Are there any robust predictors of “sudden gainers,” and how is sustained improvement in treatment outcome achieved following a gain?

2019 
OBJECTIVE: It has been widely demonstrated that the process of change many patients undergo in therapy is not linear. Some patients benefit greatly from large sudden improvements, commonly referred to as "sudden gains." It is less clear whether certain baseline characteristics make patients more prone to displaying sudden gains, as well as what mechanisms are responsible for the lasting effects of sudden gains. METHOD: In a sample of 547 patients receiving treatment in an outpatient mental health clinic, a machine learning approach was used to search for potential predictors of sudden gains. A within-patient mediation model was used to investigate whether alliance serves as a mechanism underlying the sustained effect of sudden gains. RESULTS: Twelve percent of patients showed sudden gains. Consistent with previous studies, no robust predictors of sudden gains were found, even when using an approach capable of evaluating the contributions of multiple predictors and their interactions. A significant within-patient mediation model was found, according to which sudden gains predict subsequent strengthening in alliance, which in turn predict subsequent improvement in life satisfaction and psychological dysfunction. These findings support the proposed theoretical framework whereby alliance is an important ingredient of an upward spiral that may results in sustained sudden gains. CONCLUSIONS: The findings provide first evidence of the presence of an ingredient responsible for the sustained effect of sudden gains, using a within-patient mediation model. The findings support the important role alliance may play in the consolidation and subsequent expansion of the effect of sudden gains. (PsycINFO Database Record (c) 2019 APA, all rights reserved).
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    9
    Citations
    NaN
    KQI
    []