Quantifying the association between computerised measures of attention and confusion assessment method defined delirium: a prospective study of older orthopaedic surgical patients, free of dementia.

OBJECTIVE: To determine whether attentional impairments are reliable neuropsychological markers of delirium.
METHOD: A prospective cohort of one hundred patients admitted for elective orthopaedic surgery, 70 years and over, and free of dementia were recruited from an Orthopaedic unit in a university teaching hospital. Computerized assessments of attention and the Mini Mental State Examination were administered pre and post operatively. The Confusion Assessment Method was used to evaluate signs and symptoms of delirium.
RESULTS: Over the first post operative week after surgery people with delirium scored lower on the MMSE (F = 23.53 (1, 53); p = 0.000); and performed less accurately (F = 20.02 (1, 55); p = 0.000), slower (F = 14.58 (1, 54); p = 0.000) and with greater variability of reaction time (F = 31.52 (1, 53); p = 0.000) than people without delirium. The group with delirium's neuropsychological performance was marked by a quadratic trend of accuracy (F = 10.5 (1, 8); p = 0.018) across the first post operative week. Conversely the group without delirium demonstrate quadratic trends for reaction time (F = 6.91 (1, 49); p = 0.011); and linear trends for the variability of reaction time (F = 7.06 (1, 49); p = 0.011) over this period.
CONCLUSIONS: To date the absence of any well validated neuropsychological markers of delirium has hindered progress of research of delirium. The data within this study suggest key indices of attention and in particular fluctuating cognition may offer excellent discriminative utility for this clinically important condition. (c) 2008 John Wiley & Sons, Ltd.