From Wiki Journal Club
Jump to navigation Jump to search
Rush AJ et al. "Acute and longer-term outcomes in depressed outpatient requiring one or several treatment steps: A STAR*D report". The American Journal of Psychiatry. 2006. 163(11):1905-1917.

Clinical Question

Among patients with unipolar major depression without psychotic features, how often is remission achieved when using a tiered protocol?

Bottom Line

A four-step treatment algorithm aiming for depression remission in non-psychotic depressed outpatients resulted in remission rates of 36.8% and 30.6% for first two steps, respectively, with higher relapse rates for individuals requiring the third and fourth treatment steps.

Major Points

The Sequenced Treatment Alternatives to Relieve Depression (STAR-D) trial employed a four-step treatment protocol to outpatients with non-psychotic depression with total remission as the goal. Step 1 was citalopram only. Those with remission were entered into a 12-month follow up protocol. Those who did not attain remission were randomly enrolled in step 2 treatment, which consisted of 7 different treatment options. The equipoise stratified design allowed patients to choose a switch or augmentation strategy for steps 2 and 3.

As expected, remission rates were higher in earlier steps, with step 1 achieving remission in 36.8% of patients. Among the individuals that progressed to step 2, remission was achieved in 30.6%. Steps 3 and 4 achieved remission in 13.7% and 13%, respectively.

STAR-D sheds light on the success of a specific tiered approach at treating unipolar depression.[1] However, the study was limited in part by its open-label design and lack of a placebo control, rendering it difficult to tease out the impact of pharmacological therapy and CBT from the natural history of depression.


American Psychiatric Association (2010)[2]

  • Psychotherapy is a reasonable monotherapy for patients with mild to moderate MDD
  • Patients with mild, moderate, or severe MDD can be treated with antidepressants, findings that would suggest their use as preferred modality include prior response to them, moderate to severe symptoms, significant sleep or appetite disturbances, agitation, preference of the patient, and suspicion for need of maintenance therapy
  • Suggest combining depression-focused psychotherapy with pharmacotherapy for patients with moderate to severe MDD


  • Multicenter, four tier, equipoise-stratified randomized trial
  • N=3,671
    • Step 1: N=3,671
    • Step 2: N=1,439
    • Step 3: N=390
    • Step 4: N=123
  • Setting: 41 centers in the US
    • Primary care: 18
    • Psychiatric: 23
  • Follow-up: 12 months
  • Primary outcome: Depression remission by QIDS-SR16 score


Inclusion Criteria

  • Age 18-75 with with unipolar major depressive disorder without psychotic features
  • Baseline score of at least 14 on the Hamilton Depression Rating Scale
  • Patients must have sought treatment (ie, no recruitment through advertising)

Exclusion Criteria

  • Pregnancy or breastfeeding
  • Prior treatment with a component of the first two trial steps
  • Bipolar, psychotic disorders, OCD, eating disorder, or substance dependence requiring inpatient detoxification
  • Suicidal ideations requiring hospitalization
  • General medical condition preventing protocol medications in the first two treatment steps

Baseline Characteristics

From the step 1 group.

  • Demographics: Age 40.7 years, female 62.2%, unemployed 36.3%, retired 5.9%
    • Race/ethnicity: White 76.4%, Black 17.1%, other 6.5%, Hispanic 12.3%
    • Marital status: Single 29.5%, married or cohabitating 41.6%, divorced or separated 25.7%, widowed 3.2%
    • Insurance: Private 52.6%, public 13.2%, none 34.2%
  • Care by psychiatrist: 61.8%
  • Depression-specific:
    • Age at first episode: 25.5% (<18 years 37.2%)
    • Recurrent: 74.7% (episodes 5.9)
    • Illness duration: 15.3 years
    • Family history: 54.9%
    • Suicide attempt: 16.9%
    • Duration of current episode: Mean 24.5 months (≥2 years 24.8%), median 7.8 months
  • Symptoms: Anxious 44.6%, atypical 17.1%, melancholic 19.7%
  • Additional diagnoses: agoraphobia 10.5%, alcohol abuse/dependence 11.8%, bulimia 12.1%, drug abuse/dependence 7.3%, GAD 20.6%, hypochondriasis 3.9%, OCD 13.2%, panic disorder 11.3%, PTSD 17.8%, social phobia 29.1%, somatoform disorder 2.2%
  • Comorbid axis 1 condions: None 38.5%, one 26.5%, two 15.5%, three 8.2%, ≥four 11.2%
  • Scores on symptom-specific questionnaires:
    • Quality of Life and Enjoyment Satisfaction Questionnaire: 41.8
    • SF-12: mental 26.6, physical 49.5
    • Work and social adjustment scale: 23.6
    • HRSD17: 19.9
    • IDS-C30: 35.5
    • QIDS-SR16: 15.4
    • Cumulative illness rating scale: Categories endorsed 3.0, total 4.2, severity index 1.2


Participants progressed through four different levels of treatment.

  • Step 1: All participants received flexible-dose citalopram. If in remission, enter 12-month naturalistic follow up. If no remission, proceed to step 2.
  • Step 2: 3 augmentation strategies (citalopram + bupropion; citalopram + buspirone; citalopram + CBT) and 4 switch strategies (buproprion, sertraline, venlafaxine, CBT). If in remission, enter 12-month naturalistic follow up. If no remission, proceed to step 3.
  • Step 3: 2 augmentation strategies (lithium or triiodothyronine [T3]) and 2 switch strategies (nortriptyline, mirtazapine). If in remission, enter 12-month naturalistic follow up. If no remission, proceed to step 4.
  • Step 4: randomized to treatment with either tranylcypromine or combination venlafaxine XR and mirtazapine.


Comparisons are level 1 vs. level 2 vs. level 3 vs. level 4.

Primary Outcome

Depression remission by QIDS-SR16 score
Step 1: 36.8% (higher remission rate than those in step 2; P<0.001)
Step 2: 30.6% (higher remission rate than those in step 3; P<0.001)
Step 3: 13.7%
Step 4: 13%

Secondary Outcomes

Response measured by clinician and patient self-report
Step 1: 48.6%
Step 2: 28.5%
Step 3: 16.8%
Step 4: 16.3%


  • Open label
  • Secondary outcomes were based heavily on patient report
  • Patients in psychotherapy had to pay some of it themselves
  • No placebo[3]


  • National Institute of Mental Health (NIMH)
  • No direct funding from pharmaceutical companies. Medications that were not available in generic form were provided to patients at no cost from the patent-holding company.

Further Reading