Ovarian cancer survivor

Who chooses when an operation is worth having?

EDITOR--In view of the many excellent publications recently in the BMJ on the need to involve consumers in decisions about treatment and screening,[1] I am concerned that Urban's editorial on ovarian cancer includes the statement: "Five operations for each cancer found is an acceptable surgery to malignancy ratio."[2]

To whom is five operations to find one cancer an acceptable ratio? Women? Health authorities? Surgeons? Radiologists? Surely it is time to abandon this paternalistic and rigid approach of prescribing what is or is not an acceptable trade-off of benefits and risks and then implementing that for everyone. Consumers should be given the opportunity to decide whether, for them, the benefit is worth the risk. For those of us in the health business, our role should be to support consumers in their decision making rather than to make judgments for everyone.

Alexandra Barratt senior lecturer in epidemiology Department of Public Health and Community Medicine, University of Sydney, NSW 2006, Australia alexb@pub.health.usyd.edu.au

[1] Austoker J. Gaining informed consent for screening. BMJ 1999;319:722-3. (18 September.)

[2] Urban N. Screening for ovarian cancer. BMJ 1999;319:1317-8. (20 November.)

Mass screening may not prove to be sound public health policy

EDITOR--Urban persuasively makes the scientific case for a definitive trial of ovarian screening, but it is important to consider the wider public health context.[1] A public health. policy of mass screening for ovarian cancer may not prove feasible for several reasons.

Establishing a diagnosis in patients with positive results on screening usually requires oophorectomy, a much more invasive procedure than that required for the diagnosis of breast, cervical, or colorectal cancer. There is no evidence to support the assertion that five such operations per cancer detected is "acceptable."[1] This is essentially a judgment, considering any benefit of detecting malignancy against the risks of unnecessary morbidity and mortality. At present there is insufficient evidence to permit informed judgments by either clinicians or women.[2]

Ovarian cancer develops rapidly, and an annual screening interval is proposed; at present the NHS struggles to achieve a three yearly interval for breast screening. Furthermore, CA125 screening performs adequately only in postmenopausal women. How could such a strategy be explained to older premenopausal women, who are at the same risk of ovarian cancer as their postmenopausal peers?

Ovarian cancer is a relatively uncommon disease. The theoretical maximum detection rate for annual screening is equivalent to the incidence of 1 in 2200 for women aged 50-69. The number who may benefit from screening will be considerably lower, however. About 70% of women with ovarian cancer die within five years[3]; if screening reduces mortality by about 30%, there would be 20 additional five year survivors for every 100 cancers diagnosed. This means that for each extra five year survivor, five cancers would need to be detected, 20 unnecessary operations performed, and 11 000 women screened.

It is misleading to state that computer simulation models predict screening to be potentially cost effective.[1] Maximum cost effectiveness has been estimated at about $50 000 (31 250 [pounds sterling]) per life year saved, based on private medical charges in 1990 for individual components of screening.[4] These charges are likely to bear little resemblance to the true costs of setting up a national screening programme.

A large randomised trial requires considerable investment. Even if a mortality reduction is shown, screening may not represent a good use of public funds. Research must look at the broader issues of whether ovarian screening is a cost effective and practicable public health policy, and not merely efficacious under ideal conditions.

Ruth Bell lecturer in public health medicine Centre for Cancer Epidemiology, Manchester M20 4QL ruth.bell@cce.man.ac.uk

[1] Urban N. Screening for ovarian cancer. BMJ 1999;319:1313-80. (20 November.)

[2] Bell R, Petticrew M, Sheldon T. The performance of screening tests for ovarian cancer: results of a systematic review. Br J Obstet Gynaecol 1998;105:1136-4 7.

[3] Coleman MP, Babb P, Damiecki P, Grosclaude P, Honjo S, Jones J, et al. Cancer survival trends in England and Wales, 1971-1995: deprivation and NHS region. London: Stationery Office, 1999. (Studies in Medical and Population Subjects No 61.)

[4] Urban N, Drescher C, Etzioni R, Colby C. Use of a stochastic simulation model to identify an efficient strategy for ovarian cancer screening. Control Clin Trials 1997;18: 251-70.

Authors' response

EDITOR--Barratt and Bell make important points about the editorial on screening for ovarian cancer written by one of us (NU).[1] With respect to an acceptable surgery to malignancy ratio, Barratt argues that women should make their own decisions about whether, for them, the benefit is worth the risk. Bell points out that current evidence is insufficient to permit informed judgments about the comparative benefits and risks of the ratio by either clinicians or women.

We agree that in the clinical setting screening protocols should be tailored to address the risk factors and preferences of individual women. However, we submit that without a carefully designed randomised controlled trial to obtain this information we cannot do this because, as Bell points out, currently there is insufficient evidence on which to base screening decisions.

Bell suggests that investment in a trial may not be warranted because ovarian cancer screening is unlikely to be cost effective because of the low incidence and rapid progression of the condition and the high costs of screening. We recognise the importance of these issues and submit that the cost estimates used in our published work are not underestimates of screening programme costs. We assumed there that transvaginal sonography costs $150 (94 [pounds sterling]) and that CA125 costs $40.[2] However, the costs may be lower in a high volume setting, in the range of $25 to $50 for transvaginal sonography[3] and about $25 for CA125.[4]

Of particular concern is the judgment from the editorial that a surgery to malignancy ratio of 5:1 is "acceptable." It was made in the context of the design of a screening protocol to be tested in a randomised controlled trial, in which a surgery to cancer ratio must be chosen that is likely to be tolerable for most women. The potential risks and benefits are described to each woman, and she must choose whether or not to participate. Through the informed consent process the consumer is given the opportunity to decide whether, for her, the benefit is worth the risk. This research process does not preclude consumer involvement in decision making about screening in the clinical setting. Rather, results of the research make informed clinical decision making possible.

Nicole Urban member Charles W Drescher affiliate investigator, Cancer Prevention Research Program Fred Hutchinson Cancer Research Center, Seattle, WA 98104, USA

[1] Urban N. Screening for ovarian cancer. BMJ 1999; 319:1317-8. (20 November.)

[2] Urban N, Drescher C, Etzioni R, Colby C. Use of a stochastic simulation model to identify an efficient strategy for ovarian cancer screening. Control Clin Trials 1997;18: 251-70.

[3] Pavlik EJ, van Nagell JR Jr, DePriest PD, Wheeler L, Tatman JM, Boone M, et al. Participation in transvaginal ovarian cancer screening: Compliance, correlation factors and cost. Gynecol Oncol 1995;57:395-400.

[4] Urban N, McIntosh M, Clarke L, Jacobs I, Karlan B, Anderson G, et al. Socioeconomics of ovarian cancer screening. In: Sharp F, Blackett A, Berek J, Bast R, eds. Ovarian cancer 6. Oxford: Isis Medical Media, 2000.