Psychodynamic Psychotherapies Work: The Compelling Evidence Base

In a recent volume of Psychoanalytic Psychology (37[2]), Dr. Peter Lilliengren provided a comprehensive overview of all randomized control trials (RCT) of psychodynamic psychotherapies published over the past fifty-five years (Lilliengren, 2023). RCT studies have long been viewed as the “gold standard” by which the efficacy of psychotherapy treatments are determined. Studies using RCT designs typically focus on patients carrying specific, discreet clinical diagnoses (for example, panic disorder), and assign patients randomly to either the treatment of focus (for example, Panic-Focused Psychodynamic Psychotherapy; Milrod et al., 2007) or a control group who receives a different but relevant treatment (e.g., applied relaxation techniques) or who are assigned to a waitlist. Outcomes at the end of treatment are compared between assignment groups, with the assumption that changes in one group that reflect statistically greater improvements compared to changes in the other group can be attributable to the treatment that was received, given that other potential factors have been held constant during the trial.

Lilliengren’s dedication to documenting and summarizing these findings from the past half century is important for several reasons. Perhaps most importantly, the presentation of these findings in aggregate helps to counter misguided assumptions by both non-psychoanalytic practitioners and the lay public that psychoanalytic treatments lack a scientific evidence base; the nearly 300 studies summarized by Lilliengren (2023) and others in the field (e.g., Fonagy et al., 2015) highlight decades of rigorous efforts by empirical psychotherapy researchers to utilize RCT designs to critically evaluate and better understand the strengths and limitations of a wide range of psychoanalytic treatments. In addition, his overview provides a useful reference for psychoanalytic practitioners and educators who are interested in evaluating the potential benefits of psychoanalytic treatments for patients with different diagnoses or clinical presentations (for example, determining what patient group may benefit most from mentalization-based psychotherapy as opposed to other psychoanalytic approaches). Finally, Lilliengren’s summary highlights areas that are in need of further research in the field, particularly with regards to the evaluation of psychoanalytic treatments for child and adolescent populations (Lilliengren, 2023).

Both Lilliengren (2023) and earlier review articles (e.g., Fonagy et al., 2015) describe comparisons of evidence supporting the use of psychoanalytic treatment versus cognitive-behavioral and other non-psychoanalytic approaches to a range of mental health conditions. Generally (though certainly not exclusively), psychoanalytic treatments show stronger outcomes when applied to patients experiencing more complex, chronic mental health conditions, and show longer-term stability of functional improvements in comparison to other treatments. Non-psychoanalytic approaches in contrast often show greater efficacy for clinical issues that are more limited scope (e.g., symptoms affecting more restricted domains of functioning or occurring only within certain contexts, such as specific phobias) or time duration. These general findings, while helpful, must always be considered alongside the many limitations of RCT designs, namely that the parameters set by necessity through the study design rarely reflect the ecological reality of everyday clinical practice (for example, practicing clinicians rarely treat patients exhibiting only a single psychiatric disorder; limits on treatment duration and manner of delivery are rarely constrained by non-therapist factors such as a requirement for strict compliance with a treatment manual; and relevant outcomes are rarely conceptualized in a restricted manner, for example focusing only on symptom remission at the exclusion of other relevant factors [such as improved self-esteem or quality of life] that are potentially more salient to both patients and providers). The lack of compatibility between the structure of RCT studies and everyday clinical practice in many cases has contributed to a lack of interest, or has even promoted a sense of skepticism by clinicians (particularly psychoanalytic clinicians) regarding the relevance of empirical findings for informing their practice.

While Lilliengren’s (2023) review illustrates that psychoanalytic treatments can be evaluated validly and effectively using RCT designs, important critiques have been offered that point to potential directions for improving the precision of future trials. In line with the National Institute for Mental Health’s recent shift in emphasizing global underlying mechanisms of psychopathology (through the Research Domain Criteria or “RDoC” framework), scholars have suggested reducing the emphasis in RCT studies on specific diagnostic groups and symptom outcomes, instead shifting to examine transdiagnostic constellations of issues (rather than categories of disorders) and approaches to reducing harmful repetitive cycles (instead of specific symptoms) as a primary outcome. For example, Bakker (2021) has proposed a shift in RCT designs to examine patients with specific defined “problem-maintaining circles” (PMCs) rather than formal DSM-defined disorders. PMCs include self-reinforcing cycles, such as “social anxiety->cope by using alcohol->alcohol abuse->disinhibited behavior” (see figure below), illustrating not only problematic behaviors and symptoms but also the sequences and links that serve to elicit and maintain them in daily life. The combined nuance and specificity of these links between problem areas could help to ensure greater similarity between patients meeting criteria or inclusion in RCT designs (in contrast to broader “blunderbuss” categories that are currently in use, such as major depressive disorder, which may show greater heterogeneity of appearance in everyday life), and also carry greater ecological validity for providers as they consider clinical case formulations with patients. Further, outcome measures could be used which assess not only symptom reduction in specific areas (as is currently the standard), but also interruptions to the PMC cycle that serve to meaningfully reflect subjective experiences of improvement over time. Integrating syndrome and outcome models consistent with the PMC framework could improve precision in evaluating treatment efficacy across patient populations, while improving the relevance of RCT findings to clinicians practicing across theoretical paradigms.

The field of psychoanalysis and empirical research on psychoanalytic concepts in general currently lacks adequate methodology for studying outcomes that are relevant to the unique and complex mental processes that they target. For example, while mentalization-based psychotherapy (MBT) is rooted in a rich and expansive clinical and theoretical literature on the concept of mentalization, sophisticated methods of assessing mentalization as either a trait capacity or dynamic process are currently lacking, with self-report measures representing the primary current method of assessment (e.g., RFQ, MAS, etc). Given that psychoanalytic treatments of all varieties tend to target abilities or tendencies that operate mainly outside of conscious intent or awareness (like the ability to realistically appraise and understand the experiences, feelings, and motivations of self and others, or mentalize), self-report measures are often a poor methodological approach for accurately estimating levels of functioning – whereas more subjectively accessible experiences, like severity of depression or anxiety symptoms, are more amenable to use of questionnaires which solicit direct appraisals from patients’ about their mood and experiences. In order to persuasively demonstrate the unique contribution and utility of psychoanalytic treatment for improving mental health, psychoanalytic researchers must devote energy and resources to develop the basic scientific base and assessment tools that more accurately represent the constructs being targeted by psychoanalytic treatments. Until more rigorous instruments and methods are developed for assessing phenomena like mentalization, defense mechanisms, object representations, and the like, the outcomes of psychoanalytic treatments will continue to be restricted only to symptom severity and behavioral phenomena like hospitalization rates or risk behaviors, limiting the field’s ability to present a more comprehensive and valid picture of mental health challenges and resilience. 

To summarize, Lilliengren’s (2023) comprehensive overview of efforts over the past fifty years to evaluate psychoanalytic treatments using rigorous empirical designs is an invaluable contribution. His summary helps to set a clear record of productivity, empirical excellence, and informative findings that establish psychoanalytic treatments as both evidence-based and worthy of consideration as a validated treatment modality. At the same time, limitations inherent in the underlying RCT design framework and challenges in incorporating more sophisticated and ecologically-valid concepts and outcome measures has impacted the degree to which the unique value and contribution of psychoanalytic treatments have been recognized. Our mandate as a field for addressing these issues is twofold: first, we must continue to advocate for funding, training, and support for RCT studies that may replicate and expand existing evidence for the efficacy of psychoanalytic treatments, with careful attention paid to the under-studied areas identified in Lilliengren’s review (2023); second, psychoanalytic educators and researchers must prioritize the development of new methodologies for assessing foundational treatment concepts, adding an empirical branch to the literature that is as rich and compelling as our existing clinical and theoretical canons. Achieving this second objective will require greater investment in research training during early stages of career development; advocacy within psychoanalysis for the provision of funding and institutional support for research; and addressing longstanding skepticism or cynicism by certain factions within the field towards the very idea of empirical outcome and basic science research. Many psychoanalytic researchers are already making important contributions in this area; the path forward may be challenging, but is rich with opportunity and enthusiasm from those already engaged in this work. Investing in these efforts will dictate whether a review of RCTs for psychoanalytic treatments conducted fifty years from now reflects advances in clinical research and understanding, not only within the practice of psychoanalysis itself but throughout the field of mental health at large.

Figure taken from Bakker, G. M. (2021). Psychotherapy outcome research: Implications of a new clinical taxonomy. Clinical Psychology & Psychotherapy, 1–22. https://doi.org/10.1002/cpp.2638

Bakker, G. M. (2021). Psychotherapy outcome research: Implications of a new clinical taxonomy. Clinical Psychology & Psychotherapy, 1–22. https://doi.org/10.1002/cpp.2638

Fonagy P. (2015). The effectiveness of psychodynamic psychotherapies: An update. World Psychiatry, 14(2), 137–150. https://doi.org/10.1002/wps.20235

Lilliengren, P. (2023). A comprehensive overview of randomized controlled trials of psychodynamic psychotherapies. Psychoanalytic Psychotherapy, 37(2), 117-140, DOI: 10.1080/02668734.2023.2197617