Feasibility of use of probabilistic reversal learning and serial reaction time tasks in clinical trials of Parkinson's disease.

INTRODUCTION The aim of this study was to investigate the feasibility of using two computer-administered neuropsychological tasks in a clinical trial involving participants with Parkinson's disease without dementia. The tasks, probabilistic reversal learning (PRL) and serial reaction time (SRT), target dorsolateral prefrontal cortex (SRT) and ventral striatal-orbitofrontal (PRL) functioning respectively. METHODS Participants were 53 adults with idiopathic Parkinson's disease who completed both the SRT and PRL tasks at baseline in a clinical trial. Repeated measures were examined only for the placebo group (n = 20). RESULTS No participants were removed from analyses due to inability to complete the tasks, and most had fewer than 10% of trials culled due to slow reaction times. Response accuracy on PRL was 81.98% and 66.65% for the two stages of the task respectively. Disease duration was associated with SRT relearning. Disease duration and stage were associated with initial learning on PRL, and there was a trend towards disease stage predicting greater errors on PRL. Among participants in the placebo group, practice effects were seen on PRL (Phase 1 errors) and SRT (relearning). CONCLUSIONS These results provide initial evidence for the clinical feasibility of computerized PRL and SRT tasks in clinical trials in Parkinson's disease.