Why can’t doctors be more scientific?

There is a tradition of underestimating the nastiness of measles. It has never had the bad publicity it deserves, or been represented in the canon of ‘plague literature’: it has never featured in a Decameron or a Magic Mountain or a Death in Venice or attracted a Defoe or a Camus. Its victims, mostly children, have gone to their early graves anonymously, so there have been no stories to tell. As for the researches and discoveries of the scientists who have worked on it, they have failed to stimulate celebratory writings for the general public, despite their importance, their originality and their elegance. Rabies has Louis Pasteur, smallpox has Edward Jenner. Who has heard of Peter Panum? Even the book about the institution where he spent most of his career, S.E. Stybe’s Copenhagen University: Five Hundred Years of Science and Scholarship (1979), while acknowledging his importance as a founder of modern physiology, fails to mention his study of the 1846 measles epidemic in the Faroe Islands. Not only did this investigation set the standard for all subsequent epidemiological work, it settled once and for all that measles is a specific contagion, as well as demonstrating how easily it spreads, accurately establishing its incubation period, and showing the lifelong immunity conferred by infection. Panum found that the 98 survivors of the previous Faroes epidemic were still immune, even though they had been infected in 1781, and that of the 5000 inhabitants exposed to infection, 99.5 per cent caught the disease.

John Enders and his young associate T.C. Peebles were the first to grow measles virus in the test tube in 1954, using the tissue culture techniques developed by Enders and his colleagues in the late 1940s. Much good that did him at Harvard. Even though he was awarded the Nobel Prize that year for using his tissue culture method to grow poliovirus, he had to wait another two years for promotion, at the age of 59, from the associate professorship that he had held for the previous 14. Enders shares Panum’s fate at the hands of the historians. Making Harvard Modern by Morton Keller and Phillis Keller (2001) passes him by without mention.

The popular perception that naturally occurring measles is trivial is not new. One of the first to offer an explanation for the paradox of a killer being regarded only as a nuisance was the medical officer of health for Aberdeen, William Simpson. Writing in 1882 about smallpox, diphtheria, cholera, measles, whooping cough and scarlet fever, he said:

It comes out, as a peculiar fact, that the most dreaded diseases are the least fatal, and the least dreaded diseases are the most fatal . . . measles, whooping cough and scarlet fever are the most serious, although it is usually considered they do little harm . . . their very frequency makes them less dreaded . . . the disease that comes unexpectedly, and passes over quickly, is looked upon with greater feelings of terror than the disease which may be more fatal, but more common.

The words of another Scottish MOH, Alexander MacGregor of Glasgow, illustrate why this perception was wrong for measles: ‘In 1907-08 it gave notice of its presence by appearing on the outskirts of the city in the spring, hovering there till the autumn and then descending on the city like an invading army, leaving 22,000 casualties and 1100 fatalities in the space of nine months. It reached its maximum in January 1908 with 70 deaths per week.’

Ninety years later, its impact on malnourished children in sub-Saharan Africa was the same; it killed a million there every year. But there had been an enormous change in Britain. Halfway through the First World War the number of measles deaths started to fall. It went on dropping rapidly for the next half-century, until the mortality rate had declined more than a thousandfold. Measles had ceased to cause its previously common lethal complication, pneumonia. It had not been trivialised, however. The improvements in nutrition and housing responsible for drawing its teeth had not affected its ability to cause the other, rarer complications that historically had been masked by its lethality. It now became clear that about one in a thousand children with the disease developed encephalitis, which killed 10 to 15 per cent of those affected and left a quarter of them with permanent brain damage. Encephalitis was not caused directly by the virus but by its inducing the body to mount an immunological attack on the brain. It was also evident that the virus could damage ” the brain directly. About one in a million measles sufferers went on to develop and die from subacute sclerosing panencephalitis (SSPE). Years after an apparently normal attack of measles, sometimes as long as a decade later, its onset was signalled by changes in personality and intellect, which became more and more serious. Fits then occurred and the brain became progressively dysfunctional. The patient went into a coma and died after an illness lasting many months. Abnormal forms of the virus occurred in the brain; they had mutated during the long period leading to and including the illness and could no longer be grown in Enders-type tissue culture.

Understanding what happens when viruses invade the brain is usually fairly straightforward. Many hitch lifts there in blood cells or travel up nerves, as rabies does. Once there, the viruses kill brain cells by growing in them; the illness does not last long and the virus is easy to detect. The relationship between measles and the brain is different. It is more subtle, and we are more ignorant about it. We do not know why those who get encephalitis or SSPE are singled out, for example. These subtleties and uncertainties have encouraged speculation. Whiffs of scientific evidence, such as finding that some people with multiple sclerosis make a strong immune response to measles, that cells from patients with Paget’s disease (a disorder in which bones soften and swell) contain structures resembling the component parts of measles virus, and that samples from the bowels of patients with the intestinal condition Crohn’s disease have been found by some (but not all) scientists to contain measles virus nucleic acid, remain just that – whiffs. There has been no confirmation of links between these diseases and measles.

On 28 February 1998 the Lancet published a paper in its ‘early report’ section entitled ‘Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children’. It had 13 authors, all from the Royal Free Hospital and School of Medicine; the first in the list was a surgeon, Andrew Wakefield. The paper described 11 boys and one girl (mean age six) who had been referred to a paediatric gastroenterology unit with a history of diarrhoea, abdominal pain and a loss of acquired skills, including language. Most of the children were shown to have bowel abnormalities by direct visual examination (ileocolonoscopy) or by microscopy of tissue samples. The paper’s conclusion that ‘we have identified a chronic enterocolitis in children that may be related to neuropsychiatric dysfunction’ was descriptive and uncontroversial. Not so the last two sentences of the paper: ‘In most cases, onset of symptoms was after measles, mumps and rubella immunisation. Further investigations are needed to examine this syndrome and its possible relation to this vaccine.’ Publication in the Lancet was not the only publicity that the possible autism/MMR link received. At the press conference associated with publication of the paper, Wakefield advised that measles vaccine should be given separately from the other components of the MMR vaccine.

Predictions about the impact of the paper were made in a leader in the same issue of the journal by Robert Chen and Frank DeStefano from the US Vaccine Safety and Development National Immunisation Program at the Centers for Disease Control:

Vaccine-safety concerns such as that reported by Wakefield and colleagues may snowball into societal tragedies when the media and the public confuse association with causality and shun immunisation. This painful history was shared by the UK (among others) over pertussis (whooping cough) in the 1970s . . . and it is likely to be repeated all too easily over MMR. This would be tragic because passion would then conquer reason and the facts again in the UK.

These things came to pass. MMR vaccine uptake levels fell, particularly in London. Anti-MMR pressure groups received a big boost. Newspapers campaigned for the measles component of MMR to be given separately. There have been outbreaks of measles that otherwise would not have occurred. But the story took a new and unpredicted turn in February this year. In the months before, a Sunday Times journalist had been investigating the background to the Lancet paper. Various allegations of improper conduct were made, including Wakefield’s failure to declare that, at the time he was the lead researcher, he was also conducting virological studies, funded by the Legal Aid Board, looking for evidence to pursue a multi-party legal action on behalf of the parents of allegedly vaccine-damaged children. The Lancet of 6 March responded by publishing five statements; from its editor, Richard Horton, from three authors of the original paper (including Wakefield) and from the vice-dean and campus director of the Royal Free and University College Medical School. There was also a leader by Horton, ‘The Lessons of MMR’, and a ‘Retraction of an Interpretation’ by ten of the 13 authors of the original paper (not including Wakefield).

Retractions of scientific papers are not uncommon. They usually happen because a research team has been unable to replicate or substantiate its findings – too often, sad to say, because the original results were made up. But the formal retraction of an interpretation is almost without precedent. The retractors said:

We wish to make it clear that in this paper no causal link was established between MMR vaccine and autism as the data were insufficient. However, the possibility of such a link was raised and consequent events have had major implications for public health. In view of this, we consider now is the appropriate time that we should together formally retract the interpretation placed upon these findings in the paper.

The surprise about all this is not the retraction, but that the data were even interpreted as demonstrating a possible link, that the interpretation was included in a paper drafted by 13 people, and that it survived the rigorous refereeing and editorial processes of the Lancet. The evidence supporting it was the parents’ perception of a temporal association between MMR vaccine exposure and the onset of symptoms of autism in eight children. In six of them the interval was so short (a week or less) that any direct or indirect link is highly implausible, because of the time it takes for the virus to grow and produce any pathological or immunological effects; a connection could be ruled out absolutely for the three children who developed behavioural symptoms immediately, 24 and 48 hours after immunisation. The original paper discussed none of these things, and didn’t even present any evidence that the measles component of the MMR vaccine had grown in the children. (A simple and readily available test, looking for antibodies, could have been done using the blood samples taken for other tests described in the paper.) It did not refer to any of the big studies that have shown unequivocally that measles vaccine does not damage the brain. In short, what was being retracted wasn’t science at all.

Chen and DeStefano were right to point to the public confusing association with causality. The problem here was that doctors had fallen into the same trap. It was as though the science they were practising had not moved on since Francis Bacon, whose system was criticised so memorably by Macaulay:

We are not inclined to ascribe much practical value to that analysis of the inductive method which Bacon has given in the second book of the Novum Organum . . . A plain man finds his stomach out of order. He never heard Lord Bacon’s name. But he proceeds in the strictest conformity with the rules laid down in the second book . . . ‘I ate minced pies on Monday and Wednesday, and I was kept awake by indigestion all night.’ This is the comparentia ad intellectum instantiarum convenientium. ‘I did not eat any on Tuesday and Friday and I was quite well.’ This is the comparentia instantiarum in proximo quae natura data privantur. ‘I ate very sparingly of them on Sunday, and was very slightly indisposed in the evening. But on Christmas day I almost dined on them, and was so ill that I was in great danger.’ This is the comparentia instantiarum secundum magis et minus. ‘It cannot have been the brandy which I took with them. For I have drunk brandy daily for years without being the worse for it.’ This is the rejectio naturarum. Our invalid then proceeds to what is termed by Bacon the Vindemiatio, and pronounces that minced pies do not agree with him.

New hypotheses – guesses – drive scientific progress. But most never become public. The majority are stillborn in the scientist’s mind. Only a few survive to get discussed in the laboratory. The tiny minority that get published are nearly always those that are supported by enough evidence to make them worth testing and that are persuasive enough to attract the money to pay for further studies. Even at this stage it is unusual for them to come to public attention. Two things are needed for this: an enthusiastic hypothesiser and an interesting subject. In the hard sciences it is therefore very unusual for wrong ideas unsupported by evidence to have much of a life outside the private world of the scientist. But flaky evidence can be overcome by fervour, at least for a time.

Medicine is different. It is the natural home of the untested hypothesis: acupuncture, homeopathy, osteopathy and chiropractice flourish. And it is a happy hunting ground for the enthusiast; health stories are hot news. So a heavy responsibility falls on those who publicly promulgate hypotheses whose consequences may cause alarm. This is particularly true concerning vaccines. As Roy Anderson and Robert May said in 1991, in their book on the mathematical modelling of the spread of infections, Infectious Diseases of Humans: ‘Most people have an intuitive appreciation that the best vaccine programme, from an individual’s point of view, is one where almost everyone else is vaccinated while they are not, so that they are indirectly protected without incurring any of the risks or inconvenience associated with direct protection.’ If too many people act in this way, the infection becomes commoner in the population as a whole, and returns as a real and significant threat to the unimmunised. This is a modern version of the ‘Tragedy of the Commons’ described by Garrett Hardin in his influential 1968 essay: 16th-century English peasants had free grazing on commons; their need to supplement food supplies and income was very great; the resulting overgrazing wrecked the commons for everyone.

These general considerations explain how anti-vaccination enthusiasts can cause more harm than good. Whooping cough provides the best example. A nastier disease than measles (it causes brain damage about ten times more often), it was successfully controlled in Europe, North America and Japan by vaccination programmes that started in the 1950s. Campaigning doctors – Justus Ström in Sweden, Gordon Stewart in the UK, Viera Scheibner in Australia and Galina Chervonskaya in the USSR – then said that because the disease had gone away the dangers from vaccination outweighed its benefits. In the mid-1970s vaccination levels fell, and it was virtually abandoned in Japan and Sweden. In 1979, Japan had an epidemic, with 13,000 cases and 41 deaths. The incidence of infection rose in Sweden to a hundred times that in vaccinated Norway. In Russia, diphtheria returned.

More recently, in northern Nigeria, Islamic groups have campaigned against polio vaccine, saying that it is part of a Western plot to cause infertility and HIV in Muslims. The global goal of eradicating polio by 2005 has been put at risk; almost half the world’s cases of polio in 2003 occurred in Nigeria.

What is to be done? In particular, how can doctors be persuaded to comport themselves as scientists? The great Victorian polymath William Whewell – master of Trinity, tidologist, translator of Plato, crystallographer, expert on German Gothic architecture and inventor of the word ‘scientist’ – said, regarding the scientific method: ‘An art of discovery is not possible, we can give no rules for the pursuit of truth which shall be universally and peremptorily applicable.’ So the inculcation of science into medicine cannot be by precept, but has to be by example.

Whewell’s great work of 1840, The Philosophy of the Inductive Sciences, passed medicine by. But things were changing. In 1847, one of the first journals dedicated to medical science was founded, Virchow’s Archiv für pathologische Anatomie und Physiologie, und für die klinische Medizin. Panum published his work on measles in the Faroes in its first volume. More than a century and a half on, journal editors and medical researchers could do a lot worse than to revisit the journal and the papers in it as models of medical science – in principle, they have never been bettered – but German ceased to be the scientific lingua franca a long time ago, so it is unlikely that they will. This is a pity. In modern times the exploitation of gullibility and the expectations of the sick for a cure have been so increased by the successes of science that even the most outrageous quackery markets itself with scientific jargon. Pseudo and junk science flourish with royal endorsement.

The omens for improvement aren’t good. The Lancet’s response to its Wakefield difficulties epitomises the problem. Horton’s leader on ‘The Lessons of MMR’ doesn’t discuss the scientific strength of the link made in the Wakefield paper between autism and MMR. It implies that with hindsight he would not have published it – but the reason would have been Wakefield’s apparent conflict of interest, not doubts about its truth. The only concession was: ‘We can all do better to adjust the volume of our message according to the validity of the information before us.’ So it is all right to give an unsubstantiated speculation masquerading as science the accolade and validation that comes from publishing it in one of the world’s leading medical journals, so long as it is done in a whisper. I despair.