(Scientific Session) Mississippi Aphasia Screening Test: Validity on Admission to Inpatient Stroke Rehabilitation

Background and/or Objectives: On acute inpatient stroke rehabilitation (AISR) admission, we hypothesize the Mississippi Aphasia Screening Test (MAST) will show strong correlation with assessments of language and weak correlation with non-language assessments.

Design: Retrospective review of prospective database

Setting: AISR

Participants: 283 participants at AISR admission completed the MAST: 45.9% female, 46.3% left-hemisphere-stroke, mean age=69.8+14.7, length of stay mean=14.5+/-8, ranging from 3-57, National Institute of Health Stroke Scale (NIHSS) mean=7.8+/-6.7, range 0-38.

Interventions: not applicable

Main Outcome Measures: Test battery included MAST, Boston Naming Test (BNT), Functional Independence Measure (FIM) and subscores of expression (FIMe), comprehension (FIMc), motor function (FIMm), and the Berg Balance Scale (BBS). Correlation matrix with Spearman’s rho were performed on all assessments, and Mann-Whitney U tests with effect size compared MAST score>96 and < 96.

Results: Of the 283 participants, 96 (34%) had MAST=100 (maximum score), 146 (52%) scored>96, and 0 participants scored 0-5. Ninety-seven (34%) completed MAST, BNT, FIM, and BBS. Spearman’s rho between MAST and FIMc, FIMe, BNT, BBS, FIMm was respectively 0.41, 0.40, 0.50, 0.19 and 0.18; and on Mann-Whitney, which compared high- and low-scoring MAST groups, effect sizes were respectively 0.39, 0.43, 0.45, 0.15, 0.17.

Conclusions: Concurrent and convergent validity on AISR admission were demonstrated between MAST and BNT, FIMe, and FIMc; divergent validity between the MAST and BBS, FIMm. Ceiling effect was substantial, at 34%, as noted in studies of non-English versions of the MAST screen. The MAST did not demonstrate floor effect on admission to AISR. MAST is a valid screen of aphasia which may suffer from ceiling effect.