MMSE practice effect

[nexus73]

Recently saw a patient who had MMSE score of 24/30, 6 months ago, now 28/30. College educated. No medical reason to explain lower score 6 months ago vs better score recently. Definitely has subjective memory complaints, forgetful, gets lost in unfamiliar surroundings. Recent score she missed 1 on delayed recall, 1 on orientation (wrong floor of building). This is an individual who really wants to do well and is resistant to dementia diagnosis. Is it possible practice effect would last 6 months? Also was started on Namenda six months ago, is it possible cognition could improve on a screening test from this med?

Curious if there are any articles on practice effect increasing score on MMSE.

---

Members don't see this ad.

 

[WisNeuro]

The practice effect at six weeks is pretty small, so I wouldn't expect that the 6 month would be significant. Any possibility that they had it done at other providers appointments? Test-retest for the visuoconstructional, registration and recall portions is not great in general. I wouldn't expect Namenda to really do much at 6 months unless she's a rare "responder." I'd much prefer the MoCA for screening, also has alternate forms. Has this person had a neuropsych assessment?

 

[Stagg737]

Interesting question and I'm curious about what others say. This is why tests like MMSE or SLUMS are just screens and patients with suspected neurocognitive deficits should undergo a complete assessment though. Who is to say she wasn't just having an "off" day when she got a 24/30 (poor sleep, stressful or busy day, received bad news, etc)?

 

[WisNeuro]

Interesting question and I'm curious about what others say. This is why tests like MMSE or SLUMS are just screens and patients with suspected neurocognitive deficits should undergo a complete assessment though. Who is to say she wasn't just having an "off" day when she got a 24/30 (poor sleep, stressful or busy day, received bad news, etc)?

Yeah, I see plenty of variability in screener scores quite often. Screener scores are often the least useful competent of the chart review/formulation. This is where the comprehensive neuropsych, with collateral info, comes in handy.

 

[annoyedpsychiatrist]

well namenda slows decline rather than reversing decline, but the magnitude that it slows with aricept from what ive seen, hasnt been by a large amount if i recall. Ultimately it still runs its course. My guess is that the difference in the patients scores is purely just chance/perhaps an off day, from what you said.

i personally prefer the MOCA when im concerned about cognitive impairment.

 

[WisNeuro]

well namenda slows decline rather than reversing decline, but the magnitude that it slows with aricept from what ive seen, hasnt been by a large amount if i recall. Ultimately it still runs its course. My guess is that the difference in the patients scores is purely just chance/perhaps an off day, from what you said.

i personally prefer the MOCA when im concerned about cognitive impairment.

When you look at the data, and account for the inappropriate use of things like LOCF, it really doesn't. Larger scale, real world data (e.g., AD2000 study) showed no difference in cognitive measures or time until institutionalization.

 

[mistafab]

If you are concerned about dementia, I would use a more appropriate clinical scale such as the Dementia Rating Scale. It takes some practice, and some extra time, but is much more useful specifically w/ dementing illness and its severity. It does not require neuropsych testing. Neuropsych testing is another avenue if you are specifically concerned about domains of impairment.

MOCA/MMSE are great screeners

 

[WisNeuro]

If you are concerned about dementia, I would use a more appropriate clinical scale such as the Dementia Rating Scale. It takes some practice, and some extra time, but is much more useful specifically w/ dementing illness and its severity. It does not require neuropsych testing. Neuropsych testing is another avenue if you are specifically concerned about domains of impairment.

MOCA/MMSE are great screeners

The DRS-2 is marginally better than screeners, generally speaking. It has a fairly low ceiling, so it's not great at picking up subtle early signs of dementia, or some of the non-AD dementias in early stages. If you don't have the option of a neuropsych eval, it's better than a screener, but not by much.

 

[mistafab]

The DRS-2 is marginally better than screeners, generally speaking. It has a fairly low ceiling, so it's not great at picking up subtle early signs of dementia, or some of the non-AD dementias in early stages. If you don't have the option of a neuropsych eval, it's better than a screener, but not by much.

I think to a practicing clinician, picking up subtle early signs of dementia is reading tea leaves. If we had a better system, we would be using it (such as the MAB trials targeting amyloid). For a general clinician, careful history gathering smattered onto a DRS is much more useful if going with the clinical score (1 or greater), given that what we care about is finding who has dementia or not, and where in their life it is most impairing.

Picking up early dementia is not where we are clinically. MOCA is great to identify impairment, but won't give you useful information about how it is impairing their life (such as a DRS), where a clinician can assist in targeted interventions from family, or even recommend the family hire a care coordinator.

By all means, test specific areas of neuropsych profiles. But to the treating physicians, it is more important to identify where we can intervene.

 

[WisNeuro]

I think to a practicing clinician, picking up subtle early signs of dementia is reading tea leaves. If we had a better system, we would be using it (such as the MAB trials targeting amyloid). For a general clinician, careful history gathering smattered onto a DRS is much more useful if going with the clinical score (1 or greater), given that what we care about is finding who has dementia or not, and where in their life it is most impairing.

Picking up early dementia is not where we are clinically. MOCA is great to identify impairment, but won't give you useful information about how it is impairing their life (such as a DRS), where a clinician can assist in targeted interventions from family, or even recommend the family hire a care coordinator.

By all means, test specific areas of neuropsych profiles. But to the treating physicians, it is more important to identify where we can intervene.

I get the sentiment. But, I don't see this as very helpful. We only pretty much break out the DRS-2 when someone is obviously demented, as a way to track decline over time. The ceiling is just so low. It's usefulness is determining dementia or not is pretty low in incremental validity. Also, teh DRS doesn't map on all that well to things like specific i/ADls and such. It's pretty much just an "impaired or not" ranking, you can't really interpret the "domains" there with any real sense of usefulness in terms of ecological validity. At that level, your time is much better spent just getting collateral information about function.

 

[mistafab]

I get the sentiment. But, I don't see this as very helpful. We only pretty much break out the DRS-2 when someone is obviously demented, as a way to track decline over time. The ceiling is just so low. It's usefulness is determining dementia or not is pretty low in incremental validity. Also, teh DRS doesn't map on all that well to things like specific i/ADls and such. It's pretty much just an "impaired or not" ranking, you can't really interpret the "domains" there with any real sense of usefulness in terms of ecological validity. At that level, your time is much better spent just getting collateral information about function.

Hmmm... I can see what you're saying here. There may also be some mismatch as I am not familiar with the DRS-2. Most of the DRS i've done has always been with collateral as the patient themself isn't very useful, and it is more integrated into a formal intake evaluation to make it more specific and useful.

 

[WisNeuro]

Hmmm... I can see what you're saying here. There may also be some mismatch as I am not familiar with the DRS-2. Most of the DRS i've done has always been with collateral as the patient themself isn't very useful, and it is more integrated into a formal intake evaluation to make it more specific and useful.

Are you administering the instrument with the collateral in the room?

 

[Fan_of_Meehl]

The DRS-2 is marginally better than screeners, generally speaking. It has a fairly low ceiling, so it's not great at picking up subtle early signs of dementia, or some of the non-AD dementias in early stages. If you don't have the option of a neuropsych eval, it's better than a screener, but not by much.

So, hypothetically speaking, some of the VA 'neuropsych' providers who write reports that spotlight the findings from the DRS-2 as the centerpiece of their evaluation and who don't utilize any symptom validity or performance validity elements in their evaluations...might need to be taken with a grain of salt? They also essentially recommend 'Lumosity' type 'cognitive training' in their recommendations even though these have been thoroughly debunked.

 

[mistafab]

Are you administering the instrument with the collateral in the room?

Typically yes. If the concern is dementia, an independent visit is set up for a thorough eval. Must have some collateral in the room or on speaker for the duration. Careful emphasis on domains of functioning, IADLs, ADLs, as well as red flag events (stove on, lost from store, driving, etc).

I think a thorough dementia eval takes time, typically 30-45 minutes, withthe DRS filled out at the same time. The last 15-30 minutes become addressing deficits and future care planning.

MOCA/MMSE are useful measures to quickly see if an eval is needed, but when need is possibly identified, I’m not sure how helpful someone can be without knowing the above and coordinating with interested parties (involved family). Perhaps some providers just refer anyone with “memory issues” to neuro or geriatricians to do the above, or just send for neuropsych testing and expect them to make things more clear.

 

[WisNeuro]

Typically yes. If the concern is dementia, an independent visit is set up for a thorough eval. Must have some collateral in the room or on speaker for the duration. Careful emphasis on domains of functioning, IADLs, ADLs, as well as red flag events (stove on, lost from store, driving, etc).

I think a thorough dementia eval takes time, typically 30-45 minutes, withthe DRS filled out at the same time. The last 15-30 minutes become addressing deficits and future care planning.

MOCA/MMSE are useful measures to quickly see if an eval is needed, but when need is possibly identified, I’m not sure how helpful someone can be without knowing the above and coordinating with interested parties (involved family). Perhaps some providers just refer anyone with “memory issues” to neuro or geriatricians to do the above, or just send for neuropsych testing and expect them to make things more clear.

For any of the actual cognitive/screening piece of the evaluation, I would strongly suggest that you do that part with only the patient, otherwise, your scores are likely underestimates. We have a fairly robust lit on third party observer effects. There should only be another person in that room in rare circumstances.

As for the neuro testing, quite often we actually find issues with comorbid neurodegenerative disorders, or that it's actually a different disorder than the one thought by referral, which can sometimes significantly change treatment and the feedback that the family should be given.

 

[WisNeuro]

So, hypothetically speaking, some of the VA 'neuropsych' providers who write reports that spotlight the findings from the DRS-2 as the centerpiece of their evaluation and who don't utilize any symptom validity or performance validity elements in their evaluations...might need to be taken with a grain of salt? They also essentially recommend 'Lumosity' type 'cognitive training' in their recommendations even though these have been thoroughly debunked.

The DRS-2 really only has useful capacity in the CLC, inpatient, or a research protocol as far as the VA is concerned. Anyone using it otherwise, doesn't know what they are doing in that setting.

 

[PsyDr]

DRS= "This is obvious. There is no testing necessary, but they are expecting me to administer something."

 

[nexus73]

Thanks for the responses.

I've always heard that high intelligence or educational attainment may cause false negatives on cognitive screening tests, especially MMSE being less sensitive. Is this backed up by research or just myths that get propagated in medical training? I can't seem to find any, maybe using the wrong search terms. Does anyone have a reference to back this up?

 

[nexus73]

Is there any data on observers improving scores, like if family is there maybe giving subtle guidance?

 

[WisNeuro]

Thanks for the responses.

I've always heard that high intelligence or educational attainment may cause false negatives on cognitive screening tests, especially MMSE being less sensitive. Is this backed up by research or just myths that get propagated in medical training? I can't seem to find any, maybe using the wrong search terms. Does anyone have a reference to back this up?

The MMSE is less sensitive, especially to non-AD dementia (e.g., vascular) I don't have any articles about that in my database at the moment, but it's a known quantity. One of the reasons we'd always recommend a MoCA, or SLUMS if you need it. As for an education effect, we do see a correlation between education and performance, especially on the MoCA. There are a few good community sets like the Rosetti norms, and at least one decent meta-analysis last I checked.

Is there any data on observers improving scores, like if family is there maybe giving subtle guidance?

Not aware of anything to this effect, but if you have a TPO, that's bad enough, if you have placed them within sight line of the patient, forget about any of that data being anywhere near valid.