In one study, the authors admit to false-negative rates of between 7 and 60 per cent.23
In another report, one in every five cervical cancer deaths was due to poor management or misdiagnosis by doctors. In one in every seven of these cases, the smear tests had been read as normal. Re-analysis of the slides showed that early abnormalities had in fact been present, but were missed.
Interpretation of the results varies wildly, depending on who is looking at the slides. You could even get a different interpretation from the same person looking at the same slide on separate occasions. This is particularly so, says Professor McCormick, with the minor changes that give rise to most reported abnormalities.24
A report from the National Audit Office, ‘Cervical and Breast Screening in England’, found a wide disparity in interpretations of findings and a lack of benchmarks against which to compare results. The audit found that in some areas of England, nearly a fifth of all smears were classified as abnormal, compared with 3 per cent in other areas.25 This lack of any standards is responsible for many false diagnoses of cancer.
In Scotland, some 20,000 tests done under the screening programme at the Inverclyde Royal Hospital had to be re-examined after evidence that the doctor doing the analysis may have misread the results. On a preliminary re-examination, 40 out of 1,000 smears taken since 1988 were found to be ‘inadequate’ and to require a repeat test.26
The Scottish debacle is only the latest in a series of such incidents in the UK. In 1987, in Liverpool, 45,000 tests were re-examined and 911 found to have been wrongly diagnosed. In 1988, in Manchester, a batch of 3,000 tests passed as clear were re-analysed and 60 found to be suspect.
Large numbers of smears are also technically not up to scratch. Dr Chandra Grubb, head of the Department of Cytology at the Royal Free and University College Hospital in London, estimates that some 10 per cent of all smears sent to cytology departments are useless, and a further 40 per cent are of limited usefulness because doctors haven’t taken the smear correctly or have taken it from the wrong site.27 With this kind of terrible batting average, the likelihood is that screening not only isn’t going to pick up your cancer, but could set you on the road to potentially risky treatments when you don’t need them.
The conventional treatment for early ‘precancerous’ lesions employs a colposcopy (a magnifying glass with a light) and biopsy (exploratory surgery), diathermy (burning the abnormal cells) or cytotherapy (which employs a freezing probe to freeze the outlaw cells). These procedures can all cause haemorrhage or permanently damage the cervix, resulting in an ‘incompetent’ or narrowed cervix, and thus affect a woman’s chances of carrying a baby to term.
Dr Robert Mendelsohn liked to tell the story of a colleague of his whose wife received a positive reading. She went ahead with a cone biopsy, which caused such excessive bleeding that she had to have an emergency hysterectomy, during which she almost died from the anaesthesia. All because of a test that might have been wrong in the first place.28
One of our readers, a young woman in her early twenties, had been diagnosed as having stage 2–3 abnormal cells, and was scheduled for an operation to have them frozen or burned out. At the eleventh hour, she decided to have a second smear test at another laboratory. Her new test showed that the first test was wrong; her problem turned out to be a simple infection.
Individual doctors also differ widely in their views of how to treat abnormalities. The British National Audit Office report showed that many doctors opt for radical treatment such as cervical conization for cases of mild abnormalities, which would eventually resolve themselves without intervention.29
Some reports demonstrate that getting in there early and aggressively treating cervical abnormalities doesn’t do any good, anyway. In one recent study, referring women with mildly abnormal smear test results for the more invasive examination produced no more favourable outcome than adopting a policy of watchful waiting. Referring women for a colposcopy exam, often with a biopsy, or simply giving them a repeat smear test after several months produced identical results: 1.6 per 1,000 cases developed into cervical cancer. This means 2,500 women were sent for colposcopy – with its inherent risks of causing infertility – to save one case of cancer.30
Because of the high rate of false-positives, some countries like the US and Switzerland have now introduced a new technology called ‘liquid-based cytology screening’ (LBC), also known as ‘monolayer cytology’. The specimen is collected by a special spatula, the head of which is rinsed into or broken off into a vial of preservative, so that the entire sample is immediately preserved in liquid.
The UK’s own pilot studies showed that LBC produces more accurate results, with only 1.6 samples considered ‘inadequate’, or unable to be read, compared to more than 9 per cent of traditional Pap smear slides.
Nevertheless, there is some evidence that the LBC test is even less reliable and more likely to give false-positive and false-negative results than the conventional variety: 87 per cent, compared with 91 per cent for ordinary Pap smears.31
This questionable batting average has not deterred the UK government. In October 2003, it announced that LBC would be ‘rolled out’ over the next five years, with the intention of replacing the conventional Pap smear test, at a cost of some £10 million.
The National Co-ordinating Network now specifically recommends that women with minor cell abnormalities – ‘borderline or mildly dyskarotic smears’ – adopt a path of surveillance – that is, have the smear test repeated six months later. The women should only be referred for colposcopy if the smear continues to show abnormality. At very least, according to the US CDC, waiting an interval of three years (or five years, if you are over 50) won’t affect mortality rates. If you do get a positive reading, insist on a repeat smear from another lab before going down the more invasive route of biopsies and worse.
MAMMOGRAMS
Mammography – an x-ray of the breast designed to pick up early malignancies – is the other screening test being stepped up sharply. Breast cancer, the biggest lady-killer after lung cancer, claims the lives of an estimated 40,000 American women every year;32 some 33,000 new cases are diagnosed every year in the UK – double the incidence of the 1950s. As cases continue to spiral upward (one in nine women contract it in the US, and one in twelve in the UK), the pressure is on for women, particularly those over 40, to have regular screenings, and mammography has become a huge growth industry.
Despite decades of enormous government resource and a massive money-spinner aimed at efforts to improve early detection and local treatment, the bottom line is that the level of breast cancer mortality has remained constant.
Breast cancer deaths in England and Wales fell by 12 per cent in the 1990s. Nevertheless, health officials who ascribe this sudden drop to their extensive mammogram screening programmes have no reason to be self-congratulatory. New research has discovered no evidence to link the two, although screening has helped detect more cases earlier. The National Cancer Registration Bureau believes the fall may be more likely associated with the increasing use of the drug tamoxifen, which slows cancer growth, than with any screening. Since nationwide screening was introduced in 1988, recorded incidence of the disease in the 50–64-year age group rose by 25 per cent.33 Furthermore, the fall in mortality began in 1985, but the first NHS screening units were not working until three years later, and Great Britain as a whole wasn’t sufficiently covered until 1990. As Royal Marsden Hospital breast cancer specialist Michael Baum writes, claiming