Parent-Child interaction therapy (PCIT) has been shown to improve positive, responsive parenting and lower risk for child maltreatment (CM), including among families who are already involved in the child welfare system. However, higher risk families show higher rates of treatment attrition, limiting effectiveness. In N = 120 child welfare families randomized to PCIT, we tested behavioral and physiological markers of parent self-regulation and socio-cognitive processes assessed at pre-intervention as predictors of retention in PCIT. Results of multinomial logistic regressions indicate that parents who declined treatment displayed more negative parenting, greater perceptions of child responsibility and control in adult–child transactions, respiratory sinus arrhythmia (RSA) increases to a positive dyadic interaction task, and RSA withdrawal to a challenging, dyadic toy clean-up task. Increased odds of dropout during PCIT's child-directed interaction phase were associated with greater parent attentional bias to angry facial cues on an emotional go/no-go task. Hostile attributions about one's child predicted risk for dropout during the parent-directed interaction phase, and readiness for change scores predicted higher odds of treatment completion. Implications for intervening with child welfare-involved families are discussed along with study limitations.

亲子互动疗法 (PCIT) 已被证明可以改善积极、反应灵敏的养育方式并降低虐待儿童 (CM) 的风险，包括已经参与儿童福利系统的家庭。然而，较高风险的家庭表现出较高的治疗流失率，从而限制了有效性。在N= 120 个随机分配到 PCIT 的儿童福利家庭，我们测试了在干预前评估的父母自我调节和社会认知过程的行为和生理标志物作为 PCIT 保留的预测因素。多项逻辑回归的结果表明，拒绝治疗的父母表现出更多消极的养育方式，在成人与儿童的交易中对孩子的责任感和控制感更强，呼吸性窦性心律失常 (RSA) 增加为积极的二元互动任务，而 RSA 退出以具有挑战性的，二元玩具清理任务。在 PCIT 的以儿童为导向的互动阶段，辍学几率的增加与父母对情绪上去/不去任务中愤怒的面部暗示的更大注意力偏差有关。关于一个人的敌意归因 s 孩子在父母指导的互动阶段预测了辍学的风险，并且准备改变分数预测了更高的治疗完成几率。与研究限制一起讨论了干预涉及儿童福利的家庭的影响。