How is decline assessed in clinical practice?

[neuropsychstudent2021]

Often in research decline may be assessed by asking a simple question like "do you have memory problems", which of course doesn't actually take into account if the patient has actually declined or not (what if the patient always had memory problems?!?). In clinical practice how is this criterion met?

This is in the context of mci/dementia.

---

Members don't see this ad.

 

[AcronymAllergy]

Short answer: neuropsych testing, or at least cognitive screening. The relationship between self-reported cognitive impairment and objective cognitive impairment is generally tenuous at best.

 

[neuropsychstudent2021]

Sorry, I mean how do we know that there has been a change in that individual. To establish that is part of the criteria for dementia rather than just objective impairment.

 

[WisNeuro]

Depends if it's the first time seeing someone or doing a re-eval. First time, you have to look at functioning compared with estimated premorbid function. But, in essence, the objective testing is just hypothesis testing and added predictive validity. Most of your diagnosis is weighted towards information gathering from record review, clinical interview, and collateral.

 

[Sanman]

This is why we have normed data and we estimate pre-morbid level of functioning. As @WisNeuro mentioned, this is why detailed history is important. I can't tell you how many family members have said to me, oh mom/dad has not had any memory/cognitive problems until now. Then you ask...so they pay all their own bills? Oh no, we took over that years ago because it was a mess. Ok then...

 

[neuropsychstudent2021]

Right so, in the interview, record review and collateral thats where we are getting the evidence of decline plus if there is impairment in IADL?

 

[neuropsychstudent2021]

@Sanman How do we establish premorbid? Is the discrepancy between estimated and premorbid all thats needed?

 

[Sanman]

@Sanman How do we establish premorbid? Is the discrepancy between estimated and premorbid all thats needed?

Norms are generally based on one's age and education. You can compare to a similar population and against one's general level of functioning. There is impaired compared to norms, impaired compared to history, and impaired compared to other abilities. You don't expect:

1. Anyone to be two SDs below average of people with their age and education
2. Someone with a PhD, MD, or a former CEO to have low average cognitive abilities
3. Someone with a WAIS IQ of 130 to score 1SD below the mean in delayed memory.

Is it 100% accurate all the time? No, but it is the best we got.

 

[neuropsychstudent2021]

And what if there is conflicting information from the record review,interview and collateral compared to the discrepancy shown in objective testing? For example if the collateral indicates no decline plus no impairment in adl.

 

[Sanman]

And what if there is conflicting information from the record review,interview and collateral compared to the discrepancy shown in objective testing? For example if the collateral indicates no decline plus no impairment in adl.

If there is no decline noted by anyone in any area, why are they in your office?

 

[neuropsychstudent2021]

Its more of a hypothetical really, but it could be something like they were part of research or a routine screen by a PCP (under the presumption that decline was not indicated).

 

[Sanman]

Its more of a hypothetical really, but it could be something like they were part of research or a routine screen by a PCP (under the presumption that decline was not indicated).

Don't diagnose people based on cognitive screeners alone (although physicians do it all the time), refer for further assessment. That is what neuropsych is for. Batteries of tests are given to establish a pattern of deficits not just a fluke on a few items. Also, don't make clinical diagnoses based on incomplete data from a research study.

This is like asking whether the police could solve a murder that no one reported to the police. It doesn't happen in real life.

 

[neuropsychstudent2021]

But thats what I mean, say someone was given a neuropsychological battery of tests as part of a study and they scored -2sds on a verbal learning test (but have no evidence of an actual decline), that doesnt mean they should be diagnosed with mci. Im gathering from your comment that because they have a discrepancy with their peers and their estimated premorbid that they meet all criteria for mci.

 

[Sanman]

But thats what I mean, say someone was given a neuropsychological battery of tests as part of a study and they scored -2sds on a verbal learning test (but have no evidence of an actual decline), that doesnt mean they should be diagnosed with mci. Im gathering from your comment that because they have a discrepancy with their peers and their estimated premorbid that they meet all criteria for mci.

That's not how you diagnose. What does the rest of the battery say? Consistent deficits throughout the tests? Effort? Attention? Behavioral observations? Was this person on their phone during the test? Did someone walk by and threaten to kill them (happened to someone I know while testing an inmate). Any neuroimaging available? Like I said, this is why you don't diagnose based on one test out of context. Neuropsych isn't just about test scores, but interpreting the data as a whole.

 

[AcronymAllergy]

It's not unusual for the client/patient to deny or fail to report declines, for a variety of reasons (e.g., anosognosia, "traditional" denial, etc.). It's also not entirely unusual for a collateral not to report decline (maybe they don't want to admit there's a change, maybe they aren't around enough to see one, maybe the patient has a very structured routine and is pretty good at hiding any problems), but a lot of that can relate to what questions you ask and how you ask them. In terms of diagnostic criteria specifically, MCI requires concern for change in cognition, but that can include concern from a clinician observing the patient.

We don't interpret test performances in isolation, so it's hard to come up with a simplified example. But if everything else in the evaluation points in the direction of cognitive decline despite the patient denying any changes, and the patient scores at -2SD in a cognitive domain, that's a large discrepancy from what would be expected and would generally be very unusual in the cognitively-healthy population. So the clinician very well might be considering MCI.

Now, if a person came in for a research study, went through a limited interview in which they denied any cognitive concerns or functional changes, and scored at the 2nd percentile on a single memory test, would I diagnose MCI? No. Would I be concerned and think the person needed further evaluation? Probably.

You also have to determine which MCI criteria you're using, or if you're going to instead use DSM-5's Mild Neurocognitive Disorder. Criteria are similar but not identical. NIAAA diagnostic recommendations are that MCI entails: concern for cognitive change, impairment in one or more cognitive domains (defined as "evidence of lower performance in one or more cognitive domains that is greater than would be expected for the patient’s age and educational background" and described as typically 1 to 1.5 SD below the mean of normative data), preservation of independence in functioning, and absence of dementia.

Like Sanman said, neuropsych assessment is a body of knowledge rather than being solely about administering and scoring cognitive testing. While the question you're asking is simple, the answer is not.

 

[neuropsychstudent2021]

It's not unusual for the client/patient to deny or fail to report declines, for a variety of reasons (e.g., anosognosia, "traditional" denial, etc.). It's also not entirely unusual for a collateral not to report decline (maybe they don't want to admit there's a change, maybe they aren't around enough to see one, maybe the patient has a very structured routine and is pretty good at hiding any problems), but a lot of that can relate to what questions you ask and how you ask them. In terms of diagnostic criteria specifically, MCI requires concern for change in cognition, but that can include concern from a clinician observing the patient.

We don't interpret test performances in isolation, so it's hard to come up with a simplified example. But if everything else in the evaluation points in the direction of cognitive decline despite the patient denying any changes, and the patient scores at -2SD in a cognitive domain, that's a large discrepancy from what would be expected and would generally be very unusual in the cognitively-healthy population. So the clinician very well might be considering MCI.

Now, if a person came in for a research study, went through a limited interview in which they denied any cognitive concerns or functional changes, and scored at the 2nd percentile on a single memory test, would I diagnose MCI? No. Would I be concerned and think the person needed further evaluation? Probably.

You also have to determine which MCI criteria you're using, or if you're going to instead use DSM-5's Mild Neurocognitive Disorder. Criteria are similar but not identical. NIAAA diagnostic recommendations are that MCI entails: concern for cognitive change, impairment in one or more cognitive domains (defined as "evidence of lower performance in one or more cognitive domains that is greater than would be expected for the patient’s age and educational background" and described as typically 1 to 1.5 SD below the mean of normative data), preservation of independence in functioning, and absence of dementia.

Like Sanman said, neuropsych assessment is a body of knowledge rather than being solely about administering and scoring cognitive testing. While the question you're asking is simple, the answer is not.

Quite the thorough answer there thanks for the response. I guess my lack of knowledge led me to asking this question that cant really be answered over this medium (like I say im just a first year student trying to learn as much as I can).

Does the age of the patient come into account when diagnosing these disorders?

 

[AcronymAllergy]

Quite the thorough answer there thanks for the response. I guess my lack of knowledge led me to asking this question that cant really be answered over this medium (like I say im just a first year student trying to learn as much as I can).

Does the age of the patient come into account when diagnosing these disorders?

It's definitely something taken into consideration in the context of the full evaluation, yep.

And nothing wrong with asking questions, no worries there. It's just that the answer to your question relates to a substantial body of knowledge. If it's something you're interested in, you'll definitely learn much more about it during the course of your training.

 

[Therapist4Chnge]

Like Sanman said, neuropsych assessment is a body of knowledge rather than being solely about administering and scoring cognitive testing. While the question you're asking is simple, the answer is not.

This is really the crux of neuropsych consultation. So much of my daily practice (80%+) is not formal assessment. Record review, interview & collecting collateral info, and reviewing other ppl's data (e.g. imaging, labs, etc) , which helps bring context to the case.

 

[neuropsychstudent2021]

This is really the crux of neuropsych consultation. So much of my daily practice (80%+) is not formal assessment. Record review, interview & collecting collateral info, and reviewing other ppl's data (e.g. imaging, labs, etc) , which helps bring context to the case.

What in the record review,interview and collateral leads us to consider decline? And if there is a discrepancy (like low scores on testing but interview indicates no decline) then what do we do?

 

[Therapist4Chnge]

These are all things that will get covered in training, but I can offer some basic things. Red flags can show up in annual reviews at work, friends and family reporting noticable changes, and/or from a PCP who has had the patient for a long time. Family complains of functional changes like getting lost going to common places, driving to an old house thinking they still lived there, talking about friends/family from years ago and not realizing they passed/moved/divorced/etc. Low hanging fruit would be imaging changes, but this tends to happen after the suspected concern, but sometimes is found incidentally. Someone was involved in a car accident and they had imaging done to rule out bleeds, and that imaging reveals volume loss, atrophy, etc.

Most of the time the testing is more confirming what is suspected, but not always. Cases are usually messy, so there could be co-morbid dx like hypertension, obstructive sleep apnea, substance abuse, etc. Each one can have cog impact, but they also could be added to the disease progression.

There are posters on here who handle a lot of dementia and MCI cases who can provide more insight into these areas. I work almost exclusively w. acquired brain injury...so I only see these cases here and there.

 

[neuropsychstudent2021]

These are all things that will get covered in training, but I can offer some basic things. Red flags can show up in annual reviews at work, friends and family reporting noticable changes, and/or from a PCP who has had the patient for a long time. Family complains of functional changes like getting lost going to common places, driving to an old house thinking they still lived there, talking about friends/family from years ago and not realizing they passed/moved/divorced/etc. Low hanging fruit would be imaging changes, but this tends to happen after the suspected concern, but sometimes is found incidentally. Someone was involved in a car accident and they had imaging done to rule out bleeds, and that imaging reveals volume loss, atrophy, etc.

Most of the time the testing is more confirming what is suspected, but not always. Cases are usually messy, so there could be co-morbid dx like hypertension, obstructive sleep apnea, substance abuse, etc. Each one can have cog impact, but they also could be added to the disease progression.

There are posters on here who handle a lot of dementia and MCI cases who can provide more insight into these areas. I work almost exclusively w. acquired brain injury...so I only see these cases here and there.

Thanks for the response.

And if there is a discrepancy (like low scores on testing but interview indicates no decline) then what do we do?

 

[WisNeuro]

Thanks for the response.

And if there is a discrepancy (like low scores on testing but interview indicates no decline) then what do we do?

Short answer is, it depends. Off the top of my head I can think of over a dozen scenarios where that answer to that differs depending on circumstances of the case in from of me. As T4C stated, these are all things that will come with training and supervision. At your level, I would focus on gaining the foundational knowledge in assessment and neuropsychological principles. The types of questions you are asking later on in this thread are best answered with clinical cases as there is no one size fits all answer.

 

[Fan_of_Meehl]

Short answer: neuropsych testing, or at least cognitive screening. The relationship between self-reported cognitive impairment and objective cognitive impairment is generally tenuous at best.

I think I read somewhere about research indicating that, overall, the observed correlation between self-reported memory dysfunction and actual demonstrated impairments in memory upon neuropsy testing is...zero. Anecdotally, clinical practice in the VA system confirms this.

 

[neuropsychstudent2021]

I think I read somewhere about research indicating that, overall, the observed correlation between self-reported memory dysfunction and actual demonstrated impairments in memory upon neuropsy testing is...zero. Anecdotally, clinical practice in the VA system confirms this.

If that's the case then why do we base a lot of the diagnosis on the clinical interview?

 

[Fan_of_Meehl]

If that's the case then why do we base a lot of the diagnosis on the clinical interview?

Yes. As well as consideration of all the multiple possible contributors to the observed self-report of 'memory problems,' their associated base rates, consideration of 'treatability' of the various hypothetical causes (if treatment planning is part of your workup). I would think you'd want to prioritize screening/evaluating for common causes (often very treatable) of memory complaints, e.g., things like mood/anxiety disorders, substance abuse, sleep hygiene/ apnea issues, poor diet, lack of exercise...each of which are high base-rate and highly treatable potential contributors to complaints about poor cognitive functioning. But, of course, if necessary or indicated, objective testing of cognitive functioning could be helpful as well. Of course this is all biased toward a treatment context within which the processes of assessment (and case formulation), hypothesis generation (e.g., what could be the causes here?), data gathering (e.g., observation, self-monitoring), implementation of treatment, analysis of outcome (subjective complaints improving over time or not?), and re-assessment are all interwoven and occurring across time going forward.

 

[AcronymAllergy]

If that's the case then why do we base a lot of the diagnosis on the clinical interview?

Like Fan_of_Meehl mentioned, the clinical interview entails much more than just asking a patient if they're having memory problems. And in most instances, when it comes to neurocognitive disorders, you really shouldn't be diagnosing based on the interview alone, even though it's a very important (and some would argue the most important) piece of the evaluation.

 

[Therapist4Chnge]

Like Fan_of_Meehl mentioned, the clinical interview entails much more than just asking a patient if they're having memory problems. And in most instances, when it comes to neurocognitive disorders, you really shouldn't be diagnosing based on the interview alone, even though it's a very important (and some would argue the most important) piece of the evaluation.

This.

A common interaction I have for mTBI cases is "I can't remember anything! I walk into another room and forgot what I was doing. I forget conversations."

In contrast, they can quickly and accurately recall their managing doc's name, the symptoms they reported to them, and a host of other details I can confirm independently.

Their PERCEPTION is that they are experiencing memory problems, but the data and interactions strongly demonstrate otherwise. Are mood, poor sleep, pain, and/or other factors contributing to their experience....it's very likely.

 

[Fan_of_Meehl]

And what if there is conflicting information from the record review,interview and collateral compared to the discrepancy shown in objective testing? For example if the collateral indicates no decline plus no impairment in adl.

Attached is a great article written for a smart lay audience (lawyers and law students) logically and comprehensively reviewing the literature on mTBI (in veterans) and cognitive complaints and the issues with establishing causality.

 

[PsyDr]

The clinical interview and behavioral observations provide a CONTEXT for the data (e.g., functional history, medical history, etc). A professional knows the base rates for memory complaints and the accuracy of those reports. Those reports can go either way. The true story of Rudy is VERY different than the movie (I.e., they were making fun of him, but he didn’t understand that).

 

[AcronymAllergy]

The clinical interview and behavioral observations provide a CONTEXT for the data (e.g., functional history, medical history, etc). A professional knows the base rates for memory complaints and the accuracy of those reports. Those reports can go either way. The true story of Rudy is VERY different than the movie (I.e., they were making fun of him, but he didn’t understand that).

Is nothing sacred?!

Kidding, I'd read about that before as well. Think they interviewed Joe Montana, Rudy himself, and a few other folks about it. Still a fun movie, truth be damned.

 

[clausewitz2]

Just chiming in with an example of very detailed functional history yielding valuable info not at all obvious from a screener: once worked with a retired professor of mechanical engineering (who at that point was cognitively mpaired beyond any shadow of a doubt) who first came to clinical attention because of a broken washing machine.

Y'see, it broke, and he opened it up to fix it, and realized after inspecting it he hadn't the faintest clue what any of the parts he was looking at did. He just couldn't make any sense out of the assembly. This would be any day ending in -y for many people but in context of his background it was extremely alarming for him.

As an aside is there a literature exploring any substantive hypotheses regarding the drivers of subjective cognitive impairment? People are really consistent in reporting certain kinds of subjective cognitive effects with various medical interventions, and in fact there is good agreement across patients as to the relative impairment of various AEDs, but my understanding is this is rarely detectable in more objective cognitive assessment. Clearly they are having some kind of consistent experience, even if they are mistaken about the functional impact, and it has always made me curious about what is going on.

 

[PsyDr]

Reliable Change Index

 

[AcronymAllergy]

Just chiming in with an example of very detailed functional history yielding valuable info not at all obvious from a screener: once worked with a retired professor of mechanical engineering (who at that point was cognitively mpaired beyond any shadow of a doubt) who first came to clinical attention because of a broken washing machine.

Y'see, it broke, and he opened it up to fix it, and realized after inspecting it he hadn't the faintest clue what any of the parts he was looking at did. He just couldn't make any sense out of the assembly. This would be any day ending in -y for many people but in context of his background it was extremely alarming for him.

As an aside is there a literature exploring any substantive hypotheses regarding the drivers of subjective cognitive impairment? People are really consistent in reporting certain kinds of subjective cognitive effects with various medical interventions, and in fact there is good agreement across patients as to the relative impairment of various AEDs, but my understanding is this is rarely detectable in more objective cognitive assessment. Clearly they are having some kind of consistent experience, even if they are mistaken about the functional impact, and it has always made me curious about what is going on.

As a quick reply, there's definitely been research looking broadly at what factors are correlated with report of cognitive change across various contexts, which helps provide some initial data to inform hypotheses on underlying mechanisms. If you had a specific medical intervention in mind, folks may have some knowledge and experience with that literature and/or with those patients.

 

[clausewitz2]

As a quick reply, there's definitely been research looking broadly at what factors are correlated with report of cognitive change across various contexts, which helps provide some initial data to inform hypotheses on underlying mechanisms. If you had a specific medical intervention in mind, folks may have some knowledge and experience with that literature and/or with those patients.

I think the one that is most striking to me in terms of having consistent inter-patient agreement on degree of cognitive impairment that is not reflected in validated testing would be anticonvulsants. I saw a paper (in English) from a South Korean group that had epilepsy patients rate the degrees of impairment they felt they experienced from different AEDs and these ratings seemed to map on very closely to the impressions US patients have about these meds. I found that interesting because while obvious ROK has plenty of people proficient in English they are not such contiguous societies that of course they will develop a uniform patient culture, as it were.

 

[WisNeuro]

Reliable Change Index

Yes, but I often see many people misuse the RCI. Mostly because they misunderstand the stats used to calculate the index. If they are relying test-retest stats from a 2 week interval vs a 1 year interval, your estimate will be...off. I have some papers somewhere that looked at test-retest over longer intervals (as most test manuals look at 2 to 6 week intervals) of a year-ish. And then there is work suggesting that practice effect in older and dementia individuals is greatly reduced. So, a good stat to have in some places, but only if you use the right data to compute it, and know its limitations. Unfortunately, with the de-emphasis on assessment and stats training in many programs these days, if this is even used, many times it is miscalculated.

 

[PsyDr]

Yes, but I often see many people misuse the RCI. Mostly because they misunderstand the stats used to calculate the index. If they are relying test-retest stats from a 2 week interval vs a 1 year interval, your estimate will be...off. I have some papers somewhere that looked at test-retest over longer intervals (as most test manuals look at 2 to 6 week intervals) of a year-ish. And then there is work suggesting that practice effect in older and dementia individuals is greatly reduced. So, a good stat to have in some places, but only if you use the right data to compute it, and know its limitations. Unfortunately, with the de-emphasis on assessment and stats training in many programs these days, if this is even used, many times it is miscalculated.

I agree. Someone had to mention it, though.

 

[AcronymAllergy]

Yep, RCI is great and definitely worth mentioning, but the limitations imposed by much of the available test-retest data can be tough. Although there's that 2012 meta by Calamia et al., and the more recent 2020 update/re-evaluation by Duff & Hammers, looking at practice effects that can help provide some general rules of thumb.

If there are any grad students reading, a 1- or 2-year longitudinal study looking at effects across a large number of commonly-used tests could be a great (if resource-intensive) dissertation that would almost certainly get published and widely-cited if even reasonably well done. Hint hint.