Category Archives: Points of view

EBM+, plus-epidemiology, plus-prediction

Leave a reply

Recently, I have been invited to take part in a workshop on Prediction in Epidemiology and Healthcare. The event has been organised by Jonanthan Fuller and Luis Flores at King’s College London on 20 June 2014.

The topic – prediction in epidemiology and healthcare – is well in the remit of EBM+. Actually, it is precisely one of the things philosophy of medicine, broadly construed, should be looking at. As the various talks made clear, predicting what will happen is not merely or solely solved using some sophisticated statistical models. It requires a lot of work in conceptualising the objects of prediction as well as its tools. More to the point, predictions go well beyond the problem of establishing the efficacy of treatments. Prediction likewise concerns prognosis, health policy and public health. Also, prediction concerns individuals as well as populations. So here lies an interesting connection between phil med and more classic topics in phil sci.

A dynamic and engaged group of speakers took part in the event: Jonathan Fuller (University of Toronto) and Luis Flores (King’s College London), Alex Broadbent (University of Johannesburg), Jacob Stegenga (University of Utah), Elselijn Kingma (University of Southampton), Barbara Osimani (University of Camerino), Maël Lemoine (University of Tours), Jeremy Howick (Oxford University). Many other people attended the event, such as our EBM+ bloggers Brendan Clarke and Phyllis Illari, Donald Gillies, Margherita Benzi (currently visiting UCL), Brian Hurwitz, David Papineau and many more.

I presented a paper titled “The integration of social and biological mechanisms for healthcare prediction and intervention”. This is work in progress with Mike Kelly, director of NICE, and is a follow up of some previous work on the integration of social, behavioural and biological mechanisms in models of pathogenesis. The idea of the paper is to compare models used for explaining pathogenesis and models to set up interventions. Our point is that the ‘pathogenic model’ has been by and large successful for communicable diseases, as it requires to identify the cause of the disease and to act on it to reduce exposure.

first

However this model is not as successful in explaining non-communicable diseases, as it neglects mechanisms related to human behaviour. The pathogenic model is even less successful in setting up public health interventions. The reason is that for non-communicable disease (but often for communicable diseases as well) we cannot act on the exposure directly. We instead need to take into account the complex bio-psycho-social mechanisms and we may have to act on indirect factors.

second

An important point of the paper is precisely the discussion of the integration of the ‘social’ into the biological dimension of disease explanation and prediction.

If you are interested in, here is the presentation:

 

Why association is only half the story

I want to develop a point from Jon’s earlier post. A central theme of this project is that association (a correlation found in a drug trial, for example) is only half the story about causation. As Jon mentioned, there are many reasons that an observed correlation might be non-causal (like sampling errors, confounding, and so on). Here, I want to explore a case where a non-causal correlations was taken as sufficient reason for accepting a causal claim.

Cervical cancer

Cervical cancer is caused by infection with human papillomavirus (HPV). This claim was first made in the early 1980s by Harald zur Hausen, a German virologist. You can have a look at the original paper (Durst et al, 1983), as well as some information about the half share of the 2008 Nobel prize in Physiology or Medicine which he won for this work. When I started studying medicine in the late 1990s, the causal link between HPV and cervical cancer was common knowledge. So when I began researching the history of cervical cancer for my PhD (which you can read online if you’re keen on that kind of thing), it was a shock to discover that HPV was not the only virus that had been associated with cervical cancer.

Between about 1970 and 1985, herpes simplex virus (HSV) was generally accepted as the cause of cervical cancer. For example, you can peruse the forty or so papers that make up the proceedings of the 1972 American Cancer Society conference ‘Herpesvirus and cervical cancer’ in Cancer Research, which demonstrate the existence of a thriving research program on HSV and cervical cancer. I’ll discuss the significance of this below, but for now I want to introduce the question that first bothered me when I started this research: why did anyone think that HSV might cause cervical cancer?

HSV and cervical cancer?

The roots of the claim that HSV might cause cervical cancer came from some observed correlations between certain sexual behaviours and the risk of developing the disease. In fact, cervical cancer has long been noted to behave more like an infectious disease than a typical cancer. Perhaps the most interesting series of observations of this kind was produced by Rigoni-Stern in 1842 (available in English translation as Stavola, 1987), which described a series of cases in Verona (1760-1839) that showed much higher rates of cervical cancer in married women than in nuns. One possible explanation for this difference was the celibacy practised by the nuns. Other studies during the nineteenth and early twentieth centuries found that other behaviours related to sex also seemed to modify the chance of developing cervical cancer. In general, the more sex an individual had had, the greater their risk of getting cervical cancer. So being married, having sex in adolescence, contracting other sexually transmitted infections (like syphilis) and having a large number of children positively correlated with the disease, while abstinence from sex negatively correlated with the disease.

By the time that mass population screening for cervical cancer was introduced in the mid-twentieth century, these sexual risk factors had been extensively researched. One great quote from the Aberdeenshire cervical cancer research project sums up the thinking typical at the time:

The cancer patient is characterised by more marital misadventures, divorce and separation, more pre-marital coitus and deliveries and more sexual partners. (Aitken-Swan and Baird, 1966: 656)

So perhaps cervical cancer was a consequence of a sexually transmitted disease. While the usual suspects (syphilis and the like) did not seem to account for it, research in different contexts suggested that herpes viruses might cause many kinds of cancer. The details of this are rather complicated (and probably something for another post), but the upshot was that (in the mid-twentieth century) herpes viruses seemed the most likely suspects as causes of cancer in humans. Happily for researchers at the time, this seemed to provide a causal explanation for the correlation between (sexually transmitted) HSV and cervical cancer (see, for example, Kessler, 1976).

Not much of a mechanism

So infection with HSV was an attractive explanation for these sexual risk factors. But was it also the cause of cervical cancer? Well, the lack of correlation between other sexually transmitted infections with cancer of the cervix suggested that correlation wasn’t just an accident, but was instead due to a causal relationship (Rawls et al, 1973: 1482). Other evidence, like serology, the mutagenic power of HSV, the detection of fragments of HSV DNA in cervical cancer cells, and the causal role played in other tumours by herpes viruses, seemed to support this causal claim. Yet its details remained elusive. Most of the papers from the 1973 Cancer Research volume mentioned above tried, but failed, to detect some specific evidence of a the mechanism linking the virus with the disease. And the details of this mechanism remained elusive, as we might expect. Yet the claim that HSV caused cervical cancer persisted well into the 1980s, and lead to significant resistance when other causal claims (like that involving HPV) were mooted. In conclusion, when combined with the plausibility of possible mechanisms involving HSV, the correlation between HSV infection and cervical cancer meant that it was unthinkable that HSV did not cause cervical cancer.

Conclusion

It’s pretty uncontroversial to say that we should distrust brute correlations, or mistake a correlation for a causal relation. But there are other, more subtle, issues that this case raises that I think we should be similarly mindful of. The first of these is the difference between plausible, and actual, mechanisms. HSV was linked to cervical cancer by an extremely plausible mechanism. But no actual mechanism was found. Good mechanisms in this context are specific and local: and we should be extremely cautious about mechanisms that are purely plausible. The boundaries here are pretty vague, though, and a future research goal for me is to try and come to grips with the difference between plausible and actual mechanisms.

The second issue that I’d like to raise by way of conclusion concerns being explicit about causal evidence. The HSV case is an example where, despite a great deal of research, no specific evidence mechanistically linking HSV and cervical cancer was found. However, this lack of evidence is not readily apparent from individual papers in the literature. Health researchers have recently adopted many strategies to more effectively review evidence of correlations (like meta-analysis and systematic reviews of trials). I imagine that a similar strategy for explicitly considering evidence of mechanism would have been valuable for HSV researchers as a way of detecting a persistent absence of evidence in the face of inquiry.

References

Aitken-Swan, J, and Baird, D. 1966. “Cancer of the Uterine Cervix in Aberdeenshire. Aetiological Aspects.” British Journal of Cancer, 20(4): 642–59.

Dürst, M, Gissmann, L, Ikenberg, H, and zur Hausen, H. 1983. A papillomavirus DNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. PNAS 80(12): 3812–3815.

Kessler, II. 1976. Human Cervical Cancer as a Venereal Disease. Cancer Research. 36: 783-91.

Rawls, WE, Adam, E, and Melnick, JL. 1973. “An Analysis of Seroepidemiological Studies of Herpesvirus Type 2 and Carcinoma of the Cervix.” Cancer Research, 33(6): 1477–82.

Stavola, BD, 1987. “Statistical Facts about Cancers on which Doctor Rigoni-Stern based his Contribution to the Surgeons’ Subgroup of the IV Congress of the Italian Scientists on 23 September 1842. (translation).” Statistics in Medicine, 6(8): 881–4.

Guest post by Andy Fugard: How we can be fooled into thinking a psychological therapy is effective when it’s not

If it’s tricky to decide whether pharmacological interventions are effective then it’s very tricky indeed to evaluate whether psychological therapies work. Although techniques from randomised controlled trials can help uncover causal effects, often important components such as double blinding are impossible – how could a practitioner follow a treatment manual without knowing what she or he is doing? Randomisation is not always possible, for example when evaluating routine practice (see Wolpert, Fugard, & Deighton, 2013 for examples). As Jon illustrated in a previous post, a story of mechanism is still necessary even with perfect measurement and methodology.

Perhaps it helps to have a theory of what else other than the therapy itself can cause apparent improvement in symptoms. Lilienfeld et al (2014) develop a taxonomy with 26 causes to begin to address this question. Let’s consider some examples. An intervention might provide palliative benefits; this can be understood by considering the difference between “feeling better” and “getting better”. Someone who wishes to stop excessive drinking might, thanks to a warm and understanding therapist, feel less guilty about drinking and so not change how much they drink. There may be a reduction in cognitive biases as a result of therapy; one’s perception of social skills might improve when in fact they have not changed. Both outcomes would be fine if the therapy were intended only to change perceptions, however, an intervention to reduce alcohol intake or improve social skills might be expected actually to show externally demonstrable changes. Another example given is the therapist’s office error: observed improvement in a safe and friendly therapeutic relationship might not transfer to the harsh world outside.

But what mechanisms are involved here and how should they be described? This is a much bigger question – and perhaps important for understanding therapeutic change. For example, might understanding the therapist office error lead to ideas for effective intervention? Power (2010) provides three broad examples of technique and related mechanisms. Firstly, exposure, whereby emotions are heightened in the therapist’s presence and change mechanisms involve behavioural extinction, relearning, and coping mechanisms. This is often an important component for helping reduce anxiety. Secondly, transference – a psychoanalytic concept (wait, don’t run off!) – whereby interaction with the therapist triggers “pre-existing expectations, templates, scripts, fears, and desires” (nicely explained by Shedler, 2006 p. 22) and, the idea is, they can be reworked in therapy. Thirdly, challenging assumptions, whereby emotions are heightened in relation to a person or situation, and reinterpreting or reconstructing the assumptions causes improvement.

Perhaps a next step in this interesting how-we’re-fooled-by-apparent-change programme might be to link the taxonomy to theories of mechanism like Power’s and others.

About the author

Andy Fugard is lecturer in the Educational Psychology Group at University College London. He is interested in practice-based evidence in mental health, the psychology of reasoning, and the broader autism spectrum.

References

Lilienfeld, S. O., Ritschel, L. A., Lynn, S. J., Cautin, R. L., & Latzman, R. D. (2014). Why Ineffective Psychotherapies Appear to Work: A Taxonomy of Causes of Spurious Therapeutic Effectiveness. Perspectives on Psychological Science, 9, 355–387. doi:10.1177/1745691614535216

Power, M. (2010). Emotion focussed cognitive therapy. Chichester: John Wiley & Sons.

Shedler, J. (2006). That was then this is now: an introduction to contemporary psychodynamic therapy. Retrieved from http://www.jonathanshedler.com/

Wolpert, M., Fugard, A., & Deighton, J. (2013). Issues in evaluation of psychotherapies. In P. Graham & S. Reynolds (Eds.), Cognitive Behaviour Therapy for Children and Families (3rd ed., pp. 34–47). Cambridge University Press.

 

What does philosophy offer evidence-based health care?

This was the theme of a symposium at Oxford yesterday, organised by Jeremy Howick. We covered a broad range of answers to this question, including: nothing (Ray Tallis), understanding the role of values (Elselijn Kingma), linguistic analysis (Bill Fulford), training for medical students (Alexander Bird).

My line of argument was that the recent mechanistic turn in philosophy of science has much to offer the evaluation of causal claims.

Since 2000, philosophers of science have devoted a lot of attention to mechanisms. As Machamer, Darden and Craver (2000: Philosophy of Science 67:1–25) point out:

In many fields of science what is taken to be a satisfactory explanation requires providing a description of a mechanism. So it is not surprising that much of the practice of science can be understood in terms of the discovery and description of mechanisms.

Philosophers have naturally been interested in such questions as: What are mechanisms? What are mechanistic explanations? How do mechanisms relate to causal connections?

On this last question, one can make a strong case for the following thesis:

In medicine, in order to establish that A is a cause of B we typically need to establish:
Association. A and B are correlated in the appropriate way.
Mechanism. A mechanistic connection between them is responsible for this correlation.

Association is not enough on its own, because the correlation might be due to:

  • Sampling. A statistical blip.
  • Confounding. Each variable is correlated with an unobserved common cause.
  • Time Series. Drift in each variable over time.
  • Semantic Connection. The variables have overlapping meaning.
  • Logical Connection. Particularly between logically complex variables.
  • Physical Connection. E.g., in virtue of the law of conservation of total energy.
  • Mathematical Connection. E.g., mean and variance variables defined relative to the same distribution.

One needs to rule out these alternative explanations of an observed correlation if one is to infer that the association is causal. This means determining two things:

  1. There exists some such mechanism that can explain the correlation.
  2. This explanation is the best explanation of the correlation.

Current EBM hierarchies of evidence address the question of Association, but not that of Mechanism. The rankings they advocate may well be appropriate if one wants to determine whether there is a correlation between A and B, but they are not suitable for evaluating evidence of mechanisms. This is because, while RCTs offer perhaps the best route to evidence of association, one can get high-quality evidence of mechanisms from a wide variety of sources:

  • Direct manipulation: e.g., in vitro experiments
  • Direct observation: e.g., biomedical imaging, autopsy, case reports
  • Statistical trials: e.g., RCTs
  • Confirmed theory: e.g., literature searches
  • Analogy: e.g., animal experiments
  • Simulation: e.g., agent-based models

So, philosophy of science has something useful to say about EBM:

  • Association is only half the story.
  • In order to establish causality, we also need to establish Mechanism. It is here that a thorough understanding of evidence of mechanisms is invaluable.

The mechanistic turn in the philosophy of science holds much promise, I think.

Further reading: Clarke et al. (2013); Russo & Williamson (2007).