PICO question: In adults with non-psychotic depression who failed to respond to treatment with an SSRI (citalopram), what are the remission rates when treatment is switched to a different SSRI (setraline), a dual 5HT/NE agent (vemalafaxine-XR), or a non-serotenergic agent (buproprion-SR)
Citation: Rush et al 2006. Bupropion-SR, Sertraline, or Venlafaxine-XR after Failure of SSRIs for Depression. N EnglJ Med, 354:12, 1231-42.
Design:Equipoised, stratified, randomized design--patients encouraged to accept all 7 potential treatments (including 4 switch and 3 augmentation options); however, patients could opt to exclude certain level 2 treatments. Follow up period: Up t o 14 weeks to treatment or time to drop out/go onto next level.
Setting: 18 primary care and 23 psychiatric care settings
Participants:727 adult outpt’sw/ non-psychotic MDD who did not remit on or could not tolerate citalopramand who were willing to be switched to another drug.
Intervention: Patient s switched from citalopramto another SSRI (sertraline, up to a maximum daily dose of 200mg), a dual 5HT/NE agent (venlafaxine-XR, up to a maximum daily dose of 375mg), or a non-serotonergicagent (bupropion-SR, up to a maximum daily dose of 400mg).
Outcome measures: Remission as defined by ≤ 7 on the 17-item Hamilton rating Scale for Depression (HRSD-17) and ≤ 5 on the Quick Inventory of Depressive Symptomatology–Self-Report (QIDS-SR-16)
Results:Remission rates based on the HRSD-17 and QIDS-SR-16 did not differ among the three treatments (Sertraline-17.6%/26.6%, Venlafaxine-XR-24.8%/25%, and Bupropion-SR-21.3%/25.5%). QIDS-SR-16 response rates were Setraline-26.7%, Venlafaxine-XR-28.2%, and Bupropion-SR-28.2%
Strengths: Head to head comparison s of active tx’s, txmanuals, large sample, generalizable, clinically relevant primary outcome, adequate follow-up.
Weaknesses: No placebo control, authors w/ affiliation s to Pharma, randomization limited by patient preferences, could not compare to augmentation
Was follow up adequate? Yes
Were the participants randomized? Yes*
Was allocation concealed? Yes
Intention to treat? Yes
Were the groups similar at baseline? Yes
Were pts, MDs and analysts blinded? No
Were the groups treated equally? Yes
Was there an a priori hypothesis? Yes*
Were the outcome clearly defined? Yes
Did they use validated measurements? Yes
Was the placebo/control appropriate? No
Were side effects discussed? Yes
Is the effect clinically important? Yes
Is this congruent with existing studies? No
Is this feasible in your setting? Yes
Were the investigators independent? No
Will this change your current practice? Yes