The Best of Both Worlds? General Principles of Psychopathology in Personalized Assessment

• Published in: Journal of psychopathology and clinical science 2023-10, Vol.132 (7), p.808-819
• Main Authors: Olthof, Merlijn, Hasselman, Fred, Aas, Benjamin, Lamoth, Daniela, Scholz, Silvia, Daniels-Wredenhagen, Nora, Goldbeck, Florens, Weinans, Els, Strunk, Guido, Schiepek, Günter, Bosman, Anna M. T., Lichtwarck-Aschoff, Anna
• Format: Article
• Language: English
• Subjects: Anxiety disorders; Changes; Classification; Clinical outcomes; Female; Human; Male; Personalization; Principal components analysis; Psychological Assessment; Psychological processes; Psychopathology; Psychotherapy; Questionnaires; Ratings & rankings; Systems approach; Test Construction; Time series; Treatment Outcomes
• ISSN: 2769-7541; 2769-755X
• DOI: 10.1037/abn0000858

A complex systems approach to psychopathology proposes that general principles lie in the dynamic patterns of psychopathology, which are not restricted to specific psychological processes like symptoms or affect. Hence, it must be possible to find general change profiles in time series data of fully personalized questionnaires. In the current study, we examined general change profiles in personalized self-ratings and related these to four measures of treatment outcome (International Symptom Rating, 21-item Depression Anxiety and Stress Scale, daily symptom severity, and self-reflective capacity). We analyzed data of 404 patients with mood and/or anxiety disorders who completed daily self-ratings on personalized questionnaires during psychotherapy. For each patient, a principal component analysis was applied to the multivariate time series in order to retrieve an univariate person-specific time series. Then, using classification and regression methods, we examined these time series for the presence of general change profiles. The change profile classification yielded the following distribution of patients: no-shift (n = 55; 14%), gradual-change (n = 52; 13%), one-shift (n = 233; 58%), reversed-shift (n = 39; 10%) and multiple-shifts (n = 25; 6%). The multiple-shift group had better treatment outcome than the no-shift group on all outcome measures. The one-shift and gradual-change groups had better treatment outcome than the no-shift group on two and three outcome measures, respectively. Overall, this study illustrates that person-specific (idiographic) and general (nomothetic) aspects of psychopathology can be integrated in a complex systems approach to psychopathology, which may combine "the best of both worlds." General Scientific Summary Personalized questionnaires have clinical and intuitive appeal, but the question rises how to generalize if questionnaires measure different constructs in different patients. In this article, we develop a novel method to abstract general change profiles from personalized daily questionnaire data collected during psychotherapy. We find that the way in which patients change on their personalized questionnaire over the course of treatment is related to treatment outcome as predicted by the complexity theory of psychopathology.