A 2010 meta-anlysis* by Jay Fournier and Robert DeRubeis et al (JAMA, 303, January 6, 2010) showed that for mild-to-moderate depression, medication has little effect compared to placebo. Most drug studies showing an impact for medications involve patients with more severe depressions. Therefore, past studies have been more likely to show a strong effect. This analysis shows that for the larger mass of people with less severe depression, medication has little impact. This finding has major implications for psychotherapists.
That psychotherapy alone can be highly effective with Reactive Depression, Adjustment Disorder with Depressed Mood, and mild-to-moderate Major Depressive Disorder is not surprising to seasoned therapists. But the finding the medication alone is not effective in those cases is a new finding.
Note that the severity of depression in the studies examined by Fournier was measured by the Hamilton Depression Rating Scale (which is on available online). This scales allows the professional to rate 22 symptoms. The range of scores runs from 0 to 69 points. Scores of 25 or higher, that is, high depression scores, correlated with larger differences between the treatment group and the placebo controls. That means patients with a wider array of more severe symptoms derived more benefit from drug treatments than did patients with fewer or less severe symptoms.
Fournier et al pointed out that those with mild depression are often excluded from controlled trials. Also note that most drug studies investigate treatment with patients who have depression alone and exclude those with co-occurring disorders, the type of patients most often seen in daily practice. Fournier et al found that effect sizes for medications for mild-to-moderate depression are on the order of .11 to .17, hardly any overall effect whatsoever (as scores below .20 are not considered to be significant effect sizes). The effect size with more severe cases was .47, which is considered barely a medium treatment effect, though not a robust effect. Fournier et al did point out that the research was of a short-term nature (6 weeks) and was limited to two anti-depressants of the many on the market, paroxetine and imipramine. They speculate that other medications would not produce a different result.
Note: By comparison, the effect size for the psychotherapies for which we have the best evidence runs closer to 1.0 or higher, a significant effect more than twice that of one medication used alone. One could conclude that for mild-to-moderate depression, withholding or delaying therapy in favor of medication looks to be a reckless course of action. Insisting upon medication in these cases as a first line treatment appears to be wasteful of time and resources.
Another study by Tang and DeRubeis et al (Arch of Gen Psychiatry, 66, December, 2009) came to the same conclusion. DeRubeis also noted that the long term positive outcomes revolve around a change in outlook and thinking, effects resulting from the best available psychotherapies. Those are not effects produced by medications.
A meta-analysis by Spielmans, Berman, and Usitalo (J Nerv & Mental Diseases, 199, March, 2011) demonstrated that psychotherapy is equal in impact to medications for the short-term and has better outcomes for the long term. However, the study showed only evidence-based psychotherapies (for depression) were equal or superior to medications. Other therapies had worse outcomes.
The NIMH STAR*D study (http://www.nimh.nih.gov/trials/practical/stard/index.shtml) demonstrated that Cognitive Therapy had about the same effect as medication. In addition, only about 30% of the subjects achieved an end to the depression from one round of medication. This is consistent with the finding of a mere .47 effect size for medication. Medication also has a downside – there is a rebound effect when the medications are discontinued.
A number of studies show psychotherapy affects the brain in a number of positive ways which are not unlike the benefits from medication.
A review by Antonuccio and Danton (Professional Psychology: Research and Practice, 26, December 1995) determined cognitive-behavioral therapy is at least as effective as medication in lowering depression scores, including scores for vegetative symptoms (e.g., sleep, appetite, energy/fatigue). That means CBT is effective even for more severe depressions.
1. Start Psychotherapy or Combined Treatment Immediately
Although primary care guidelines recommend two trials of medication by the PCP before referring the patient to a mental health professional, the data clearly show that for the majority of depression patients, a very timely referral for psychotherapy is indicated. For more severe cases, the patient should be started on both psychotherapy and anti-depressants.
Also remember, these are group data. A patient with moderate depression, but who continues to be symptomatic with a few sessions of psychotherapy, should still be referred promptly for medications.
In addition, a patient who comes into the clinic with a larger range of symptoms, for instance, Hamilton scores over 25 or Beck Depression Index scores over 26, should usually start on both medication and psychotherapy immediately. Persistent difficulties with sleep, appetite, low energy level, fatigue, and unyielding low self-esteem would be amongst the symptoms which suggest a referral for medications earlier rather than later.
Those patients with long-lasting depression which has not abated and recurrent depressions should also be started on both treatments.
The severity and type of symptoms is not correlated with whether psychotherapy will be effective. So therapists should be optimistic that therapy will help 80-90% of depression patients, often in a 6-12 week period.
2. Evidence-Informed Treatments
The Spielmans et al study noted that to produce the effect in which recovery is achieved through psychotherapy without medication, psychotherapies which have been proven effective with depression must be used, not general counseling. These include, for example, Interpersonal Psychotherapy, Cognitive Therapy or Cognitive Behavioral Psychotherapy, and focused Pychoanalytic Psychotherapy.
3. Start Quickly
These findings are consistent with an approach of having a therapist screen patients first before making a referral for medications. Because studies show significant relief can be achieved in as little as six weeks, the screening should take place very rapidly and psychotherapy or combination therapy should begin quickly. Having data that we can have a rapid effect creates some obligation to get the treatment started and the patient’s distress and impairment reduced.
From this data, understand the responsibility we have due to the importance and value of the psychotherapeutic work we do. But also be aware of thresholds, such as high symptom scores, for making the decision not to delay accessing medications. For severe depression, some evidence suggests the newer anti-depressants may work more quickly than psychotherapy.
4. Offer Hope and Confidence
Psychotherapists can be confident that most patients will benefit from treatment. Patients with depression can be informed that most people** feel better in a short period of time. Those who have started on medications before coming for psychotherapy, but who have not had much relief, may be on too low a dose, may have been taking the medication inconsistently, may have experienced unwanted side effects, or may lack confidence in medication. So a therapist should check on all the medications a patient is taking, including dosages, and should check on patient adherence.
These data are indeed important for therapists. In essence, ‘the buck stops here.’ Given these findings, it is a therapeutic responsibility to deliver evidence-informed treatment for depression and to take the responsibility for a beneficial outcome.
* The data in this article is based upon meta-analysis. This is a methodology for comparing the findings of many studies. The method calculates an effect size, the average difference (in standard deviation units) between the treatment group and the comparison, control, or placebo group across numerous studies. Effect sizes near zero mean no difference between the groups. Effect sizes for the best psychotherapies are well above 1.00.
**This essay does not address repetitive and treat-refractory depression or Bipolar Disorder.