Schiff et al [1] reported that two severely brain-injured patients
monitored with fMRI showed brain responses to auditory “personal
narratives” but not when the input was presented in reverse, meaningless
form. A similar study [2] with equally-remarkable results was carried
out more than 20 years earlier, ironically also by researchers at Columbia
University. That it was overlooked by Schiff et al is an indication of the obscurity into which this study has undeservedly fallen.
In the earlier report of Boyle and Greer, three brain-injured
patients with disabilities at least as severe as those in the Schiff et
al study were studied. In place of monitoring by fMRI, operant
conditioning was used to increase small spontaneous responses made by each
patient including eyelid, finger, or mouth movements. The reinforcer in
each case was the playing of a 15-second sample of music identified as the
patient’s favorite, presented contingent on an occurrence of the target
response. A multiple-baseline design with reversal, [3] a
technique of the field of applied behavior analysis, was used to assess
causal relationships. The authors reported good evidence of learning for
the first patient, lesser but suggestive evidence for the second, and
little evidence of learning for the third.
The first patient subsequently recovered from the vegetative state to a
limited extent, the second did not recover, and the third died within
one week after the end of testing. Thus successful operant conditioning
may possibly predict recovery and may even help produce it.
Schiff et al concluded in their later study that their findings of
changes in brain activity in response to meaningful auditory input “raise
important questions related to whether MCS [minimally conscious state]
patients have a greater capacity to experience subjective states but also
to benefit from therapeutic intervention”. The results of the
Boyle and Greer study raise the same questions, but go one step further by
showing an actual change in behavior in response to reinforcement by
emotionally-meaningful stimuli.
Although the methods differ, the results
of the two studies complement each other. Both methods should be exploited
as valuable tools towards the goals of understanding and
enhancing the capabilities of the brain-injured, minimally conscious patient.
References
1. Schiff ND, Rodriguez-Moreno D, Kamal A, et al. fMRI reveals large-
scale network activation in minimally conscious patients. Neurology 2005;
64: 514-523.
2. Boyle ME, Greer, RD. Operant procedures and the comatose patient.
J Appl Beh Anal 1983; 16: 3-12.
3. Kazdin, A. Behavior modification in applied settings, 6th ed.
Belmont, CA: Wadsworth, 2001.
Conflict of interest: none
The authors had the opportunity to respond to these Correspondences but declined.