Key Takeaways
Extended Takeaways
- This was an exceptionally high-risk TRD cohort: participants averaged approximately 29 years of depression, 13 failed antidepressant treatments, 71.0% had failed at least 1 interventional treatment in the current MDE, and 40.4% had attempted suicide.
- The suicidal ideation outcome was based on a prespecified composite suicidal ideation score that integrated MADRS item 10, QIDS-C item 12, and QIDS-SR item 12 into a 0 to 9 scale, allowing clinicians to interpret change across both clinician-rated and self-reported ideation.
- For patients entering with meaningful suicidal ideation, the benefit signal favored persistence of improvement over time rather than a large separation in mean scores at single visits: active VNS showed more months with clinically meaningful improvement across months 3-12 (OR = 1.43; 95% CI, 1.004–2.021; P = .047), even though mean CSI differences at individual time points were not significant.
- Symptom improvement in suicidal ideation was gradual in both groups, consistent with VNS not being an acute antisuicidal intervention; by month 12, the mean (SD) drop in CSI score was 2.42 (0.19) for active VNS and 2.12 (0.19) for sham VNS, and the visit effect across treatments was significant (P<.001).
- The treatment effect was directionally consistent regardless of how suicidal ideation was measured: across single-item and 2-item combinations, effect sizes ranged from 1.31 to 1.49 for meaningful or better improvement and from 1.27 to 1.46 for remission.
- Applicability is limited for patients with acute suicidal intent because RECOVER excluded participants with severe suicidal intent/behavior at baseline; these findings are more relevant to managing chronic suicidal ideation in markedly treatment-resistant unipolar depression.
