The Rise and Fall of Modern Medicine (7 page)

The ramifications of this paradigm shift were to be of great importance, but surprisingly that is not the main reason for its inclusion in the pantheon of ‘definitive moments'. Rather it is the manner in which it was brought about. Prior to 1950, the cornerstone of reliable knowledge in medicine was the cumulative wisdom acquired through everyday practice. The notion that the validity, or otherwise, of specific treatments might be objectively tested was hardly ever raised. But the demonstration of the curability of tuberculosis and the role of smoking in lung cancer changed all this, for both relied on statistical methods of proof that soon permeated every aspect of medicine to become the main – indeed the sole – arbiter of ‘scientific truth'. This was almost entirely due to the influence of one man, Austin Bradford Hill, Professor of Medical Statistics at the London School of Hygiene and Tropical Medicine. Bradford Hill was not medically qualified and had no formal training in statistical methods, at which, on his own admission, he was ‘not very proficient', viewing them rather as being ‘common sense applied to figures'. Nonetheless he was to guide statistics to a dominant position within medicine whose subsequent indiscriminate application would eventually exert a most baleful influence.

This intellectual ascendancy of statistics is essentially the story of Bradford Hill's life. He was born in 1897 into a distinguished Victorian family, at least one of whose members in each of the preceding four generations had featured in the
Dictionary of National Biography
, including his father, Sir Leonard Hill, Professor of Physiology at the London Hospital, who,
inter alia
, developed a machine to measure the blood pressure and in a series of self-experiments conducted by himself and his junior lecturer, Dr Major Greenwood, showed that ‘the bends' in divers, caused by the formation of bubbles of nitrogen in the blood, could be prevented by slow uniform decompression.
²
Powerfully influenced by the stimulating atmosphere of his home life, Bradford Hill decided to follow his father into medicine but, when the time came to enter medical school, Britain was at war with Germany so instead he joined the Royal Naval Air Service as a pilot. In January 1917 he was posted to the Aegean and joined a party of a dozen officers at Charing Cross Station to travel by train to the toe of Italy. ‘It was on this exhausting, overcrowded and unhygienic journey, I would guess, I picked up the tubercle bacillus,' he subsequently recalled. Based on the tiny island of Tenedos just off the Turkish coast, he had a quiet time, other than the occasional flying accident, in the last of which his engine failed at 11,000 feet, leaving him no alternative other than to glide down to the narrow airstrip. ‘I misjudged by about 10 yards and landed on the edge of a muddy lake. The plane stood on its nose and broke its propeller.' Within five months he had become seriously ill with a cough and a fever, the tubercle bacilli that had been multiplying in his lungs were identified in his sputum and he was ‘invalided home to die'. To his own astonishment, and that of his doctors, he responded to the only two treatments for tuberculosis available at that time – bed-rest and an artificial pneumothorax (the introduction of air into the pleural cavity to collapse down the lung and thus slow the spread of the infection). In 1919 he was discharged from hospital with a 100 per cent disability pension (only given to those whose disability is deemed so severe as to preclude them from any further gainful employment), which he continued to draw for the next seventy-four years till his death at the age of ninety-three.
³

Though Bradford Hill had survived tuberculosis, a medical career was now out of the question. At the suggestion of his father's erstwhile physiology lecturer, Dr Major Greenwood, he opted for a correspondence course in economics at London
University, which he passed with second-class honours. In 1928 Major Greenwood was appointed the first Professor of Vital Statistics at the recently opened London School of Hygiene and Tropical Medicine, and in 1931 he appointed Bradford Hill as Reader in Epidemiology. Thus began their long collaboration.

These details of Bradford Hill's early life illuminate what was to follow. His frustration at being unable to fulfil his childhood ambition of following his father into medicine only heightened his fascination with every aspect of the medical sciences. It may have been the fascination of an outsider, but this was only to his advantage. From this Olympian vantage point he was able to take a detached and critical view of medical developments. He was particularly lucky that his mentor, Greenwood, was that rarity in medical circles, a full-blooded intellectual, whose perception of the contribution statistics could make to medicine was driven by his strong historical sense of past achievements, particularly in the field of public health, reinforced by his contact with the mathematical polymath Karl Pearson (of whom more later), who in his turn had been a protégé of one of the greatest of all Victorian intellectuals, Francis Galton. It is to these roots of medical statistics that we now turn before taking up again the thread of Bradford Hill's career.

For many, statistics are numbers to which complex mathematical formulae can be applied to produce conclusions of dubious veracity and from which all wit and human life is rigorously excluded. Certainly, any single statistic by itself is a dreary thing, but when they are linked together over months and years then patterns begin to emerge and it is possible to see things that previously were hidden. An unarguable event such as death lends itself particularly well to the statistical method, and when the numbers in any town or region are recorded over a brief period
it is possible to appreciate that in the aggregate they represent the distinct biological phenomenon of an epidemic. This is the simplest form of epidemiology – literally the study of epidemics – which, nonetheless, has the power to change the world for the better.

This beneficent capacity of statistics was seen most strikingly in the great movement for sanitary reform in the mid-nineteenth century, when William Farr – ‘a very great Englishman' in Greenwood's words – was compiler of abstracts in the General Register Office. In his thirty-fifth annual report Farr drew attention to the yawning differential in childhood mortality rates between the rich and poor and asked: ‘What are the causes? Do they admit of removal? If they do admit of removal, is the destruction of life to be allowed to go on indefinitely?'
⁴

The crucial point to inspire twentieth-century epidemiologists was that statistical enquiry, by determining the underlying causes of ill health such as poor sanitation, provided the means for the prevention of disease on a massive scale, so ‘statisticians' potentially had an infinitely greater effect in improving the health of the nation than white-coated doctors with their airs and graces.

There was, as already mentioned, a second and very different use of statistics. The mathematical techniques invented by Karl Pearson of University College to interpret biological variations in height, blood pressure or any other physiological characteristic made it possible to infer general rules about groups of people rather than individuals, and were relevant to experiments to test the efficacy of treatments from which it was possible to deduce whether the results were ‘significant'. In an illustrative example famous for its inconsequentiality, Ronald Fisher – a pupil of Pearson's – poured a cup of tea

In his classic book
The Design of Experiments
, published in 1935, Fisher used this example ‘to state the terms of the experiment minutely and distinctly; predicted all possible results, ascertaining by sensible reasoning, what probability should be assigned to each possible result under the assumption that the woman was guessing'.
⁵

Thus Greenwood's main intellectual legacy, which he was to pass on to Bradford Hill, was essentially two-fold: the historical contribution of statistical methods to elucidating the cause of substantial public health problems; and the importance of conducting properly designed experiments to test whether a new treatment was effective.

In 1945 Greenwood retired and Bradford Hill was duly appointed as his successor. The paradigm shift in which he was to play so important a role was just five years away. Its two components – the trial of streptomycin and PAS in the treatment of tuberculosis and the elucidation of the causative role of smoking in lung cancer – are here described separately, though they were occurring at the same time. Accordingly the rise of the power and influence of statistics in medicine began to appear inevitable.

Bradford Hill had personal reasons for being interested in the treatment of tuberculosis, having himself only just escaped from being yet another mortality statistic for the disease at the cost of two years' convalescence and an artificial pneumothorax. In 1946 he joined the Tuberculosis Trials Committee, which had been set up to evaluate a new drug, streptomycin, that two years earlier in the United States had been shown to be capable of killing the tubercle bacillus. It had been unnecessary to formally test whether or not penicillin worked in humans because the results were immediate and dramatic. But the efficacy of streptomycin in tuberculosis was not quite so straightforward because the tubercle bacillus, in its waxy shell, is very resilient and the damage it causes the lungs and other organs is more chronic. Accordingly, streptomycin had to be given for several months before there was any obvious improvement. Nonetheless there was no obvious reason at the time why doctors should not, as they had done in the United States, give streptomycin to patients and see what happened. If it worked, that was fine; if not, then nothing was lost. Bradford Hill was, however, determined that streptomycin should first be put to the test in a properly conducted trial, comparing the outcome with those not given the drug. His view prevailed, but only because of the fortuitous circumstances that streptomycin was extremely difficult to acquire in Britain at that time and so for the foreseeable future many would be unable to benefit from it. Bradford Hill resolved to make a virtue out of this necessity, as he subsequently recalled:

Bradford Hill's position was to be more than vindicated, though not by showing whether streptomycin worked (which was in a sense predictable), but for showing that after a while, and for important reasons, it stopped working. Further, the fact that the treatment of tuberculosis was (almost) the first treatment to be formally tested in this way is highly significant. Tuberculosis was after all much the commonest lethal infectious disease in the West. The fact that this new drug was being tested in the context of a formal experiment designed by Bradford Hill would so enhance his authority as to make his position virtually unassailable. Once again it is necessary to interrupt the narrative, first to briefly examine the origins of this new treatment for tuberculosis, and second to look at the intellectual provenance of the clinical trial, before rejoining Bradford Hill for the
dénouement
of this part of his career.

It is almost impossible nowadays to imagine what it must have been like to live in a world before tuberculosis was curable. Everyone was vulnerable: a cough and a temperature would precipitate a visit to the doctor, who would confirm the presence of tubercle bacilli in the sputum, and an X-ray would reveal how widespread was the infection in the lungs. This
signalled a profound change in patients' lives, which became, as it were, suspended for the year or two spent in the sanatorium from which they might emerge, like Bradford Hill, as one of the lucky ones – or not.
⁷
In Thomas Mann's
The Magic Mountain
the young hero, when visiting his cousin in a tuberculosis sanatorium in the Alps, becomes

Tuberculosis could kill quickly in a matter of weeks, or, like a lingering death sentence, slowly over many years. It did not only affect the lungs but also the brain, to cause tuberculous meningitis, from which no one ever recovered, as well as the joints and bones, to cause a relentless, progressive crippling arthritis. The bacillus that caused tuberculosis was identified in 1885 by an obscure country doctor, Robert Koch, since when treatment had remained elusive. There was a vaccine of moderate efficacy, BCG, but, besides convalescence and an artificial pneumothorax, there were no other breakthroughs until the discovery of the two drugs, streptomycin in 1944 and para-amino salicylic acid or PAS in 1946, whose origins, though utterly different, lay in two of the greatest insights of modern medicine.
⁸
The first insight, already described in the story of penicillin, was antibiotics: the notion that non-toxic chemical compounds produced by bacteria and fungi could destroy the bacteria that
caused disease in humans. The second, which led to the discovery of PAS, was ‘competitive inhibition': a bacterium will die if its metabolism is ‘jammed' by a compound that is necessary for its growth but has been chemically slightly modified.