The Marburg Declaration: Advancing Psychological Treatment

The provided document, “The future of psychological treatments: The Marburg Declaration,” published in Clinical Psychology Review in 2024, presents a comprehensive overview of the current state of psychological treatments and offers recommendations for their future development. Authored by a consortium of clinical psychology experts, including Winfried Rief, Gordon J.G. Asmundson, Richard A. Bryant, David M. Clark, Anke Ehlers, Emily A. Holmes, Richard J. McNally, Carmem B. Neufeld, Sabine Wilhelm, Adam C. Jaroszewski, Max Berg, Anke Haberkamp, and Stefan G. Hofmann, the declaration stems from a Think-Tank held in Marburg, Germany, in August 2022.

The core motivation behind the Marburg Declaration is the recognition that despite psychological treatments being widely acknowledged as evidence-based for various mental disorders, significant challenges persist. These challenges include:

• A substantial proportion of patients not achieving full recovery.
• Obstacles hindering the dissemination, implementation, and training of psychological treatments.
• The considerable societal burden of mental health problems, which are a primary reason for work disability and a significant expense for healthcare systems.
• The fact that many psychological interventions are rooted in theoretical concepts that have evolved little, and some lack connection to basic science.
• A general societal need to optimize mental health care, requiring improvements in training, implementation, and research.

To address these issues, the experts reviewed evidence and analyzed barriers, proposing specific recommendations for both research and practice. The declaration aims to stimulate a significant shift in understanding mental disorders and their treatments, thereby initiating a new era of evidence-based psychological interventions.

The recommendations and discussions are organized into several key areas:

• Bridging the Gap Between Basic Science and Clinical Applications
  • Challenge: While psychological treatments are effective, many patients don’t achieve full remission, with some reviews indicating only half improve. There’s a need to translate findings from basic experimental studies on cognitive, biological, and emotional mechanisms into clinical treatments to improve success rates.
  • Recommendations:
    • Focus studies on mechanisms and processes relevant to individual mental challenges rather than heterogeneous “disorders,” enabling personalized treatment planning linked to basic science.
    • Gain greater clarity on specific symptoms and problems to be targeted, including their phenomenology and patient perspectives.
    • Identify missing links between symptoms and relevant concepts in basic science (e.g., attention, memory, emotional processes, neurobiology).
    • Enhance the relevance of experimental paradigms to clinical symptoms and confirm disorder-relevant mechanisms and change mechanisms experimentally.
    • Improve information exchange between basic findings and clinical applications (“From bench to bedside – and back”) through conferences, symposia, and special issues.
    • Investigate mechanisms implicated in poor treatment response and adverse effects.
    • Utilize sophisticated methodological and computational tools, such as adaptive and Bayesian randomized trial designs, network analysis, machine learning algorithms, and causal inference methods, to analyze relevant processes.
• Optimal Deployment of Psychotherapy: Lessons from the English IAPT Experience
  • Challenge: Despite effective treatments, few people benefit due to issues like limited availability, long waitlists, and high costs.
  • IAPT as a Model: The Improving Access to Psychological Therapies (IAPT) program (now NHS Talking Therapies for Anxiety and Depression) in England serves as a prominent example. Launched in 2008, it provides psychological therapy to around 670,000 people annually, with economics playing a key role in its development.
  • Key Learnings and Recommendations:
    • IAPT implemented a session-by-session outcome monitoring system, ensuring data collection from 99% of patients, which is crucial for accurate assessment and improvement.
    • Outcomes steadily improved from less than 40% recovery to meeting the 50% target for seven consecutive years, driven by continuous quality improvement and data reflection.
    • Clinical guidelines (e.g., NICE recommendations for CBT) and structural aspects of service organization (e.g., timely treatment, SMS reminders, sufficient sessions) are vital for success.
    • It’s crucial to identify and address the needs of subgroups of patients who do less well, as IAPT did with minority ethnic groups and specific latent profiles, leading to improved outcomes for all.
    • The large datasets from services like IAPT allow for unprecedented data completeness for RCTs and can reveal positive impacts on physical health (e.g., reduced cardiovascular disease risk, lower physical health treatment rates).
    • International collaboration using IAPT-like services has created a broader evidence base, demonstrating improved labor market participation, which has helped secure further funding.
• New Ways for Understanding Mental Health Problems: Network Analysis
  • Challenge: Traditional categorical diagnostic models (e.g., DSM) face discontent due to widespread comorbidity and difficulty in identifying biomarkers for discrete disorders.
  • Network Perspective: This alternative conceptualizes mental disorders as emergent phenomena arising from dynamic interactions among elements like symptoms (nodes). The strength of association between symptoms is represented by edges.
  • Applications:
    • Cross-sectional networks visualize symptom relations at the group level, while time series networks capture temporal dynamics within individuals.
    • Gaussian graphical models (GGMs) estimate non-spurious partial correlations between symptoms, adjusting for other symptoms in the network, though they don’t show direction of influence.
    • Bayesian network analysis can return a directed acyclic graph (DAG), indicating the direction of prediction between symptoms, offering complementary insights.
  • Recommendations:
    • Utilize dynamic network models at both nomothetic (group) and idiographic (person-centered) levels to capture psychopathology complexity and analyze patterns of change.
    • Employ network approaches for individualized case conceptualization and treatment planning, suitable for new frameworks like process-based therapy.
    • Use intensive time-series networks (e.g., via Ecological Momentary Assessments – EMA) to tailor interventions to target key symptoms dynamically.
• Using Digital Health to Scale Evidence-Based Mental Health Care (DMHIs)
  • Potential: DMHIs, delivered via technology, have shown efficacy for various mental disorders and can address key structural barriers such as limited clinician availability, long waitlists, high costs, and logistical issues. They can be self-guided or supported by coaches/clinicians in stepped or stratified care approaches.
  • Precision and Data Collection: DMHIs provide a platform for collecting multi-modal data with high temporal resolution, enabling personalized (N=1) machine learning models to guide prevention and treatment decisions, including just-in-time adaptive interventions (JITAIs).
  • Challenges and Recommendations:
    • Issues include lack of clinician/patient consultation in development, complexity, privacy concerns, low evidence basis, poor engagement/adherence, and lack of integration into healthcare systems.
    • Recommendations:
      • Develop DMHIs using iterative user-centered design, incorporating all stakeholders (clinicians, patients, insurers, regulators).
      • Maximize inclusion of evidence-based content and rigorously test DMHIs in controlled and real-world settings.
      • Include patients from diverse backgrounds to improve personal relevance.
      • Improve engagement and adherence through coaches, peer-support, online communities, and game design principles.
      • Facilitate integration into healthcare systems by developing pragmatic regulatory frameworks, evidence standards, reimbursement pathways, and training for providers.
• Challenges for Research Methods in Clinical Psychology
  • Issues: Clinical psychology research suffers from problems in robustness and replicability, influenced by factors like allegiance effects, poor therapist training, varying assessment tools, and insufficient reporting of trial details. RCTs often use fixed “one-size-fits-all” treatment packages, limiting optimization and understanding of active components.
  • Innovations in Trial Design:
    • Sequential Multiple Assignment Randomized Trial (SMART): Optimizes adaptive interventions by tailoring treatments based on individual characteristics and context, using “decision rules” and periodic assessment.
    • Micro-Randomized Trial (MRT): Optimizes JITAIs by leveraging within-person heterogeneity over time, randomizing participants frequently to assess impact on momentary and distal outcomes.
  • Recommendations:
    • Conduct more large-scale, multicenter trials with high treatment fidelity and therapist competency.
    • Standardize methods of measuring clinical problems using psychometrically sound, freely available instruments to foster comparability and “learning systems”.
    • Investigate variables predicting inadequate treatment response and adverse side effects.
    • Publish treatment protocols and manuals (preferably before trial publication) and research methods for training and fidelity.
    • Endorse open science practices, including preregistration, publication of non-significant findings, and de-identified data sharing.
    • Provide more information on individual trajectories and the robustness of effects (e.g., using Bayesian methods).
    • Include societal relevance and cost-effectiveness information in research protocols and reports.
    • Utilize technical, methodological, and statistical innovations (e.g., EMAs, machine learning for outcome prediction, feedback systems) to improve research and link it to patients’ everyday lives.
• The Special Case of Meta-Analysis
  • Issues: Meta-analyses, which greatly influence treatment guidelines, can yield discrepant results due to heterogeneity of included samples, differing study quality, merging results from diverse measurement tools, and miscategorization of treatments. The “Dodo bird verdict” (lack of efficacy differences between treatments) can arise from a lack of precision in categorizing by mechanisms.
  • Recommendations:
    • Continuously improve guidelines for meta-analyses, focusing on clinical decisions and categorization by relevant disorder mechanisms and change processes, rather than solely statistical procedures or theoretical frameworks.
    • Achieve consensus on treatment arm categorization with developers and classify interventions by focused processes of change instead of historical roots.
    • Conduct more meta-analyses focusing on specific mechanisms maintaining symptoms and linking with new classification options (e.g., RDoC, HiTOP, network approaches).
    • Perform component meta-analyses to understand essential vs. non-essential treatment components.
    • Publish meta-analyses with a focus on critical evaluations and robustness analyses (e.g., jackknife procedure).
    • Preregister meta-analyses.
• Cultural Adaptations and Socioeconomic Variables
  • Challenge: The evidence base for psychotherapy is primarily from developed countries, leading to an underrepresentation of over 60% of the world’s population and a lack of generalization to diverse cultural and socioeconomic contexts. There is a significant gap in treatment coverage in low- and middle-income countries (LMICs).
  • Recommendations:
    • Integrate cultural aspects for understanding symptoms and adapting EBTs according to cultural values and principles, using tools like the DSM-5 Cultural Formulation Interview (CFI).
    • Consider dimensions like language, persons, metaphors, content, concepts, goals, methods, and context for cross-cultural treatment adaptations.
    • Develop international efforts to address variations in culturally adapted programs and training.
    • Balance fidelity to evidence-based principles with inclusivity for all people, considering cultural values, socioeconomic status, gender, ethnicity, language, and religious beliefs, to avoid disparities and social exclusion.
    • Adapt interventions for LMICs to their limited resources, including task-sharing approaches where lay providers deliver simple, transdiagnostic interventions.
• Developing a Common Framework and Language for Psychological Interventions
  • Challenge: The field is characterized by “therapy schools” with disjointed ideologies, theoretical frameworks, and languages, contrasting with a unified scientific discipline.
  • Recommendations:
    • Move towards “speaking the same language,” using concepts applicable across diverse clinical interventions, not tied to specific therapy models.
    • Develop meta-models (e.g., the extended evolutionary meta-model, which conceptualizes psychopathology as failed adaptation processes) to serve as umbrellas for different effective treatments, facilitating comparability and personalized decisions.
    • Establish alternatives to school-based thinking through transtheoretical advocacy in various media and groups.
    • Classify interventions based on their focused processes of change, rather than historical roots, to improve comparisons and stimulate improvements.
• Education and Training
  • Challenge: Training in many countries is often disconnected from scientific education systems, focuses on single theoretical orientations, and does not adequately consider healthcare needs or EBTs. There’s a lack of research on effective training methods and the cost-effectiveness of training. Suboptimal delivery in routine care leads to a drop in effectiveness.
  • Recommendations:
    • Establish dynamic learning systems and continuous quality improvement circles for training, focusing on acquired competencies for optimal patient outcomes.
    • Define national curricula to teach evidence-based, guideline-recommended interventions and define competence profiles for psychotherapists.
    • Establish continual feedback systems for individual therapists and encourage their use.
    • Make treatment protocols, manuals, and assessment materials freely available, complemented by teaching materials like video clips, and continuously update them.
    • Establish continuing education programs and networks to bridge the scientist-practitioner gap and promote EBTs.
    • Integrate culture-sensitive interventions into training programs.
    • Improve and evaluate supervisor training, encouraging the use of random session samples in supervision.
    • Offer a system of different training levels for providers, including shorter, less-expensive training for a wider range of people (e.g., trained para-professionals) to address cost and dissemination needs, especially in LMICs.
• Public Outreach
  • Challenge: The field needs to better communicate its findings and the value of EBTs to the public, government representatives, policymakers, and healthcare authorities.
  • Recommendations:
    • Improve access to EBTs and training opportunities for EBTs.
    • Improve treatment fidelity using monitoring systems.
    • Establish networks to disseminate information and raise professional and public awareness of EBTs.
    • Apply implementation science frameworks (e.g., RE-AIM, COM-B) to understand and facilitate the transition of evidence-based interventions into routine healthcare, focusing on factors that impede or facilitate uptake.
    • Engage in effective communication tactics such as podcasts, blogs, feature stories, targeted editorials, and strategic collaboration with science media to make scientific findings accessible and meaningful to the public.

In conclusion, the Marburg Declaration highlights the urgent need for innovation in clinical psychology across conceptualizations of mental disorders, research methodologies, training, and implementation, all while emphasizing the societal responsibility to make effective psychological treatments broadly accessible. It underscores that overcoming current barriers requires a shift towards improved translational science, new conceptual models (like network models), a common scientific language, and updated clinical trial designs, with successful initiatives like IAPT serving as inspiring examples for future developments.

Reference: Rief, W., Asmundson, G. J., Bryant, R. A., Clark, D. M., Ehlers, A., Holmes, E. A., … & Hofmann, S. G. (2024). The future of psychological treatments: The Marburg Declaration. Clinical Psychology Review, 110, 102417.