Anybody who questions the value of screening tests for cancer risks provoking the wrath of powerful vested interests.
Dr Michael Fitzpatrick
Anybody who questions the value of screening tests for cancer risks provoking the wrath of powerful vested interests. Recent controversies in the USA around mammography and tests for prostatic cancer show how these interests keep the screening programmes going despite growing doubts over their efficacy.
In January 2002 the independent screening and prevention panel for the National Cancer Institute’s (NCI) online database announced that there was not enough evidence to show that mammography prevents deaths from breast cancer (1). This conclusion followed a review of a ‘meta-analysis’ of seven randomised trials of mammography that was published in the Lancet last year (2). According to Ole Olsen and Peter Gotzsche of the Nordic Cochrane Centre in Copenhagen, Denmark, these trials confirmed that mammography did not save lives, though it did lead to more aggressive treatment, including a 20 percent increase in mastectomies.
These findings were immediately unpopular. According to Richard Horton, editor of the Lancet, ‘when the Nordic investigators submitted their systematic review to the editors of the Breast Cancer Group, they found that their conclusions were unwelcome’. However, while statistical disputes raged over the interpretation of different studies, the results were published by the Cochrane Library (3).
The latest studies confirm long-standing doubts about the efficacy of mammography (4). Ever since the widespread introduction of this screening test in the USA in the 1980s and in Britain in the 1990s, several authorities have questioned the value of early diagnosis of tiny tumours, as well as pointing out the high level of ‘false positive’ results, leading to invasive and distressing investigations. Yet a combination of government desire to display a concern for women’s health, the growing influence of voluntary organisations dedicated to raising awareness of cancer, and the resulting public anxiety about cancer, have ensured the continuation of the screening programme.
In the USA in particular, commercial interests have a major stake in screening. As Dr Donald Berry, a cancer biostatistician based in Houston, Texas and a member of the NCI panel, told the New York Times on 24 January 2002, ‘screening brings in patients – and it isn’t just the mammography, but it’s the biopsies, the surgeries and the like’.
Squeezed between its expert panel and pressures from the cancer lobby, the NCI website now faces the prospect of offering contradictory positions on mammography – the old view that it reduces mortality by up to 30 percent and the new line that it does not reduce it at all, while causing further problems.
The controversy over screening for prostate cancer has turned even nastier. In January 2002 campaigners for screening tests set about trying to silence the editors of the Western Journal of Medicine, Michael Wilkes and Gavin Yamey. Their offence was that they argued, in an article in the San Francisco Chronicle that screening may cause more harm than good. This is, as several contributors to the ensuing debate have pointed out, a relatively uncontroversial view in Europe (see extract from The Tyranny of Health, below).
It is widely recognised that the blood test for the prostate specific antigen yields a high proportion of both false negative and false positive results, often leading on to invasive investigations with a high rate of complications. Furthermore, and most critically, there is no evidence that early intervention improves outcome.
Whatever the inadequacies of screening tests for prostate cancer, it has become the focus of lobbying and mobilisation for the men’s health movement, a growing force on both sides of the Atlantic. Commercial interests are again a significant influence. Activists from this movement have launched an email campaign accusing Wilkes and Yamey of ‘geriatricide’ and demanding that they be fired and that their work be suppressed.
The screening controversies reveal the invidious consequences of the politicisation of health issues – by government, by doctors, by commercial interests and by pressure groups. The case against screening for cervical cancer with smear tests is, if anything, stronger than that against mammography (5). A rational approach to the problem of cancer would be to divert the vast resources wasted in these programmes into research – and into providing better care and treatment for people with cancer.
(1) See the website National Cancer Institute
(2) Lancet 2001; 358:1340 linkable text
(3) See the Cochrane Library website
(4) See extract from The Tyranny of Health, below
(5) See extract from The Tyranny of Health, below
What’s wrong with screening?
This is an edited extract from Dr Michael Fitzpatrick’s The Tyranny of Health: doctors and the regulation of lifestyle, published by Routledge in 2001.
The feminist campaign of 20 years ago to seize control over women’s health from the medical profession, has given way to state-sponsored, doctor-led systems of vaginal examination and cervical surveillance today. It is doubly ironic that within the same period, male resistance to medical regulation was replaced by the demand, under the banner of ‘men’s health’, for invasive screening tests analogous to cervical smears.
Whereas the early women’s movement rejected medical inspection of the cervix as an act of symbolic domination, the modern men’s health movement invited rectal penetration as a symbol of its subordination to medical authority.
Given the failure of modern medicine to discover the causes of most forms of cancer, which might lead to a strategy of prevention, an obvious alternative was to devise some means of early detection, leading to prompt treatment, and, hopefully, a better prognosis. For cancer of the cervix (neck of the womb) and the breast, screening tests have become popular over the past two decades, especially over the 1990s, when they were made available through national programmes and taken up by a large majority of the eligible population.
These screening programmes claim substantial benefits in terms of reduced mortality – though in both cases these claims have been questioned by experts in the field. In recent years both programmes, but particularly the cervical smear scheme, have been subject to exposures of poor standards in some areas, leading to scares, scandals and litigation. As each screening test has specific features, let’s take them in turn.
‘By offering screening to 250,000, we have helped a few, harmed thousands, disappointed many, used £1.5million each year, and kept a few lawyers at work.’ (1)
Cancer of the cervix is a fairly rare form of cancer in Britain, accounting for less than half a percent of cancer deaths and around four percent cancer cases in women. Deaths from cervical cancer slowly declined over the past 50 years, from 2500 in 1950 to 1150 in 1997 (2). In 15 years as a GP I have had two patients who have died from cervical cancer, which is probably over the career average; typically, neither had ever had a smear test.
The smear test was introduced in Britain in the mid-1960s, following a famous study in British Columbia, Canada. This study appeared to show a dramatic reduction in cancer following the introduction of smear tests, which allowed the early detection and treatment of ‘pre-cancerous’ areas. There was considerable controversy at the time over whether the decline in death rate could be attributed to smear tests (it had declined elsewhere without such tests) and over whether cells labelled as ‘pre-cancerous’ might return to normal without treatment, rather than progressing to invasive malignancy.
In addition, smear tests failed to meet two of the standard criteria for screening programmes laid down by the World Health Organisation: cervical cancer is uncommon and its natural course is not well understood (3). Though many experts were sceptical, a powerful lobby of cancer specialists, prevailed upon the Labour government to introduce a cervical smear service in 1966 (4).
Two years later, the eminent epidemiologist Archie Cochrane caused a furore when he claimed that there was no evidence that smears would reduce the death rate from cervical cancer. He particularly objected to the use of a screening test for a condition for which there was no effective treatment (an authoritative review in 1999 conceded that there had been ‘no significant improvements in treatment for cervical cancer over the past 20 years’) (5). Reflecting some years later on the ‘uproar, abuse and isolation’ he experienced as a result of his questioning of the cervical smear programme, Cochrane commented that, because of the introduction of this programme without proper evaluation, ‘we would never know whether smears were effective or not’ (6).
In 1988, following criticisms of the haphazard character of the cervical smear system, a National Coordinating Network for the NHS Cervical Screening Programme was established. In 1990, the new contract imposed on GPs by the government offered substantial incentives, now worth around £65million a year, tied to smear rate performance targets. As a result of these measures, coverage of the target age group rose from 42 percent in 1988 to 85 percent in 1994, a level subsequently maintained (7). The claim by cancer specialist Brian M Quinn that in women under 55 ‘screening may have prevented 800 deaths in 1997’ was contested by surgeons Jayant S Vaidya and Michael Baum, who noted that the data presented could equally well support the conclusion that screening caused a similar increase in mortality (8).
The contrast between the high level of public faith in the cervical smear programme and the private recognition among medical authorities of its unsatisfactory character is remarkable. In their reply to Vaidya and Baum, Quinn and his colleagues admitted that they remained ‘deeply concerned about the well known problems with cervical screening’, which they listed:
‘cervical cancer is a comparatively rare disease and its natural course is not well understood; the smear test has both low sensitivity and low specificity; many tests are technically unsatisfactory and the proportion of such tests varies across the country; the mix of three-year and five-year screening intervals is inequitable; too many smear tests are opportunistic; and the programme costs four times as much as breast screening.’
The low sensitivity of the smear test means that every year women are diagnosed as having cervical cancer which had been missed on previous smears. The fact that some such cases have resulted in litigation has led to calls for doctors to make clear that smears may miss between five and 15 percent of abnormalities and to ensure that patients are giving properly informed consent to this procedure (10).
The low specificity of the smear test means that it yields a relatively high proportion of false positive results: that is, it suggests that a woman has malignant or pre-malignant cells when more invasive procedures (involving the removal of a wider area of tissue in a ‘loop’ or ‘cone’ biopsy) confirm that this is not the case. In day to day practice, this is by far the biggest problem arising from smear tests, causing enormous anxiety and distress, often continuing for weeks or months pending delays in further investigations.
Bristol public health consultant Angela Raffle noted the tendency of staff, in response to publicity over missed cases, to over-diagnose minor abnormalities (11). While patients suffered needless anxiety, staff lived in fear of failing to identify potentially malignant cases. As a result, ‘much of our effort in Bristol is devoted to limiting the harm done to healthy women and to protecting our staff from litigation as cases of serious disease continue to occur’. As Raffle recognised, many healthy women are left with worries about cancer and difficulties in obtaining life insurance. Those who receive treatment may experience considerable discomfort, bleeding and sexual problems – as well as long-term anxieties about fertility.
Meanwhile women in that 10 to 15 percent of the female population which has never had a smear, who are likely to be (like my two patients), older, poorer and from ethnic minorities, will ensure that the mortality figures remain fairly steady. Health promotion propaganda which characterises cervical cancer as a sexually transmitted disease (on the dubious grounds of an association with the wart virus) has undoubtedly deterred many women from having smear tests.
The annual cost of the cervical cancer screening programme is £132million (12). This is about four times the cost of the breast screening programme – though the death rate from breast cancer is around 10 times greater.
Breast cancer is not only much more common than cancer of the cervix, but the number of cases has gradually increased over the past 20 years. After rising slowly through the 1970s and 1980s, the death rate declined in the 1990s. There are currently around 30,000 cases a year, accounting for one third of cancer in women; breast cancer kills around 11,000 women every year, causing around one fifth of cancer deaths. In our surgery we see several new cases of breast cancer every year and one or two deaths. We see many more women who turn out not to have breast cancer but are understandably terrified by the appearance of a lump or other breast symptoms.
Trials of mammography – X-ray examination of the breast – for early detection of malignancy were carried out in the USA in the 1960s. Early results showed a resulting reduction in mortality among women over the age of 50, but no benefit in younger women (13). More extensive research in the 1970s confirmed the earlier results and mammography became established as a screening test for breast cancer. In Britain a national screening programme became operational in 1988; now women between the ages of 50 and 64 are invited for free mammography every three years. The combination of mammography with ultrasound and the microscopic study of cells extracted from a suspicious lump through ‘fine needle aspiration’ has greatly improved the diagnostic sensitivity of this process in the 1990s.
In 1995, the organisers published the results of the first five years of the mammography programme and claimed some credit for an 11 percent drop in mortality from breast cancer in the target age group (14). In response, Professor Michael Baum, who had helped to set up the screening service, pointed out that though the mammography programme could not be expected to have an effect on mortality before 1997, the decline in the death rate began in 1985. Suggesting that a more likely explanation was the introduction of the drug Tamoxifen for the treatment of breast cancer, he argued that ‘to claim that any part of this 11 percent fall is attributable to the screening programme is intellectually dishonest’ (15). In protest, he resigned from the Department of Health’s breast cancer screening advisory group.
Baum also pointed to the high level of false positive results generated by mammography, causing anxiety and leading to further investigations, either aspiration cytology or excision biopsy. He concluded that mammography was ‘not worth doing’ because it saved too few lives at too high a cost, while causing needless anxiety among thousands of healthy women by incorrectly suggesting that they have the disease (16). He suggested that the money spent on screening might be better spent on research and specialist treatment for women diagnosed with breast cancer.
But breast cancer screening had acquired high political prestige; only three months earlier a parliamentary select committee had commended the mammography programme as a model of excellence in preventive health care and had called for it to be extended to cover women up to the age of 69. Baum’s proposals were ignored.
The controversy over breast screening flared up again five years later. A study by a team from Denmark reviewed major trials of mammography in Sweden, Scotland, Canada and the USA, involving 500,000 women, and concluded that there was ‘no reliable evidence that screening decreases breast cancer mortality’ (17). Prominent representatives of the government screening programme and the leading cancer charities immediately rejected this conclusion and asserted their conviction that mammography saved lives. Delyth Morgan, chief executive of Breakthrough Breast Cancer, insisted that ‘we must not be deterred from continuing our screening programmes until we have seen categorically that they are ineffective’ (18).
This ethical imperative to prove a negative stood in dramatic contrast to the one imposed 20 years earlier by Cochrane in what has become recognised as a classic paper (19). These authors distinguished between ‘everyday medical practice’, in which a patient asks for help and the doctor ‘does the best he can’, and the ‘very different position’ when the doctor ‘initiates screening procedures’. In this situation, the doctor should ‘have conclusive evidence that screening can alter the natural history of disease in a significant proportion of those screened’.
By the turn of the millennium the science and ethics of screening were subordinated in the mammography programme to a combination of the government’s need to maintain a high profile of concern for both health and women’s issues, medical vested interests, the demands of the powerful cancer charities – and women’s anxieties.
Public anxieties about breast cancer were encouraged by the official campaign to raise ‘breast awareness’. October 1996 was designated Breast Cancer Awareness Month and, following the style set by the adoption of a red ribbon by AIDS activists, a pink ribbon became a badge of breast awareness. Breast cancer won celebrity endorsements and became fashionable. One of the main effects of the promotion of breast cancer awareness is that it generates an exaggerated sense of risk. The Cancer Research Campaign promoted the estimate that ‘1 in 12’ women will develop breast cancer, which featured on a nationwide poster campaign. According to an authoritative review, this was ‘correct only for women who have escaped a number of equally serious but more likely threats to life at an earlier age’ (20). The authors concluded that ‘for most women, the lifetime risk of dying of breast cancer is only 1 in 26; the other 25 will die of something else’.
Most of the women who come into the surgery worried about breast lumps are young, that is, under 50 – though the vast majority of deaths from breast cancer are in women over 65. Only one woman in 136 in Britain dies of breast cancer before the age of 50. Though the risk of dying from breast cancer increases with age, it appears to progress more slowly in older women, so that they often live long enough to die from some other cause. One of the ironies of discussing the risks of breast cancer is that, if the woman smokes, she has a greater risk of dying from lung cancer; even if she is a non-smoker, she is far more likely to die of heart disease.
Public awareness of breast cancer has intensified the demand for screening tests which promise early diagnosis. The most basic is the technique of breast self-examination, which is generally recognised to be much more effective in generating anxiety than it is at detecting tumours (21). Women’s magazines and health-promotion leaflets are still offering detailed diagrams and earnest advice about how to detect lumps – resulting in a steady flow of frightened women, some scarcely out of their teens, who are more likely to win the national lottery than to have breast cancer.
Another consequence of greater breast cancer awareness is the demand to extend mammography to women in their forties. According to one commentator, this has provoked a debate in the USA ‘out of proportion to its potential impact on public health’ (22). Despite the fact that numerous trials have failed to confirm the efficacy of this technique in younger women – and despite concerns that the radiation exposure involved might do greater harm – political pressures resulting from disease awareness campaigns have resulted in younger women having mammograms.
Women who have had breast cancer are perhaps the greatest casualties of breast awareness. It is not only that they are reminded of their disease every time they turn on the television or open a newspaper or magazine – and every time they see a pink ribbon on the bus or train. The popular discussion of the role of lifestyle factors in predisposing women to breast cancer compounds women’s worries about their future with guilty reflections on their past behaviour. This is encouraged by epidemiological surveys which report the loosest of associations as causal influences.
Thus the risk of breast cancer appears to be increased in women who have no children or who have them after the age of 30; in women who have taken the oral contraceptive pill or hormone replacement therapy; in women who drink alcohol and have a high fat diet. The relatively strong influence of family history on chances of getting breast cancer provides further scope for recriminations about genetic destiny and fatalistic ruminations about dying a premature and disfiguring death.
During breast awareness week, a patient of mine who has survived mastectomy, radiotherapy and chemotherapy and now has a good prognosis, came in to ask me what she had done to deserve breast cancer. I don’t know who benefits from breast awareness, but I know many of its victims.
Carrying on screening
Despite all the problems of the cervical cytology and mammography programmes, the demand for more screening tests for other cancers continues to rise. One of the main sources of such demands in the late 1990s was the burgeoning men’s health movement, associated with the wave of men’s magazines, one of the publishing successes of the decade.
Though it lacked the early radical impulse of feminism, the men’s health movement adopted the later preoccupation of some feminists with health as their model. Far from challenging medical authority, men’s health promoters urged men to submit themselves to it on a greater scale than ever before. In choosing campaigning issues, they proceeded by analogy with the feminists: they had cervical smears – we demand prostate examinations; they can do breast self-examination – we can feel our testicles. ‘Perhaps’, mused one urologist, ‘there was a subliminal male desire to have a disease all of our own, even if it had to be a cancer’ (23).
Though prostate cancer is relatively common in older men (95 percent of 15,000 cases a year occur in men over 60), testicular cancer is a rare disease of younger men (causing around 100 deaths a year). Though treatment is often effective for both cancers, screening tests for early detection are generally considered unreliable. To detect prostate cancer it is possible to have a regular digital rectal examination, a blood test for the Prostate Specific Antigen (PSA), and a local ultrasound scan, but the predictive value of all these tests is low. Urologist Peter Whelan suggested that ‘Promotes Stress and Anxiety’ was an accurate description of the effect of the blood test. Given the rarity of testicular tumours, a high rate of false positive results is the inevitable outcome of any promotion of self-examination (24).
It is, however, striking that, long after medical authorities have accepted the ineffectiveness of screening tests like the PSA, or self-examination of breasts and testicles, pressure groups and popular magazines continue to promote them. The extent of popular approval of these techniques, which is grossly disproportionate to any value they might have in reducing the impact of cancer, is a potent indicator of the pathological preoccupation with health that now prevails in society.
It is ironic that young women are often advised to examine their breasts every month – an arbitrarily selected frequency that happens to coincide with the menstrual cycle – though the large majority of women with breast cancer are post-menopausal. Similarly, young men now turn up at the surgery after reading about prostate cancer in their men’s magazines and request screening for a condition that only rarely appears before retirement age. The parallel between screening tests for cervical and prostatic cancer is symbolic. Just as the smear test exposes women not merely to the medical gaze but to vaginal penetration, so the palpation of the prostate involves digital penetration of the male rectum. The slippery finger may be less impressive than the metal speculum, but it is no less significant as an instrument of symbolic domination.
For its advocates, screening has become an article of faith. Rejecting the evidence of the ineffectiveness of mammography, Delyth Morgan of Breakthrough Breast Cancer insisted that ‘what we should be debating is how best to screen women’ (25).This response provides striking confirmation of the observation made 15 years earlier in another critique of screening: ‘In “keeping the faith”, screening advocates may find themselves forced to accept or reject evidence not so much on the basis of its scientific merit as on the extent to which it supports or rejects the stand that screening is good’ (26). The danger of this approach is not only that it leads to the continuation of costly and ineffective programmes. It also means that the harms of screening are passed over in silence: to mention them could discourage people from taking up the offer of testing.
Indeed this was the first concern of the cancer charities in response to reports of the Danish study of mammography quoted above; public reassurances about the quality of the national cervical screening programme accompany every exposure of poor standards. Yet the harms resulting from screening are substantial: for every woman who benefits, tens of thousands undergo testing and hundreds receive unnecessary treatment. In presenting screening as an unequivocal benefit to women, doctors become advocates of state policy rather than of their patients’ interests.
Dr Michael Fitzpatrick is the author of MMR and Autism, Routledge, 2004 (buy this book from Amazon (UK) or Amazon (USA)); and The Tyranny of Health: Doctors and the Regulation of Lifestyle, Routledge, 2000 (buy this book from Amazon UK or Amazon USA). He is also a contributor to Alternative Medicine: Should We Swallow It? Hodder Murray, 2002 (buy this book from Amazon (UK) or Amazon (USA)).
(1) ‘Informed participation is essential’, AE Raffle, BMJ 314: 1762-3, 1997
(2) ‘Effect of screening on incidence of and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics’, M Quinn et al, BMJ, 318: 904-8, 1999
(3) Principles and Practice of Screening for Disease, JMG Wilson and G Junger, WHO, 1968
(4) The Diseases of Civilisation, B Inglis, Hodder and Stoughton, 1981, p66-69
(5) ‘Effect of screening on incidence of and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics’, M Quinn et al, BMJ, 318: 904-8, 1999
(6) ‘Some reflections’, A Cochrane, in A Question of Quality: Roads to Assurance in Medical Care, G McLachlan, OUP, 1976
(7) ‘Effect of screening on incidence of and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics’, M Quinn et al, BMJ, 318: 904-8, 1999
(8) ‘Screening and mortality from cervical cancer’, JS Vaidya and M Baum, BMJ, 319:642, 1999
(9) ‘Effect of screening on incidence of and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics’, M Quinn et al, BMJ, 318: 904-8, 1999
(10) ‘Smear tests were not on trial but should have been’, CM Anderson 1999, BMJ, 318: 1007; ‘Women must be given fully informed information about cervical screening’, J Nottingham, BMJ, 318: 1555-6, 1999
(11) ‘Detection rates for abnormal cervical smears: what are we screening for?’, Lancet 345: 1469-73, Raffle et al, 1995
(12) ‘Effect of screening on incidence of and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics’, M Quinn et al, BMJ, 318: 904-8, 1999
(13) ‘Mammography and the politics of randomised controlled trials’, J Wells, BMJ, 317: 1224-30, 1998
(14) ‘Sudden fall in breast cancer rates in England and Wales’, V Beral et al, Lancet, 345: 1642-3, 1995
(15) ‘Screening for breast cancer, time to think – and stop’, M Baum, Lancet, 346:426-7, 1995
(16) ‘Pioneer resigns over “useless” breast cancer tests’, L Rogers, Sunday Times, 3 September 1995
(17) ‘Is screening for breast cancer with mammography justifiable?’, PC Gotzsche and O Olsen, Lancet, 355: 129-34, 2000
(18) Guardian, 7 January 2000
(19) ‘Validation of screening procedures’, AL Cochrane and WW Holland, BMJ, 27:3-8, 1971
(20) ‘Putting the risk of breast cancer in perspective’, JP Bunker et al, BMJ, 317: 1307-09, 1998
(21) ‘Cancer prevention in primary care: screening and self-examination for breast cancer’, J Austoker, BMJ, 309: 168-74, 1994
(22) ‘Mammography and the politics of randomised controlled trials’, J Wells, BMJ, 317: 1224-30, 1998
(23) ‘Are we promoting stress and anxiety?’, P Whelan, BMJ, 315: 1549-50, 1997
(24) ‘Cancer prevention in primary care: screening for ovarian, prostatic and testicular cancer’, J Austoker, BMJ, 309: 315-20, 1994
(25) Guardian, 7 January 2000
(26) ‘Controversy in the detection of disease’, DL Sackett and WW Holland, Lancet, ii: 357-9, 1975