1 - (SYM 19) Comparative Effectiveness of Varying Levels of Coach Contact on Distressed Relationships: A Randomized Trial of the Online Ourrelationship Program

Introduction. Online programs that reduce relationship distress (e.g. OurRelationship) fill a critical need; however, their scalability is limited by their reliance on coach calls. Support from coaches or other paraprofessionals is associated with online program adherence and treatment gains, but there is heterogeneity in who benefits from coach support. To measure the effectiveness of the online OurRelationship program with varying levels of coach support, we conducted a comparative effectiveness trial with 740 low-income couples in the United States June 2018 to Sept. 2020. 

Method. Couples were randomly assigned to full-coach (ndyads = 226; as originally designed, 1 hour contact), automated-coach (ndyads = 145; as a stand-alone program with tailored automated emails only), contingent-coach (ndyads = 145; as an adaptive program where tailored automated emails is followed by up to 1 hour coaching if couples did not meet progress milestones), or a waitlist control condition (ndyads = 224). Intervention couples completed the two-month online OurRelationship program plus coaching per random assignment, and all couples completed an assessment of relationship satisfaction at baseline, 1-, 2- (post program), 4-, and 6-months post baseline. Analyses were conducted within a Bayesian framework. 

Results. Individuals were on average 33 years old (SD=8), 53% female, 31% non-White, and 13% Hispanic/Latino, and 39% had a high school degree or less. Couples were 92% mixed-gender and 85% were within 200% of the federal poverty line. Completion rates were comparable across conditions (full: 65%, automated: 59%, contingent: 54%) but couples in the full-coach condition were reliably more likely to complete the program than those in the contingent-coach condition. All intervention couples reported reliable gains in relationship satisfaction post program compared to waitlist control couples (dfull = .46, dcontingent = .46, and dautomated = .40) with no reliable differences across intervention conditions. Over 2-months post intervention, couples in full- and contingent-coach conditions maintained gains in relationship satisfaction and couples in the automated-coach condition continued to improve. 

 
Discussion. Given the comparable completion rates and minimal differences in effect sizes across intervention conditions in a diverse, nationwide sample of low-income couples, all three coaching models appear viable. The choice of model for implementation can thus vary depending on available resources as well as couple and stakeholder preferences.