These strategies are probably broadly equivalent in terms of the detection of cervical intraepithelial neoplasia (CIN), but represent very different approaches in terms of policy, in risk of non-adherence to the protocol and in women’s perception of the treatment. This paper by Fredericksen et al. explores women’s perceptions of the treatment strategies by attempting to determine which approach the majority of women would prefer.2 In order to do this, they have conducted a systematic review of the literature. They have drawn on a rather heterogeneous group of studies in order to reach their conclusion, which is that broadly the majority of women would prefer an immediate referral to colposcopy over follow-up with repeated cytology. The studies they have selected as their data sources range from those in which women make a real-time choice when faced with the need to have minor cytological abnormalities managed, to studies where women have been asked to state their preference in the hypothetical situation that they had been found to have minor cytological abnormalities requiring management.
The authors report that although the majority of women would prefer immediate referral, a significant minority, probably around one-third would prefer surveillance. There are probably a number of reasons why different women have a different preference. Immediate colposcopy offers a way forward for women who are very anxious about the abnormality, and want a decision as to whether treatment is required or not. It is also likely that many more practically minded women will prefer to have the matter resolved than to have to rely on follow-up, which could have a more indeterminate outcome. On the other hand, women who prefer to have surveillance could include those who are anxious about having to undergo a colposcopic examination, and others who if told that they are at lower risk, are relaxed about having surveillance in the hope that things will be resolved naturally.
There are a number of key areas that are relevant to a preference for colposcopy referral over surveillance, or vice versa: (i) the proportion of affected women who are eventually diagnosed with CIN; (ii) the psychological consequences of the abnormality; (iii) does the fact that women have different preferences mean that a choice should be offered between surveillance and immediate referral? These are worthy of brief comment. The first point is that the randomised study by the TOMBOLA group, a large and recent controlled study, showed that both approaches resulted in equivalent numbers of women being diagnosed with CIN2 or worse.3 The surveillance policy carries the potential benefit of low-grade abnormalities resolving naturally in some women, but also carries the risk that these abnormalities will simply persist in others. This may require eventual treatment, and there is always a risk in surveillance of non-adherence to follow-up protocol, which creates its own difficulties. On the other hand, immediate colposcopy might lead to over-treatment of disease that was destined to regress, and therefore unnecessarily expose some women to the potential long-term sequalae of treatment.4 A second difficulty can be the associated psychological consequences. In one of the papers used for the systematic review, women were randomised between cytological surveillance and being able to choose between surveillance and immediate colposcopy in managing mild abnormalities.5 The primary outcome measure was the general health questionnaire, completed after 12 months. As reported in Fredericksen’s paper, a small majority chose colposcopy, but the key finding was that whether the women had been randomised to surveillance or to no choice, psychological distress fell from baseline to 12 months, and there was no statistical difference between each strategy at each time point. It therefore appears that even though women are not able to achieve their preference, this does not impact adversely on psychosocial outcomes, and whichever approach is used, women are less concerned over time. Thirdly, we live in an age where individual decision-making in medical care is becoming more prominent; however, in a national screening programme choice would create a level of complexity that is probably not feasible to manage. The study performed by Fredericksen indicates a majority preference, i.e. referral for colposcopy, and this should be borne in mind in determining policy; it seems, therefore, that immediate referral is generally preferred by women over a policy of surveillance.
As things turn out, human papillomavirus (HPV) triage of these mild abnormalities is being increasingly introduced into clinical practice and into national programmes, because this allows a very low-risk group, who do not require referral (i.e. women who are HPV negative), to be distinguished from a higher risk group, for whom colposcopic referral is appropriate (i.e. women who are HPV positive).4 This process of triage is probably highly acceptable to women because the concept of separating those who do not require further management from those who do is easily understood, even though some women may have individual difficulties with knowing that they are HPV positive.
Although not relevant to the study itself, the authors go on to discuss the implications of preference for colposcopy referral in the context of primary HPV screening, picking out the group of women who are HPV positive but cytology negative. Such women are at risk of developing a higher proportion of CIN than the screened population overall, and there will be some prevalent disease undetected by cytology. Introduction of HPV in primary screening will bring challenges in this regard, and these will be resolved by different approaches in different countries. There may be some scope for triaging or stratifying this group, with restricted HPV genotyping or using other biomarkers, so that a small proportion who are considered at greater risk are referred for colposcopy, and the remainder are offered an interval screen prior to the next routine screening round.
To conclude, we welcome studies that address women’s preferences in the way in which they are managed. Allowing the prevailing view of women to be fed into policy development is likely to result in a high level of acceptance. Although individual choice is feasible for a single woman, national screening programmes require more uniform and simple protocols.