In late 2011, a University of Western Ontario study rocked the neuroscience community by reporting that inexpensive and portable handheld electroencephalogram (EEG) scanners had detected signs of consciousness in three people thought to be in persistent vegetative states. Now a team of researchers from Weill Cornell Medical College (that’s Cornell University’s medical school in NYC) is saying “not so fast.” In this week’s issue of the journal Lancet, the Weill Cornell team says the study suffers from statistical errors. More to the point, the Western Ontario researchers misinterpreted noise in the data.
In the original study, researchers took EEG readings of 16 patients in persistent vegetative states while asking them to perform simple physical tasks, like wiggling their toes. Of those, the study claimed, three showed neural activity when given the command that is consistent with understanding and responding to aforementioned command. The findings were huge, or so it seemed at the time. Not only did they demonstrate that it’s possible to perceive consciousness in those thought to be in vegetative states, but it’s possible to do so with a cheap, bedside diagnostic.
The reanalysis team from Weill Cornell is careful to note that it did not set out to refute the earlier study, but rather to confirm it. And while it doesn’t criticize the idea of using bedside EEG for consciousness detection, it does criticize the statistical methods the Western Ontario researchers used to arrive at their conclusions. The Western Ontario team failed to account for noise in the data, the reanalysis says, like “contamination of EEG signals with muscle activity and the random characteristics of the EEG over extended periods of time.” The report goes on to note that the signals the original researchers pointed to as proof of conscious thought don’t look the same as the signals that were observed in the control group.
Of course, the parameters of consciousness are very important to the people they most effect. Families often have to make life-and-death decisions based on a doctor’s opinion of whether loved ones in persistent vegetative states can regain consciousness or not.
While this is a setback for the original study findings and perhaps for the U. of Western Ontario team, it’s not necessarily a setback for neuroscience at large. This is how peer review works, and the occasional refutation of scientific claims proves exactly that–that it works.