A bitter pill to swallow

The sketchy evidence for the effectiveness of homoeopathic medicine has no scientific basis, and pos

There was an outcry in September when we learned that children in Scotland were being given a homoeopathic "MMR vaccine", a product that offered no protection against the serious dangers posed by measles, mumps and, for pregnant women, rubella. This had echoes of the discovery a few years ago by Sense About Science, Simon Singh and Newsnight that some pharmacists were offering homoeopathic pills for protection against malaria to people travelling to Central Africa. Such practices may be disturbing, but they occur because we tend to think there is no harm in indulging the clamour to maintain the alternative health market.

Reading the 11 October issue of the New Statesman, I was shocked by an advertisement in the accompanying supplement, "Social Care: Who Pays?", referring to me and my work. Rarely had I seen an advert so inaccurate and borderline libellous in a respected publi­cation. The advert, which appeared to breach the British Code of Advertising, was by a lobby group called Homeopathy: Medicine for the 21st Century (H:MC21). It contained unjustified attacks on myself and colleagues, including statements that gave a dangerously false impression of homoeopathy's therapeutic value.

As the advert questioned my own competence, I should address this first. I started my medical career in a homoeopathic hospital, where I was trained in homoeopathy for several months. Many years later, it became my job to apply science to this field and I felt I had a duty to keep an open mind - open but not uncritical.

A critical mind would notice that the two basic principles of homoeopathy fly in the face of science, logic and common sense. The first assumption is that "like cures like". For instance, if onions make my eyes and nose water, homoeopathic remedies derived from onions can be used to treat my patients' hay fever, which sometimes causes runny eyes and noses. The second assumption proposes that diluting remedies homoeopathically makes them not less but more potent, even if the final preparation no longer contains a single molecule of any active substance. These theories are not based on anything that remotely resembles fact. Like does not cure like, and endlessly diluting remedies certainly does not render them stronger, but weaker. But is there some entirely new energy to be discovered that we do not yet comprehend? Not understanding homoeopathy does not necessarily mean that it is useless.

The best way to find out is to determine whether homoeopathic remedies behave differently from placebos when patients use them. In other words, we need clinical trials.

Data gap

About 150 such studies (mostly conducted by homoeopaths) and well over a dozen syntheses of this research are available. Their results are sobering: the totality of the most reliable evidence fails to show that homoeopathic remedies work better than placebos. So, after about 200 years of research, there is no good data to convince non-homoeopaths that homoeopa­thic remedies are any different from pure sugar pills. Pro-homoeopathic lobby groups such as the one that placed the advertisement therefore have to employ propaganda to try to convince consumers who may not know better. This is perhaps understandable, but surely not right.

What of patients' experience, some might ask. Thousands of people across the world swear by homoeopathy. Are they all deluded? Clearly not. People undoubtedly do get better after seeing a homoeopath. There are many observational studies to show that this is true. Homoeopaths therefore keep telling us that their treatments work, regardless of the implausibility of homoeopathy's principles and the largely negative trial evidence.

When we rationally analyse this apparent contradiction of evidence versus experience, it quickly dissolves into thin air. The empathic encounter with a homoeopath is just one of many factors that provide ample explanation for the observation that patients can improve even when they receive placebos. A case in point is Bristol Homoeopathic Hospital's 2005 study, cited in the offending advert. The 6,500 chronically ill patients might have im­proved because of the concomitant use of conventional treatments, or because of the attention they experienced, or because of their own expectation to improve, or because the disease process had come to an end. In fact, they might have improved not because of, but despite, the homoeopathic remedies they were given.

Still, some people ask what is wrong with using placebos as long as they help patients feel better. The answer is that it prevents clinicians telling the truth to patients. Being honest would defeat any placebo effect: if I tell my patient, "Take this remedy; it contains nothing and the trial data shows nothing," she is unlikely to experience a placebo response. Hence, homoeopaths, knowingly or unknowingly, deprive patients of informed consent. This paternalistic approach is recognised as unethical. Also, placebo effects are unreliable and normally short-lived; they happen occasionally but often do not. Even if placebo responses are generated, they are usually small - certainly too small to compete with effective therapies.

Twin-track effect

Endorsing homoeopathic placebos would mean that people might use them for serious, treatable conditions. In such circumstances, homoeopathy can even cause (and has caused) the death of patients. Furthermore, if we allow the homoeopathic industry to sell placebos, we must do the same for "Big Pharma". Imagine a world where pharmaceutical companies could sell us placebos for all sorts of conditions just because some patients experience benefits through a placebo response.

Crucially, and paradoxically, we don't need placebos to generate placebo effects. If I, for instance, prescribe an antihistamine for a patient suffering from hay fever, with empathy, time and understanding, that patient benefits from a placebo effect as well as the pharmacological action of the antihistamine. If, by contrast, I prescribe a homoeopathic remedy, I deprive her of the latter, crucial benefit. It is difficult to argue, as most homoeopaths try to, that this approach would be in the interest of my patient.

What follows is straightforward: there is no good evidence that homoeopathy does more good than harm. This is not just my conclusion after 17 years of researching the subject, but a fact based on the best available evidence, which is supported by virtually all experts who are not homoeopaths. The recent decision by the coalition government to continue homoeopathy on the NHS is thus puzzling, to say the least.
The advertisement that prompted this article is misleading about the work of experts which has conclusively shown that homoeopathy can have no place in evidence-based medicine. It is an insult to our intelligence.

Edzard Ernst is professor of complementary medicine at the Peninsula Medical School, University of Exeter, and co-author, with Simon Singh, of "Trick or Treatment? Alternative Medicine on Trial" (Corgi, £8.99)

Here comes the non-science

Homoeopathy was developed in 1796 by the German physician Samuel Hahnemann. He based his treatments on the twin ideas that "like cures like" and "less is more". The latter notion was implemented by taking a substance and diluting it over and over again, so that the final product generally contains not a single molecule of the original active ingredient.

Homoeopaths accept that most of their remedies are devoid of pharmacologically active principles, but they argue that the pills contain a "memory" of the original ingredient. The memory is supposedly imprinted in the diluting agent, which is used to moisten sugar pills.

Although homoeopathy defies the laws of physics, chemistry, biology and therapeutics, there have been numerous attempts to test its impact on patients through clinical trials. In 2005, Aijing Shang and seven colleagues from the University of Berne published an analysis of the best trials in the Lancet.

Their findings confirmed many other such published assessments. Commenting on the paper, they wrote: "This finding is compatible with the notion that the clinical effects of homoeopathy are placebo effects." An accompanying editorial entitled "The end of homoeopathy" said: "Doctors need to be bold and honest with their patients about homoeopathy's lack of benefit."

This article first appeared in the 08 November 2010 issue of the New Statesman, Israel divided

Show Hide image

Why don’t we learn from our mistakes – even when it matters most?

Juan Rivera served ten years of prison time until DNA evidence overturned his sentence. But even now, some maintain his guilt.

On the afternoon of 17 August 1992, an 11-year-old girl called Holly Staker walked from her home to a neighbour’s apartment in Waukegan, a small town in Illinois. She had been asked to babysit two children, a girl aged two and a boy of five. By 8pm Holly was dead. An unidentified intruder had broken into the apartment, violently raped her, and stabbed her 27 times. The local police force pursued 600 leads and interviewed 200 people, but within a few weeks the trail had run cold.

Then, through the testimony of a jailhouse informant, police happened upon a new suspect: Juan Rivera, a 19-year-old man who lived a few miles south of the murder scene. Over four days, Rivera, who had a history of psychological problems, was subjected to a gruelling examination by the Lake County Task Force. At one point it seemed to get too much for Rivera. Officers saw him pulling out a clump of hair and banging his head on the wall.

On the third day, as the interviewers’ tone became aggressive, Rivera finally nodded his head when asked if he had committed the crime. By this time, his hands and feet were tied together, and he was confined to a padded cell. On the basis of his confession, police prepared a statement for Rivera to sign.

There were no witnesses to the crime. There was no physical evidence. But there was a brutally murdered young girl, a community still in mourning, and that signed confession. The jury didn’t take long to make up their minds. Rivera was convicted of first-degree murder and sentenced to life in prison.

Many journalists who covered the case were uneasy. They could see that the case hinged on the confession of a disturbed young man, who had retracted hours after signing it. But the police and prosecutors felt vindicated. It had been a troubling crime. A man had been sentenced. Holly’s family could find closure.

Or could they?


Eight years earlier, in 1984, Alec Jeffreys, a British scientist, had the insight that would lead to DNA fingerprinting – a technique that would transform the criminal justice system. In the absence of contamination, and provided the test is administered correctly, the odds of two unrelated people having matching DNA is roughly one in a billion. In certain cases, such as rape, it would in time be possible to identify the attacker conclusively, based on DNA extracted from the crime scene. After all, if the police swabbed the sperm found in the victim, they could narrow the number of potential suspects to just one. This is why DNA fingerprinting has helped to secure thousands of convictions – it has a unique power in establishing guilt.

But it also has implications for cases that have already been tried: the power to exonerate. If the DNA from the sperm in a rape victim has been stored, and if it does not match that of the person serving time in prison, the conclusion is difficult to deny: it came from a different man, the real criminal.

In August 1989, just months after DNA fingerprinting technology became advanced enough for use in the criminal justice system, Gary Dotson, who had been convicted of a 1977 rape in Chicago, was released from jail, having consistently proclaimed his innocence. Underwear worn by the victim had been sent for DNA testing, which showed that the semen belonged to a different man. Dotson had served more than ten years in jail.

The first DNA exoneration in the UK involved Michael Shirley, a young sailor who had been convicted of the rape and murder of Linda Cook, a barmaid working in Portsmouth, in December 1986. A simple DNA test performed in 2001, after police finally admitted they had stored swabs from the crime, proved that the semen found in the victim did not belong to Shirley. He had served 16 years at the time of his release.

By 2005, in the United States alone, the convictions of more than 300 people had been overturned following DNA tests. It seemed that the criminal justice system was getting it wrong repeatedly. In situations where evidence had been kept, clients of the Innocence Project, an American charity that helps prisoners who say they were wrongfully convicted, were exonerated in almost half the cases. A study led by Samuel R Gross, a professor at the University of Michigan Law School, concluded: “If we reviewed prison sentences with the same level of care that we devote to death sentences, there would have been over 28,500 non-death-row exonerations [in the US] in the past 15 years rather than the 255 that have in fact occurred.”

At that point, Juan Rivera, the man convicted of killing Holly Staker, had been in jail for almost 13 years. His lawyers decided to apply for a DNA test. It hadn’t been performed at the trial because techniques were not, at that time, sufficiently advanced to analyse very small samples of tissue. When the results came back they were conclusive: the semen found inside Holly Staker did not belong to Rivera.

He was yet another victim of wrongful conviction.


When a system produces an error, it is a tragedy for the person on the wrong side of the mistake. But it is also a precious learning opportunity, a chance to figure out what went wrong, and make reforms to the system to ensure that the same mistakes don’t happen again.

Aviation is a powerful model in this respect. Every plane is equipped with two black boxes, which record vital information. If an accident occurs, these boxes are recovered, the failure analysed, and reforms implemented. Pilots report near-miss events, too, which are statistically analysed in order to help prevent accidents.

In 1912, eight of the 14 qualified US army pilots died in crashes. In 2014, the accident rate for major airlines had dropped to just one crash for every 8.3 million take-offs. That’s the power of learning from mistakes.

Broadly speaking, this is how science works. Scientists make testable predictions; when these fail, the theories are reformed. As the philosopher Karl Popper put it: “The history of science . . . is a history . . . of error. But science is one of the very few human activities – perhaps the only one – in which errors are systematically criticised and fairly often, in time, corrected. This is why . . . we can speak clearly and sensibly about making progress there.”

But there is a practical problem with adopting this model of progress. Many people don’t like to admit to their mistakes. They do not meet their failures with intellectual honesty and a willingness to understand what went wrong, but with back covering and defensiveness. This has a simple consequence: if mistakes are not learned from, they will be made again and again. And again.

In his famous studies of “cognitive dissonance” in the 1950s, the sociologist Leon Festinger discovered the array of tactics used by people to spin, reframe and cover up their mistakes. A classic example involved a small group of cult members. They had left their families to live with a housewife called “Marian Keech” (real name: Dorothy Martin), who had predicted the world would end before dawn on 21 December 1954. She had also prophesied that the members of the cult would be saved from the looming apocalypse by a spaceship that would land at Keech’s small house in suburban Minnesota at midnight.

Neither the apocalypse nor the spaceship materialised (Keech’s husband, who was a non-believer, had gone to his bedroom and slept through the whole thing). This was an unambiguous failure. Keech had said the world would end, and that a spaceship would save true believers. Neither had happened. The cult members could have responded by altering their beliefs about Keech’s supernatural insights.

Instead, they altered the “evidence”. At first, the cult members kept checking outside to see if the spaceship had landed. Then, as the clock ticked past midnight, they became sullen and bemused. Ultimately, however, they became defiant. Just as Festinger had predicted before he infiltrated the cult, the faith of hardcore members was unaffected by what should have been a crushing disappointment. In fact, their faith seemed to grow stronger.

How is this possible? As Festinger recounted in his book When Prophecy Fails, first published in 1956, they simply redefined the failure. “The godlike figure is so impressed with our faith that he has decided to give the planet a second chance,” they proclaimed (I am paraphrasing only a little). “We saved the world!” Far from abandoning the cult, members went out on a recruitment drive. As Festinger put it: “The little group, sitting all night long, had spread so much light that God had saved the world from destruction.” They were “jubilant”.

This may sound bizarre, but it is, in fact, commonplace. Festinger showed that this behaviour, though seemingly extreme, provides an insight into psychological mechanisms that are universal. When we are confronted with evidence that challenges our deeply held beliefs we are more likely to reframe the evidence than we are to alter our beliefs. We simply invent new reasons, new justifications, new explanations. Sometimes we ignore the evidence altogether.

Think, for example, of health care. In her book After Harm, published in 2005, the American researcher Nancy Berlinger investigated how doctors reframe their mistakes. “Observing more senior physicians, students learn that their mentors and supervisors believe in, practise and reward the concealment of errors,” she wrote. “They learn how to talk about unanticipated outcomes until a ‘mistake’ morphs into a ‘complication’. Above all, they learn not to tell the patient anything.”

This is partly about fear of litigation, but also about the protecting ego. Think of the language associated with senior physicians. Surgeons work in a “theatre”. This is the stage where they “perform”. How dare they fluff their lines? As the safety expert James Reason put it: “After a long education, you are expected to get it right. The consequence is that medical errors are marginalised and stigmatised.”

But if doctors do not learn from their mistakes, they are destined to repeat them. A report by the UK National Audit Office in 2005 estimated that up to 34,000 people are killed each year as a result of human error in the health-care system. It put the overall number of patient incidents (fatal and non-fatal) at 974,000. In the US, annually, 400,000 people die because of preventable error in hospitals alone. That is the equivalent of two jumbo jets crashing every day.

Another example is economics. In 2010, a group of influential thinkers, including Michael J Boskin, a former chairman of the US president’s Council of Economic Advisers, and the historian Niall Ferguson, wrote an open letter to the chairman of the Federal Reserve. The signatories were worried that a second tranche of quantitative easing would “risk currency debasement and inflation” and “distort financial markets”. The letter was published in the Wall Street Journal and New York Times.

At the time, inflation was running at 1.5 per cent, but what happened in the aftermath? Did inflation soar? In fact, by December 2014, inflation had not merely stayed at historically low levels, but had fallen to 0.8 per cent (soon after that, it fell into negative territory). The US economy was creating jobs at a fast rate, too, while unemployment had dropped and companies were reporting record profits.

This was a stark error in prediction, an opportunity to revise theoretical assumptions (just as mistakes in medicine provide an opportunity to revise procedures). But the signatories saw it differently. David Malpass, a deputy assistant secretary to the Treasury under Ronald Reagan, said “the letter was correct as stated”. John Taylor, an economics professor, said: “The letter mentioned many things – the risk of inflation . . . to employment . . . – and all have happened.”

Again, like revered clinicians, these prestigious thinkers could not bear that they had got things wrong. So, instead of learning from their mistake, they spun it. Some of the signatories came up with detailed graphs and tables to show that they were right all along.

Consider that if inflation had soared, they would have claimed this as a vindication for their prediction (as the cult members would have said the apocalypse had materialised). Yet these eminent thinkers also claimed vindication when inflation fell (just like the cult members when the spaceship didn’t turn up). Heads I win; tails I don’t lose.

And this is why it is the social pundits with the most prestigious reputations – as measured by how often they tour the television studios – who often make the worst predictions. They have the most to lose from mistakes and are therefore the most likely to come up with ex post explanations for why they were right all along. And because these pundits are clever, their rationalisations sound plausible.

But it is also why they don’t learn from their mistakes. After all, if you can’t admit when you are wrong, how can you ever put things right?


This brings us back to the criminal justice system. Many of these DNA cases show unequivocal failures. They demonstrate almost beyond doubt that mistakes were made by the police, prosecution and courts. This should have led to an acknowledgement of error, and meaningful reform. In fact, the astonishing thing is not that the system – or at least the people behind it – learned, but the extent to which it continued to evade and deny.

When the DNA results came back for Juan Rivera, state prosecutors offered a new story to account for the evidence – a very different story from the one that they had presented at the original trial. Holly, an 11-year-old child, had had consensual sex with a lover a few hours before the attack, prosecutors claimed. This accounted for the semen. And Rivera? He had happened upon Holly after intercourse had taken place, and murdered her.

“It was a grotesque way of squaring the new evidence with their unshakeable belief that Rivera was guilty,” Steven Art, one of the jailed man’s lawyers, told me. “But it was also totally inconsistent with the overwhelming evidence that Holly had been raped, quite brutally. There were signs of vaginal and anal trauma and stab wounds in her genitals.”

The prosecutor’s new story may have seemed outlandish, but the consequences were very real. Rivera was not released from prison for another six years. “I got stabbed twice and endured three attempted rapes,” he told me during a phone interview last year. “People wanted to hurt me; they thought that I was a child rapist. But perhaps the toughest thing of all was knowing that I was innocent.”

Rivera’s experience was far from unusual. The vast majority of exonerations through DNA evidence were met with breathtaking obstruction. Nothing seemed to budge prosecutors from their conviction that the man who had been sent to prison was guilty. Even after the test had been performed. Even after the conviction had been overturned. Even after the prisoner had been released from jail.

The problem was not the strength of the evidence, which was often overwhelming; it was the psychological difficulty in accepting it.

The process often took a distinctive path. First the prosecutors would try to deny access to DNA evidence. When that strategy was batted away, and the test excluded the convict as the source of the DNA, they would claim that it had not been carried out correctly. This didn’t last long, either, because when the test was redone it would invariably come back with the same result. The next stage in rape cases was for the prosecutor to argue that the semen belonged to a different man who was not the murderer. In other words, the victim had consensual sex with another man, but was subsequently raped by the prisoner, who used a condom.

The presence of an entirely new man, not mentioned at the initial trial, for whom there were no eyewitnesses, and who the victim often couldn’t remember having sex with, may seem like a desperate ploy to evade the evidence. But it has been used so often that it has been given a name by defence lawyers: “the unindicted co-ejaculator”.

In an adversarial system you would expect any new evidence secured by the defence to be looked at with healthy scepticism by prosecutors. You would expect them to give it scrutiny and to look at the wider context to be sure it stacks up. But in case after case, the sense of denial from prosecutors and police went a lot further.

And this explains, in turn, why the system was making so many mistakes. Proper investigation into wrongful convictions by the Innocence Project in recent years has discovered profound weaknesses in forensic science, police detection methods, eyewitness testimony and more. If these investigations had taken place earlier, and reforms been introduced, untold suffering could have been averted. But these opportunities were missed repeatedly, on both sides of the Atlantic.

This is why legal campaigners argue that the most robust way to improve the system, so as to reduce wrongful convictions and simultaneously increase rightful convictions, is to establish innocence commissions. These are independent bodies, mandated to investigate wrongful convictions and to recommend reforms, along the lines of air accident investigation teams. As of publication, only 11 states in the US had such commissions.

In the UK, a reform commission of sorts was set up in 1995 after a series of spectacular miscarriages of justice, including the Birmingham Six and the Guildford Four. The Criminal Cases Review Commission, an independent body, has the authority to refer questionable verdicts to the Court of Appeal. Between 1997 and the end of October 2013 it referred 538 cases in total.

Of these, 70 per cent succeeded at appeal.


More than 20 years after Juan Rivera was sentenced, a DNA test was conducted on a bloodstained piece of timber that had been used in a different murder. In January 2000, a man called Delwin Foxworth, who also lived in Lake County, was beaten savagely with a plank, doused with gasoline and set on fire. He died of his injuries, having suffered burns over 80 per cent of his body.

The DNA of the blood found on the wood matched that of the semen found in Holly Staker. Police are now almost certain that the man who got away with the rape and murder of an 11-year-old girl in 1992 went on to commit another murder eight years later. Foxworth may be yet another victim of the wrongful conviction of Juan Rivera; it allowed the real culprit to get away with it, and kill again. (The murderer has never been found.)

But what about those who were responsible for sending Rivera to jail? How do they feel about it today? Perhaps it should come as no surprise that even now many remain convinced of his guilt. In October 2014, Charles Fagan, an investigator who helped obtain Rivera’s confessions, was asked if he still believed that he committed the murder. “I think so,” he told a local newspaper.

And what of the prosecutors? Even after Rivera was released, Lake County lawyers wanted to put him back on trial. Only with a further conviction would they be able to say that they had been right all along. Only with a conviction could they quell their dissonance. Rivera walking around free was like an accusation against their competence.

It was left to the Illinois Appellate Court to take what might otherwise seem to be an astonishing step: it barred Lake County from ever prosecuting Juan Rivera for the murder of Holly Staker again.

Matthew Syed’s latest book is “Black Box Thinking: the Surprising Truth About Success” (John Murray)

This article first appeared in the 12 November 2015 issue of the New Statesman, Isis and the threat to Britain