Classification accuracy of the Portland Digit Recognition Test in traumatic brain injury: results of a known-groups analysis.

The present study used a known-groups design to examine the accuracy of the Portland Digit Recognition Test (PDRT) in the detection of malingering in traumatic brain injury (TBI). Data were derived from 262 TBI patients who were classified as not malingering, possibly malingering, and malingering based on the Slick, Sherman, and Iverson (1999) criteria. The original PDRT cutoffs detected between 20 and 50% of malingering TBI patients with a false positive error rate of 5% or less. When the false positive error rate was held at 5%, across all item sets, sensitivity was as high as 70%. The results show that the original PDRT cutoffs are conservative and that higher scores detect more MND patients without causing the false positive error rate to become unacceptably high. Clinical application and future research needs are discussed.