Abstract
This systematic review and network meta-analysis (NMA) sought to compare different antidepressant treatments for treatment-resistant depression (TRD) in order to facilitate evidence-based choices. A literature search of PubMed, Cochrane Library, and Embase from inception until April 13th, 2023 identified randomized, controlled trials (RCTs) of adults with depression who had not responded to at least two antidepressant trials; all RCTs had ≥10 participants per study arm, and participants with bipolar or psychotic depression were excluded. The Cochrane Risk of Bias Tool-2 was used to assess study quality. Response rate was the primary outcome measure. Odds ratios (ORs) using a random effects NMA are reported. From 8234 records, 69 RCTs were included in this analysis, encompassing 10,285 participants (5662 F/4623 M) and 25 separate treatments. Six of the 25 treatments demonstrated a higher response rate versus placebo or sham treatment: electroconvulsive therapy (ECT), minocycline, theta-burst stimulation (TBS), repetitive transcranial magnetic stimulation (rTMS), ketamine, and aripiprazole. ORs ranged from 1.9 (95%CI = [1.25; 2.91]) for aripiprazole to 12.86 (95%CI = [4.07; 40.63]) for ECT. Moderate heterogeneity of the model was observed (I(2) = 47.3% (95%CI [26.8-62%]). Of the included studies, 12.5% were rated as having high risk of bias, 28.13% as having low risk, and 59.38% as showing some concerns. Several effective treatments for TRD showed robust treatment effects across outcomes (ECT, TBS, rTMS, and ketamine), and others showed promising results for some, but not all, outcomes (minocycline, aripiprazole). These findings may help guide evidence-based treatment choices for TRD. Study Registration: PROSPERO (#CRD42023420584).