The Centers for Disease Control and Prevention released a bombshell alcohol recommendation to women on Feb. 2 that led to an explosion of responses. I was among those who commented on the fray, and I primarily addressed how the evidence itself about alcohol and pregnancy was obscured by the resulting backlash.
I also mentioned that I had previously interpreted the evidence differently over several years of covering periodic studies about light drinking and pregnancy. I didn’t go into a great deal of detail, however, on how I made that switch, and I thought that process might be instructive for other health journalists covering such controversial issues in which the science can be confusing. Writing about risk, in particular, can be incredibly thorny. Continue reading
It’s just about impossible to report on medical research without becoming intimately familiar with PubMed. But just because a reporter uses the database site doesn’t mean they’re getting the most out of it.
How often have you used a service, such as an email client or a social media site, for years when someone suddenly points out to you a shortcut or a feature you’ve never used and didn’t know existed? Chances are that there’s at least one new skill you can pick up in Hilda Bastian’s tip sheet on using PubMed. Continue reading
In the years since its inception, Retraction Watch has documented hundreds of troubled scientific papers that were eventually retracted, as well as other related controversies. Founders Ivan Oransky and Adam Marcus have learned a lot in that time about following up on retractions, errors or other problematic aspects of scientific research.
Two years ago, they thought they had come up with some good advice for others who wanted to investigate concerns about a particular paper and published a piece on how to report alleged scientific misconduct in Lab Times. Continue reading
In the past several decades, an explosion of research has looked more closely at how exposure to certain substances during pregnancy affect the child after birth. One of the biggest challenges of this research is that nearly all of it is based on epidemiological/observational studies. Therefore, the studies can show an association between an exposure and an outcome, but they cannot show evidence that the exposure actually caused the outcome — and the difference is crucial. When looking at exposures during pregnancy, so many confounders enter the equation that it’s hard to think of all of the possible covariates, much less actually account for them all. Continue reading
Photo: Rama via Wikimedia Commons
A common type of bias that plagues medical research across all journals is publication bias: studies that find positive results are considered more interesting and therefore more likely to be published.
Positive findings about drugs in particular tend have a higher chance of ending up in a journal than those that didn’t – especially among industry-funded studies – but publication bias tends to appear across the board.
That’s what makes the open-access Journal of Negative Results in Biomedicine so interesting, and helpful for journalists. The most common word you’ll find in the titles of these studies is “not.” Continue reading