The story’s language swings back and forth in a confusing way over whether this is an association or a cause and effect finding. Reuters story was better in small but significant ways.
As strongly suggestive as a large, long-term observational study may be – and as important as the research may be as a piece of the puzzle of answering research questions – such a study cannot establish cause and effect. So news stories that use causal language (“benefit…risk reduction”) should be edited more carefully. Read our primer on association versus causation.
It matters because we need to start educating the public that not all studies are equal. Also, people are exhausted by (and suspicious of) the constant stream of information about what will make them healthier and thinner, only to find out later that the research was debunked.
The cost of aspirin is not in question.
We’re going to take an admittedly hard-line stance on this because we think such a story can easily do a better job.
The story acknowledges at one point that the study “found an association, not cause and effect.” Yet it uses cause and effect language in various places, such as:
“Benefit” and “risk reduction” and “reduced risk” convey more than an association. Those terms convey cause and effect. Take aspirin, get the “benefit” of “reduced risk.” This study, strongly suggestive though it may be given the numbers of people involved, cannot establish cause and effect.
So the suggestion of benefit needs to be couched in more careful language than this story did.
The competing Reuters story gave absolute risk reduction and was more cautious in general in interpreting results.
The story acknowledged that “even low-dose aspirin can lead to problems such as serious gastrointestinal bleeding.”
The story had two key lines:
Unfortunately, the story contradicted these two points with some of the language used to describe “benefit” and “risk reduction” as already noted. Nonethless, we won’t ding it in two different places for the same flaw.
Reuters did a better job, though, of itemizing limitations of the work.
No disease mongering at play.
Again, we’ll take a hard-line approach.
Yes, the story had an additional source, an editorial writer. And yes, it disclosed the industry ties of that editorial writer and the fact that he holds a patent for chemoprevention use of aspirin.
It is understandable that the story would turn to Dr. Baron’s comments, since he wrote the accompanying editorial and since he has published extensively on this topic.
But, by definition, since he was the only other source used, there was no independent source used in the story.
At least the story included this line: “Many other strategies will also reduce the risk of cancer. Among them: engaging in regular physical activity, maintaining a healthy diet and not smoking cigarettes.”
The availability of aspirin is not in question.
The story at least acknowledges “a previous study by others,” making at least passing reference to the fact that this is not the only research on this question. Barely satisfactory, though, as the Reuters story gives additional context to be clear what contribution the new study makes/
There’s no evidence this story relied on a news release.