by Burch, Druin
The Cochrane Collaboration was formed to pursue Archie Cochrane’s goal of making medicine more evidence-based. A non-governmental organisation with the aim of publishing comprehensive analyses of the available data on different treatments, it has undertaken several reviews of antidepressants. One review published in 2004 paid particular attention to the way in which studies of these drugs might be misleading when compared with inactive placebo tablets.2 Given that antidepressants cause side effects, it is reasonable to think that many people are able to tell whether they have been given a placebo instead – and since the placebo effect rests on believing that you are not actually taking a placebo, this matters. Three Cochrane Collaboration reviewers looked at trials that specifically used an ‘active’ placebo in order to overcome this, a placebo designed to produce similar side effects to the active drug with none of the main effects. (They were looking at studies of a class of antidepressants called tricyclics that commonly cause effects like a dry mouth, dry nose and constipation.) The review found nine such studies, containing a total of 751 patients, and showing much smaller than expected differences between the active placebos and the tricyclics. ‘This suggests that unblinding effects may inflate the efficacy of antidepressants in trials using inert placebos,’ they concluded. The bulk of the good properties that antidepressants were believed to have, in other words, were probably fictional – reflecting only the consequence of carrying out badly designed studies.
Early in 2008, a paper was published that drew widespread attention to the issue of whether antidepressants do far less than most doctors and patients believe (Kirsch et al.). Perhaps one of the most curious aspects of the paper, was that much of what it had to say was not new. Irving Kirsch, of the Department of Psychology at the University of Hull, had published a very similar study in 2002 while at the University of Connecticut. Then he had looked at the data submitted to the FDA for approval of the six most popular antidepressants between 1987 and 1999. He found that almost all (80 per cent) of the drugs’ beneficial effects came from their power as a placebo. The 20 per cent left over appeared to be definitely there, but not necessarily to matter very much. On the Hamilton Depression Scale, the method commonly accepted for measuring depression, it made a difference of about 2 points. In Britain the National Institute for Clinical Excellence (NICE), a governmental body set up to independently assess the effectiveness and safety of medical interventions, set a cut-off of 3 points as being clinically useful. (An effect can be statistically measurable but too small to be worthwhile to a patient, particularly if it comes at the expense of a known chance of side effects.) Kirsch’s 2008 study looked at four of the newer antidepressants, and attempted to see if they were particularly effective or ineffective in relation to the severity of someone’s depression. It found the drugs made a similar difference overall, again less than the minimum that NICE believed was genuinely useful. Among the small minority of patients who were most severely depressed, the difference was slightly larger. The fact that their 2008 study drew the bulk of its public attention for findings that were already in their 2002 paper shows another problem: disseminating information can be as difficult as gathering it to begin with.
There are worse things than the widespread use of generally ineffective drugs. Antidepressants are not a scandal like the poisonous elixir of sulphonamide, or thalidomide. But they are a reminder that regulatory frameworks currently fail to make sure that, as doctors or as patients, we have the information we need about the full impacts of many medical interventions.
In the past regulatory reforms have usually followed on from the sort of tragedies that do leave people dead or permanently disabled. Currently we have the opportunity to improve them voluntarily, without waiting for the next medical catastrophe to come to light and force us into action.
The story of our development of testing is a tale of mental advances that have showered riches upon the world, and are hardly known about.
Testing and experiment and trial have always been a part of human life, and their failure to protect us from deluded cures and poisonous remedies tells us something vital. They cannot be relied upon unless they have sufficient method to them, and the quality of that method determines the quality of their results. It is not enough to use words like ‘test’ and ‘experiment’ and ‘trial’ without understanding that they can mean something no better than a guess, or something so organised and reliable as a double-blind, randomised, controlled trial.
Paying attention to the errors that those before us have made, noticed and reported can be profoundly useful. Not only is our understanding of their care and trouble deepened, but our chance of repeating their mistakes is a little reduced. Historians often argue that it is ‘sympathetic’ to view people the way they viewed themselves, and to look for ways of excusing them for not doing better than they did. That seems to me a condescending approach. Those who thought seriously about how to help their sick fellows were not trying to do ‘as well as could be expected’. They did not mean for their theories to be ‘perfectly reasonable given the way people understood the world at the time’. They meant to do their patients some good and to uncover the truth, and we take them most seriously when we recognise the ways they often failed.
In his book Effectiveness & Efficiency, first published in 1971, Archie Cochrane wrote:
Two of the most striking changes in word usage in the last twenty years are the upgrading of ‘opinion’ in comparison with other types of evidence, and the downgrading of the word ‘experiment’. The upgrading of ‘opinion’ has doubtless many causes, but one of the most potent is, I am sure, the television interviewer and producer. They want everything to be brief, dramatic, black and white. Any discussion of evidence is written off as lengthy, dull, and grey. I have seldom heard a television interviewer ask anyone what his evidence was for some particular statement.
Lewis Thomas was born in 1913, four years after Cochrane, and qualified at Harvard Medical School in 1937. Despite sulphonamides, he found contemporary hospitals offered little more to the sick than hotel accommodation. ‘Whether you survived or not depended on the natural history of the disease itself,’ he wrote. ‘Medicine made little or no difference.’
Bad as things were, when he looked back at his father’s medical training, he was impressed at how much they had improved. A quarter of a century before him, his father had qualified from Columbia. The cutting-edge training that Thomas’s father received contained far more truth about the human body than had ever been known before. By the start of the twentieth century, knowledge of the sciences underpinning medicine was full of sophistication – in pathology, microbiology, physiology, chemistry and even pharmacology. In practice, however, medicine was scarcely different from how it had been for thousands of years. ‘Paper after paper’, found Thomas when he looked at his father’s books, ‘recounts the benefits of bleeding, cupping, violent purging, the raising of blisters by vesicant ointments, the immersion of the body in either ice water or intolerably hot water . . . Endless lists of botanical extracts cooked up and mixed together under the influence of nothing more than pure whim.’ It is strange what people were, and were not, capable of. At the same time as Galen was recommending ineffective potions and leeches, the Romans were performing architectural and engineering marvels. By the time Thomas’s father was being taught largely identical potions and leeches, Einstein had his theory of relativity and aeroplanes were taking to the skies.
What was missing in medicine was experimental method. People could apply it to the hard sciences, but doctors believed in the power of their intuition. They relied on trial and error, and in such an unstructured way that it was of little use. As Lewis Thomas wrote:
My first hope is for removal of substantial parts of the curriculum in the first two years, making enough room for a few courses in medical ignorance, so that students can start out with a clear view of the things medicine does not know.
There is a bitter joke in modern medicine: the violence with which
someone makes an argument is inversely proportional to the amount of evidence they have backing it up. The more people are left without reliable experiments, the more they seem to fall back on strongly held opinions, as though confidence was a starch that could stiffen ideas into facts simply by being applied with enough fervour.
History shows that there is a better way of behaving. What could do us more good than understanding that many of our opinions can be tested, and that those which can be tested should be tested? The medical mistakes of the past tell us that beliefs based on theories that are untestable, or simply yet to be tested, need to be held humbly. Randomised controlled trials have swept away much suffering and error from hospitals and homes, ushering in comfort and healing instead. Trials can be full of statistics; difficult to understand and laborious to undertake. They have a loveliness to them all the same, and it comes from their power to uncover parts of the reality we live in.
It seems to be our nature to prefer credulity to doubt, confidence to scepticism. We share a tendency to simplify and confuse things, to slip into mental habits that let us down. But by acknowledging this we can be on our guard against it. When it comes to certain questions about the world, questions where the possible answers can be experimentally tested, ‘scientific’ becomes synonymous with ‘rational’. Without demanding evidence, without understanding the qualities that make it reliable, we are vulnerable.
Within the medical profession, many people feel threatened by the rise of clinical trials. They are frightened by the statistics, or by the feeling that others understand such numbers better than they. A clinical trial, these people say, will not replace the rich complexity of an interaction between a patient and a doctor. Yet no one ever suggested it could, or should. Trials tell you certain truths about the world, but not others. They add to your ability to make decisions. They do not wipe out the importance of making them.
The technique of the randomised controlled trial has a splendour to it, as do the parts of human life to which it is wholly inapplicable. No statistical test or trial design will help tell you if you are in love, or if you are loved in return. That does not mean that trials and statistics are useless, only that they have their place. That place is a fine one. Experiments and numbers reveal truths: they are tools for understanding the world, and for making it better.
* * *
1 See http://adc.bmj.com/content/76/1/50.full.pdf
2 There has been relatively little effort made to compare different antidepressants with each other, an expensive business that drug companies do not voluntarily pursue. That they fail to do suggests that they have little conviction that any one agent is better than another, as well as being a reminder that governments do nothing to demand such studies.
Index
The page references in this index correspond to the printed edition from which this ebook was created. To find a specific word or phrase from the index, please use the search feature of your ebook reader.
Aberdeen 131
abortifacients 20
acetanilide 110, 138
acetyl groups 117
acetylsalicylic acid see Aspirin
acids 51
adenoids 240
agriculture 11
ague 39
AIDS 103, 274, 277–9
Akers, Elva 149
al-Haytham, Ibn 24–5
Albert, Prince 81, 85
alchemy 64, 65
alcohol 60, 122, 125, 204
Alexander the Great 37
alkalis 68–9, 82
alkaloids 15, 117
almonds 12
alternate allocation 130, 131, 177–8, 183
America, Americans 5, 26, 47, 71, 76, 113, 119, 126, 149, 150, 151, 152, 155, 231, 264–5
American College of Cardiology 264
American Journal of Public Health 155
American Review of Tuberculosis 156
American War of Independence 280–1
ammonia 83, 105
ammonium purpurate 83
amphetamines 8–9
amputation 122
anaesthetics 20, 32
Anales de Chimie et de Physique 49, 68
analgesics 20, 296
anatomy 49, 89
angina 237
angina pectoris 254
aniline 80–1, 83, 110
aniline dyes 84, 90, 91, 93, 95, 104, 111
animal fat 51, 290
animalcules 86, 88
aniseed 12
Annals of Western Medicine and Surgery 240, 242
anopheles atroparvus 42
Anson, Commodore 157–8
anthrax 91
anti-cholesterol drugs 209
antiarrhythmics 263–70
antibacterial compounds 152
antibiosis 143
antibiotics 132–4, 140, 141, 141–56, 153, 165, 258–62, 296
antibodies 93–4, 219, 221
anticoagulants 255, 259
antidepressants 9, 297, 298, 299–301
Antifebrin 110, 112, 138
antimony 21
Antipyrine 108–9, 110, 112
antisepsis 113
antiseptic 113, 114, 141
appendicitis 295
Aquinas, Thomas 24
Arabs 28
árbol de calenturas (fever tree) 40
Archives of Internal Medicine 223
Aristotle 24
arsenic 97–9, 100, 101
asepsis 113
Aspirin 12–13, 114, 115, 117, 119, 148, 167, 208, 222, 223, 224, 231–4, 238, 239, 240, 241–52, 253–8, 259, 260
aspirin trials 247–51, 253–5, 256
asthma 104–5
Athens 23
Atlantic Monthly 75
Atoxyl 101, 103
Aubrey, John 26, 42n
Aurelius, Marcus 125
Auschwitz 234–5
Australia, Australians 99, 147, 210
Avicenna 28
AZT 277–9
B. influenzae 144
Bacon, Francis 25–6, 34–5, 45, 140–1
Bacon, Roger 25, 76
bacteria 21, 47, 91, 100, 114, 140, 141, 143, 144, 258, 296
bacterial infection 128, 150, 170, 233
Badische Anilin & Soda-Fabrik-BASF 90
Baeyer, Adolf von 90
Baghdad 52, 53, 117
balsam bark see Peruvian balsam bark
Baltimore 80
Baraldini, V. 294
barbiturates 206, 207
Basra 25
Bayer, Friedrich 111
Bayer (Friedrich) & Co. 111, 112, 114, 115, 116, 117, 119, 126, 231, 232, 233, 235–6
bed-rest 9
beer 11, 13, 51, 84, 85, 200
beeswax 290
beetroot 65, 68
Behçet’s Disease 221, 222
Behring, Emil 128
Belchier, John 87
Bell, Julian 160
benzene 80, 81 and note, 85, 108
Berlin 105, 114–15, 128, 139
beta blockers 224, 260
Bevan, Aneurin 186
Biometrika 172
Birmingham 236, 294
Birmingham General Hospital 121
Bismarck Brown 85, 90
bladder 19
bladder stones 83
Blair, Eric Arthur see Orwell, George
bleeding 240–1, 244, 251, 255, 259, 269, 282, 284, 286, 296
blood 95, 136, 138, 143, 154, 216, 220, 233, 236, 239, 244, 258
blood cells 91
blood circulation 26
blood clots 239, 240, 245, 246, 252, 258–9, 263
blood letting 20, 26, 38, 43, 47, 51, 52, 53, 72, 74, 121, 149, 284–5, 288, 289, 290, 303
blood pressure 236
blood thinning 264, 295
BMC Women’s Health 294
boa constrictor 83
body temperature 47
boils 220
Bonnet, Charles 87–8
Recherches sur l’usage des feuilles dans les plantes 88
Boston (Massachusetts) 74, 87, 258
bowels 19
Boyle, Robert 30
Brandt, Karl 216
brandy 88
bread 13, 21
breast cancer 296–7
breath, shortness of 18, 19, 55, 118, 119
Breslau 91
Breurn 60
Britannia Violet 84, 85
British Association for the Advancement of Science 85
British Board of Trade 202
British Heart Foundation 258, 260
British Journal of Ophthalmology 295
British Medical Association 201
British Medical Journal 97, 129, 185, 228, 253, 292, 294
British Pharmacopoeia 250
Brooklyn 104–5
Browne, Thomas 26
Bruce, David 99–100 and note
Bryant, William Cullen 75
bubonic plague 41
Buchner, Johann 106
Bulletin of Narcotics (1967) 16
Bumstead, John 150–1
Bunsen, Robert Wilhelm 90, 105
Bureau of Chemistry see Food and Drug Administration (FDA)
burns 57–8
Burroughs, William 118n
Burton beer 85
Cahn, Arnold 109–10
California 240
calomel 97
Cambridge 159, 178
Cambridge (Massachusetts) 71
Camp Detrick (Maryland) 152, 272n
camphor 88
Canada 282, 292
Canadian Medical Journal 155
canary wine 31, 32
cancer 119, 149, 193, 218–19, 272–5, 296–7
carbon dioxide 136
cardamom 97
cardiology, cardiologists 192, 196, 245, 252, 255, 259–60, 261, 263, 266–7
carmine 89
Caro, Heinrich 90
Cassella Dye Works 102, 126
Castellani, Count Aldo 100
castor oil 19
cathartics 20
Caventou, Joseph Bienaimé 69