ILLUSTRATIVE TIMELINE

Hippocrates (5th century BCE)

Hippocrates (5th century BCE)

More than two millennia ago, the Greek physician Hippocrates stressed that a competent researcher must ensure that his starting point is knowledge of what has already been discovered.

View the JLL record →
al-Razi  (10th century CE; 4th century AH)

al-Razi (10th century CE; 4th century AH)

In the 10th century, the Persian physician Abu Bakr Muhammad ibn ZakariyaʾAl-Razi  (Rhazes) recognized the need for an untreated control group when assessing the effects of a treatment.

View the JLL record →
Ibn Sīnā (c.1012 CE; c.402 AH)

Ibn Sīnā (c.1012 CE; c.402 AH)

In the 11th century, the Persian physician Ibn Sina  (Avicenna) stressed that both the theoretical and practical parts of medicine are science, one dealing with the principles of medicine, the other with how to put those principles into practice.

View the JLL record →
Bacon, Roger (1266)

Bacon, Roger (1266)

In the 13th century, the English Franciscan monk Roger Bacon, often regarded as England’s first scientist, asserted that ‘without experiment nothing can be sufficiently known’.

View the JLL record →
Petrarca F (14th century)

Petrarca F (14th century)

In 1364, in a letter to his fellow poet Bocaccio, the Italian Francesco Petrarca proposed an experiment to assess whether people would be better off avoiding rather than seeking medical treatment.

View the JLL record →
Paré A (1575)

Paré A (1575)

In the 16th century, the French military surgeon Ambroise Paré exploited natural and planned experiments to test treatments for burns.

View the JLL record →
Hidalgo de Agüero, Bartolomé (1604)

Hidalgo de Agüero, Bartolomé (1604)

At the end of the 16th century, the Spanish surgeon Bartolomé Hidalgo de Agüero collected statistics showing that dry treatment of wounds resulted in fewer deaths than the orthodox, wet method.

View the JLL record →
Bacon F  (1645)

Bacon F (1645)

In the 17th century, the English scientist Francis Bacon, pointed out that positive and negative results of experiments should be accorded equal weights.

View the JLL record →
Van Helmont JB (1648)

Van Helmont JB (1648)

In the 17th century, the Flemish physician Jean-Baptiste van Helmont proposed casting lots to decide which patients should be treated by orthodox physicians with blood letting and purging, and which by him without these treatments.

View the JLL record →
Nettleton T (1722)

Nettleton T (1722)

Early in the 18th century, Thomas Nettleton, a physician in Halifax, England, compiled figures to compare the death rate from natural smallpox with that following inoculation.

View the JLL record →
Clifton F (1732)

Clifton F (1732)

At the beginning of a century during which British use of quantified, tabulated data became widespread, Francis Clifford designed tables to record illnesses, treatments, and outcomes.

View the JLL record →
Cheselden W (1740)

Cheselden W (1740)

In the mid-18th century, the English lithotomist William Cheslden drew attention to the importance of taking account of the age of patients when assessing survival after operations for bladder stones.

View the JLL record →
Lind J (1753)

Lind J (1753)

In 1747, James Lind, a Scottish naval surgeon faced with uncertainty about which of many proposed treatments for scurvy to use, compared six of them in a prospective controlled trial.

View the JLL record →
Cullen W (1772)

Cullen W (1772)

In manuscript notes of his clinical lectures, the Scottish physician William Cullen describes how he used ‘a pure placebo’ in an attempt to comfort a dying patient.

View the JLL record →
Millar J (1779)

Millar J (1779)

The Scottish physician John Millar showed the results of treatment for fevers in tables comparing death rates.

View the JLL record →
Mesmer FA (1781)

Mesmer FA (1781)

Franz Mesmer, the Austrian doctor who claimed to be able to heal patients by manipulating ‘animal magnetism’, proposed a fair comparison of his methods with those of his detractors, by drawing lots to decide which patients should be allocated to each of the comparison groups.

View the JLL record →
Alanson E (1782)

Alanson E (1782)

Edward Alanson, a surgeon in Liverpool, described and evaluated a new method of performing limb amputations using a skin flap to cover the stump.

View the JLL record →
Commission Royale. Bailly A (1784)

Commission Royale. Bailly A (1784)

Antoine Lavoisier, Benjamin Franklin and others in Paris assessed the effects of Franz Mesmer’s animal magnetism by blindfolding patients to whom it was applied.

View the JLL record →
Black W (1788)

Black W (1788)

As the 18th century drew to a close, the British physician William Black reviewed how collecting, analyzing and presenting evidence about the effects of treatments had evolved.

View the JLL record →
Fordyce G (1793)

Fordyce G (1793)

In the Transactions of a Society for the Improvement of Medical and Chirurgical Knowledge, the British physician George Fordyce reviewed the progress made during the 18th century in collecting and publishing data about medical care.

View the JLL record →
Sutton D (1796)

Sutton D (1796)

During the second half of the 18th century, Daniel Sutton became an extremely successful smallpox inoculator, who used experiments to inform the development of his technique.

View the JLL record →
Haygarth J (1800)

Haygarth J (1800)

At the end of the 18th century, John Haygarth, an English physician, did placebo controlled trials to test the  therapeutic claims made by Elisha Perkins for his ‘metallic tractors’.

View the JLL record →
Cooper A (1801)

Cooper A (1801)

At the beginning of the 19th century, the British surgeon Astley Cooper demonstrated the dramatic effect of tympanotomy in restoring hearing in certain kinds of deafness.

View the JLL record →
Percival T (1803)

Percival T (1803)

Percival 1803: Percival noted that, when no treatment is known, the interests of the public require a search for new remedies, “scrupulously and conscientiously governed by sound reason, just analogy, or well authenticated fact”.

View the JLL record →
Lesassier Hamilton A (1816)

Lesassier Hamilton A (1816)

In 1816, the British army surgeon Alexander Lesassier Hamilton reported a controlled trial done during the Peninsular War to assess the effects of bloodletting for fever.

View the JLL record →
Hawkins FB (1829)

Hawkins FB (1829)

In a book entitled The Elements of Medical Statistics published in 1829, Francis Bisset Hawkins welcomed the replacement of unquantified clinical impressions with statistical analyses assessing the effects of treatments.

View the JLL record →
Löhner G, on behalf of a Society of truth-loving men (1835)

Löhner G et al (1835)

In 1835, a Society of Truth-Loving Men in Nuremberg conducted a randomized, blinded trial to assess whether homeopathic provings could be distinguished from snow water.

View the JLL record →
Louis PCA (1835)

Louis PCA (1835)

In 1835, Pierre-Charles-Alexandre Louis compared mortality rates among Parisian patients with pneumonia who had received prompt treatment with bloodletting with other patients who had been treated later in the course of their disease.

View the JLL record →
Gavarret LDJ (1840)

Gavarret LDJ (1840)

In 1840, the French physician and statistician Jules Gavarret published a book on statistical analysis of treatment tests, stressing the importance of estimating uncertainty and calculating ‘limits of oscillation’ associated with estimates of treatment effects.

View the JLL record →
Forbes J (1846)

Forbes J (1846)

In 1846, the founding editor of British and Foreign Medical Review, John Forbes, was unable to find any evidence that homeopathy was useful, but concluded that the evidence supporting orthodox medicine was little better.

View the JLL record →
Tessier JP (1852)

Tessier JP (1852)

In the mid-19th century, Jean-Paul Tessier used the Parisian hospital system to organize a carefully controlled comparison of homeopathic treatment with allopathic treatment.

View the JLL record →
Balfour TG (1854)

Balfour TG (1854)

In the mid-19th century, Thomas Graham Balfour, a British army surgeon in charge of an orphanage, used alternate allocation - ‘to avoid the imputation of selection’ – in research to assess claims that belladonna was protective against scarlet fever.

View the JLL record →
Semmelweis I (1861)

Semmelweis I (1861)

In 1861, Ignaz Semmelweis, a Hungarian doctor working at the Vienna Maternity Hospital, published his analyses of statistics from ‘a natural experiment’ demonstrating the importance of hygiene in preventing death from ‘childbed fever.’

View the JLL record →
Kocher T (1883)

Kocher T (1883)

In 1883, the Swiss surgeon Theodor Kocher reported his follow-up of patients from whom he had removed goitrous thyroid glands, showing that several of them had developed cretinism.

View the JLL record →
Rayleigh, The Lord (1885)

Rayleigh, The Lord (1885)

In his presidential address to the British Association for the Advancement of Science, Lord Rayleigh, a British physicist, observed that “the work that deserves most credit is that in which not only are new facts presented, but their relation to old ones is pointed out”.

View the JLL record →
Bettencourt R, Serrano J-A (1891)

Bettencourt R, Serrano J-A (1891)

In 1890, the Portuguese investigators Bettencourt and Serrano reported the successful and dramatic treatment of myxœdema with a hypodermic transplant of sheep thyroid gland.

View the JLL record →
Vorderman AG (1897)

Vorderman AG (1897)

At the end of the 19th century, Adolphe Vorderman, a prison medical officer in the Dutch East Indies, used blinding to obtain unbiased estimates of the effects of unpolished rice in preventing beri-beri.

View the JLL record →
Fibiger J (1898)

Fibiger J (1898)

At the end of the 19th century, French studies using historical controls to assess the effect of anti-diphtheritic serum were controversial. Johannes Fibiger, a Danish physician, used serum only in child patients admitted to hospital on alternate days, and confirmed that serum reduced mortality.

View the JLL record →
Haffkine WM (1900)

Haffkine WM (1900)

At the end of the 19th and beginning of the 20th centuries, Waldemar Haffkine,a Russian physician working in India, organized alternate allocation trials to test serum treatments for cholera and plague.

View the JLL record →
Celli A (1900)

Celli A (1900)

In 1900, Angelo Celli reported his controlled trials of barrier methods to protect Italian railway workers and their families from malaria.

View the JLL record →
Choksy NH (1900)

Choksy NH (1900)

At the end of the 19th and beginning of the 20th centuries, Nasarwanji Hormusji Choksy  used ‘the alternate method’ to generate comparison groups in assessments of the effects of serum treatments for plague and cholera in Bombay.

View the JLL record →
Park WH  (1900)

Park WH (1900)

In 1900, the American physician William Park described his use of alternation  to create comparison groups in a test of the effects of anti-diphtheritic serum in children.

View the JLL record →
Lustig A, Galeotti G (1901)

Lustig A, Galeotti G (1901)

In 1901, the Italian investigators Lustig and Galeotti used ‘The Alternate System’ to create comparison groups for testing plague serum in India.

View the JLL record →
Pearson K (1904)

Pearson K (1904)

Karl Pearson, one of the founders of the British school of statistics, assessed the effects of inoculation on enteric fever (typhoid) in the British army using ‘meta-analysis’ to combine statistics from separate but similar studies.

View the JLL record →
Cousin M (1905)

Cousin M (1905)

In 1905, the French investigators Cousin and Netter used ‘la système alternante’ to generate comparison groups to test the effects of calcium chlorate in reducing serum sickness among children receiving anti-diphtheritic serum.

View the JLL record →
Fletcher W (1907)

Fletcher W (1907)

In his 1907 report of alternate allocation to polished or unpolished to the inmates of a lunatic asylum in Kuala Lumpur, William Fletcher demonstrated that unpolished rice was protective against beri-beri.

View the JLL record →
Bingel A (1918)

Bingel A (1918)

In 1918, Adolf Bingel, a German physician in Brunswick, reported a large, blinded, alternate allocation trial that concluded that normal horse serum was as effective as activated horse serum in treating diphtheria.

View the JLL record →
Park WH, Bullowa JGM, Rosenbluth NM (1928)

Park WH, Bullowa JGM, Rosenbluth NM (1928)

In the late 1920s, William Park, Jesse Bullowa and Milton Rosenbluth used ‘meta-analysis’ to combine the results of alternate allocation trials of serum for lobar pneumonia in three New York hospitals.

View the JLL record →
Wagner-Jauregg J (1931)

Wagner-Jauregg J (1931)

The Austrian physician Julius Wagner-Jauregg, a Nobel Prize Winner in 1927, had first reported using alternation to generate treatment comparison groups in 1926. In a further paper in 1931 he named this methodological principle ‘Simultanmethode’.

View the JLL record →
Martini P (1932)

Martini P (1932)

In 1932, the German physician Paul Martini published a book setting out the methodological principles for assessing the effects of treatments.

View the JLL record →
Hill AB (1937)

Hill AB (1937)

In a series of articles published in The Lancet in 1937 and republished in a book the same year, the British statistician Austin Bradford Hill introduced the principles of medical statistics to medical doctors.

View the JLL record →
Kendrick P, Eldering G (1939)

Kendrick P, Eldering G (1939)

During the late 1930s, Pearl Kendrick and Grace Eldering, working in the Michigan Department of Health, did an important controlled study to evaluate the effects of the whooping cough vaccine they had developed.

View the JLL record →
Bell JA (1941)

Bell JA (1941)

In 1941, the American physician Joseph Bell, in an exceptionally clearly written report of his trial of whooping cough vaccine, reported his use of random sampling numbers to generate unbiased comparison groups.

View the JLL record →
Medical Research Council (1944)

Medical Research Council (1944)

During World War 2, the British physician Philip d’Arcy Hart, supported by his wife Ruth and Joan Faulkner, designed, conducted and reported the UK Medical Research Council’s first well controlled multicentre trial.

View the JLL record →
Medical Research Council (1948)

Medical Research Council (1948)

The UK Medical Research Council’s 1948 report of a controlled trial of streptomycin for pulmonary tuberculosis was a methodological landmark because it provided detailed information about the trial, and in particular, the steps taken to prevent foreknowledge of treatment assignments.

View the JLL record →
Doll R, Pygott F (1952)

Doll R, Pygott F (1952)

As trial design became more sophisticated, the factorial trial of treatments for gastric ulcers reported in 1952 by the British physicians Richard Doll and Frank Pygott provided an early illustration of how more than one treatment could be compared with placebo in a single trial.

View the JLL record →
Leitch I (1959)

Leitch I (1959)

Isabella Leitch was one of the first people to recognise that reviews of research evidence (sometimes referred to as ‘secondary research’) need to respect the same scientific principles as those that were becoming accepted for the design of ‘primary’ research.

View the JLL record →
Hill AB (1960)

Hill AB (1960)

By the end of the 1950s, the British Medical Research Council had accumulated substantial experience of designing, running, analysing and reporting controlled trials. Austin Bradford Hill was invited to plan and chair a meeting in Vienna in 1959 to report this experience.

View the JLL record →
Chalmers TC (1968)

Chalmers TC (1968)

From the 1960s until his death in the 1990s, the American physician Thomas Chalmers was one of the most vocal promoters of clinical trials, suggesting that inadequately evaluated new treatments should only be used in randomized trials until their effects became known.

View the JLL record →
Schwartz D, Flamant R, Lellouch J (1970)

Schwartz D, Flamant R, Lellouch J (1970)

In 1970, the French statisticians Daniel Schwartz, Robert Flamant and Joseph Lellouch drew attention to the different design requirements for explanatory trials (to assess whether a treatment worked under the most favourable circumastnaces) and trials to assess whether treatments actually worked, in practice.

View the JLL record →
Stjernswärd J (1974)

Stjernswärd J (1974)

In 1974, Jan Stjernswärd, a Swedish oncologist, used statistical synthesis (‘meta-analysis’) of the results of five controlled trials of radiotherapy for breast cancer,  and showed that  it inceased the risk of death.

View the JLL record →
Glass GV (1976)

Glass GV (1976)

In 1976, in his presidential address to the American Educational Research Association, Gene Glass introduced the term ‘meta-analysis’ to denote statistical synthesis of the results of similar studies.

View the JLL record →
Chalmers TC, Matta RJ, Smith H, Kunzler A-M (1977)

Chalmers TC, Matta RJ, Smith H, Kunzler A-M (1977)

In 1977, the New England Journal of Medicine published a meta-analysis of controlled trials of anti-coagulation for myocardial infarction, by Thomas Chalmers and his colleagues.

View the JLL record →
Hemminki E  (1980)

Hemminki E (1980)

Using drug research done during the late 1970s, Elina Hemminki, a Finnish health services researcher, showed that studies of new drugs submitted to licensing authorities were less likely to be published subsequently if they had looked for adverse effects.

View the JLL record →
Gruppo Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI) (1986)

GISSI (1986)

In the mid-1980s, through widespread collaboration, the Italian Gruppo Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI) published  the first report of a ‘mega-trial’ of a treatment. This , confirmed  the beneficial effects of thrombolysis after heart attack.

View the JLL record →
Simes RJ (1986)

Simes RJ (1986)

Simes 1986:  John Simes proposed international registration of all clinical trials after he showed that conclusions about treatments for ovarian cancer differed depending on whether the results of unpublished trials had been taken into account.

View the JLL record →
Mulrow C (1987)

Mulrow C (1987)

In an analysis of the quality of review articles in major general medical journals, the American physician Cynthia Mulrow, showed that most review articles were based on analyses that had ignored basic scientific principles.

View the JLL record →
Early Breast Cancer Trialists' Collaborative Group (1990)

EBCTCG (1990)

The multinational Early Breast Cancer Trialists' Collaborative Group (EBCTCG ) showed that, by sharing individual patient data, scientifically more robust evidence about the effects of treatments could be obtained.

View the JLL record →
Antman EM, Lau J, Kupelnick B, Mosteller F, Chalmers TC (1992)

Antman EM, Lau J, Kupelnick B, Mosteller F, Chalmers TC (1992)

The 1992 comparison of textbook recommendations with the results of clinical trials reported by Elliott Antman, Joseph Lau, Bruce Kupelnick, Frederick Mosteller and Thomas Chalmers showed that patients had died avoidably because textbooks had not reviewed reliable research evidence systematically.

View the JLL record →
Dickersin K, Min YI, Meinert CL  (1992)

Dickersin K, Min YI, Meinert CL (1992)

This study is the first of several led by Kay Dickersin which have provided the rationale for her lifelong commitment to reducing biased under-reporting of research.

View the JLL record →
Standards of Reporting Trials Group (1994).

Standards of Reporting Trials Group (1994).

This paper reports the first international collaboration to reach agreement on what details should be provided in reports of controlled trials.

View the JLL record →
Sackett D, Straus S; for Firm A of the Nuffield Department of Medicine (1998)

Sackett D, Straus S (1998).

This study reports the use of a mobile "Evidence Cart" to access the best available evidence from systematic research during hospital rounds in support of patient care and clinical teaching.

View the JLL record →
Kunz R, Vist G, Oxman AD (2002)

Kunz R, Vist G, Oxman AD (2002).

Regina Kunz, Gunn Vist and Andy Oxman showed in their 2002 Cochrane Review that comparisons of the results of randomised trials with those of non-randomised (observational) studies sometimes reveal concordance and sometimes discordance. Discordance seems likely to reflect inadequately controlled allocation biases in non-randomised studies. Unfortunately, it is not currently possible to predict which non-randomised (observational) studies will yield results that will be discordant with those of randomised trials.

View the JLL record →
Chan A-W, Hròbjartsson A, Haahr M, Gøtzsche PC, Altman DG (2004)

Chan A-W et al. (2004)

By comparing trial protocols with trial reports, a Canadian physician, An-Wen Chan and his colleagues, demonstrated frequent, biased switching of primary outcomes based on whether differences in outcomes were statistically significant.       

View the JLL record →
Perel P, Roberts I, Sena E, Wheble P, Briscoe C, Sandercock S, Macleod M,  Mignini LE, Jayaram P, Khan KS (2007)

Perel P et al. (2007)

In 2007, Pablo Perel and his colleagues addressed longstanding questions about the validity of animal models for developing treatments for patients. They demonstrated both concordance and discordance between the results of studies in animals and humans.

View the JLL record →
Savović J, Jones HE, Altman DG, Harris RJ, Jüni P, Pildal J, Als-Nielsen B, Balk EM, Gluud C, Gluud LL, Ioannidis JPA, Schulz KF, Beynon R, Welton NJ, Wood L, Moher D, Deeks JJ, Sterne JAC (2012)

Savović J et al. (2012)

Meta-epidemiology applies multivariable analysis to large numbers of studies to identify associations. The example reported in 2012 by Jelena Savović and her colleagues examined the relationships between study design features and estimates of treatment effects.  The findings emphasized the importance of blinded outcome assessment to reduce observer/measurement bias.

View the JLL record →
Jefferson T, Jones MA, Doshi P, Del Mar CB, Hama R, Thompson MJ, Spencer EA, Onakpoya I, Mahtani KR, Nunan D, Howick J, Heneghan CJ (2014).

Jefferson T et al. (2014)

In 2014, in a major methodological development, Tom Jefferson and his colleagues conducted a systematic review of 46 trials of anti-flu drugs, drawing on information reported in more than 160,000 pages of Clinical Study Reports prepared for drug regulatory authorities.

View the JLL record →

Browse the library

Fair tests

Abu Bakr Muhammad ibn Zakariyya al-Razi (portrait courtesy of the Wellcome Trust)

Despite acting with the best of intentions, health professionals have sometimes done more harm than good to the patients who have looked to them for help. Some of this suffering can be reduced by ensuring that fair tests are done to address uncertainties about the effects of treatments.

Sub-topics:

 

Biases

John Haygarth

Biases in tests of treatments are those influences and factors that can lead to conclusions about treatment effects that are systematically different from the truth.

Sub-topics:

Play of chance

Jules Gavarret

When treatments are compared, any differences in outcome events may simply reflect the play of chance.

Increasing the number of events studied in research reduces the likelihood of being misled in this way.

Sub-topics:

Serving patients

A patient

The interests of patients can be served by: improving reports of research, preparing and updating systematic reviews of reliable studies, and using these to inform decisions about treatment.

Sub-topics:

 

SPECIALIST COLLECTIONS

You can also browse the Library by choosing topics of interest from the drop-down menus.