Optimizing Detection of Very Mild Aphasia With Letter Fluency Tests.

Abstract

Purpose: The purpose of this study was to find an efficient measure for clinicians to diagnose very mild aphasia.

Method: Twenty-seven individuals with very mild aphasia who tested as not aphasic by the Western Aphasia Battery-Revised (WAB-R; NABW) were compared to 68 matched controls on 28 tests. Group-level differences were determined with independent-samples t tests. A univariate receiver operator characteristic (ROC) curve analysis was constructed to determine optimal cutoff points for each test. Sensitivity and specificity were calculated for the optimal cutoffs. Boolean operators of two or more tests were used to determine if combinations of tests were more sensitive and specific than single tests in classifying the NABW participants versus neurotypical controls.

Results: Twenty out of 28 behavioral tests and subtests had group-level differences, with 18 surviving Bonferroni correction. Only scores derived from letter fluency and a two-letter fluency test had sensitivity and specificity ≥ .70. The Boolean combination of letter fluency and two-letter fluency tests gave a sensitivity of .89 and specificity of .81.

Conclusions: The combined letter fluency and two-letter fluency tests were sensitive in detecting very mild aphasia beyond what is captured in the WAB-R. Administration takes an estimated 11 min, making it accessible to clinicians in various settings. While letter fluency tests rely on phonological and orthographic knowledge, they are also frequently used to measure executive control. Thus, this combined letter fluency test should be used in conjunction with other language and cognitive assessments to determine treatment targets.