Sally Keat

Abstract

Background: Cognitive impairment is common amongst acute geriatric hospital admissions but detection is often poor and this is associated with worse outcomes. The four-question abbreviated mental test (AMT4) has previously been promoted nationally in the acute setting as a succinct assessment tool. However, a recent national dementia Commissioning for Quality and Innovation (CQUIN) goal recommends a single screening question followed by the tenquestion abbreviated mental test (AMT10). We aimed to evaluate the negative predictive value of the AMT4 within the acute setting by comparing it to three other validated tools.

Methods: We identified 100 acute medical admissions (>60 years old) with a negative AMT4 and administered the AMT10, six-item cognitive impairment test (6CIT) and confusion assessment method (CAM) within 24 hours of admission.

Results: Forty-six percent of the participants scored positively on at least one of the additional tests despite a negative AMT4. Forty-four patients had a positive 6CIT, 23 had a positive AMT10 and six had a positive CAM. Using the 6CIT as a diagnostic standard, tests of short-term memory had the greatest sensitivity and specificity for cognitive impairment.

Conclusions: Nearly half of the participants had signs of cognitive impairment despite a negative AMT4. Consequently, there is a risk of under diagnosis with potentially serious consequences for morbidity and mortality. Tests of shortterm memory were strongly associated with cognitive impairment. We propose the addition of such a test in order to increase the sensitivity of the AMT4 without compromising its brevity and utility in the acute setting.