High-quality RCTs needed to estimate effectiveness of youth suicidality psychotherapies

Methodological problems and high risk for bias prevented a consistent estimate of the comparative performance of certain psychotherapeutic modalities for youth self-harm and suicidality, according to study results.

“At present, there are insufficient data from randomized clinical trials (RCTs) to recommend targeted pharmacological treatments for self-harm or suicidal behavior in youths,” Anees Bahji, MD, of the department of psychiatry at the University of Calgary in Canada, and colleagues wrote in JAMA Network Open. “However, some nonpharmacological interventions, including psychotherapies, appear to improve some aspects of suicidal behavior. Several meta-analyses have synthesized data from RCTs examining psychotherapies for self-harm and suicidality in youth populations.”

The researchers conducted the current systematic review and network meta-analysis to re-assess the comparative efficacy and safety of psychotherapies for treating self-harm and suicidal behaviors among children and adolescents. They utilized a network meta-analysis for its ability to mitigate biases related to specific studies, which may not be identifiable via head-to-head RCTs. They searched PubMed, MEDLINE, PsycINFO and Embase between inception and September 2020 and included RCTs that compared psychotherapies for child/adolescent suicidality and/or self-harm with control conditions. Dichotomized self-harm and retention in treatment served as the primary outcomes. Dichotomized all-cause treatment discontinuation and scores on instruments that assessed suicidal ideation and depressive symptoms served as secondary outcomes.

Bahji and colleagues included 44 RCTs with 5,406 total participants, of whom 76% were women. The included studies were published between Jan. 1, 1995, and Dec. 31, 2020. Median treatment duration was 3 months and median follow-up period was 12 months. Results showed no increases in study withdrawals or improvements in retention in treatment among the investigated psychotherapies compared with treatment as usual. The researchers noted a link between dialectical behavioral therapies and self-harm reductions (OR = 0.28; 95% CI, 0.12-0.64) and suicidal ideation (Cohen d SMD = 0.71; 95% CI, 1.19 to 0.23) at treatment’s end. Further, they observed an association between mentalization-based therapies and self-harm decreases (OR = 0.38; 95% CI, 0.15-0.97) and suicidal ideation (Cohen d SMD = 1.22; 95% CI, 2.18 to 0.26) at follow-up’s end. High risk for bias overall, heterogeneity, publication bias, inconsistency and imprecision led to a downgrading in the quality of evidence.

“Although the findings of this review suggest that some psychotherapies are well tolerated and have some efficacy for specific measures of self-harm or suicidality, the estimates indicated that the evidence quality was low to very low for most psychotherapies,” Bahji and colleagues wrote. “A lack of consistent evidence precludes a definitive hierarchy of treatments and suggests a need for additional high-quality RCTs.”