An exploratory study of using polygraph to detect deception in patients with traumatic brain injury.

To investigate the accuracy of the polygraph when used on patients with traumatic brain injury (TBI) and to analyze the factors influencing accuracy to enable provision of some field data for expert witnesses. Sixty patients with TBI (experimental group) and sixty normal individuals (control group) were tested by the polygraph with a Stimulation Test; responses of skin conductance, respiration, and finger pulse were recorded. The intelligence quotient of the experimental participants was measured and the factors influencing the accuracy of the polygraph test were analyzed. According to the computer automatic scoring, the overall accuracy was 76.7% for the control group; the accuracy of skin conductance, respiration, and finger pulse was 76.7, 35, and 18.7%, respectively, whereas for the experimental group, the overall accuracy was 56.7%. The accuracy of skin conductance, respiration, and finger pulse was 55.6, 23.3, and 25%, respectively. There was a significant difference in the overall accuracy between the two groups (P=0.020). According to the examiner's manual scoring, the overall accuracy was 86.7% for the control group and 61.7% for the experimental group (χ=9.786, P=0.002). The difference in accuracy obtained from the examiner and the computer was not significant. In the experimental group, the intelligence quotient of participants who had the right outcome was 63.2±10.9 and it was 55.3±9.5 for those who had the wrong outcome; this difference was significant (t=2.68, P=0.010).The accuracy of a polygraph test is much lower when it is used on patients with TBI compared with that of normal individuals. In actual application, full communication with the examinees should take place and their comprehension of the protocol should be evaluated. If participants can understand and co-operate with the test, the outcome can be considered reliable; if not, the outcome should not be relied on.