Saturday, 30 January 2010

Is Depression Undertreated?

Neuroskeptic readers will be familiar with the idea that too many people are being treated for mental illness. But not everyone agrees. Many people argue that common mental illnesses, such as depression, are undertreated. Take, for example, a paper just out in the esteemed Archives of General Psychiatry: Depression Care in the United States: Too Little for Too Few.

The authors looked at the results of three large (total N=15,762) surveys designed to measure the prevalence of mental illness in American adults. I've described how these surveys are conducted before: they took a randomly selected representative sample of Americans, and asked them a standardized series of questions (the CIDI interview) about their mood and emotions, in order to try to diagnose mental illness. The interviewers, while trained, were not clinicians.

What did they find? The rate of people experiencing Major Depressive Disorder (MDD), as defined in DSM-IV, in the past year, was 8.3%. When they examined ethnicity, this ranged from 6.7% in African Americans to 11.8% in Puerto Ricans. The average severity of the depression was roughly the same in all ethnic groups.

Of those with MDD, 51% reported that they'd had treatment in the past year, either antidepressants, psychotherapy, or both. This ranged from 53% for Whites down to just 29% of Caribbean Blacks and 33% of Mexican Americans. Therapy was somewhat more popular than drugs in all ethnic groups, although a lot of people used both. However, few of the treatments were classed as "guideline-concordant", i.e. long enough to do any good, which they defined as
use of an antidepressant for at least 60 days with supervision by a psychiatrist, or other prescribing clinician, for at least 4 visits in the past year. For psychotherapy...having at least 4 visits to a mental health professional in the past year lasting on average for at least 30 minutes each.
Only 21% of depressed people were getting such treatment, even though these strike me as very lenient guidelines, especially in the case of psychotherapy - how much good is 2 hours per year doing to do?


So depression's undertreated, especially in minorities. Too little, for too few. But this rests on an assumption: that we should treat Major Depressive Disorder.

That might not seem like an assumption, but assumptions generally don't. It seems like common sense, almost a tautology - it's a disorder, of course we should treat it! Yet it's not so simple. DSM-IV criteria for MDD require you to have 5 or more out of a list of 9 symptoms, including either depressed mood or a loss of interest in activities, lasting at least 2 weeks, and causing significant distress or impairment in social, occupational, or other important areas of functioning.

Fair enough. That's quite useful as a way of ensuring that psychiatrists in different countries are talking about the same thing when they talk about depression. But to think that depression is undertreated because only half of people meeting DSM-IV criteria for Major Depressive Disorder are being treated, is to put absolute faith in DSM-IV as a guide to who to treat. This is not what the DSM was meant to be, and there's no evidence it works for that purpose.

Is it really true that people with 5 symptoms need help, and those with 4 don't? Why not 6, or all 9? Why 2 weeks - why not 3 weeks, or 3 months? It's not as if there are loads of studies showing that treating people who have 5 symptoms for 2 weeks, and not treating people who don't, is the best strategy. I'm not aware of any such research. In particular, there's no evidence that people from the general population who meet these criteria when interviewed, but don't seek treatment, would all benefit from treatment as opposed to being left alone. Certainly some would, but they may be a minority.

This is not to say that any other criteria would be better than DSM-IV as guides to treatment, or that there is anything identifiably wrong with the DSM-IV criteria (although there is evidence that antidepressants are not useful in people with relatively "mild" MDD). The point is that doctors don't strictly apply textbook criteria when diagnosing and treating mental illness; they also use clinical judgement.

I don't know any psychiatrist who would prescribe treatment for someone solely on the basis that they met DSM-IV criteria for MDD. They would also want to know about the severity of the symptoms, whether they're related to any stresses or life events, how far they're "out of character" for that individual, etc. In general, they would deploy their training and experience to try to judge whether this person would benefit from treatment. This is why the DSM-IV carries a cautionary statement that "The proper use of these criteria requires specialized clinical training that provides both a body of knowledge and clinical skills."

So, it's far from clear that we should be treating everyone who answers interview questions in such a way that they meet DSM-IV criteria for Major Depressive Disorder. That's an assumption.

This isn't to say that everyone who needs depression treatment gets it. Sadly, there are many sufferers who would benefit from help and don't get any, or don't get it as early as they should. We need to do more to help such people. In this respect, depression is undertreated, although it's hard to know the extent of the problem. Yet it's quite possible that depression is also overtreated at the same time.

H/T Thanks to The Neurocritic for drawing my attention to this paper.

ResearchBlogging.orgGonzalez, H., Vega, W., Williams, D., Tarraf, W., West, B., & Neighbors, H. (2010). Depression Care in the United States: Too Little for Too Few Archives of General Psychiatry, 67 (1), 37-46 DOI: 10.1001/archgenpsychiatry.2009.168


pj said...

Bizarrely the implication is that people who had symptoms meeting the criteria for MDD for 2 weeks but did not seek help, and who recovered after those two weeks should have received treatment including longterm antidepressants.

I won't repeat my comments on the previous post about these prevalence studies, but I really think failure to seek help should be considered at least an indicative marker of severity.

Jumping Jack said...

I don't know any psychiatrist who would prescribe treatment for someone solely on the basis that they met DSM-IV criteria for MDD.

I do. More than one or two.

As useful as it may sometimes be, there are days when I think that generally speaking the DSM is a huge mistake.

Is depression undertreated?

I am fairly sure it is over diagnosed.

Anonymous said...

Given your comments, why do we have a DSM-IV at all? Or a looming DSM-V? Why, it's all a gut feeling, right? This is why Psychiatry is such an unscientific field.

Neuroskeptic said...

pj: Heh, that's a great point. And we know that a great many people experience "MDD", recover without treatment, and go on to forget it ever happened (at least in NZ).

Failure to seek treatment is a tricky one. On the one hand, people with sub-clinical symptoms are less likely to seek treatment. But people with very severe symptoms (and no insight) don't either. In mania or psychosis, seeking treatment would be considered a good sign. I think they key point is that we just have no idea whether people who don't seek treatment would be better off if they did, because no-one's done the trials, because they don't volunteer for trials.

Neuroskeptic said...

Jumping Jack: Well, I'm glad I don't know those psychiatrists. It makes you wonder what they consider their role to be; if they're just applying criteria mechanically, we might as well hand out DSM-IVs on the street with attached prescription pads, and let people prescribe themselves whatever they think they need...

pj said...

Remember that, certainly in the UK and even from my experience in the US, DSM-IV (and ICD-10) are research tools, they are not the basis for clinical management.

Bernard Carroll said...

pj: Actually DSM-III and ICD-10 are not research tools. They were created as administrative tools for the psychiatric profession, with a sharp eye on justifying payment for psychiatric services.

When they were introduced there was some posturing about DSM-III disorders having the status of provisional hypotheses. The profession, instantiated in the American Psychiatric Association, did not live up to that promise, and any changes, like the original criteria, were largely decided in a data-free manner.

In the US, once DSM-III appeared the forces of orthodoxy prevailed over alternative diagnostic ideas, so the new criteria became de rigueur in research studies. In the field of mood disorders, at least, we are still paying the price for that blunder. The hijacking of the DSM-III criteria by the marketers in Pharma for the purpose of disease mongering is another part of the story, but I will not go there today.

Neuroskeptic said...

BC: Thanks, that's interesting. But what disturbs me about this paper is that even the DSM warns that you can't diagnose mental illness simply by applying DSM criteria "in a cookbook fashion" (it says so in the disclaimer at the start).

Yet this is, essentially, what this paper did.

This isn't a criticism of the authors themselves, because clearly the editors of Arch Gen Psych and the peer reviewers saw no problem with it either. What's disturbing is that the assumption that DSM criteria can be applied cookbook-style to indicate clinical need, has become implicitly accepted. Something that even the DSM, flawed as it is, says not to do (at least that's how I read the disclaimer).

Bernard Carroll said...

You are quite correct, Neuroskeptic. The paper to which you called attention is just more evidence of what I call the epistemological quagmire that confounds contemporary studies of major depression.

The essential point is that operational diagnostic criteria are necessary but not sufficient in themselves for arriving at a diagnosis. In practice, diagnoses are made by pattern recognition and clinical heuristics. Inspection of the DSM criteria is a further step, designed to validate the clinician's diagnosis. The notion that we could make diagnoses by counting nonspecific depressive symptoms is naive, but it has been embraced in many research studies from psychiatric epidemiology to clinical drug trials.

The boilerplate caveat at the front of DSM-III-IV is regularly disregarded.

wichitarick said...

.Hi NOT High
I have enjoyed reading these pgs. Thanks.
I have never understood where anyone can get "real" numbers in research when so much of the population is just an unknown?
Meaning simply a large number of people live and die their whole lives and have "never" even spoke to a relative about their "depression" or p.t.s.d. or add,adhd, manic,MDD level 1,2,3,4,5or even the fact the dog really does talk back AND listens.
Crazy aunt Sally is really crazy or depressed or what ever may be right or wrong but how can this just be a percentage or with in so many degrees type of thing and be even close to accurate.?
OOPS my comment was supposed to about this QUOTE...
"we might as well hand out DSM-IVs on the street with attached prescription pads, and let people prescribe themselves whatever they think they need"
We do have this I believe the politically correct term is self diagnoses and the med ,phych 101 manuals and the web.
Followed by the very nice anti-? pharmaceutical commercials I just saw during the news and the pop ups from my screen.
Combine this with the rare person who pays the doctor in cash or check and you have an instant script pad at your request 24/7 if needed.
If not, I take the first part of the post and say no sweat! I can show not one or two but at least 6 different m.d,s and phychs who have written just myself scripts for at least 10 drugs and most of these meds are the high level type pharmaceuticals.
It is funny I see the over treated part on my T.V. but can show the under treated part in my own family .I bet those lines will get my scripts re-filled . R.C

Neuroskeptic said...

"The boilerplate caveat at the front of DSM-III-IV is regularly disregarded."

Quite - although when you think about it, this is not surprising because DSM is a cookbook, so telling people not to use it as a cookbook is hardly very sensible. "Don't try to diagnose mental illness by mechanically applying criteria - but if you do, here's several hundred pages of criteria"

Anonymous said...

God Almighty: With all of the foregoing in mind, just how is one to make a diagnosis of major depression or any other psychiatric condition for that matter! Is this a pseudo-science???

Bernard Carroll said...

Anonymous: no, it is not a pseudo-science but it is a science under commercial pressure. Here is a commentary on some of the issues.