Austin Bradford Hill was born into a family in which he took justifiable pride, for each of the four preceding generations of Hills had had at least one member commemorated m The Dictionary of National Biography. Two were fellows of the Royal Society; his great-great uncle Sir Rowland Hill, who invented the rotary press and reformed the postal service - introducing the penny post in 1840 - and his father, Sir Leonard Hill, professor of physiology in the University of London, who contributed to the understanding of the cerebral circulation and the hazards associated with the rapid decompression of divers.
From childhood, Austin - or Tony, as he came to be universally called - had wanted to be a doctor but the outbreak of the first world war intervened and in 1916 he volunteered for training as a pilot in the RNVR. He was soon posted to the Greek islands in support of the attack on the Dardanelles but by November 1917 he had developed pulmonary tuberculosis and was sent home to die. The progress of the disease changed after an artificial pneumothorax and by 1919 he was sufficiently recovered to think again about his future. Medicine was out of the question and he opted to study economics as an external student of London University, something that he could do with the aid of a correspondence college while convalescing in bed. Three years later he obtained a BSc (Econ) with second class honours - having attended the university only twice to take examinations.
Tony had no desire to make a career in economics and with the support of Major Greenwood, a long-term friend of the family [Munk's Roll. Vol.IV, p.592], he obtained a grant from the MRC to investigate the reason for the high mortality of young adults in country districts which enabled him, inter alia, to extend his knowledge of statistics by attending part of the course for the BSc in Statistics at UCL. The success of his research enabled him to obtain further appointments with the MRC’s industrial health research board. He remained a member of the board’s scientific staff until 1933 when he was appointed reader in epidemiology and vital statistics at the London School of Hygiene and Tropical Medicine, where Major Greenwood held the chair of medical statistics.
In 1945 he succeeded Greenwood in the chair and in the directorship of the MRC’s statistical research unit. In this dual capacity, he rapidly came to be accepted as the doyen of medical statisticians in the English speaking world - a remarkable achievement for a man who held no degree in either medicine or statistics.
That Bradford Hill occupied this position was not because of the importance of any particular piece of his research but because of the effect his teaching had on the way medical research developed in the two decades after the second world war. Three contributions stand out: his articles on medical statistics which appeared serially in The Lancet in 1937 and were reprinted in book form as his Principles of medical statistics, London, The Lancet Ltd, 1966; his development of epidemiological methods for investigating the causes of non-infectious diseases, and his introduction of randomization for the conduct of clinical trials.
He never thought of himself as a statistician but rather as an arithmetician. The impact of his teaching was so great because he was, as Sir John Simon said of William Farr a hundred years before, ‘. . . a master of the methods by which arithmetic is made argumentative.’ His exposition of statistics emphasized the need to compare like with like, to avoid potential sources of bias and to allow for the play of chance, but it eschewed the use of algebraic formulae and set out the procedures that needed to be adopted in plain English. By so doing he secured the attention of what had been a largely innumerate profession and persuaded the members of the need to present their research results both logically and quantitatively.
The same critical but unencumbered approach, based on logical principles, enabled him to design epidemiological studies and to assess their results in such a way that it was possible to conclude not only that an observed association was real but that it did, or did not, imply a relationship of cause and effect. His logical analysis was first applied effectively to the study of the causes of lung cancer which he began with the assistance of Richard Doll, later Sir Richard, in 1948. The study was initiated at the request of the MRC which had been alerted to the great increase in mortality attributed to the disease over the previous 20 years by Percy Stocks [Munk's Roll, Vol.VI, p.423]. It sought to determine what environmental exposures or behavioural factors more clearly distinguished patients with or without the disease by means of a case-control study of patients admitted to 20 London hospitals. In discussion of the results, Hill set out clearly - for the first time - the various possible explanations of an observed association that had to be taken into account (bias, confounding, chance, and cause and effect) and the characteristics that would enable a conclusion to be reached that the disease was due to cause and effect [Brit. med.J.,1950, 2,739]. In this case that ‘. . . smoking is a factor, and an important factor, in the production of carcinoma of the lung . . .’, subsequently formalized in Bradford Hill’s presidential address to the section of occupational medicine of the Royal Society of Medicine in 1965 and summarized under nine heads in the later editions of his Principles of medical statistics.
Meanwhile, to check the validity of the conclusions drawn from the case-control study, Hill had designed the first large scale prospective study of people with defined exposures by obtaining information about the smoking habits of some 40,000 British doctors and had initiated the cohort study which revealed that smoking also caused myocardial infarction, chronic obstructive lung disease and many other diseases [Brit.med.J.,1964, 1, 1399-1414;1460-1467], which has been continued by his pupils for over 40 years.
But perhaps his most important contribution was the introduction of a method for assessing the effects of treatment by the random allocation to patients of different treatment schedules. Until the second world war the standard method of assessing new treatments had been to treat a consecutive series of patients and to compare the results obtained with those previously obtained by other investigators or the same investigator previously. By the late 1930s Hill had pointed out that selective factors commonly made such series incomparable and had advised, in his lectures on statistics, that it was essential to have a concurrent series of controls and that this could be achieved by allocating different treatments to alternate patients. He did not at that time suggest randomization as he was trying to persuade doctors first to use concurrent controls and the suggestion of randomization might have scared them off. He was also concerned that knowledge of which treatment the next patient would receive might bias a doctor in deciding whether or not the patient was suitable for admission to the trial and in 1946 he persuaded two MRC committees to adopt the technique of randomization; first the MRC’s whooping cough immunization committee (1951) and then a few months later the MRC’s streptomycin in tuberculosis trials committee. The results of the latter’s work were published first, in 1948, and the trial of streptomycin has undeservedly received the accolade of being the first randomized trial. In recommending randomization Hill was acutely aware of the importance of maintaining the standard of medical ethics and he always stressed that in any controlled trial the interests of the patient, as judged by the treating physician, must come first. In the streptomycin trial the first issue that had to be faced was whether it was ethical to withhold the drug from any patient. There was, however, only a small amount available in Britain at the time and it was agreed that the limited supplies should be used first to treat patients with miliary tuberculosis and tuberculous meningitis as these conditions were otherwise invariably fatal. The amount left over was insufficient to treat more than a tiny portion of the many desperately ill patients with other types of tuberculosis and the committee agreed with Hill’s guidance that ‘ ... it would . . . have been unethical not to have seized the opportunity to design a strictly controlled trial which could speedily and effectively reveal the value of the treatment.’
Two other questions were posed: whether the physician in charge could modify the schedule, ana whether the control patients should be given placebos. The doctor, it was agreed, must always ‘ ... do for his patient whatever he really believes to be essential for that patient to return him to health.’ This meant that if any patient seemed likely to benefit by collapse therapy - the only specific treatment available for pulmonary tuberculosis before the introduction of streptomycin - the treatment must be given irrespective of whether it upset the balance of the streptomycin and the control groups, as in fact it proved to do. As for the use of a placebo, this was ruled out in the patient’s interest as it would have required an ineffective intramuscular injection four times a day for four months. The response to treatment could be assessed objectively without it, psychological factors would have little impact on such a serious disease and there was, as Hill said later, ‘ ... no need in the search for precision to throw common sense out of the window.’ That Hill, without a medical qualification, should have delivered a lecture to the College on medical ethics in controlled trials, which was regarded as setting the standard for many years to come, illustrates perhaps better than anything else the respect in which he was held by senior members of the profession and must have been a major factor in bringing about his election as one of the first non-medical honorary Fellows of the College in 1963.
Hill was a quiet, unassuming, private person who sought to lead but not to drive. In committee he expressed his opinion clearly and cogently but never sought to impose it on others. As a result he was listened to with respect and his advice was nearly always accepted. In public he sought to avoid controversy; although distressed by Sir Ronald Fisher’s attacks on his judgement in concluding that smoking was a principal cause of lung cancer he preferred to let the facts speak for themselves rather than to respond in kind. He took immense trouble over his lectures which he embellished with witty asides. He rehearsed and read them so well - without the use of visual aids - that his audiences frequently thought that he was speaking without a text. As a departmental head, his door was always open to any junior who sought his advice. He saw his job as being to provide the conditions under which his junior colleagues could be most productive; no one who worked in his department ever wanted to leave and it was only with the greatest difficulty that anyone could be prized away to take up a professorship elsewhere.
He served on the research and experimental department of the Ministry of Home Security and in the medical directorate RAF during the second world war; he maintained his relationship with the RAF as an honorary civil consultant in medical statistics and as a member of its flying personnel research committee until 1978. He also served as dean of the London School of Hygiene and Tropical Medicine from 1955-57, as a civil consultant to the RN from 1958-77, as a member of the committee on safety of medicines from 1964-75 and as secretary of the Royal Statistical Society from 1940-50, being president from 1950-52.
His many academic honours included fellowships of University College London and of the LSH&TM, honorary fellowships of the Faculties of Community Medicine and Occupational Medicine of the RCP, the Royal Society of Medicine, the American Public Health Association and the Faculty of Medicine of the University of Chile. He was awarded honorary degress from the University of Oxford, 1963, and Edinburgh, 1968, and many medals - including the gold medal of the Royal Statistical Society, 1953, and the Galen medal of the Society of Apothecaries, 1959.
He retired voluntarily in 1961, a year before he needed to, to make way for younger men and have more time to spare with his wife. His wife died in 1980 and for many years afterwards he lived with his married daughter, incapacitated by frequent transient ischaemic attacks. Eventually he had to retire to a nursing home in Cumbria, where he died having just failed to achieve his ambition of being recorded in every census in the 20th century.
Sir Richard Doll
[Brit.med.J., 1991,302,1017;1992,305,1521-26; The Lancet, 1991,337,1154; Times, 22 Apr 1991;The Independent, 7 May 1991;The Guardian, 23 Apr 1992; The Daily Telegraph, 23 Apr 1991;MRC News, Sept 1991,52,35,Dec 1987,37,5]