Thursday, December 25, 2014

Exaggeration of Medical Experiments

In a recent paper in BMJ the authors present an analysis of exaggeration in the Medical Literature of the results obtained. They state:

For our analysis of advice we found that 40% of the press releases contained more direct or explicit advice than did the journal article (bootstrapped 95% confidence interval 33% to 46%). For our analysis of statements based on correlational results (cross sectional or longitudinal) we found that  33% of primary claims in press releases were more strongly deterministic than those present in the  associated journal article (bootstrapped 95% confidence interval 26% to 40%). For studies on animals, cells, or simulations, 36% of press releases exhibited inflated inference to humans compared with the journal article (bootstrapped 95% confidence interval 28% to 46%). Given the likelihood that some statements in journal articles themselves would be considered exaggerated by other scientists in the specialty, our levels of measured exaggeration are likely to be underestimates.

We have commented on this frequently. Typical of the exaggerations are claims to therapeutic implementations using just a mouse model, claims to targets of specific genes, miRNAs and the like.

The problem has become pandemic and claims are made which truly extend the limits of credibility. The result of this may be significant if not managed better.