Therapeutic alliance and change in suicide ideation among psychiatric inpatients at risk for suicide

Background: Therapeutic alliance is thought to be essential in the treatment of suicidality. Surprisingly, studies about the association between therapeutic alliance and change in suicide ideation over the course of treatment are sparse  and reported inconsistent results. Furthermore, theoretically important moderators were rarely explored empirically. Methods: We investigated the association between therapeutic alliance and change in suicide ideation (difference between intake and discharge), unadjusted and adjusted for potential confounding variables (diagnoses,  sociodemographics etc.) in a sample of 351 inpatients treated in a psychiatric department specialized in crisis intervention and suicide prevention. We also explored if the association was moderated by suicide ideation at intake, history of  suicide attempts, and borderline personality disorder (BPD). We ran sensitivity analyses for different diagnostic subgroups, history of suicide attempts, and a quantitative measure of BPD symptoms. Results: We found a moderate association between therapeutic alliance and change in suicide ideation (r = 0.30, p < 0.01). This association remained robust after accounting for potentially confounding variables. Suicide ideation at  intake, history of suicide attempts, and BPD were not statistically significant moderators. Sensitivity analyses led to similar results. Limitations: Therapeutic alliance was assessed only at the end of treatment and causality cannot be inferred from our study method. Conclusions: Therapeutic alliance was a robust correlate of improvement in suicide ideation among psychiatric inpatients at risk for suicide, independent from diagnostic groups and other patient characteristics. Our results support the crucial role of therapeutic alliance in the treatment of patients at risk for suicide.