DSM-IV, love or hate

[Psyclops]

---

Members do not see this ad.

I wanted to get peoples thoughts on the DSM as a document. Love it or hate it, and why. Thanks.

 

[Forensic M.S.]

I wanted to get peoples thoughts on the DSM as a document. Love it or hate it, and why. Thanks.

i like it except for the personality disorders. A lot of them are culutrally biased and do not work well for those who come from different culutures. i also think they need to implement the "spectrum" scale to rate the level of each disorder rather than just saying you meet the criteria. overall i like it.

 

[JatPenn]

Not a fan. Arbitrary dichotomous classification schemes don't work for mental disorders.

Forensic - do you have a study to point us to for cultural biases in Axis II? The five factor model seems to work across cultures, I would expect that points us to stability to Axis II diagnoses.

 

[Psyclops]

I would also be interested in knowing more about the culture bias, I'm not familiar with it. But I do agree with you Axis II is a mess. I personally am not a fan either. I think it does a poor job of focusing on the eitiology of disorders, I think that is a result of it being an atheoretical document created by the other APA. Only classifying things by thier symptomology although doesn't seem to be the best in my opinion.

 

[Forensic M.S.]

I would also be interested in knowing more about the culture bias, I'm not familiar with it. But I do agree with you Axis II is a mess. I personally am not a fan either. I think it does a poor job of focusing on the eitiology of disorders, I think that is a result of it being an atheoretical document created by the other APA. Only classifying things by thier symptomology although doesn't seem to be the best in my opinion.

i don't have a particular study but i am quite positive it is culturally biased, there is a lot of research and interest in the area (not just personality disorders, i mean everything)

http://www.sfsu.edu/~multsowk/title/38.htm

that is an abstract on the old DSM 3

 

[clinpsychgirl]

I agree that the categorization of what constitutes a mental disorder is somewhat problematic...

However, I think DSM categorication is helpful when used in conjuction with the development of (evidence-based) treatment protocols. The symptom guidlines allows for a systematic method to measure treatment outcome (by measuring symptom reduction).

 

[Psyclops]

That's a good point, and naturally symptomology shouldn't be ignored. But sometimes the data show some interesting results, I recently read a study, it wasn't a very good one, but I could look it up if you'd like, that spoke about how certain depressed patients show a faster response to norepinefrine agonists (or reuptake inhibitors not sure which) and others show a faster response to serotonergic agents. That would suggest to me that there might be a couple of ways to reach the same disorder. I.e., two etiologies. (I know that is a large leap, and it's a little off topic but I wanted to illustrate the point.) I'm just not sure symptomology is the best way to classify the disorders, at least not idealy, maybe for now it is.

 

[JatPenn]

I agree that the categorization of what constitutes a mental disorder is somewhat problematic...

However, I think DSM categorication is helpful when used in conjuction with the development of (evidence-based) treatment protocols. The symptom guidlines allows for a systematic method to measure treatment outcome (by measuring symptom reduction).

To me, if you'd like to measure symptom reduction, that alone tells me you are considering symptomatology a continuous variable (which of course, it is). So why not have a classification system where symptoms are measured on a dimensional scale and disorders are determined based on symptom severity profiles?

 

[Psyclops]

The dimensional model certainly makes sense to me if you intend to be measureing and treating overt abnormal behavior or merely intrapsychic issues that are leading to some form of distress. I think one of the problems lies in that it is constructed by psychiatrists who naturally follow the medical model. As long as they continue to be the major contributors to the main classiffication system for psychopathology that will be the cases. That being said, I think there will be a big overhaul to Axis-2 come DSM-V. Will it be a dimensional model? That remains to be seen.

 

[JatPenn]

Unfortunately, I think we'll need to wait for VI to see any considerable overhaul. From what I understand, V is just about ready to be published.

 

[psychanon]

I think that the DSM is an imperfect and overvalued volume, but it can be useful, and I think that the inception of criterion based diagnosis has been beneficial for the field as a whole. BUT it's a shame that so many overrely on it, treating it as gospel rather than a collection of heuristics.

i like it except for the personality disorders. A lot of them are culutrally biased and do not work well for those who come from different culutures.

Axis II is a mess-- in that case, the categorical system is clearly inappropriate, as the diagnoses are rarely reliable. Plus, I question their clinical utilitiy--in my (albeit relatively limited) experience, people rarely use PD diagnoses except for borderline & antisocial. I think a dimensional system in Axis II would make a lot of sense. But on Axis I? I don't know. I think it would complicate decision making, and more research is needed. What would the underlying dimensions be?

As for the cultural comment, I think it could be applied to the whole DSM. It's basically a cultural document. That doesn't take away its utility within that culture, but it should be recognized as such.

that spoke about how certain depressed patients show a faster response to norepinefrine agonists (or reuptake inhibitors not sure which) and others show a faster response to serotonergic agents. That would suggest to me that there might be a couple of ways to reach the same disorder. I.e., two etiologies.

Are you talking about evidence for atypical depression responding better to MAO inhibitors (I think that was it)? That may be true, but drawing conclusions about etiology based on treatment is backwards science.

To me, if you'd like to measure symptom reduction, that alone tells me you are considering symptomatology a continuous variable (which of course, it is).

I guess that's another philosophical question...is the goal of treatment to linearly reduce symptoms, or is it to restore functioning above a certain threshold? Just a thought.

 

[Psyclops]

Psychanon,

Good post, but I have to disagree with you that making inferences about etiology based on responses to tx is backwards. I don't see how it would be backwards? I see it as making inferences of etiology based on different responses to different chemicals. If you think that something is affecting the dopamine system for example, give it a dopamine agonist or antagonist and see if it affects the thing you are looking at, no? But please let me know your further thoughts on the issue. Frankly I didn't think anyone would take that part of my post seriously.

About your philasophical question, I suppose I would vote for linearly reducing symptoms. Many times it would be trying to reduce them beyond a threshold, even a set threshold like the DSM. But I think for many disorders a persons' own tolerance for the symptoms would determine where the threshold would lie. Maybe this is just semantics.

 

[docjohng]

It is what it is -- a "good enough" system that attempts to look at empirical research in each disorder category and make reliable enough so that professionals can use it for practice and for research.

Could another, "better" system be devised? Sure, and some have. But it is doubtful that anything else will ever supercede the DSM, because it is as entrenched within the U.S. mental healthcare system as much as the ICD is in medical practice.

John

 

[clinpsychgirl]

To me, if you'd like to measure symptom reduction, that alone tells me you are considering symptomatology a continuous variable (which of course, it is). So why not have a classification system where symptoms are measured on a dimensional scale and disorders are determined based on symptom severity profiles?

I understand your point. I still think that if you are using standardized assessment tools (DISC, DPS, ADIS, whatever) to warrant diagnoses in treatment outcome studies, it's important to note that these measures are literally based off the the DSM classification system. Moreover, if you are using a CRS (clinician rating of severity) when assigning diagnoses, you are sort of imbedding a "symptom severity profile" within your diagnoses. Granted the threshold between a clinical and subclinical disorder is a bit arbitrary. However, in the context of a research setting ,and with a more neatly defined and treatable disorder (say anxiety), the DSM and the assessment tools which capture its classification system have some relevance.

On the other hand, as someone who is broadly interested in dissemination research, I wonder if the DSM has much, if any, relevance in "real world settings". I have been involved with several studies that have provided evidence-based diagnostic tools in school-based mental health clinics. What do you think the most common diagnosis is? Invariably, its some kind of "adjustment disorder." I don't know if this issue is a can of worms worth opening on this thread... but I think clinical applicability (or lack thereof) of the DSM is another important facet to consider in terms of its utility as an diagnostic instrument.

 

[Psyclops]

I agree, clinical applicability is something that should be strived for, and qestioned vis-a-vis the DSM. I think you only have to look at the SCID to find that it lacks clinical utility. I think that the SCID is another document that lacks utility if it is used as it was intended. It certainly is a valiant effort, and in the right spirit (standardised diagnostic tool) but in practice it is unwieldy, confusing, and far too rigid. Of course there will always be give and take in utility.

I certainly don't think that we should be striving for a "good enough" system. And diagnoses such as adjustment disorder, in my opinion, a joke. Another thing to think about is how much useful are the diagnostic codes. What does a DSM diagnotic code really tell you about an incoming patient, especially if it something like adjustment disorder?

 

[Pterion]

I wanted to get peoples thoughts on the DSM as a document. Love it or hate it, and why. Thanks.

Not a fan, not a hater. As Docjohng said, "it is what it is".

There was an interesting series of views on this very topic expressed in JAMA 294 (15), Oct 2005 in the letters to the editor section. It appears that psychiatrists wrestle with this issue, as well. Given the state of the science, I wonder what the final compromise will be. I see the advantages of moving to more of an ICD or SNOMED classification, but I am also leery of the inherent reductionism. On the other hand, too much focus on symptom reduction as the gold standard misses the forest for the trees. No, I don't have a better idea.

And let's not forget that there will be some impact on the diagnostic system by what is billable and what is not.

I think the fundamental problem is that mental illness and the convergence of neuroscience, psychology and medicine is far too complex at this point to adequately be addressed by any single discipline.

 

[LM02]

The dimensional model certainly makes sense to me if you intend to be measureing and treating overt abnormal behavior or merely intrapsychic issues that are leading to some form of distress. I think one of the problems lies in that it is constructed by psychiatrists who naturally follow the medical model. As long as they continue to be the major contributors to the main classiffication system for psychopathology that will be the cases. That being said, I think there will be a big overhaul to Axis-2 come DSM-V. Will it be a dimensional model? That remains to be seen.

The answer is most likely - yes. Primarily, this is because the data more strongly support a dimensional approach - see the Durbin & Klein paper in the most recent issue of the Journal of Abnormal for some interesting data.

Moreover, unlike some of the other disorder classes, the DSM task force for Axis II is comprised primarily of psychologists (vs. psychiatrists). So that, again, most likely points to a dimensional reconfiguration.

Unfortunately, I think we'll need to wait for VI to see any considerable overhaul. From what I understand, V is just about ready to be published.

Nope - DSM-V is scheduled for 2011, and the task forces are scheduled to start the meat of their work in 2007.

 

[JatPenn]

Nope - DSM-V is scheduled for 2011, and the task forces are scheduled to start the meat of their work in 2007.

wow, I stand corrected. I must get my info from some uninformed faculty 🙄

 

[psychanon]

Psychanon,

Good post, but I have to disagree with you that making inferences about etiology based on responses to tx is backwards.

The traditional progression of science is from description to theory to treatment. When the treatment comes first (e.g., the first evidence that serotonin might be involved in depression came from the efficacy of SSRIs), then you run the danger of making fallacious assumptions about description and theory. Aspirin is an effective treatment for headaches, but headaches are not caused by a lack of aspirin. The fact that different subtypes show some differential treatment response might suggest different underlying chemical etiologies, but it also might mean a number of other things. And if you are going to take that approach as evidence for the validity of the DSM-IV, then you'd have explain why SSRIs are effective not only for depression but also for social anxiety, OCD, panic disorder, etc.

 

[Psyclops]

Ohhhh, that kind of backward science, yes I agree it's not the best way to do things. Results like these should be taken into consideration when making future theoretical updates.

I would like to think that theories of depression, SA, OCD, and Panic DO are more intricate than just a serotonin disregulation issue. But maybe they should just be reduced to a biological issue.

 

[50960]

psychanon, to understand that one would have to understand the biochemistry of SSRI drugs and most psychologists do not. It is really not a secret why SSRI's treat depression, PTSD, social anxiety, GAD etc...if you understand what you are doing when you add a SSRI into the CNS.

 

[soaringheights]

The DSM IV is a good starting point for beginners. However, I felt that disgnosing through the DSM did not give me a "feel" of the disorder, as all disorders have their criteria very clear cut. The essential aspects of most disorders are also given in the 'criteria to be satisfied format', I think it'd have been better if a good introduction sans following these criteria was included at the introduction of each disorder, rather than making the intro a mere elaboration of various criteria.
When I read the F chapter of the ICD 10, I was so used to the categorical approach of teh DSM,that it took me a bit getting used to. However, now I prefer the ICD to the DSM.

 

[Neuro-Dr]

The point of making an atheoretical model for psychopathology is to allow for a common language to be used across disciplines (psychiatry, social work, LMFT, etc.). In addition an attempt was made to move away from rediculous beliefs like the "schizophrenic mother". While this annoys some people who believe that the person should be considered as a whole. nothing changes the fact that you are still a person even if depressed, anxious, what have you.

A model based on onset, duration, prognosis and sx presentation also allows for the clean-up of errors caused by sx alone. I can't tell you how many times I have seen people give onset schizophrenia to someonw in there late 40's when dopaminergic tracts are already dying out. I have seen people blow psychotic disorders becasue they believed olfactory hallucinations are common in schizophrenia. People who use dysthymia as depression light and other aggregious errors that demonstrate their limited understanding of psychopathology.

Contrary to popular belief, the accuracy of medical dx is about 67% and psych is 65%, so there is not a great deal more innacuracy. In addition, those of you who like research, what are you basing your inclusion criteria on? Is someone depressed because they watch a sad movie, have an MMPI-2 scale 2 of 65 or higher, BDI of 12, say they are sad or do you want to use a standard set of criteria. This is the major issue..do we really expect all of these patients to present and respond to treatment the same? Show the same changes on fMRI?

Using rating scales is completely problematic, those who have tried to use even decent measures like the BDI show horrible reliability over sessions, even with the same rater.

While problematic, the DSM-IV-TR is still the most useful resource in the MH field. Most of the errors I have seen clinicians make involve the "broken leg" theory of psychology rather than application of the DSM, they do far more damage with there views that neurological conditions such as autism are due to parenting styles, poor diets (leaky gut syndrome) and other non-science based evidence.

Having said all this, I'm glad someone brought the topic up, because these types of debates are productive.

 

[clinpsychgirl]

I agree, clinical applicability is something that should be strived for, and qestioned vis-a-vis the DSM. I think you only have to look at the SCID to find that it lacks clinical utility. I think that the SCID is another document that lacks utility if it is used as it was intended. It certainly is a valiant effort, and in the right spirit (standardised diagnostic tool) but in practice it is unwieldy, confusing, and far too rigid. Of course there will always be give and take in utility.

I certainly don't think that we should be striving for a "good enough" system. And diagnoses such as adjustment disorder, in my opinion, a joke. Another thing to think about is how much useful are the diagnostic codes. What does a DSM diagnotic code really tell you about an incoming patient, especially if it something like adjustment disorder?

I agree with your assessment of adjustment disorders. Perhaps, in some cases, it is an accurate diagnosis; however, in the school-based settings I liase in, it is really abused by clinicians who are afraid of "stigmatizing" children. It's too bad that those clinicans are denying those children of adequate treatment in the process of denying them an accurate diagnosis.

 

[LM02]

The point of making an atheoretical model for psychopathology is to allow for a common language to be used across disciplines (psychiatry, social work, LMFT, etc.). In addition an attempt was made to move away from rediculous beliefs like the "schizophrenic mother". While this annoys some people who believe that the person should be considered as a whole. nothing changes the fact that you are still a person even if depressed, anxious, what have you.

A model based on onset, duration, prognosis and sx presentation also allows for the clean-up of errors caused by sx alone. I can't tell you how many times I have seen people give onset schizophrenia to someonw in there late 40's when dopaminergic tracts are already dying out. I have seen people blow psychotic disorders becasue they believed olfactory hallucinations are common in schizophrenia. People who use dysthymia as depression light and other aggregious errors that demonstrate their limited understanding of psychopathology.

Contrary to popular belief, the accuracy of medical dx is about 67% and psych is 65%, so there is not a great deal more innacuracy. In addition, those of you who like research, what are you basing your inclusion criteria on? Is someone depressed because they watch a sad movie, have an MMPI-2 scale 2 of 65 or higher, BDI of 12, say they are sad or do you want to use a standard set of criteria. This is the major issue..do we really expect all of these patients to present and respond to treatment the same? Show the same changes on fMRI?

Using rating scales is completely problematic, those who have tried to use even decent measures like the BDI show horrible reliability over sessions, even with the same rater.

While problematic, the DSM-IV-TR is still the most useful resource in the MH field. Most of the errors I have seen clinicians make involve the "broken leg" theory of psychology rather than application of the DSM, they do far more damage with there views that neurological conditions such as autism are due to parenting styles, poor diets (leaky gut syndrome) and other non-science based evidence.

Having said all this, I'm glad someone brought the topic up, because these types of debates are productive.

While I mostly agree with the statement above, I still believe that a limitation is the somewhat arbitrary nature of the some of the criteria.

For example, why 5 and not 6 Sx for MDD? Or why is it that, if a person goes without depressive symptoms for 2 months they wouldn't meet criteria for dysthymia, even if all other criteria are met? Why isn't it 1 month or even 3 months? When someone is symptomatic for years, what do these distinctions even mean?

That's my problem - the purpose of the DSM is clear, and I agree that it vastly improves communication between mental health providers and establishes standards for our research. But it's in the process of applying some of these arbitrary criteria that the DSM loses its appeal.

 

[Neuro-Dr]

The answer is actuarial. Those cut points determined a different outcome, response to treatment, long term prognosis, etc. These points of differential where not picked out of a hat. The reason why we have brief psychotic disorder, schizophrenaform and schizophrenia is becasue the cut points of 1 month and 6 months were associated with higher percentages of variable outcomes.

As for MDD, the same was true, that when sx of 6 but not 5 were put together it changed the duration of the episodes, lielihood of reoccurance, response to treatment. I'm not saying that the changes were so dramatic as to always be of clinical relevance, but nor were they arbitrary.

 

[LM02]

The answer is actuarial. Those cut points determined a different outcome, response to treatment, long term prognosis, etc. These points of differential where not picked out of a hat. The reason why we have brief psychotic disorder, schizophrenaform and schizophrenia is becasue the cut points of 1 month and 6 months were associated with higher percentages of variable outcomes.

As for MDD, the same was true, that when sx of 6 but not 5 were put together it changed the duration of the episodes, lielihood of reoccurance, response to treatment. I'm not saying that the changes were so dramatic as to always be of clinical relevance, but nor were they arbitrary.

I'm not asking this to be snarky, but would you mind providing some citations for this work? I'm a full-time mood disorders researcher working in a Dept of Psychiatry, and I have never seen these data published.

 

[Neuro-Dr]

I can't say that I'm positive that I know what "snarky" means. But, if you ever want data regarding what someone says you should feel free to ask for it.

With regard to the accuracy of medical and MH dx - ther references are from Monahan's 1982 prediction data and Lawler 2003.

For the DSM stats, you can use the sourcebook or refer to the appendices regarding the field trials of 6,000 patients and over 100 sites sponsored by the NIMH, NIDA, and NIAAA. These are the work groups used in every DSM printed since III, I think. As to whether they seperately published the findings, they say "no", but I can't say that no one ever has.

The psychotic quote cam from Benazzi 2003, that linked schizophreniform more to a mood disorder tham schizophrenia. With mood disorders the data is rather pervasive, there are numerous publications indicating factors such as early onset, frequency of episodes, partial recovery from episode are all major indicators of prognosis.

I think you wanted to know why associated findings such as sleep, appetite, attention and concentration were included and why you need 5 (I said 6 in the previous post). The prognosis and onset indicators change with these factors:
for instance (Vermetten, Geuze & Mertens 2004) - insomnia of more than 1 year's duration predicted depression at 40% with insomnia predating the depressive episode at 38% and anxiety at 38%. Increase stage 2 sleep was also predictive of onset depressive episodes at higher rates. Possibly these findings could indicate a link betwee 5HT and sleep onset and depression. In bipolar, the relation to sleep should be obvious, but the duration may be related to "kindling" in the ventral-tegmental dopaminergic tracts.

With appetite, remember that you are looking at a qualifier for "with vegitative signs" or "catatonic features" - it is not necessarily that the appetite itself changes the prognosis, but rather leads to the qualifier. These have different treatments (Peselow, Sanfilipo & Defiglia 1992) - there are probably more curent references, but this is for illustration.

Is that enough - I know I painted some broad strokes, but I hope this shows where I'm going with this.

 

[LM02]

I can't say that I'm positive that I know what "snarky" means. But, if you ever want data regarding what someone says you should feel free to ask for it.

With regard to the accuracy of medical and MH dx - ther references are from Monahan's 1982 prediction data and Lawler 2003.

For the DSM stats, you can use the sourcebook or refer to the appendices regarding the field trials of 6,000 patients and over 100 sites sponsored by the NIMH, NIDA, and NIAAA. These are the work groups used in every DSM printed since III, I think. As to whether they seperately published the findings, they say "no", but I can't say that no one ever has.

The psychotic quote cam from Benazzi 2003, that linked schizophreniform more to a mood disorder tham schizophrenia. With mood disorders the data is rather pervasive, there are numerous publications indicating factors such as early onset, frequency of episodes, partial recovery from episode are all major indicators of prognosis.

I think you wanted to know why associated findings such as sleep, appetite, attention and concentration were included and why you need 5 (I said 6 in the previous post). The prognosis and onset indicators change with these factors:
for instance (Vermetten, Geuze & Mertens 2004) - insomnia of more than 1 year's duration predicted depression at 40% with insomnia predating the depressive episode at 38% and anxiety at 38%. Increase stage 2 sleep was also predictive of onset depressive episodes at higher rates. Possibly these findings could indicate a link betwee 5HT and sleep onset and depression. In bipolar, the relation to sleep should be obvious, but the duration may be related to "kindling" in the ventral-tegmental dopaminergic tracts.

With appetite, remember that you are looking at a qualifier for "with vegitative signs" or "catatonic features" - it is not necessarily that the appetite itself changes the prognosis, but rather leads to the qualifier. These have different treatments (Peselow, Sanfilipo & Defiglia 1992) - there are probably more curent references, but this is for illustration.

Is that enough - I know I painted some broad strokes, but I hope this shows where I'm going with this.

Thanks - I didn't want you to think that I was putting you on the spot (which is why I qualified my statement).

I was only interested in the specific cut-points of 5 vs. 6 symptoms (or why 2 months is the magic number for a differential of dysthymia vs. MDD - or even DD NOS).

As I previously stated, I'm a full-time PhD researcher in a Medical School, and I specialize in mood disorders. I have actually done quite a bit of independent research on the neurovegetative symptoms associated with both MDD and bipolar disorder, and work with people who have served on the DSM task forces. So I'm fully on board with the symptoms (I wasn't asking about that), but would love to have a good reference to justify the thresholds and time durations. At least in the empirical literature, I feel these diagnostic issues receive less attention than those related to actual symptomatology. I'm going to check out some of the relevant references you posted...

Thanks again.

 

[Neuro-Dr]

No worries, I really think these discussions are useful. My recollection to your specific point was the point of "Modality" in that, when does a sx become a modal deficit for a disorder. In MDD, these are dysphoria and anhedonia. The others are associated findings that have predictive validity. I'll call a colleague and ask about the number of sx relative to outcome versus prognosis.