Evidence for causation

Berkovitch, M., Pope, E., Phillips, J., and Koren, G. (1995) Pemoline-associated fulminant liver failure testing the evidence for causation. Clini Pharmacol Ther 57 696—698. 

Causation of adverse event by medical management is jndged on a 6-point scale, where 1 indicates virtually no evidence for causation and 6 indicates virtually certain evidence for causation. Only adverse events with a score of 4 or higher, requiring evidence that cansation is more likely than not, are reported in the resnlts. 

Case-control studies retrospectively compare the frequency of breast implantation in women with and without the outcomes of interest. If, for example, connective tissue disorders were more likely to occur among women with breast implants, this would constitute some evidence for causation. Case-control studies typically require less time and resources than cohort studies. However, they are susceptible to many more biases than cohort studies [62]. The primary reason not to perform a case-control study here, however, is that a separate case-control study would be required for each of the outcomes of interest, i.e., a case-control study of systemic sclerosis, a case-control study of rheumatoid arthritis, etc. The... 

Heyland D, Mandell LA. Gastric colonization by gram-negative bacilli and nosocomial pneumonia in the intensive care unit patient evidence for causation. Chest 1992 101 187-193. 

One of the origins of this view of error and accident causation is the theory of accident proneness, which tried to show that a small number of individuals were responsible for the majority of accidents. Despite a number of studies that have shown that there is little statistical evidence for this idea (see, e.g., Shaw and Sichel, 1971) the belief remains, particularly in traditional industries, that a relatively small number of individuals accoimt for the majority of accidents. Another element in the emphasis on individual responsibility has been the legal dimension in many major accident investigations, which has often been concerned with attributing blame to individuals from the point of view of determining compensation, rather than in identifying the possible system causes of error. 

With some possible exceptions (to be discussed in Chapter 9), dose-response relations identical or similar to those shown in Figure 3.1 are observed for all expressions of chemical toxicity. Indeed, the absence of such a dose-response relationship is often used as evidence that a chemical has not caused a particular response. Criteria for causation in... 

Another type of biological evidence that supports a case for causation derives from experimental work with animals if the substance under consideration is carcinogenic in animals, associations seen in epidemiology studies become biologically supportable. Sometimes it is possible to conduct a study after intervening to remove the suspect substance if risk declines following such an intervention, the case for causation strengthens (recall Pott s chimney sweeps). 

There is, I think, also a psychological preference for simplicity that is related to comprehensibihty, but which also may have an esthetic component. I will return to the psychology of explanation after finishing this brief review of the principal proposals for scientific explanation from the community of professional philosophers. In completing the summary of Boyd s position, though, I would say that the evidence for the important role of nonexperimental factors that Boyd cites as the reason to overthrow Humean causation is not clear-cut. I guess I remain a Humean at heart. 

Here then are nine different viewpoints flom all of which we should study association before we cry causation. What I do not believe—and this has been suggested—that we can usefully lay down some hard-and-fast rules of evidence that must be obeyed before we can accept cause and effect. None of my nine viewpoints can bring indisputable evidence for or against the cause-and-effect hypothesis and none can be required as a sine qua non. What they can do, with greater or less strength, is to help us to make up our minds on the fundamental question—is there any other way of explaining the set of facts before us, is there any other answer equally, or more, likely than cause and effect ... 

Despite the strong supportive evidence for animal and mechanistic studies, there have been major difficulties in assessing the precise role of aflatoxin in the causation of liver cancer in humans. 

On the other hand, evidence for a causal relationship between cannabis use and psychoses in otherwise low-risk persons is much more scarce (Johns, 2001 Hall and Degenhardt, 2000 Mass et al., 2001). Convincingly arguing against the causation hypothesis is the finding that in Australia no increased rate of incidence of schizophrenia has been reported over a period of several decades despite a dramatic increase in cannabis use during that period (Degenhardt, Hall and Lynskey, 2003). 

The issues of nature of the event and plausibility need to be considered with some caution - these factors may add to the arguments for causation but a clinical event that is not normally known to be drug-related or the absence of any information supporting plausibility is not strong evidence against it. 

For this reason, other techniques are required to provide unequivocal evidence for the effect of a food or a food component in the causation or prevention of disease. One important techniqne is epidemiology, in which factors such as serum levels, adipose tissue levels, or dietary intake of a component can be correlated with the incidence of, or death from, major diseases such as cancer and coronary heart disease. Serum and adipose tissue levels of a component such as CLA can be measured accurately however, they indicate only recent intake and may not be indicative of intake in previous years or at a time close to the initiation of the disease. Although subject to assessment error, dietary intake at various periods can overcome these difficulties. Final proof of efficacy for a component in disease prevention is provided by the randomized, double-blinded, clinical intervention study. 

The remainder of this chapter discusses the evidence for the various causal relationships between Pb exposure and toxicity and disease. As set forth in Table 21.2, adverse human health effects of lead are placed in causal context using the nine proofs-of-causality criteria first enunciated by British medical statistician Sir Austin Bradford Hill in 1965 and subsequently labeled Hill s Criteria of Causation. These criteria are relevant and of particular value in the environmental epidemiology of Pb, given the many contentious arguments over association versus cause that have been lodged against Pb s human health impacts, particularly in earlier eras of lead health research and in the evolution of Pb as a major public health factor. 

Guzelian PS, Victoroff MS, Halmes NC et al (2005) Evidence-based toxicology a comprehensive framework for causation. Elum Exp Toxicol 24 161-201... 

Howick J, Glasziou P, Aronson JK. The evolution of evidence hierarchies what can Bradford Hill s guidelines for causation contribute J R Soc Med 2009 102 186-94. 

---