Thanks again for your complete answer.

MMSE is a rough estimate to screen for cognitive deficits. The loss of autonomy depends heavily on the person itself as well the conditions of living. For quantification, on should account for living status (alone/couple/large family), premorbid resources, cognitive and emotional requirements in the current situation... This is not addressed by most of the cognitive scales we depose of.

Thanks Dirk for your answer.

I completely second Jordi!

The MMSE is a great screening tool often used to give a baseline and then re-evaluate at a later date or next presentation to see if there is any further decline. 

The MOCA or the ACE are much more comprehensive and often used as evidence when deciding on capacity and future planning.

The MMSE is not to be disregarded but it is also not to be used to dictate capacity because results can vary dependent on person completing it, clients educational background and cultural background.

Hope you have enough information

The MOCA (there are three versions to help in obtaining cleaner comparisons without bleed-over upon repetition) is what I use in my practice along with other tests. I rely heavily upon the clinical interview as well. The Independent Living Scales assessment is very robust in a practical sense in determining if a person has the skills to be safe living in their home environment.

It is a test of mental status best used to screen for dementia, and  MCI. It has been described as a test of general cognitive function which is a bit of an exaggeration as it is a test of mental status with a low ceiling. It is claimed that it measures different cognitive abilities by some, but these are individual items not composite scores.

Having said that we do find risk factors related to MCI scores. It is interesting how many investigators use the continuous distribution of MMSE scores in statistical analyses  when it clearly has two distributions. 

It has been used in place of a dementia screen, a practice that should not be encouraged.  It should be used to prompt a screen, not as a screen. When screening longitudinal change is important and scores considered in the impaired range need to be seen consistently over time.

I like to think of it as a measure of general mental status, not a measure of global cognitive performance.

Thanks Merril

You can read : Juva K et al - Acta Neural Scand 1994 ; 90 (4) : 293-298 or Aske D - Rehabil nurs 1990 ; 15(3) : 140-142 

Michele Sammicheli

Hi Michele, I am gerontologist Jan Vinita White from the U.S. I Use the MMSE and other cognitive assessments in revised formats and with caution.  I was approved by PAR [the copyright holder] to use the MMSE Alternate Word Version.  Practice effects have made the MMSE and other assessments inaccurate.  I have assessed hundreds of older adults in clinical and community settings.  When I administer the MMSE, I have had people respond with the words I did not ask them… they have memorized the online version!  I never ask them to spell WORLD backwards, or count by sevens, or repeat the phrase, “No if, ands, or buts”  because they come prepared with the answers.  That is why I never administer the original version from PAR.

I often substitute the  MMSE with SLUMS, GPCOG, Mini-Cog, MIS, and BAS.  However, some of those also require an adjustment for practice effects.  For example, the GPCOG, questions 1, 4, 6 are always changed.  For the Mini-Cog, I always substitute the words Banana, Sunrise, and Chair.  I also change the clock time.  I rarely use the MIS because clients practice it beforehand.  The BAS also requires changes for practice effects.  Specifically, questions 2 and 3.  I have had people come to an assessment and when I administer the BAS with changed words, they say,” Apple Table Penny” but those were not the words I gave to them.  Then, they get grumpy and ask why I changed it.  Once, after I administered the BAS, the client said, “Well, what about the clock? I’ve been practicing.”  Here is an article by Buckingham et al. (2013) comparing MMSE and SLUMS.  I hope you find it useful.  Below is an article from my blog.  As you can see from the comprehensive GeroAssessment, below, autonomy measurement requires a complete battery and medical follow-up to determine if the older adult is able to remain independent.  There are at least 20 medical conditions that mimic dementia and they should never assume that dementia is normal aging.  It is not.  Thank you.  

Jan Vinita White, PhD, Gerontologist

Hi Readers, many people have asked me about a comprehensive GeroAssessment.  What exactly is that?  A GeroAssessment is a CONFIDENTIAL battery of assessments to provide an overall baseline of how an older adult is functioning both physically and cognitively.  A comprehensive gerontological assessment must be performed/evaluated by an aging expert, preferably with a doctoral degree and many years of experience.  By evaluating all of these assessments and appropriately interpreting the findings, the older adult then can take the assessment to a primary care physician and/or geriatric psychiatrist for follow-up treatment.  The GeroAssessment is also requested for older adults who are moving in to a Continuing Care Retirement Community [CCRC]. 

There is a problem with MMSE bootleg copies being passed around by practitioners.  A bootleg copy of an assessment is one that has been photocopied.  Assessments used by the aging expert CANNOT be copies.  Why?  Only people who are authorized to administer them are authorized to buy them.  

It is very important that during an assessment, behavior and affect are measured.  Example.  A man started pawing me and laughing inappropriately during an assessment and I knew how to handle that situation and avoid a confrontation.  That observed behavior and his reaction to my request was part of the report.  

What is an informant report?  It is given to the caregiver to ensure accuracy of the information.  The caregiver completes the assessment to determine similarities and differences in the responses and they are discussed with the geriatrician.  By getting information from two sources, the findings have more credibility.  

Below is a list of the assessments I use in conducting a GeroAssessment.  Not necessarily in this order.  Older clients must bring all of their prescription medicines to the appointment. 

1. Nestle Food/Nutrition Assessment + Informant Report

2. Katz ADL

3. KATZ IADL

4. Drug Inventory + Informant report

5. Geriatric Depression Screening [GDS-SF] + Informant Report

6. Mood Inventory + Informant Report

7. Trail Making Test, B [TMT-B]

8. Freund Clock Drawing Test [CDT]

9. Epworth Sleep Assessment

10. A cognitive assessment.  Examples are MMSE Alternative Word Set version [I have been approved to modify it.  The MMSE has practice effects that interfere with accurately interpreting the results.] and   modified versions of BAS, SLUMS, GPCOG, Mini-COG, and MIS.

Ok, thanks Jan Vita White for your answer. 

Really thoughtful answers from some experts here - I have to say that I spent a LOT OF TIME analyzing decline on MMSE and BMC (but also many other specific cognitive measures) in both normal elders and those with dementia. My real goal was predicting the rate of cognitive decline and activities of daily living (ADLs, which indeed might serve as a proxy measure of "autonomy"). We got very bogged down and side-tracked with the very basic question of the best metric of decline - not just the best measure of cognition, but how to measure the rate of decline... BELOW is a paper from ADRC (AD Research Center) data collected at Johns Hopkins (where Folstein & Folstein developed the MMSE) a while back that addresses PREDICTORS of RATE of COGNITIVE DECLINE  using the MMSE. I know it is not a perfect measure - but "practice effects" are really only present in non-demented healthy elders, or vert mildly-effected elders who "study up" for the "test". I have literally 100s of analyses I never published because we were never 100% sure how best to quantify the rate of decline... one twist is that we cannot assume linear decline over time.

https://www.researchgate.net/publication/223252324_Predicting_Rate_of_Cognitive_Decline_in_Probable_Alzheimer%27s_Disease

Article Predicting Rate of Cognitive Decline in Probable Alzheimer's Disease

Thank you Xeno.

Absolutely not.  The reason being there are many measures with superior test-retest reliability and sensitivity to decline in dementia.  The measure you choose to track decline will also depend on the stage of dementia you are looking at.

The question is : is the MMSE test  correct in the evaluation of cognitive decline ?

As Jordi wrote, depends on the type of cognitive decline (max in Alzheimer for example ; minimum in FTD ). In Lewy-body dementia it's characteristic a strong variation in the cognitive performance tested by MMSE .  

MMSE is a screening test. When it shows critical elements , other evaluation tests must be considered .

But the point is a number and other diagnostic opportunities must be considered. Obviously.....a point cannot made a diagnosis .....

You must also consider that in Italy MMSE point is useful (and only MMSE point) for verifing the utility of ACHE-i and their prescription  by note (note 85 by AIFA ).     

No The MMSE is NOT valid, I say after 42 years experience assessing & treating subtle cognitive decline.   This screen, though widely reported by lazy researchers, is seriously misleading.     Paul

Wow - This is getting pretty good, and is an on-going debate. I can safely say that the MMSE is not the most reliable nor valid measure of cognitive ability - it has many flaws; for example ceiling & practice effects for mild dementia or highly educated people and floor effects for severe dementia, so no wonder Paul does not like it. But all cognitive measures have limits - there are certainly better tools for many aspects of measuring cognitive ability, but it can indeed be used to track cognitive decline... one of the GREAT advantages is it's simplicity and it's MINI nature allows it to be administered rapidly. I have compared longitudinal slopes in normal aging and dementia on many measures - MMSE slope over time does not look much different from more sophisticated cognitive measures or ADL measures. Paul, if working mild very mild decline, the MMSE would be nearly useless, but if tracking overall cognitive impairment in progressive diseases, MMSE is pretty useful, but still not the best... I do not take offense to the lazy researcher comment, as SOME have chosen MMSE out of laziness or ignorance. But I do wonder which form of validity you have the most concerns about? In my past "labs", we gave extensive batteries of cognitive and other measures, amazing how often the MMSE and other measures gnerally agreed. Often, the problem is the MISUSE or mis-interpretation of the MMSE rather the MMSE itself. Screening and staging are it's best use, should not be given more than a few times per year and should be standardized - that was the very point of Folstein & Folstein in the first place - standardize & score an age old approach to assessing mental status by clinical interview, since 1975, others have been developed, but in many ways MMSE is still the standard for THAT type of tool. BIMC, ADAS, CamCog.... whatever, those are all fine... 

In Italy, geriatric examinations used for medicolegal evaluations of disabilities are usually based on MMSE, ADL and IADL.

MMSE index less than 12-13 usually is considered an index of inability of the patient to make activities of living without the help of other person.

Xeno, my bad.    Thus I amend: If the MMSE is NOT EVER used as a stand-alone, and IS restricted to n=1 observations, It may be useful. However, with several better (eg, fewer false-pssitives) brief screens, why ever use MMSE?

 In my practice, I use MMSE, ADL and IADL, as well as comprehensive medical history, personal screening, and interactive evaluations.  Holistic evaluations of mind, body, and emotion (spirit) allow for more effective treatment planning.  Range of cognitive decline is primarily based on the expertise of the clinician assessing ranges from mild to moderate to severe.  For medical, legal, or insurance purposes considerations determine care, housing, et.al.  Where a cognitive disorder or AD, the stage of decline affects treatment.  

To answer your question "is the MMSE correct in evaluating cognitive decline?"

It is a tool, a resource in a range of therapeutic assessments and not to be exclusive.  It may be a baseline but is limited by external factors as time of day, medication, comorbidity symptoms, infections disturbing cognitive functioning as UTI.  This is a larger issue for geriatricians administering the MMSE and diagnosing for insurances.  It is inconclusive and, in my opinion, not a clearly defined diagnostic tool.

Thank you Andrea. I also think so!

I am a gerontologist and have assessed hundreds of older adults in practitioner and community settings for almost 20 years.

The MMSE is not reliable, due to practice effects.  It means that they have been administered the MMSE before on multiple occasions....or have memorized it from online practicing, that it is ineffective.  I have been approved to revise it by the copyright owner, PAR, and I use the revision.  I have had "complaints" from clients, who balked because they knew the "other" answers.  For example, instead of spelling WORLD backwards, I use EARTH.  I do not use the pencil or watch.  I have a different phrase than, "No if, ands, or buts."  I ask them to count backwards by fives.

I prefer the Mini-Cog [see link below], as there are six versions to offset practice effects. I have also been given permission for a "clock" revision.  On the clock, I use the time 8:20.  Funny story: When I conducted free memory screening in the community for Alzheimer's Foundation of America, one man was thrown off by my use of the revised MMSE instead of the Mini-Cog I used the previous time.  He said, "Where's the clock?" and I explained that this time, there was no clock.  His response was, "Oh heck, I have been practicing."  

A major problem with the MMSE is bootleg copying.  We are not authorized to copy it, we must use the tricolor original form.  Practitioners who use a copy are usually not authorized to administer it.  PAR approves the practitioner, and only after their approval, it can be purchased.  They do not sell it to anyone they do not authorize. Malpractice could be the result of using a bootleg copy.   

The ADL and IADL assess independence, not cognitive functioning.  The DRS-2 is probably the most accurate, but cumbersome.  That is why I use a combination of cognitive assessments instead of relying solely on the MMSE.

Michele, I am not sure of what you mean by autonomy.  Older adults have autonomy until such time that they cannot make informed decisions.   Dr. White

http://www.alz.org/documents_custom/minicog.pdf

Thank you Jan for your analized answer.

 It is correct for some applications and not others. It is a low ceiling test of mental ability. It is sometimes used as a measure of cognitive ability in large epidemiological samples where they have no other measures or few. It is valuable in identifying people with clinical cognitive impairment, or better state: in screening for cognitive impairment. It is not a high ceiling test of cognitive ability or cognitive function in persons with normal ability. It should not ever be used alone to diagnose dementia or mild cognitive impairment, but rather in a screening context. Despite its limitations it is widely used and thus there is no harm in including it along with other tests. It should not be used to measure specific abilities because too few items measuring these abilities are available. It is a marvelous clinical tool but very limited as a measure of cognitive ability in normal functioning persons or where the sample is largely normal in cognitive ability. 

Many investigators ignore its distributional properties, often bi-modal in statistical analyses.

Michele, your question has raised a storm of useful answers, AND has revealed a continuing inertia in cognitive assessment. Very puzzling that an old screen, developed by well-meaning researchers when not much else was available, is still being used JUST due to its age. MMSE has been cited in medical & psychological literature so many times, for so long, that  it wears the unearned mantle of wisdom.  MOCA, MiniCog, and other screens, that can be administered in about the same amount of time, are valid & reliable for both test & retest; they should be used instead of the MMSE.     Good fortune in your work,  Psychologist Paul McGaffey

Thanks Paul. In fact, a fruitful discussion has developed concerning the use of the MMSE in the medico legal field.

Of course the scale is still use. It is a very valuable tool. Why do folks try to make it food or bad [present company excluded]. It is good for some applications and not for others. It is a very valuable screening instrument and very useful in low performing samples. Please see my previous comment. Merrill

Paul's comment is very helpful and I agree.  We often see the MMSE in studies relating CVD and cognition because folks use archival data and that is the only measure they have. 

I would be very helpful if psychometric methods were to be taught in Medical Schools, but how much can one add to the curriculum? On the other hand, cognitive testing is emerging as a major outcome measure in medical research, e.g MCI, AD and VAD.

Unfortunately  there are still researchers that accept, whiteout challenge, that clinical test measure what they say they measure and too many psychology programs that no longer offer courses in psychometric medicine.

Merrill Elias, PhD, MPH

Thank you Elias

Any test, including the MMSE can be used in evaluating cognitive decline if you mean this in the sense of longitudinal decline [decline over time]. The issue is how this is done via statistical analysis. The test has a bi-modal distribution [this must be taken into consideration]. If used in its primary context, a screening instrument then thee, at least longitudinal measurements are needed as scores can vary as much as 3 points across time as minor changes affect the score. One does not define BP at a single session in the clinic, nor should possible cognitive deficit be decided on the basis of a one time score on the MMSE. 

Decline is of course a change over time and need not indicate impairment unless change is in the impaired range based on normative data.

Merrill F. Elias, PhD, MPH

thanks

Hi Michele, I am a gerontologist and have used the MMSE in the field to assess in the community for the Alzheimer's Foundation of America.  The MMSE is a quick assessment.  It has reduced validity due to "practice effects" and people finding the answers online.  I have found unqualified persons administering it, and using bootleg copies is a major problem.  It is copyrighted and using bootleg copies is an ethical issue because PAR grants permission to buy it and administer it.  Administering it without permission, and/or using copies is malpractice and it could harm people.  

I have been approved by PAR to revise it, and I do.  Instead of WORLD, I use EARTH.  I substitute a button and an ink pen for the watch and pencil.  "No ifs ands or buts" is changed to "It is a lovely day in the neighborhood."  I change "close your eyes" to "blink your eyes."  Jan Vinita White, PhD

Thank you Jane for the answer.

There is much literature on this topic. It is a low ceiling  measure of cognition but very useful in SCREENING for dementia and MCI; there seems to be and unfortunate trend toward using  to diagnoses dementia and mild cognitive impairment. It absolute cannot be used in this context. It is a screening measure. One merely needs to go to an immense available  literature to answer this question.  

thank you Merril for your answer

MMSE is good tool for screening, therefore is of immense value in prevalence studies related to dementia and MCI wherein it can effectively be used in a two step process to identify patients with dementia and MCI. Generally the second step in this process is clinical assessment. 

MMSE is a screening tool for dementia. As with all screening tests, it has its limitation. So for diagnostic purposes clunical eval with MMSE will be useful. I personally belief MMSE and its modifications are of great value in population based screening surveys to identify suspects of dementia and MCI. I have realised one limitation in its use among illiterate individuals though, maybe because MMSE relies heavily on being literate.

Thanks Flavia both for starting this conversation and for your own contribution to establishing usefulness. I think bottom line dementia and its assessment is a complex issue for which there is not a simple measure. We should neither dismiss the MMSE,... it's a useful starting point but we should not use it alone to establish a diagnosis. Also we should be clear what are question really is in the use/choice of a measure and many uses are beyond the intent of the measure. Finally even in screening the MMSE can quickly lose sensitivity as cognitive impairment advances, and, as you note, may not be sensitive enough in very early early stages again suggesting we should be clear about who as well as for what we plan to use it. All of that said there is an ease of use and an acceptability to patients that should not be discounted.

Thank you Flavia for your reference.

MoCA test is useful in the evaluation of mild cognitive impairment.