Education And Debate

Controversies in Management: Immediate referral to colposcopy is safer

BMJ 1994; 309 doi: (Published 03 September 1994) Cite this as: BMJ 1994;309:591
  1. W P Soutter
  1. Royal Postgraduate Medical School, Hammersmith Hospital, London W12 0HS.

    Cervical cytological screening is effective in reducing the incidence of and mortality from cervical cancer.1 However, no screening test is perfect, and invasive squamous cancers do occur in screened women.2 As coverage of the population increases, management of mild cytological abnormalities will become more important. Currently, about 2% of all smears in England and Wales are reported to show mild dyskaryosis, although it varies among regions.3 In 1987, the intercollegiate working party on cervical cytology screening recommended immediate colposcopy for all women with dyskaryosis, where resources permitted,4 but others have subsequently suggested various formulas for cytological surveillance.

    A retrospective study in the United Kingdom in 1986 showed that 48% of women with mild dyskaryosis had cervical intraepithelial neoplasia grade II or III.5 The rate of abnormality did not depend on the number of mildly dyskaryotic smears before referral. This high prevalence of grade II or III neoplasia has been confirmed by prospective studies.*RF 6,6a* Some people have suggested that most of these lesions are small and inferred that the risk of progression to invasive disease will be less than with larger lesions.7 But there are no data to support that contention.

    Surveillance is inadequate

    Cytological surveillance is often said to allow most women with mild dyskaryosis to avoid colposcopy. However, after two years only a quarter will have an abnormal smear result.*RF 6,6a* An analysis of all the recent studies of cytological follow up in the United Kingdom showed that the cumulative referral rate after about four years ranged from 14% to 64%.8 The two studies with the lowest referral rates had the highest rates of invasion.

    One of the main measures of the success of cervical screening is the incidence of invasive cancer. In one retrospective study of cytological surveillance that has been widely quoted as reassurance of the safety of this approach, 10 of the 1781 patients developed invasive cancer.9 Excluding the three carcinomas that occurred in the 434 women who were lost to follow up, this represents an annual incidence of invasive cancer of 143 per 100 000 women8 - hardly reassuring. A similar study in north London found an annual incidence of invasive disease of 420 per 100 000, a rate 60 times greater than the age adjusted background rate.10 In the large British studies of cytological surveillance of women with mild cytological abnormalities the annual incidence of invasive cancer ranged from 143 to 420 per 100 000.8 The average rate of all six studies was 208 per 100 000.

    The women in these studies were mainly aged 15-34 years. They had a 16- 47 times greater incidence of invasive disease than women of the same age in the general population, in whom the incidence of invasive cancer was 9 per 100 000 in England and Wales in 1985. These data do not include cancers that may have occurred in women lost to follow up. Given that all of these women had been screened and that 80% of invasive cancers at that time occurred in women who had never had a smear, the real increased risk is likely to be even greater.

    A recent decision analysis of this problem concluded that repeating the smear would be almost as effective as referral to colposcopy in reducing the rate of invasive cancer.11 However, the cancer risks were calculated in a very indirect way that underestimates the rate in women followed cytologically. Furthermore, prevalence data rather than incidence rates were used in the calculations. Substitution of five year cumulative incidences for prevalences in their calculations, suggests that immediate referral to colposcopy would result in a 54-84% reduction in the risk of invasion.8


    Average annual incidence invasive cancer in women with mild cytological abnormalities followed cytologically and in women in England and Wales aged 15-34 years in 1985. Error bars represent 95% confidence intervals. The age group 15-34 was chosen as an approximation to the age distribution of populations included in studies.

    Immediate or deferred colposcopy?

    Because three quarters of women with mild dyskaryosis will eventually need colposcopy, the choice for most is immediate colposcopy or deferred colposcopy with the risk that a quarter may be lost to follow up.*RF 6a* The objections to immediate referral to colposcopy are the cost, lack of resources, and the psychological impact on women. One comparison of the cost of colposcopy with that of multiple repeated smears has suggested that immediate colposcopy would be cheaper.11 In my experience many women with mildly abnormal smears ask for colposcopy even when cytological surveillance has been recommended. Colposcopy does induce a great deal of anxiety but this can be reduced substantially by explanatory leaflets sent out with the appointment.12 I do not know of any published study comparing the levels of anxiety in women being followed cytologically with those attending for immediate colposcopy. This obviously requires investigation.

    Commentary: immediate colposcopy is not justified

    Cervical screening reduces the incidence of cervical cancer. Optimising the take up rate for cervical screening and the reliability of laboratory analysis is therefore an extremely worthwhile aim. In contrast, the advantages of immediate colposcopy in those with mild dyskaryosis are unclear. The bottom line is that no prospective information from randomised trials is available. On the basis of the information presented here it seems that immediate colposcopy cannot be justified in terms of clinical need or cost. A repeat cervical smear four to six months later with colposcopy if the result is still abnormal is a better way forward, with resources put primarily into optimising the efficiency of cervical screening. -- PETER C RUBIN, professor of therapeutics, University of Nottingham


    Women with mild dyskaryosis often have cervical intraepithelial neoplasia grade II or III. If followed cytologically most need colposcopy eventually, many are lost to follow up, and even those who return for all their smear tests are at high risk of developing invasive cancer. Though there has been no prospective randomised comparison of cytological follow up with immediate colposcopy, it seems prudent to refer this high risk group of women for colposcopy after the first abnormal smear. Many will have no abnormality, will not require biopsy, and may be reassured quickly and returned to the screening programme. Those who do have abnormalities can be treated effectively in the outpatient clinic with only mild discomfort and little or no risk to their fertility or fecundity.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 6a.
    8. 7.
    9. 8.
    10. 9.
    11. 10.
    12. 11.
    13. 12.
    View Abstract

    Sign in

    Log in through your institution