Editor's note: The original online version of this story was posted on September 20, 2009.

In patients who have survived severe brain damage, judging the level of actual awareness has proved a difficult process. And the prognosis can sometimes mean the difference between life and death. New research suggests that some vegetative patients are capable of simple learning—a sign of consciousness in many who had failed other traditional cognitive tests.

To determine whether patients are in a minimally conscious state (in which there is some evidence of perception or intentional movement) or have sunk into a vegetative state (in which neither exists), doctors have traditionally used a battery of tests and observations. Many of them require some subjective interpretation, such as deciding whether a patient’s movements are purposeful or just random. “We want to have an objective way of knowing whether the other person has consciousness or not,” says Mariano Sigman, who directs the Integrative Neuroscience Laboratory at the University of Buenos Aires.

That desire stems in part from surprising neuroimaging work that showed that some vegetative patients, when asked to imagine performing physical tasks such as playing tennis, still had activity in premotor areas of their brains. In others, verbal cues sparked language sectors. A recent study found that about 40 percent of vegetative state diagnoses are incorrect.

To explore possible tests of consciousness in patients, Sigman and his colleagues turned to classical conditioning: they sounded a tone and then sent a light puff of air to the patient’s eye. The air puff would cause a patient to blink or flinch the eye, but after repeated trials over half an hour, many patients would begin to anticipate the puff, blinking an eye after only hearing the tone.

If two stimuli are delivered at exactly the same time, even snails will equate the stimuli. But the team actually delayed the puff after the tone by 500 milliseconds. To associate two stimuli separated by that time gap, “you need conscious processing,” says lead study author Tristan Bekinschtein of the Impaired Consciousness Research Group at the University of Cambridge. In fact, delaying the second stim­ulus by more than 200 milliseconds is enough to demonstrate some learning, he adds. By comparison, people under general anesthesia, considered to be entirely lacking awareness, showed no sign of such learning when given the tone and air-puff test.

The detection of learning, described in the September 20 Nature Neuroscience (Scientific American is part of the Nature Publishing Group), also opens up questions about when patients should be classified as being in a persistent vegetative state, in which emergence isn’t predicted to be likely. (Terri Schiavo, the center of a heated national debate in 2005, was determined to be in such a state.) Decisions to end life support often depend on predictions of recovery and assessments of consciousness. If “someone shows the patients can learn,” Bekinschtein says, “I think it would be a very clear argument.”

Indeed, the researchers found that learning ability accurately predicted the extent of recovery within the next year about 86 percent of the time. The neural reorganization that bypasses damaged parts of the brain “implies that there’s room for at least some recovery,” Bekinschtein notes.

The findings do not surprise everyone. Research using functional MRI on vegetative patients had already led John Whyte, principal investigator at the Neuro-Cognitive Rehabilitation Research Network at Thomas Jefferson University in Philadelphia, to question the designation system. It may be that “there is a firm line” between vegetative and minimally conscious patients,” he observes. “But our tools are too crude to tell us who is on which side of the line.” Or it may be that categories of consciousness are not so easy to define.

The learning tests probably could not completely supplant fMRI. Joy Hirsch, a neuroscientist at Columbia University, says that “functional imaging is by far the tool of choice” in determining consciousness because it can reveal “cognitive processes that are latent in these patients that aren’t visible through [traditional] bedside tests.” But neuroimaging can be expensive and hard to come by in many places, note Bekinschtein, Sigman and their study collaborators. Much of the testing was completed in Argentina, where imaging capabilities can be less available than in the U.S. or U.K. For this test, “you just need two wires, and it costs $100,” Sigman states. “In practical terms, it has strong implications.”

Note: This article was originally printed with the title, "Conditional Consciousness."