I noticed that this story has gotten picked up by a number of members of the medical-science blogosphere and it seems as though there are a number of differing opinions about the interpretation of this study. If there is confusion amongst us, then I'm certain that there is even more confusion among the general public about what exactly this all means. Here, I would like to offer some of my own thoughts on the matter and how it illustrates a number of key points that I often see arising in the field and how it can lead to disagreement.
1) Source: Before we even get to the results of the survey, and even before we make conclusions, we have to examine the source of the information. As a charity driven by donations from people concerned about cancer, I doubt that Cancer Research UK is an unbiased arbiter in the information collection business. I am not surprised that the survey they conduct results in data that supports the need for their own organization. This is why polling services are used by a majority of interests groups, so that they can not be blamed for influencing the results.
Moreover, the original source is a press release and the coverage is based upon the scant information provided in it. This brings up two important faults: a) we do not have access to the original data and methods (see below) and that there was no oversight in the way the survey was conducted (no peer-review) and b) the information has been processed by two people who probably have very little experience in reporting scientific results. I don't know about you, but most of the humanities majors I knew in college loathed their science work [the rant on teaching science in college will be saved for a later time] and would probably have a hard time fairly making conclusions about a report that has an unclear experimental method (again, see below).
2) Experimental Design: What was the exact question posed to Brits in this study? We don't have this information. The wording of such surveys is extremely important, especially when the wording can influence the responses of the participants. Without the wording, we can not know for sure if this report was clear and unbiased. For this and the reasons above, I usually don't read these kinds of report expecting to learn something.
3) Results are open to interpretation: In perhaps my biggest gripe about this report, I feel that the results, as described, are ambiguous. For this, I will draw on a hypothetical example, illustrated below:
In this scenario, we have a total population of 200 people. We assume that the occurrence of spontaneous cancer is low, about 1%. Splitting the groups into smokers (ie lifestyle choice) and non-smokers of 100 each, we would expect that 1 person from each group would get cancer, regardless of any lifestyle choices. We can say that fate was not on their side because there was nothing they could do.
Assuming that smoking raises your risk 10 times, we then would expect 10 out of 100 smokers to develop cancer for a total of 11.
This is where the results become a point of debate. What makes the 1 person in the non-smoking group get cancer - is it fate? In this case, lets assume yes. The difference between her and the 99 other non-smokers was simply a stroke of bad luck. And the 1 smoker who gets cancer who would have gotten cancer anyway? Again, lets say that fate was responsible. Then there are the 10 additional people who got cancer in the smoking group. What made these 10+1 people instead of the other 89 "healthy" smokers is unclear and attributed to chance, or fate* in this case.
As other bloggers correctly indicate, no physician can tell a patient with certainty whether they will get cancer, smoking or not. In fact and to be fair, there is a greater chance that a smoker will not develop cancer (89) than actually get it (11). Making the assumptions above, it is fate that separates the 1 from the 99 in the non-smokers and the 11 from the 89 in the smokers group. Looking at it from this angle, it is valid to claim that fate is responsible for who gets cancer and who doesn't.
But to say that it is fate that dictates the 10 smoking-caused cancers vs. the 1 spontaneous tumor in either group is simply not true. It is the lifestyle choice itself that has given rise to the 10 additional malignancies. Therefore, we can no longer state that fate alone is responsible for cancer, irrespective of behavior.
Conclusion:
These kinds of studies may lead to confusing results. We can go back and forth on this all day depending on our interpretation, but ultimately it is the experiment and conclusions drawn are flawed from the get-go. We need to be cautious of our source of information and not let these press releases do more damage than good when their results are not scientifically sound.
* A note on "fate": Simply because something is unknown, does not mean it is due to fate or chance. There are many phenomena that we do not quite understand completely and tumorigenesis is one of them. Many things that used to be considered fate or luck now have scientific explanations. Take, for instance, the people who can have repeated unprotected sexual intercourse with an HIV+ partner without acquiring the virus (akin to not developing cancer despite smoking). An uninformed observer would say that these people are not fated to get AIDS. However, in reality, these many of these people are descendants from survivors of the Bubonic Plague in Europe. The disease selected for a genetic polymorphism that made some people resistant to viral infection and this allele remains today. I suspect that the susceptibility of some people (the 10) over others (the 89) will soon become clear in time with a biologic rational rather than pure fate.
No comments:
Post a Comment