Interpreting the evidence: Balancing truth, facts, values and interests
- Paul Komesaroff
- Early release
- 1 September 2013
Decisions taken in scientific and clinical settings are subject to complex influences. The ethos of science as a disinterested pursuit of objective truth is at best imperfectly realised in practice. As a starting point, clinicians are not scientists; there is always a major contribution of clinical judgement to all health encounters. Ethical and social values, and personal, context-dependent characteristics are all taken into account in any interaction.
Two examples will be used to illustrate these points – the case of Galileo and that of hormone ‘replacement’ therapy (HRT) after menopause.
The concept of evidence has been discussed widely. It can be taken for granted that there are two levels of evidence, which are associated with very different logical processes:
Both approaches are associated with a mix of facts and interpretations. Local, individual clinical decision making is where actual decisions are made and actions are taken in consultation with the patient.
There are separate activities and concepts of knowledge with different standards for ‘truth’. Clinical decision relies on two sets of transitions to make sense of (or convert) the data that are relevant. This is complicated. The clinician must move between the data and their interpretations, and between the level of scientific evidence and the local, individual decisions that are made every day in the clinic.
Clearly, clinical decisions are influenced by the interpretation of the data collected in the controlled setting of scientific experimentation. These interpretations are not, as they are often presented to be, objective, disinterested, a pure and simple reflection of the facts, unbiased or unaffected by personal or local cultural beliefs or prejudices. Rather, the process of scientific interpretation is always deeply affected by such factors. Scientific facts always exist in a social context and their validity depends on this context.
Galileo Galilei lived from 1564 to 1642. A great Italian scientist, he was the originator of modern physics. Many scientists believe that Galileo's work was plainly ‘true’, and that through the power of his intellect and logic he was able to supplant the erstwhile Aristotelian view of the world with a new, mathematical one. The view of succeeding generations was to seek objective truth, but this is not how it happened.
Galileo was a great innovator and truly one of the great imaginative heroes of human history. However, his theories came to be accepted not primarily on the basis of the strength of his thought and observation, but on abstraction, a new language, the question of empirical validation. Galileo published a number of articles in the early 17th century strongly supporting the Copernican theory that the earth circled the sun (heliocentrism). These ideas were seen by church authorities as profoundly disruptive of the existing intellectual status quo and the church's teachings – and they therefore had a wide set of social implications. From the time Galileo wrote The Starry Messenger in 1610 the church took steps to shut him up and he was subject to harassment from the authorities (especially the ecclesiastical authorities) for the rest of his life. The authorities understood that the new ideas Galileo was writing about were dangerous, that his propositions were mere representation of objective facts, but that they carried profound, potentially uncontainable social implications for political authority.
In 1633 Galileo was tried before a formal inquisition for promoting the views of heliocentrism. He was found guilty, placed under house arrest for the rest of his life, and made to recant. They showed him the instruments of torture and told him to imagine what would happen if he did not recant. They knew their man and this was all they needed to do. Galileo recanted, even though he continued writing in secret and was able to arrange for his manuscripts to be smuggled to safety, where they were subsequently published.
Today's school students are taught that the strength of Galileo's theory lay purely in the fact that, compared to the Aristotelian orthodoxy, it carried greater empirical validity and was more closely in accord with objective reality. In reality, the opposite was the case. Galileo's theory – at least in its early stages – was actually empirically inferior to the Aristotelian system. Everyone knows, after all, that the first law of motion – that a body in motion tends to remain in motion – is never borne out in fact. Rather, the Aristotelian idea of inertia – that if you push a body it will stop – is what we observe on earth in every real instance. In reality, Galileo's theory was not empirically valid. It was abstract and contrary to the facts. Indeed, this was its great innovation. Galileo discovered not the concept of empirical science, but a new idealised language, the language of mathematics, in which the empirical questions could be formulated. The success of the theory was dependent not on its ‘objective’ validity, but on other factors related to knowledge, philosophy, society and politics.
Galileo was attacked, but he fought back in his writings. He took the struggle to the streets, writing popular works, not just in Latin, but also in Italian, the language of the common people. In a book on the arguments for and against heliocentrism (Dialogue Concerning the Two Chief World Systems) he sought to obtain the support not of the effete intelligentsia, but of the common people. There are two main characters in the book: Sagredo, who is Galileo himself, and a second person called Simplicius in Latin or Simplicio in Italian, who represents the Aristotelian view (that the sun revolved around the earth) and speaks in phrases used by the Pope. The name Simplicio in Italian has the connotation of a simpleton, and the portrayal of this character served to cast greater doubt on the established ideas amongst the local readers of the book and ensured readers were on Galileo's side.
Galileo's struggles to get his ideas accepted show how science, culture and society are intertwined. They draw attention to the fact that the success of a scientific theory – its acceptance or rejection, its capacity to generate approval or hostility – is not determined merely by its philosophical or empirical value, but also crucially by the complex social and political frame within which it is embedded.
Scientific meanings are subject to and dependent on social attitudes, prejudices and belief systems. People can find ways to convince themselves that a particular view is true depending on their personal attitudes. It is not the case that the process of understanding science, data and the world, is detached and objective. It is always rooted in local vested interests. Science and medicine cannot be separated from the contemporary cultural debates. This is the essence of what Galileo discovered. The success of a theory depends on social struggles, which can be ruthless and even violent. In Galileo’s case, he fought hard and, despite the setbacks, was eventually victorious, even if in his case the full victory came only posthumously.
In the 1930s scientific studies showed that administering oestrogen to mice could promote the development of breast cancer. Despite this, there was a strong interest in promoting oestrogens for women, starting in the 1960s. The American psychologist Robert Wilson initiated the modern emphasis on the use of hormone therapy in his book Feminine Forever. In today's terms this book is misogynistic, but at the time it was very popular, selling more than 100 000 copies, with the claim that menopause was ‘preventable’:
Every woman alive today has the option of remaining feminine forever… No longer need she fret about the cruel irony of women aging faster than men. It is simply no longer true that the sexuality of a woman past forty necessarily declines more rapidly than that of her husband.
All postmenopausal women are castrates… [but with HRT] a woman’s breasts and genital organs will not shrivel. She will be much more pleasant to live with and will not become dull and unattractive.
An oestrogen-rich woman capable of being physically and emotionally fulfilled by her husband … is least likely to go afield in search of casual encounters.
Robert Wilson, Feminine Forever (1966)
Over the succeeding decades a great deal of scientific evidence has emerged about the actions of oestrogens and their effects on bones, the cardiovascular system, the neurological system, breast and uterus. We now know a lot about the way these hormones work.
During the 1990s HRT became widely accepted as a key treatment for menopause. The number of prescriptions for oestrogen-based therapies skyrocketed in the industrialised world. Guidelines issued by menopause societies in Australia, North America and Europe actively promoted HRT.
June 1990: ‘The cardiovascular benefits of Premarin may outweigh the risks depending on the individual patient’s risk profile for various oestrogen-related diseases and conditions.’1
1991: ‘Epidemiological evidence is accumulating that postmenopausal oestrogen therapy reduces the risk of cardiovascular disease and stroke by between 30 and 70% even in the presence of established risk factors.’2
1992: ‘Oestrogen is cardioprotective for women. Not only does it have a beneficial effect on the circulating blood lipid fractions, but it is now established that oestrogen has a positive influence in preventing the deposit of cholesterol in the arterial endothelium. It also induces vasodilation, increases peripheral blood flow and leads to a fall in blood pressure. The use of oestrogen to reduce cardiovascular disease far outweighs any potential adverse changes.’3
1992: ‘All women…should consider preventive hormone therapy.’ 10 to 20 years of therapy were recommended for ‘maximum benefit’.4
1993: ‘All postmenopausal women, barring a medical contraindication like breast cancer, should take HRT for life.’5
1994: ‘HRT can reduce the incidence of CHD in postmenopausal women by 50%… HRT reverses the increased fat distribution that results from loss of ovarian function at the menopause. HRT may also (result) in a reduction in arterial thrombosis.’6
1996: ‘There is now good population-based evidence that HRT in postmenopausal women reduces the incidence of CHD, perhaps by up to 50%. HRT should therefore now be considered for use in postmenopausal women with established CHD risk.’7
2000: ‘Today, HRT may be used for the primary prevention of CVD. (T)here is no compelling evidence for discontinuing – or indeed not initiating – HRT in women without CVD because of concern about cardiovascular risk.’8
The comfortable consensus was exploded in 2002 with the publication of the first results of the Women's Health Initiative (WHI). The WHI was, in its time, the largest and most expensive clinical trial ever, involving a total of 160 000 participants and running for more than 10 years. It began in the early 1990s as a response to rising discomfort about the lack of clinical data and concern that some of the benefits of HRT were exaggerated. The WHI evoked considerable opposition when it started, with some advocates for hormone therapies even arguing that it was ‘unethical’ to continue testing these treatments because so much was already known about them and the agreement about their effectiveness and safety was so strong.
The WHI hormone therapy study had two arms – an oestrogen plus progestogen arm that ceased in July 2002, and an oestrogen-only arm that ceased in February 2004. The oestrogen plus progestogen arm was stopped because the trial showed an increased risk of invasive breast cancer and an increase in coronary heart disease, stroke and pulmonary embolism in the study participants. Two years later the oestrogen only arm was also stopped because it increased the risk of stroke, decreased the risk of hip fracture and did not affect coronary heart disease incidence. By this time several other large trials had been published that suggested the assumptions about the benefits and risks of hormone therapy may have been misplaced.
The conclusions of the WHI generated a remarkable situation. Guidelines produced by experts had been widely propagated supporting the use of hormone therapies. However, now data were available that called into question the basic assumptions on which those guidelines had been founded. The experts, of course, were scientists so one would expect that they would respond to the new knowledge by revising the guidelines to incorporate the current insights and introduce appropriate notes of uncertainty and caution. Correct? Unfortunately, no. This did not happen. The guideline developers mostly did not change their attitudes. Instead, they mounted a vigorous campaign – which even continues today – to attack and refute the study conclusions.
To avoid misunderstanding, it is important to emphasise that debates about the validity of clinical studies and the interpretation of their results are part of the legitimate discourse of science. However, in this case what is remarkable is that despite the significant size and rigour of the WHI trial and the clarity of its outcomes, and the subsequent accumulation of a great deal of supporting data, many of the committed advocates of hormone therapy refused to alter their positions to any significant degree. Instead their response was not to question their own prior assumptions, but to find arguments to support why the WHI had to be wrong. For example, it was argued that the wrong population was used (e.g. the study participants were too old, only North American, and included smokers), that the wrong hormonal preparation was administered, that the interpretation of the data was wrong, and that the wrong methods were used in the design of the trial.
In a 10-year review of the WHI results published in 2012, the original view about breast cancer in the oestrogen plus progestogen arm was validated by the long-term follow-up. In the oestrogen-only arm the breast cancer outcomes are reported as equivocal, the cardiovascular outcomes were negative, and oestrogen was shown to be harmful in Alzheimer's disease prevention.
However, even now, few of the original proponents of hormone therapy have changed their minds. Instead, they promote arguments not based on evidence, or based on studies that provide limited (often flawed) data that appear to support their case. For example, a recent Danish study9 (which included 1000 people with questionable recruitment strategies and an idiosyncratic end point) suggested possible cardiovascular benefits in a particular population and was widely promoted as refuting the results of the entire WHI program. At the same time, the popular media – the same media that actively supported and encouraged the use of HRT – have continued to cast doubt on the conclusions of the WHI and a large number of other studies that now support its broad conclusions, and to promote new studies of lower quality or study design in order to demonstrate to the public that there is positive news for the use of HRT. Headlines abound, such as ‘Flaws in Major US Study on HRT’, ‘HRT gets another chance’, ‘Doctors to rethink benefits of HRT after study shows that oestrogen may protect women against breast cancer’, ‘Call for NIH to revise recommendations on HRT’, ‘A wasted decade: how one HRT scare has caused countless women ten years of needless suffering’ and ‘Expert calls for inquiry into the NIH’.
I do not want to suggest that the evidence, one way or the other, about hormone therapy is clear-cut. On the contrary, there is much that remains uncertain and there are key issues that await clarification from future scientific studies, involving both laboratory experiments and clinical studies. However, the point I want to make is that the original uncritical support for hormone therapy was not based on evidence, and the appearance of new evidence was insufficient in many cases to change this support. Many of the most zealous advocates of the therapy have retained their commitment, even in the face of considerable – if not completely conclusive – data to the contrary.
I hope that the parallel with the case of Galileo is clear. Even today, the informed opinion of experts and the debates around contending theories are not subject purely to the careful, dispassionate and disinterested assessment of the facts. Rather, they are dependent on other factors operating outside of scientific discourse. They are subject to the influences of culture and ideology, of personal beliefs and reputation, of the vested interests of the pharmaceutical industry and medical practitioners, of prevailing social prejudices, and of politics and power. Of course, these arguments apply to both sides. The debates about hormone therapy after menopause – just like the debates about the two world systems four hundred years earlier – are not fought out merely in the domain of science.
These are culture wars. The battle, as in the case of Galileo, has been – and is still being – contested not just in the halls of academia, but also in the streets, in the newspapers, on television, through the internet. Every new finding is accompanied by a press release, often tendentious.
As with Galileo, with hormone therapy there are many interests and influences at play. The field is not one of pure science in the idealised sense. The ‘ethos’ of science is imperfectly realised in practice. Despite the proclaimed commitment to evidence and an open scepticism, the reality is very different.
This is the sad and sobering moral of my story, but it should not be a cause for despair. Rather, those who work (through guidelines or other means) to improve clinical practice and to effect changes in policy merely need to recognise the complexity of the environment within which they are operating and to add the factors I have been discussing as additional variables to be taken into account. There is no pure science. There is no simple, incontestable objective truth, the facts never speak for themselves. Rather, we always have to make judgments from interest-laden points of view. Whether we like it or not, the process of responsible decision making – ours and everyone else’s – always involves the establishment of a careful balance between truth, facts, values and interests.
Professor Komesaroff provided a detailed declaration of his interests (see Appendix 2 for details)
Professor of Medicine and Director of the Centre for Ethics in Medicine and Society, Monash University, Melbourne