A Preliminary Precision Treatment Rule for Remission of Suicide Ideation

Abstract

Objective There is growing interest in the development of composite precision treatment rules (PTRs) to guide the selection of the treatments most likely to be helpful for individual patients. We present here the results of an effort to develop a preliminary PTR for Collaborative Assessment and Management of Suicidality (CAMS) relative to enhanced‐care as usual based on secondary analysis of the Operation Worth Living (OWL) randomized controlled trial. The outcome of interest is eliminating suicide ideation (SI) within 3 months of initiating treatment. Method A state‐of‐the‐art ensemble machine learning method was used to develop the PTR among the n = 148 U.S. Soldiers (predominately male and White, age range 18–48) OWL patients. Results We estimated that CAMS was the better treatment for 77.8% of patients and that treatment assignment according to the PTR would result in a 13.6% (95% CI: 0.9%–26.3%) increase in 3‐month SI remission compared to random treatment assignment. Conclusions Although promising, results are limited by the small sample size, restrictive baseline assessment, and inability to evaluate effects on suicidal behaviors or disaggregate based on history of suicidal behaviors. Replication is needed in larger samples with comprehensive baseline assessments, longer‐term follow‐ups, and more extensive outcomes.

Predicting treatment failure in regular care Internet-Delivered Cognitive Behavior Therapy for depression and anxiety using only weekly symptom measures.

“…This is important in the same way as it is important to evaluate a therapy against an active control rather than a waiting list. For instance, Kessler, Chalker, Luedtke, Sadikova, and Jobes (2020) used a rather advanced machine-learning ensemble to allocate optimal treatment for suicidal ideation among 148 U.S. soldiers. They found that this could increase 3-month remission of suicidal ideation with 13.6 % (albeit with a wide confidence interval between 0.9 -26.3%).…”

Consensus Statement on Ethical & Safety Practices for Conducting Digital Monitoring Studies with People at Risk of Suicide and Related Behaviors

“…Still absent from all such studies are data on the most effective intervention for a particular person at a particular time point-a longstanding question in mental health research and practice (17). Efforts to address questions such as this are currently underway (18,19) and hopefully will provide guidance that can be used to update the current consensus at that time.…”

The effectiveness of the Collaborative Assessment and Management of Suicidality (CAMS) compared to alternative treatment conditions: A meta‐analysis

“…Such sample sizes are, for example, necessary to have sufficient statistical power to better understand the effects of CAMS on low base-rate outcomes like suicide attempts and to detect significant moderators of intervention effectiveness. Relatedly, recent research identifying significant moderators of CAMS effectiveness (e.g., marital status and baseline distress; Huh et al, 2018) that may be used to predict treatment response to CAMS (e.g., Kessler et al, 2020) underscores the need to further identify when, with whom, and under what conditions CAMS is most effective. Studies may also explore whether utilizing methods that leverage such moderators (e.g., Kessler et al, 2020) to determine optimal approaches for individual patients (e.g., CAMS versus an alternative intervention like DBT) improve treatment outcomes for those who are matched to their optimal approach.…”

“…Relatedly, recent research identifying significant moderators of CAMS effectiveness (e.g., marital status and baseline distress; Huh et al, 2018) that may be used to predict treatment response to CAMS (e.g., Kessler et al, 2020) underscores the need to further identify when, with whom, and under what conditions CAMS is most effective. Studies may also explore whether utilizing methods that leverage such moderators (e.g., Kessler et al, 2020) to determine optimal approaches for individual patients (e.g., CAMS versus an alternative intervention like DBT) improve treatment outcomes for those who are matched to their optimal approach. Researchers should also seek to incorporate repeated assessments of potential mechanisms of change (e.g., therapeutic alliance, increasing hope) in future studies, given that it is unclear the degree to which CAMS components individually contribute to enhanced treatment outcomes, which is also a concern with research on other suicide interventions (Fox et al, 2020); dismantling studies may also help clarify which components of CAMS are most effective in producing change.…”