Small is Beautiful Emergency Medicine
Medical school taught me to use the 10 point Abbreviated Mental Test (AMT10) to assess for delirium and possible dementia or cognitive impairment in our older patients. I've always struggled to remember all 10 questions; I can't be alone here! Increasingly, it's not just me who can't remember the dates of World War 1, but also my patients, when they are definitely cognitively intact! While the memory of the significance of WW1 must be held, the actual dates are perhaps less significant in the 21st century.
I've been using a more abbreviated 4 - point AMT scale (AMT4) - Place, Age, Date of Birth and Year (helpfully abbreviated to PADY) for the past few years having heard that it was 'as good'. And because I can remember it.
As an ED clinician I want a screening test that is quick, memorable, easy to score, and ideally validated in an ED population of elderly patients. I want to be able to selectively apply it to help me identify delerium and dementia or other cognitive impairment. I don't necessarily need to be able to discriminate between them - in ED, the two are often superimposed anyway. The main use would be as part of capacity assessment, but it is generally valuable; flagging the need for urgent collateral history, for example.
I thought I'd better look this all up. Here goes...
The AMT10 was developed for hospital use by a geriatrician in 1972(1), and has not been validated as a screening test in ED. I think it was derived from the tool that became the MMSE - it's not completely clear in the original paper.
The AMT4 has, I've read, been recommended in DoH guidelines, though I couldn't find specifics on a brief search. It has been validated - though these studies are debatable, as there is no clearly accepted 'gold standard' test to detect dementia, cognitive impairment or delirium.
One study of 100 older patients in an acute ward setting(2) compared the 4 point AMT with the 6-CIT (I'll explain), the 10 point AMT, and the CAM (confusion assessment method) and found the AMT4 of limited sensitivity. Their results suggested some kind of short term memory test was required as an addition to P.A.D.Y.
Another well constructed study of 196 older patients (3) found the AMT4 test could screen for delirium in older patients in the ED (92% PPV), but wasn't great at identifying those with probable dementia (48% PPV) (validated against MMSE, CAM-ICU, and the AD-8). Overall, Using AMT4 to screen for any kind of 'impaired mental status' gave a PPV of 94.6% and an NPV of 73.3%. It's specific (0.96) but not very sensitive (0.53). It misses likely dementia more than delirium, and was very weak at detecting cognitive impairment that isn't dementia (PPV 22.2%).
Enter two more abbreviated tests:
The 6CIT test - derived by regression analysis from an older more complex test, it is fairly memorable and user friendly (6 questions), though the scoring is not entirely straightforwards - you need the (free) app. It's been validated in a population where some had mild dementia, and found to have been as specific as, and more sensitive than, the MMSE. It wasn't designed to be a delerium screen, however.
The 4AT test - this was designed for delerium and cognitive impairment screening. It takes the AMT4, and adds 3 more bits - objective alertness, attention (months of the year backwards), and evidence of acute change or fluctuating course. This last bit needs collateral history in some form. Scoring is straighforwards (see Word document on their website).
The biggest and most relevant study I found (419 patients) study (4) looked at the 6-CIT and the 4AT for combined dementia and delirium screening in the ED. They used expert diagnosis as their gold standard, arguing this is superior to using other, varyingly validated, screening tests; I agree. The 4AT had a PPV of 0.68, and an NPV of 0.99. The 6-CIT had a PPV of 0.35 and an NPV of 0.98.
The bottom line?
The 4AT is probably the test to go for. It appears to perform better in an ED population. It's easy to remember, and score; there's a simple, free document that could easily be stored for use in our ED. It does depend on some sort of collateral history however - which may mean we cannot complete it entirely. That's not the end of the world - the ED starts many things it doesn't finish in a patient care episode! I have always argued we 'set the trend'. Get it right at the front door, and everything falls into place. Start the 'ceiling of care' discussion. Begin a brief cognitive assessment. It informs and influences further care significantly.
Discussing this (not so) 'quick dip', I've become aware of organisation-wide work going on in this vein - watch this space.
What do you think about cognitive screening in the ED? Comments please!
References and useful resources (not open access - those that are have hyperlinks above)
1) Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment. Age & Ageing, 1972 Nov;1(4):233-8.
2) Locke, T et al. Assessing the performance of the four question abbreviated mental test in the acute geriatric setting. Acute Med. 2013;12(1):13-7. (abstract only obtained)
3) Dyer, AH et al. The Abbreviated Mental Test 4 for cognitive screening of older adults presenting to the Emergency Department. European Journal of Emergency Medicine 2017, 24:417–422
4) O'Sullivan D, et al. Validation of the 6-Item Cognitive Impairment Test and the 4AT test for combined delirium and dementia screening in older Emergency Department attendees. Age and Ageing 2018; 47: 61–68