Creating interpretation guidelines for the Hebrew trail making test.

The Hebrew language version of the Trail Making Test (TMT; Army Individual Test Battery, 1944) was administered to a group of normal control participants as well as a sample of outpatients approximately 1 year following a head injury. A ratio of TMT-B to TMT-A performance was computed in an effort to establish usable cutting scores for the Hebrew TMT. A ratio of 2.26 was observed to result in 63% sensitivity and 77% specificity for the sample, with positive predictive power of 71% and negative predictive power of 70%. A more stringent cutoff of 3.09 reduced sensitivity (22%) and negative predictive power (44%). In contrast, specificity (97%) and positive predictive power (86%) were greatly improved once the speed of TMT-A performance was also considered. The more conservative cutoff is considered appropriate when a finding is indicative of pathological performance.