Table 2. Comparisons Among Antidepressant Types and Reporting Years.
| Measures | All drugs | TCAs | MAO inhibitors | SRIs | SNRIs | Atypicals | Early (1983–1997) | Late (1998–2010) |
|---|---|---|---|---|---|---|---|---|
| Trials (n): | 124 | 31 | 5 | 47 | 30 | 11 | 57 | 67 |
| Responder RR | ||||||||
| Pooled RR | 1.42 | 1.62 | 1.39 | 1.37 | 1.40 | 1.25 | 1.63 | 1.32 |
| 95% CI | 1.38–1.48 | 1.47–1.78 | 1.11–1.48 | 1.27–1.48 | 1.30–1.51 | 1.15–1.35 | 1.49–1.78 | 1.26–1.38 |
| z-Score | 15.7 | 9.86 | 2.88 | 8.28 | 8.60 | 5.50 | 10.9 | 12.6 |
| p-Value | <0.0001 | 0.0001 | 0.004 | 0.0001 | 0.0001 | 0.0001 | 0.0001 | 0.0001 |
| Responder RD | ||||||||
| Pooled RD | 16.3% | 21.4% | 12.1% | 14.6% | 16.4% | 11.9% | 20.7% | 13.4% |
| 95% CI | 14.4–18.2 | 17.7–25.1 | 3.58–20.5 | 11.5–17.7 | 12.3–20.5 | 8.15–15.7 | 17.5–23.8 | 11.1–15.6 |
| z-Score | 16.6 | 11.3 | 2.79 | 9.21 | 7.81 | 6.19 | 12.9 | 11.7 |
| p-Value | <0.0001 | 0.0001 | 0.005 | 0.0001 | 0.0001 | 0.0001 | 0.0001 | 0.0001 |
| Improvement RD | ||||||||
| Pooled RD | 12.5% | 16.2% | 16.0% | 11.5% | 9.80% | 12.8% | 16.8% | 9.80% |
| 95% CI | 11.0–14.1 | 13.3–19.1 | 0.98–33.0 | 8.70–14.2 | 7.14–12.5 | 8.19–17.4 | 14.5–19.2 | 7.17–10.2 |
| Paired-t | 16.1 | 11.4 | 2.62 | 8.40 | 7.54 | 6.18 | 14.5 | 11.3 |
| p-Value | <0.0001 | <0.0001 | 0.05 | <0.0001 | <0.0001 | 0.0001 | <0.0001 | <0.0001 |
| NNT | 8.0 | 6.2 | 6.2 | 8.7 | 10.2 | 7.8 | 6.0 | 10.2 |
| 95%CI | 7.1–9.1 | 5.2–7.5 | 3.0–102 | 7.0–11.5 | 8.0–14.0 | 5.7–12.2 | 5.2–6.9 | 9.8–13.9 |
Abbreviations: MAO, monoamine oxidase; NNT, number-needed-to-treat (reciprocal of RD); SNRI, serotonin-norepinephrine reuptake inhibitors; SRI, serotonin-reuptake inhibitor; TCA, tricyclic antidepressants.
Based on meta-analytic computation of ratios of responder rates with antidepressants/placebos (RR) or their differences (RD), and on differences in percentage-improvement in initial depression ratings with drug—placebo for 19 antidepressants tested for efficacy in 124 trials summarized in Table 1. Note that most CIs overlap between agents, and that ranking by apparent potency varies among the three outcome measures, but that TCAs appear to be consistently more effective than other types of antidepressants, including SNRIs, SRIs, MAO inhibitors, or atypical agents (bupropion, mirtazapine, and trazodone). Also, early trials (reported in 1983–1997 vs 1998–2010) yield consistently greater drug–placebo differences.