Friday, December 07, 2007

How This Psychiatrist Thinks About Psychological Tests

First a big thank you to Gerbil for giving me the idea for this post. In a comment on my post "What Good Are Psychologists?" she mentioned psychiatrists who refer patients for diagnostic psychological testing. It got me thinking---this is a good thing---about why I do (or don't do) what I do.

I have to say I hardly ever request psychological tests. Even before I started working in prison, it just wasn't something I routinely did with my patients. When I was in residency we had lectures from psychologists about the different types of tests and what they're indicated for and a few things about interpretation, and later psychologists I've worked with have told me that I have a better understanding of testing than the average psychiatrist, but I'm not sure what that means.

For the lay readers among us, there are some general categories of psychological tests. There are personality tests that measures different character traits. There are intelligence tests that measure IQ. There are projective tests that are used to get an understanding of the person's interpersonal dynamics and style of thinking. There are neuropsychological tests that measure a person's cognitive capacity---ability to learn and remember, use language, coordinate eye-hand movements and so forth.

In general, when thinking about tests you have to consider what it is you're trying to figure out. If you have a patient who is failing in school you might want to order IQ or other achievement tests to see if the personal has a developmental disability or learning disability. If the patient has had a head injury or you think he or she might be getting demented you'd order neuropsychological testing. If you have a patient in therapy and you want material that might be useful to help the patient understand his own inner workings, you'd request projective testing and/or personality testing. Some tests are used as tools to predict certain things: whether or not someone would perform well on a certain job or whether or not they will re-offend as criminals.

It's important for tests to be used as part of an overall patient assessment. Test results fit into a whole database of information that a psychiatrist considers when making a diagnosis or putting together a treatment plan, in addition to a good patient history and a review of available treatment records.

It's also important to know whether or not the given test has been validated for your particular patient because 'normal' test results can be different for different groups of people. A test is only as good as the patients it's been based upon. For example, normal results for the original Minnesota Multiphasic Personality Inventory (MMPI, a test used to diagnose psychiatric disorders) was originally based upon only 500 people living in Minnesota. Much as I would like to think that Minnesota should be the gold standard for normalcy, this just isn't realistic. I mean really---Baltimoreans would end up looking pretty depressed compared to them. This is where a big limitation comes into play for me working in prison---many psychological tests have not been validated for use in prisoners.

Similarly, predictive tests only give group predictions and aren't necessarily reliable for the individual. A certain score on the Hare Psychopathy Checklist might give you a result that the patient has a 15% chance of re-offending, but that just means that out of 100 people with an identical score 85 will not re-offend and 15 will. The score doesn't tell you which of the two groups your patient will actually be in. The other trick with using tests to predict things is that many tests used for prediction have never been proven to have predictive value---there is no test to predict 'good parenthood', for example, yet psychological tests are used constantly in custody evaluations. It's important to know the limits of the test.

So...which tests do I actually use?

In prison the most common test I use is the Mini-Mental State Examination (MMSE). It was invented by two psychiatrists as a quick bedside test of cognitive functioning. You can give it in about ten minutes and it's a great way of measuring how brain impairment changes over time. You use it to check to see if someone's delirium is resolving, or as a screening tool for problems that should be investigated further.

In free society the most common tests I used, besides the MMSE, were general symptom inventories. For example, the Beck Depression Inventory (BDI) is a nice tool for measuring the severity of depressive symptoms. It's used to screen people for depression and also to monitor reponse to treatment.

I never use projective testing, mainly because I've never found it particularly useful for anything---maybe I would if I had a psychotherapy practice but even then I don't know too many psychiatrists who use them. I'd probably use neuropsychological tests if I could, but these are very specialized tests that have to be administered by a neuropsychologist (a subspecialty of psychology) and I don't know of any prisons who have one of those.

So that's my take on psychological tests. Thanks, Gerbil.

24 comments:

Gerbil said...

You're welcome, Clink. And yes, you do have a pretty good understanding of psychological testing (and the dangers thereof)... for a psychiatrist. ;)

Just a random piece of testing lore, because I am a dork: did you know that graduate students in psychology tend to exhibit the same "psychotic valley" (high 6 [Paranoia], low 7 [Psychasthenia], and high 8 [Schizophrenia]) on the MMPI-2 as do people with psychotic disorders? That either says something about the validity of the test, or it says something about the sorts of people who are drawn to graduate study in psychology.

ClinkShrink said...

Gerbil, I would never consider you a dork.

When I applied to medical school we were all required to take the MMPI. Everybody came out very defensive with high K scores. We were the last class that had to do that. I don't know of anybody who was denied medical school admission on the basis of the MMPI; goodness knows what it was used for.

Your example shows how you never know for sure if you're measuring what you think you're measuring. Definitely not psychoticism; maybe intellectual openness? Imaginative ability?

Once upon a time every inmate was given an MMPI for classification purposes, to predict adjustment to incarceration (thus needed security level). That was back in the good old days when they had time to give two hour tests to everybody.

Gerbil said...

Clink, but I am a dork! And proud of it, too. :)

I had to take the MMPI-2 when I applied to the University of Minnesota's psychology PhD program. (I didn't get in, but hopefully it wasn't on account of my rather pronounced psychotic valley.) My scores probably ended up in some giant validation database. They have to balance out all the Norwegian bachelor farmers somehow!

Anonymous said...

SLP's give tests too...
I much prefer the SLUMS (St. Louis University Mental Status exam)over the mini mental, for one thing most patients I see have done the mini mental so often that they're really bored with it, and possible have some parts memorized ("D...L...R...O...W").
There's some research (sorry I don't have the reference) that indicates the SLUMS more accurately differentiates the milder forms of dementia. Also it's fun to write about it in my notes.
Just throwing that out there...

FooFoo5 said...

In a forensic population it's not unusual to see violence, head trauma, and dementia related to a medical condition (and I'm thinking of paint & glue huffers). So if an inmate (generally young & under-exposed to testing) does poorly on the MMSE (which I also like), I'll request more formal testing. I'll also request it in reverse (e.g. rule out "self-reported" TBI) when I suspect malingering. But I emphasize request and not order because I'm a contractor. To my knowledge, not a single patient yet has been actually tested!

I actually like the PCL-R because it is valid in this specific male population when there really is nothing else for CA's "Mentally Disordered Offender" status. And practically speaking, when I see a significant PCL-R, I always grab a vest before examining an inmate!

yay said...

My med school once used the med student body as guinea pigs for a test someone there was developing for OCD. Apparently 60% of students screened positive (ie. further investigation would be needed in symptomatic individuals) for OCD traits.

When told of the results, the dean of the faculty apparently wanted to know what was wrong with the other 40% and how did they get into med school. Heh.

ClinkShrink said...

Foo: When I think about neuropsych testing in inmates I immediately remember the axiom I was taught in med school: never order a test if you don't know what you'd do with the results. The results would be useful for documenting impairment for the sake of disability ratings, but it wouldn't mean much as far as what you'd do with the prisoner. His functional adaptation (or lack of it) would be more telling than neuropsych results. And I agree with you about the PCL-R. So far I've never needed a vest, thank goodness (and keeping my fingers crossed).

Yay: When we were learning about the Rorschach I kept seeing nothing but body parts in the ink blots. I was reassured this was normal for health professionals. If they only knew....

Gerbil said...

Ah, but Clink... which body parts? ;)

Rach said...

Clink, stupid question - what's a projective test?

Roy said...

Anon- Thanks for letting us know of the SLUMS. I like it. It mirrors some of the questions I ask to test cognition, which the MMSE isn't sufficient. Like, making change and recalling bits from a story. It also doesn't place too much emphasis on orientation (one-third of the MMSE is orientation).

I found the .pdf form itself here, and the article comparing it to the Folstein MMSE is here.

FooFoo5 said...

Since Dinah apparently hasn't finished today's Times, there is an article in the Magazine about a new 50-question screening for Alzheimer's developed specifically to used over the phone. I wonder how you bill for that...

Dinah, I hope you're planning on writing about that Martini-infused cucumber.

ClinkShrink said...

Rach: Not a stupid question at all. A projective test is one in which the subject is shown some type of ambiguous stimulous (like an inkblot, or a drawing where it isn't clear exactly what's going on) and the test subject is asked to describe it. The theory is that the subject will fill in the gaps with his or her own personal issues, which reveals how they perceive or think about the world. That's my unofficial description. Our psychologist readers may have a more formal definition.

Gerbil said...

Clink, that is an excellent definition. This psychologist reader has nothing to add!

NeoNurseChic said...

Clink - Your subjective test definition has just given me a new way to view aspects of my own psychotherapy. My psychiatrist is a blank slate towards me - I mean that he doesn't tell me details of his life. Yet he asks me to describe what I think or imagine about him and his life outside of what I know. And then he can take that information and use it to see how I view the world and other relationships. I am very resistant to this "test" because I am uncomfortable talking about what I do not know with the person who it pertains to, but I can see the function of it. Anyway - it related somehow!

Take care,
Carrie :) (who is still happy to be in sunny Florida!)

NeoNurseChic said...

I meant projective tests not subjective - but you know what I meant!

Anonymous said...

Carrie,

I have never heard of a shrink asking anyone to do this unless it is in response to a question from the patient (the shrink may not answer but will ask what the client is imagining or thinking about). Otherwise it sounds a little too much like a shrink who is bringing the discussion around to himself and that is not what it should be about.

NeoNurseChic said...

Anon - I hear what you are saying, and it's been discussed a number of times between my psychiatrist and myself. I think a lot of it has to do with my reluctance to ask questions of anyone at all - and he knows that there is a lot I do actually think about, but do not say. I think the reason we are discussing it is not at all to discuss the real answers - in fact I have not yet ever had him answer something I've asked except one time when he did answer where he went on vacation. For example, I said to him that I wondered a great deal about his interaction with his siblings and parents, and the reason I think so much about his family is because family is more important to me than anything else, and I look up to him and hope very much that he has good family relationships - even though I really don't know what his real family is like.

I'm not explaining this very well.... He doesn't turn the conversation towards himself, though - and that's what I want to say. These things come up in discussion, but it is very difficult for me to articulate how I feel towards someone or to ask questions about their lives - and by discussing our relationship and what I think about him, we are learning a lot about how I view other relationships and other people. There's nothing inappropriate about it - or I wouldn't do it - I don't push boundaries or even want to go anywhere near boundaries as I've had boundaries pushed quite a few times in my own life, so I don't even let him be the keeper of the boundaries - so if there was even a hint of it not being appropriate or it being about him, I wouldn't discuss it with him at all.

Anonymous said...

You have written a post on what you think of psychological tests. You need a psychologist to come and delineate the how in the whole process.

Dr X said...

My practice has been about 70-30, psychotherapy to psychological testing. Just a clarification on referrals for testing. Psychologists shouldn't accept orders for specific tests. I always have a discussion with a referrer to explore and develop questions that would best assist the referrer. Stand alone testing data and test scores are useless without interpretation in light of specific clinical questions.

In this regard, single tests are rarely of value. Instead, examiners construct test batteries designed to best answer the referal questions once those questions have been developed and clarified.

In the course of conducting an assessment, tests may even be added as new information comes to light either in the interview or because the initial testing results point to the need for further investigation. For example, I've conducted personality testing that led to neuropsychological testing because the patient's report and/or data gathered in personality testing pointed to organic issues relevant to a referral queston.

Sarebear said...

When we were having our daughter evaluated because something was "off", and had been for a long time, but was functional. Anyway, the school psychologist observed, they sent home from school some formally named psychological test things for us to answer questions on about her, and then we took these materials to a mental health place, where a psych intern evaluated her over the course of 4-5 appointments and multiple tests, some of those appointments, the first two, more taken up with us answering her questions as she filled out a purplish test thingie. We had so much to answer with it took twice as long as a usual appt for that. Just because I'm so sensitive to picking up on stuff since, as a child, I was not . . . suited to the world, and also bumped up against it in ways.

Anyway, a bunch of psychological tests, and I still can't believe they think she'd make a good architect (has some math problems, but when I hit the right angle to explore a concept with her, BAM she picks it up INSTANTLY, her verbal reasoning was on the level of a 10th grader or something, they said, not sure if that's accurate or how that could be . . . . a couple of her numbers were WAY high or off the chart or something lol!)

Anyway. I also observed in her office, that the smooth rocks piled in the sink of a play kitchen, were probably there to see if tactile attraction to smooth type things like that was obviously evident.

In 10th Grade Social Science class, we had to come up with a survey and then gather data, put it together, from the results. Well, turns out, I didn't ask my questions the right way, and it was difficult to impossible to get much meaningful data. So that sucked, but at least he gave me a B- for writing up what I learned about where we went wrong and why and what we learned from that.

Graham said...

Excuse my ignorance, as I'm just an interested layperson. I understand the point made about projective tests and agree. But what about something like the HAM-D. Is that considered a test? And if so, isn't it a valuable tooling for measuring response to treatment/therapy?

Nervous Applicant said...

What do you experts know about the "Caliper Test"? Just heard I'm going to have to take it for a job application next week, and my research mostly turned up horror stories like postings I saw at http://www.holyshnikes.com, that make it sound like torture! Can anyone reassure me, please, so I don't literally psych myself out?

Roy said...

Caliper test: never heard of it. Looks like some sort of management Q&A. No idea here.

ClinkShrink said...

The only caliper test I know is the one used to measure percent body fat. It's the easiest and cheapest body fat test, but not as accurate as impedance measurement (found in some digital scales) or immersion (a research methodology where they dunk you in water. I always imagine a Monty Python skit when I think about that.)