Improving Assessment of Cognitive Impairment after Spinal Cord Injury: Methods to Reduce the Risk of Reporting False Positives.

ABSTRACT

Adults with spinal cord injury (SCI) are reported to have heightened risk of cognitive impairment, notably mild cognitive impairment (MCI). Reports of increased risk of MCI are almost exclusively based on cross-sectional assessments of cognitive function using norm-referenced scores. Norm-referenced single-point assessments do not reflect cognitive decline at the individual level but rather represent between group differences in cognitive function. The practice of relying solely on norm-referenced assessment to study MCI after SCI is therefore problematic as it lends to potential misclassification of MCI. Premorbid intelligence estimates permit comparison of people's actual versus expected cognitive function and thereby can be used to validate the presence of genuine cognitive decline. These are not utilized in the assessment of MCI after SCI. This study simulated data for 500,000 adults with SCI to compare norm-referenced and premorbid-intelligence methods of screening for MCI to examine the potential extent of MCI misclassification after SCI resulting from the overreliance on norm-referenced methods and exclusion of premorbid intelligence methods. One in five to one in 13 simulated adults with SCI were potentially misclassified as having MCI showing that measures of premorbid cognitive function must be included in assessment of cognitive function after SCI.