This week, I taught the @UCSF IM residents about tools for cognitive assessment in hospitalized patients. For my very first #tweetorial – I’m sharing my pearls with #medtwitter! #FOAM #geriatrics #MedEd 1/

Throughout my training, I was taught to use the MoCA (Montreal Cognitive Assessment). But after a few months seeing patients at @VABostonHC, I noticed that most of my patients were getting a score of 21-23. Everyone especially seemed to think this was a hippo. 2/

I wanted to know why! So I looked into the development and validation of the MoCA a bit. Turns out, a cognitive test developed in a mostly white, highly educated, bilingual population (aka the people of Montreal) hasn’t really panned out for use among other groups. 3/

I found one study by @KSinkMD et al, in which out of 414 African Americans WITHOUT dementia, 93.5% scored under 26 – the standard cutoff used to diagnose cognitive impairment.

I stopped used the MoCA after that.

(Image adapted from https://www.ncbi.nlm.nih.gov/pubmed/26618003 ) 4/

I started using a test developed at the St. Louis VA by @drjohnmorley – the (unfortunately named) SLUMS. It has much better data for patients with varying levels of education. This is now my go-to cognitive test for patients from the US.

https://www.slu.edu/medicine/internal-medicine/geriatric-medicine/aging-successfully/assessment-tools/mental-status-exam.php

5/

I like the SLUMS because it’s free, easy to use, tests executive function, and has scoring ranges for mild disease, and varying by education level.

Pro tip: always print out the second page of the PDF with the shapes. Giving the whole thing to the patient is distracting. 6/

What I don’t like about the SLUMS: the story. Every woman I work with who has used it agrees. @drjohnmorley – how about an update? 7/

The SLUMS isn’t my go-to test for non-native English speakers. For those patients, I use the RUDAS (Rowland Universal Dementia Assessment Scale) out of @DementiaAus. Developed in a multi-lingual, multi-ethnic population, and validated in patients with less formal education. 8/

I like the RUDAS because it’s practical, but I haven’t yet figured out how to administer the “judgment” question despite reading the administration guide. Is the patient allowed to name looking for a crosswalk or a light? If any #dementia experts have tips, send ‘em my way! 9/

All of these tests are great, but they have limitations in the hospital. #delirium in hospitalized older adults is super common, and under-diagnosed. It’s the brain’s manifestation of acute illness, with many contributing factors. 10/

In order to diagnose dementia, you have to exclude delirium, to ensure that the patient is in their usual cognitive state. My favorite tool for doing so is the 3D-CAM, by @sharon_inouye and @MarcantonioEd.
https://www.ncbi.nlm.nih.gov/pubmed/25329203  11/

You can use the 3D-CAM both to make the initial diagnosis of delirium and to track the patient’s progress day-by-day. It is very fast, but you can make it even faster by skipping questions if the patient answers incorrectly. For example, if they miss 1, you can skip to 4. 12/

But what if your patient is delirious, and you still want to know whether they have dementia? What if they’re in the MRI? Or in a coma? That’s when my absolute favorite tool, the #IQCODE (Informant Questionnaire of Cognitive Decline in the Elderly), comes in handy! 13/

The IQCODE is essentially a structured way of taking a collateral history. It helps you understand both the patient’s cognition and their functional status, all in one handy tool! It’s endorsed by @alzassociation and available for free. https://rsph.anu.edu.au/research/tools-resources/informant-questionnaire-cognitive-decline-elderly 14/

I usually ask the informant the questions. My colleague @Geri_IPE suggests handing the sheet to them. I do find that many folks need a lot of redirection when using this test – people get side-tracked with the story of when Mom put the keys in the fridge. 15/

So there you have it – some tools to assess your patients’ brains! I’d love to hear your questions and any feedback on improving my tweet-teaching technique. Thanks for your eyeballs! 16/end