Bereavement dropped as an exclusion in diagnosis of depression: protecting reimbursement, but hurting science

“The elimination of the bereavement exclusion shows that psychiatry has no idea how to define what’s normal, what’s abnormal, and how to differentiate between them…One of the essential ways that we show our humanity is to grieve after the death of an intimate. Amazingly, psychiatry now sees this as a mental disorder.”—Allan V. Horowitz, author of The Loss of Sadness

Among the many changes announced for the new fifth edition of the Diagnostic and Statistical Manual of the American Psychiatric Association is that persons who otherwise meet criteria for major depression will no longer be excluded if they are bereaved, as in recently having lost a spouse. Eliminating the bereavement exclusion is good for psychiatrists seeking reimbursement for their services, but bad for the scientific study of depression.

If APA preserved the bereavement exclusion, psychiatrists might have trouble getting reimbursed for providing medication to bereaved persons. For instance, suppose the wife of a man with a history of depression has died.  After a few days, this man is lying awake all night with eyes-wide-open insomnia, refuses to eat, and is markedly slowed down in his movements (psychomotor retardation). Maybe he even has irrational thoughts of having caused his wife’s death. Based on this past history and constellation of symptoms, this man is probably experiencing a treatable major depression. The recommendation should be to consider intervening early in such an atypical reaction to death of a spouse and maybe to start with an evaluation for treatment with antidepressants. But would this be reimbursable? Maybe not if the bereavement exclusion is kept.

But in eliminating any threat to be reimbursed for treating such a person– resolving might be termed the acute compensation reaction of psychiatrists–bereavement is medicalized, made into a mental disorder. The change avoids psychiatists having to make individualized clinical decisions whether to treat grieving persons,  but couldn’t the APA simply have just defined some further criteria for when grieving becomes abnormal, even shortly after a loss?

In terms of its impact on the science of depression, the change inflates estimates of the current incidence of depression in epidemiological studies of the general population by counting persons as depressed who are simply bereaved. Untreated bereaved persons are not only counted as depressed, they are entered into the pool of “untreated depression in the community.” Their added numbers lend support to calls for screening, detection, and providing psychiatrists and primary care physicians’ preferred mode of treatment, antidepressant medication.

This comes at a time when rates of antidepressant prescription exceed the estimated prevalence of depression in the elderly population. This also comes at a time when the modal person receiving a new antidepressant prescription is a person just barely attaining or missing the minimum number of required symptoms for diagnosis and who gets the prescription without formal diagnosis or follow-up. Under these conditions, response to treatment with antidepressant can be expected to be minimal, and no better than what is  obtained with a pill placebo in a clinical trial, where there’s a lot more attention and support. It is estimated that 40% of patients being treated antidepressants in primary medical care are getting no benefit over remaining on a waiting list, and this estimate may be increasing as prescribing gets even more casual.

The dropping of the exclusion criteria also bedevils research evaluating new treatments for depression. Many researchers concede that the diagnostic criteria for depression used to enter patients in such randomized trials are already way too loose and set too low a threshold for defining a clinical disorder. The result is that many of the patients entered into the studies are going to recover, independent of whether they are assigned to the active treatment or control condition. When treatments are tested in samples where such persons predominate, the treatments might be discarded as ineffective when they  would have proven effective in a more carefully selected sample of clearly depressed persons.

Expanding the criteria of major depression to include persons were merely grieving also threatens biological research aimed at identifying neurohormonal and genetic factors in depression. If people are recruited for such research when they have never been depressed other than during a current bereavement, they add noise  to samples where it is already difficult to detect a clear signal of undoubtedly complex biological factors. Already the dexamethasone suppression test that up into the mid 80s was thought to distinguish between the clinical disorder of depression and mere misery no longer distinguishes depressed psychiatry patients.

We cannot expect psychologists to rise to the occasion and protest psychiatrists’ expansion of diagnostic criteria. DSM-5 allows psychologists to bill for the treatment of newly bereaved patients who might otherwise have their reimbursement for treatment question. Yet, the benefits of bereavement therapy are at best questionable and there’s even evidence that indiscriminate  immediate intervention can be harmful. Interestingly, treatment studies that depend on bereaved persons to self-select and apply for treatment tend to have better outcomes than those that make active outreach. Apparently there is some wisdom to bereaved persons’ own assessment that they need therapy versus their being approached and been told they might need therapy and entering treatment on that basis.

But why limit the bereavement to death of a loved one anyway? If anything, rather than being eliminated, the exclusion for bereavement should be extended to other devastating losses, like loss of all of one’s belongings to a fire or hurricane, or loss of a significant role, like being suddenly fired. Shouldnt there be an allowance for it being  normal to be sad and depressed,  after such a loss?

We need to keep in mind that an episode of depression is seldom an isolated occurrence. Best evidence is that major depressive disorder is a recurrent episodic condition with an onset in adolescence or early adulthood. It’s relatively unusual for persons to experience a first episode of recurrent depression in later life. Again, our ability to make this observation is hampered if we count persons as depressed whose only episode of “depression” occurs in the presence of loss of a spouse or other significant disruption in their life.

And finally, the removal of the bereavement exclusion for major depression infiltrates our culture and confuses our understanding of normal adult development. What we take to be resiliency is distorted if we assume that normal and even self-limiting sadness and other symptoms following a loss represents the absence of resiliency, an abnormality.

Across cultures, even in modern culture, there are established rituals for the family and social environment to respond to the newly bereaved person that can be disrupted if bereavement is considered a mental disorder. Re-labeling a person as having major depression can be stigmatizing and invites seeking professional intervention for what might reasonably be handled with an informal supportive response from the family and community.

Stretching the definition of abnormal, we compromise our ability to understand and respond to the normal.

9 thoughts on “Bereavement dropped as an exclusion in diagnosis of depression: protecting reimbursement, but hurting science”

  1. I’m going to start this comment with caveat that I’m not a psychiatrist and I don’t know anything.

    “The change avoids psychiatists having to make individualized clinical decisions whether to treat grieving persons, but couldn’t the APA simply have just defined some further criteria for when grieving becomes abnormal, even shortly after a loss?”

    Is this true? It would seem like removing the exclusion would give psychiatrists more freedom to make individual decisions. Does removing the exclusion mean psychiatrists have to diagnose patients with major depression two weeks after bereavement? Or does it mean that they just don’t have to wait a month before making a diagnosis when a patient is displaying the symptoms of major depression?


    1. With or without the exclusion of bereavement, psychiatrists are free to make decisions whether to treat a bereaved person. The issue is, however, that if the bereavement exclusion were retained, insurance companies could conceivably refuse reimbursement for treatment of a bereaved person with an antidepressant.

      I certainly could conceive of instances in which psychiatrists ought to be treating a bereaved person for major depression, as in the case example that I gave. However, what I take issue with is the administrative decision to remove the bereavement exclusion for diagnosis of depression. What’s fine for reimbursement causes problems for the scientific study of depression.


  2. “Eliminating the bereavement exclusion is good for psychiatrists seeking reimbursement for their services”

    In that context, wouldn’t the more honest amendment be to remove depression from the DSM completely, since, as we have already discussed, currently available treatments are no more effective than placebo?


    1. Colin, I am not sure what you mean by “as we have already discussed, currently available treatments are no more effective than placebo,” but that view is not supported by data. Antidepressants ave about the same advantage as psychotherapy does over pill-placebo, in the studies that allow comparison of the three. But let’s hold this discussion until my next post.


      1. Will do but just to explain further – I was concatenating two ideas – your point that psychotherapy is no more effective than antidepressants and my point, made in the comments section of your previous blog (hence ‘already discussed’) and detailed in my recent paper “The failure of the antidepressant drug discovery process is systemic” that antidepressants are no more effective than placebo – sometimes in individual studies yes, but in overall response rates no

        if a=b and b=c then a=c – hence the concatenation that psychotherapy is no more effective than placebo


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s