Magazine article Clinical Psychiatry News

Length of Therapy and Depression

Magazine article Clinical Psychiatry News

Length of Therapy and Depression

Article excerpt

A recent study found that patients with depression who had more than 12 sessions did not progress as fast as did those who had fewer. Have you found that some depressed patients do better with fewer sessions?

Least Improvement = Most Therapy

The study makes good clinical sense and seems generally rational. It seems to me that there is only one phrase that would stir controversy: "Patients who received the most therapy tended to have the worst outcomes."

The statement would be better put: Patients with the least improvement received the most therapy. Put this way, the finding is unremarkable. One could make the same sensationalized statement about hospital-based medical care: Patients who receive the most care tend to have the worst outcome. This lends itself to the misunderstanding that the care causes the worst outcome. Acutally, it reflects the fact that patients with the worst prognosis receive the most medical attention. This is probably the case among the depressed patients in the study cited.

Peter Kotcher, M.D.

Cincinnati

Confusing Correlation and Causality

The "study," published in the August 2004 issue of the journal "Employee Benefit Plan Review," is not a scientific paper. There are few data. There is no evidence that this "study" was actually a rigorous, scientific study that conforms to the standards of modern, peer-reviewed psychiatric research.

The report stated that "... findings of the study do not support the notion that more care is always better" and that "patients who received the most therapy tended to have the worst outcomes." Beware whenever anyone says "always." Few issues in life are "always" one way or another. As it concerns this study, the statement I quoted suggests that more care is not better than less care. Actually, sicker patients logically require more sessions, and they improve more slowly. The difference between correlation and causality is being confused here.

The number of sessions and treatment outcome are correlated, precisely because sicker patients (who improve less than healthier patients) require more sessions. There is no evidence here, certainly not in this "study," that an increased number of sessions caused the worse outcomes.

The typical example of this logical error is that alcohol purchases rise as the income of preachers increases. Increased preacher income does not cause the rise in alcohol consumption. These are only correlated, inasmuch as the income of preachers tends to rise as overall population income (and therefore overall population alcohol consumption) rises.

Joseph Annibali, M.D.

McLean, Va.

Note History, Monitor Improvement

The study investigators developed a 30-item questionnaire that assessed symptoms to recommend continued treatment or to flag patients who are severely ill among patients in a managed care company. Once you use rating scales to determine how severely ill patients are and when they fully recover, the entry level severity of illness is a good clue as to what kind of treatment might be needed: psychotherapy alone, medication alone, electroconvulsive therapy, or combination treatment. Without measuring recovery, one can't really ascertain whether a patient is still symptomatic or not and still in need of treatment.

The study does not account for a patient's history. The therapeutic approach to patients who present with their first episode of depression is very different than for patients who present with their nth episode. Patients with recurrent episodes of depression or chronic forms of depression need maintenance treatment.

The bottom line is that monitoring improvement of symptoms in depressed patients would seem to be of benefit in determining how to proceed with treatment. In the study, some patients seemingly required very little treatment; others required more extensive treatment because they had more severe illness. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.