Abstract
Ethical approval is not required for this study since it will be based on anonymised data from RCTs that have already been completed. The findings of the present study will be disseminated through a peer-reviewed journal or conference presentation.
Introduction
Depression treatment currently faces a significant clinical challenge: the "average" deadlock. While large-scale meta-analyses consistently demonstrate that Cognitive Behavioural Therapy (CBT) and Interpersonal Psychotherapy (IPT) are equally effective on average, clinicians recognize that individual patient responses vary significantly. This protocol outlines a massive Individual Participant Data (IPD) meta-analysis designed to move beyond aggregate treatment effects and toward personalized medicine.
The core objective is to determine "what works for whom." Previous attempts to identify treatment moderators—factors such as personality traits or baseline symptom severity—have been limited by the small sample sizes of individual trials and the inherent limitations of aggregate meta-analyses, which rely on study-level summaries rather than individual patient trajectories. By pooling raw data, this study aims to achieve the statistical power necessary to detect subtle moderator effects and examine diverse outcomes, including anxiety, social functioning, and long-term well-being.
Methodology
This study utilizes a one-stage IPD meta-analysis, widely considered the gold standard for synthesizing clinical trial data. The research process involves several critical phases:
Data Acquisition and Standardization
The research team will conduct a systematic search of major repositories (PubMed, PsycINFO, EMBASE, Cochrane Library) and the Metapsy database to identify every Randomized Clinical Trial (RCT) comparing CBT and IPT for adult depression through January 2026. Rather than relying on published summaries, the authors are contacting original investigators to obtain raw, anonymized participant data. To ensure comparability across diverse studies, validated algorithms will be used to convert various depression scales into a single common metric: the Beck Depression Inventory (BDI-II).
Statistical Framework
The analysis will employ three-level mixed-effects models to account for the nested structure of the data: measurements within participants and participants within specific studies. This approach addresses the "ecological fallacy" common in standard meta-analyses, where group-level trends are incorrectly assumed to apply to individuals. The moderator analysis will specifically test three-way interactions (Time × Treatment × Moderator) to identify baseline characteristics—such as age, gender, or comorbidities—that predict superior outcomes for one therapy over the other.
Expected Outcomes
As a study protocol, this paper establishes the framework for a project scheduled to conclude in May 2026. The anticipated results include:
- A definitive, high-powered comparison of CBT and IPT efficacy across multiple psychological domains.
- A validated list of patient characteristics (moderators) to guide clinical decision-making.
- An assessment of long-term effectiveness and durability (6 months to 1.5 years post-treatment) to determine which therapy better prevents relapse.
Discussion and Limitations
While the IPD approach offers unprecedented statistical power, it is subject to certain limitations. The success of the project depends on data availability; a bias may emerge if authors of "negative" trials are less likely to share raw data. Additionally, the process of standardizing different scales and recoding categorical variables may result in the loss of some original nuance.
Finally, it is important to note that moderators themselves are observational rather than randomized. Consequently, any identified predictors of treatment success will require prospective validation in future clinical trials. The ultimate goal of this work is to provide the empirical foundation for "precision medicine" algorithms that match patients to the most effective treatment for their specific profile.
Summary updated