Systemic therapy is a widespread psychotherapeutic approach. There are numerous definitions of systemic therapy, proposing multiple theoretical approaches and settings. Consequently, reviews on systemic therapy differ substantially in their scope. We adopt the definition by von Sydow and colleagues (1), whose definition has served as a base for several systematic reviews and meta-analyses on systemic therapy (2–4). According to their definition of systemic therapy, mental disorders are understood within the context of social systems (e.g., families, couples, peers), as well as the relations and interactions of those who compose these social systems. Systemic therapy assumes a recursive causality between symptoms, (social) interactions and social construction of realities. Important others are included directly or virtually in the therapy (e.g., by systems-oriented questions and genogram work) (5). According to this definition, not all kinds of family and couple therapies are part of systemic therapy. Both settings can be found in other theoretical orientations as well (e.g., cognitive behavioral therapy (CBT)) (1, 6–8). Based on von Sydow and colleagues’ (2010) definition, several systematic reviews and meta-analyses confirmed the efficacy of systemic therapy for adults (1–3) and adolescents as well as children with regard to symptom change (1, 4, 9). However, the search strategy of the aforementioned reviews was often insufficiently efficient with regard to the modern standards of psychotherapy research. For the purpose of this systematic review and meta-analysis, we will review, integrate and thus optimize the search strategies applied in previous reviews. Given the growing recognition of systemic therapy, we believe that the development of an efficient search strategy will be a helpful contribution to the field.
Previous meta-analyses mainly concentrated on symptom change as the most prominent outcome variable in psychotherapy treatment research. Primary studies that reported only system functioning were usually excluded. Since systemic therapy views symptoms within the context of social systems, the focus on symptom change contradicts the central idea of systemic therapy. System functioning measures how relevant members of a social system communicate and interact with one another (10). It can measure relationship quality and collective efficacy (11) as well as climate of a social system or burden of family members (12). Social systems themselves are defined as intimate social structures including all members (e.g., families, couples) significant for the development, maintenance, and change of e.g., an individual’s symptomatology (12, 13).
As systemic therapy draws its focus on social systems, one would assume that it is more effective referring to system functioning as compared to therapy schools who primarily focus on changing the symptoms of individuals. This was illustrated by a Cochrane review and meta-analysis focusing on couple therapy for depression. The authors confirmed that couple therapy had stronger effects on system functioning compared to individual symptomatology (14). To our knowledge, no meta-analytic evidence has been provided on the effects of systemic therapy on symptom as well as system outcomes.
The previous meta-analyses showed that systemic therapy had medium effects on symptom change compared to no treatment (for adults: g = .51; for adolescent/children: g = .59) or pharmacotherapy (for adults: g = .57), and small effects compared to alternative active treatments (for adults: g = .25; for adolescents/children: g = .32). Systemic therapy had equal effects compared to bona fide psychotherapies for adults (g = .12) as well as compared to CBT for adolescents and children (g = .15) (2, 9). Although disorder-specific differences were discovered, they need to be interpreted with caution due to the high heterogeneity of effects based on a limited number of included studies (2, 9). We will take a closer look at these factors contributing to this heterogeneity by taking into account additional moderators (e.g., setting) and outcomes (e.g., system functioning) that have not been considered in previous reviews.
Several experts in the field (e.g. 2, 15, 16) propose a more distinguished analysis of different treatment approaches for systemic therapy. Previous meta-analyses compared the efficacy of therapy approaches that were exclusively related to systemic theory to those which integrate systemic therapy with other approaches (e.g., CBT): ‘Pure’ systemic therapy had equal effects compared to integrative systemic therapy for both adults (2) as well as adolescents and children (9). However, there was an integrative wave in systemic therapy research in recent decades (17–19), increasing the number of studies on integrative systemic therapy (15, 16). There was a comparable integrative wave in CBT where a variety of treatment approaches were developed and scientifically evaluated. Therefore, we aim to replicate the comparison between “pure” vs. “integrative” systemic therapy. Apart from integrative treatments, combined treatments add different treatments such as medication or social skills training to systemic therapy in order to create a comprehensive treatment program. We also aim to compare studies where systemic therapy was the central intervention within a combined treatment program to studies, which only offered systemic therapy.
In order to evaluate heterogeneous treatment approaches, modern evidence-based psychotherapy research suggests manualization and measuring adherence. Adherence refers to the extent to which a therapist adheres to a manual (20). While the number of manualized systemic treatments has considerably increased in recent decades (15), it has been questioned if manualization can fully reflect the theoretical foundations of systemic therapy. Several studies have cast doubt on if and how adherence can be measured in systemic therapy (21, 22). These questions are addressed in other psychotherapeutic approaches as well, when there are mixed effects for adherence on symptom outcome (23, 24). However, there are some promising results of measuring manualization and adherence for systemic family therapy (25, 26). These findings indicate that manualization and adherence is positively associated with outcomes. In addition to this, assessing researcher allegiance was also suggested (27). Consistent with a large body of meta-analytic findings, there is strong evidence confirming the substantial effects of allegiance on treatment outcomes for other psychotherapeutic approaches (28–30). Thus, in the present meta-analysis we will consider manualization, adherence and allegiance as moderators that have not been incorporated sufficiently in previous reviews.