Reliable serial measurement of cognitive processes in rehabilitation: the Cognitive Log.

Abstract

To evaluate the reliability and utility of a brief quantitative measure of cognitive recovery, the Cognitive Log (Cog-Log), developed for daily use with rehabilitation inpatients to provide information about the recovery of higher neurocognitive processes including verbal recall, attention, working memory, motor sequencing, and response inhibition. Descriptive study of the Cog-Log's normative scores, reliability (interrater, internal consistency), and validity as shown by its relationship to standard neuropsychologic measures. Inpatient rehabilitation hospital affiliated with a large university medical center. One hundred fifty neurorehabilitation inpatients with acquired brain injury; 83 young adults without acquired brain injury were included to provide normative data. Not applicable. The Cog-Log; standardized neuropsychologic measures of memory (Wechsler Memory Scale-Revised, Rey Auditory Verbal Learning Test), language, attention (Wechsler Adult Intelligence Scale-Revised), and reasoning (Trail Making Test). Reliability analysis showed strong interrater reliability across items (Spearman r, .749-1.00) and high internal consistency (Cronbach alpha=.778). Factor analysis of the Cog-Log using principal components extraction revealed a unitary factor (eigenvalue=3.48). Cog-Log items designed to measure working memory and immediate and delayed verbal memory were most strongly predictive of performance on similar standardized neuropsychologic measures administered on the same day. The Cog-Log appears to be a reliable and efficient tool for measuring ongoing neurocognitive recovery during inpatient rehabilitation.