Therapeutic Alliance and Intervention Approach Among Acutely Suicidal Patients.

Abstract

Decades of research have supported a positive correlation between the therapeutic alliance and a reduction in psychological symptoms, but studies investigating the correlation of therapeutic alliance with suiciderelated outcomes are sparse. Preliminary studies suggest a negative correlation of alliance with suicidal thoughts and occurrence of nonsuicidal self-injury following suicidefocused treatments (Bedics, Atkins, Comtois, & Linehan, 2012; Gysin-Maillart, Schwab, Soravia, Megert, & Michel, 2016) but no correlation in treatments that are not suicide focused (Bedics et al., 2012; Bryan et al., 2012). Interpersonal dynamics and intervention approaches may therefore interact with each other. Unfortunately, studies to date have not explicitly tested this possibility. To address this knowledge gap, we conducted a secondary analysis of a recently completed randomized clinical trial (RCT) finding that crisis response planning (CRP) significantly reduced suicidal thoughts and behaviors among 97 high-risk soldiers as compared to treatment as usual (TAU; Bryan et al., 2017). After providing consent, soldiers reporting suicide ideation during the past week and/or a lifetime history of suicide attempts were randomly assigned to either a CRP or TAU. Immediately postintervention, soldiers completed the Working Alliance Inventory—Short Form (Hatcher & Gillaspy, 2006). Suicide attempts were assessed with the Suicide Attempt Self-Injury Interview (Linehan, Comtois, Brown, Heard, & Wagner, 2006), and severity of suicide ideation was assessed with the Scale for Suicide Ideation (Beck & Steer, 1991). Although CRP alliance ratings were somewhat higher (M = 58.9, SD = 12.2) than TAU ratings (M = 53.3, SD = 15.6), this difference was not statistically significant (F (1, 91) = 3.19, p = .077, d = 0.42). Results of generalized linear mixed effects modeling with a robust sandwich estimator and baseline suicide ideation entered as a covariate indicated the interaction of therapeutic alliance with treatment group was a statistically significant predictor of follow-up suicide ideation (F (1, 41) = 6.83, p = .013) and suicide attempt (F (1, 87) = 4.97, p = .028). The forms of these interactions are displayed in Figure 1. In CRP, suicide ideation (B = −0.10, SE = 0.05, p = .044, R = 0.42)