Medicine is only as good as its evidence.

People talk about "evidence-based medicine" in obstetrics, and in the holistic birthing community as well, as if it's a sleek, shiny package tied up with a bow and ready for Christmas morning.

But evidence-based medicine is a shambling, messy thing, always on the move, only as reliable as the researchers who conduct it -- and as the research that has gone before and serves as its foundation.

The Lancet's retraction earlier this month of a 1998 study that appeared to establish a link between the development of autism in young children and the vaccine for measles, mumps and rubella is a good demonstration of the fallibility of evidence-based medicine.

One reason the original Lancet study collapsed was that other researchers could not duplicate lead author Andrew Wakefield MD et al.'s results. But even while 10 of the 13 original co-authors withdrew their support for the study, the public absorbed the idea that a cause-effect relationship exists between vaccination and autism.  That idea has not been eradicated by the Lancet's retraction.

And while such a retraction is unusual, it is unfortunately not unheard of.

Last July, anesthesiology researcher Scott S. Reuben MD of Baystate Medical Center in Springfield, Mass., was revealed to have falsified results in at least 21 studies that appeared in several peer-reviewed journals. Anesthesia and Analgesia alone retracted 10 of Dr. Reuben's articles.

Dr. Reuben specialized in multimodal analgesia, a drug-delivery system designed to control the pain of surgery.

In January, Dr. Reuben pleaded guilty to fraud for such audacious creations as a 2005 trial of Pfizer's drug Celebrex as an agent in a multimodal analgesia model. Dr. Reuben accepted $74,000 from Pfizer to conduct the study, simply did not do the work, and then published fabricated results in Anesthesia and Analgesia.

Unfortunately, other researchers based their own work on those phony results.

The collapse of Reuben's data has left multimodal analgesia “in shambles....The big chunk of what people have based their protocol on is gone,” said Jacques Chelly,  an anesthesiologist at the University of Pittsburgh Medical Center, in Anesthesiology News last March.