The Three Cities Test: preliminary validation of a short bedside memory test in persons with acute stroke.

Screening tests for memory can be administered more quickly than standard tests of memory. They can be particularly useful with patients with acute medical illness or with the elderly who are unable to tolerate complex or lengthy memory testing, such as patients with acute stroke. However, screening measures for memory often lack validation and may have significant psychometric limitations. The purpose of this study was to validate and determine the psychometric properties of the Three Cities Test (TCT), a short test of memory that uses a selective reminding paradigm and the names of well-known cities as stimuli. The TCT was administered to 115 subjects: 60 patients with acute cerebrovascular accidents (Stroke group) and 55 age-matched orthopedic control patients (Ortho group). Results show that the TCT was significantly correlated with general measures of cognition (MMSE), another well-validated measure of learning and memory (HVLT-R), and clinical variables such as length of hospitalization and functional recovery. Compared to the Ortho group, the Stroke group had significantly worse performance on the TCT in terms of number of trials to criterion, delayed recall, and recognition discrimination. Preliminary results suggest that this instrument is well-received by patients with acute medical illness and cognitive impairment and that it possesses good construct and discriminative validity. Sensitivity and specificity performance as well as recommended cut scores are offered for the TCT.