Ongoing pain as a mental workload indexed by P300 depression: discrimination of real and feigned pain conditions.

Two groups of subjects were formed. One, a real pain group, was subjected to aversive pressure on the fingernail surface during pain-only and pain-tracking conditions. The second group (feigned pain) was instructed to feign pain during corresponding runs in which fingernail pressure was applied at a just noticeable (not painful) level. Both groups also simultaneously performed an auditory oddball (.8/.2) task during their pain-only and pain-tracking conditions, as well as during an initial (baseline-1) and final (baseline-2) baseline condition in which only the oddball task was performed. Oddball-evoked P300 amplitudes were significantly different for the groups during the pain-tracking and baseline-2 conditions, with the amplitudes smaller in the real pain groups in both conditions. The pain-only and pain-tracking conditions in both groups caused significant P300 reductions in comparison with baseline-1 values. P300 latencies did not differ as a function of either groups or conditions. 100% of the real pain subjects and 67% of the feigned pain subjects could be correctly classified using P300 amplitude-derived indices.