Route learning in amnesia: a comparison of trial-and-error and errorless learning in patients with the Korsakoff syndrome.

OBJECTIVE: To examine the errorless learning approach using a procedural memory task (i.e. learning of actual routes) in patients with amnesia, as compared to trial-and-error learning.
DESIGN: Counterbalanced self-controlled cases series.
SETTING: Psychiatric hospital (Korsakoff clinic).
SUBJECTS: A convenience sample of 10 patients with the Korsakoff amnestic syndrome. INTERVENTION: All patients learned a route in four sessions on separate days using an errorless approach and a different route using trial-and-error. MAIN MEASURES: Error rate was scored during route learning and standard neuro-psychological tests were administered (i.e. subtest route recall of the Rivermead Behavioural Memory Test (RBMT) and the Dutch version of the California Verbal Learning Test (VLGT)).
RESULTS: A significant learning effect was found in the trial-and-error condition over consecutive sessions (P = 0.006), but no performance difference was found between errorless and trial-and-error learning of the routes. VLGT performance was significantly correlated with a trial-and-error advantage (P < 0.05); no significant correlation was found between the RBMT subtest and the learning conditions.
CONCLUSION: Errorless learning was no more successful than trial-and-error learning of a procedural spatial task in patients with the Korsakoff syndrome (severe amnesia).