Target trials for talking and digital enabled therapies to estimate their real-world effectiveness, cost-effectiveness, and inequalities in outcomes using causal inference techniques applied to NHS Talking Therapies real-world data: The Target Therapies quasi-experimental study

Research question.

Among referrals to NHS Talking Therapies for anxiety and depression (TTad) services, how effective and cost-effective are different talking therapies allocated for the same presenting mental disorder for achieving patient- and service-focussed outcomes, and are there outcome inequalities? Background.

Randomised controlled trials (RCTs) provide gold-standard causal treatment effect evidence to inform decision-making.

However, RCT issues include limited external validity, insufficient sample sizes to explore outcome inequalities, time and cost implications.

Emulating RCTs using real-world data (RWD) are a viable and cost-efficient option to inform evidence-based decision-making if conducted appropriately.

Common mental disorders (CMDs), like anxiety and depression, affect ≈15%-population, with higher mortality levels and years lived with disability then the general population. NHS TTad-services offer evidence-based psychological therapies with 1.8-million referrals during 2022/23. However, TTad doesn t work for everyone.

Cost-efficient identification of what works for whom is needed to support TTad-service reform to tackle unmet service-user needs. Aims and Objectives. To address RCT evidence-gaps and real-world uncertainties by answering our research question using RWD.

Our head-to-head comparisons include: Group versus individual-based low-intensity cognitive behavioural therapy (CBT) for depression and anxiety disorders Low-intensity digitally-enabled interventions (with support) versus therapist-guided self-help for depression and anxiety disorders CBT versus counselling for depression Eye Movement Desensitisation and Reprocessing (EMDR) versus trauma-focussed CBT for post-traumatic stress disorder (PTSD) We will adhere to robust quality assurance and reporting processes for conducting real-world comparative effectiveness studies, alongside exploring the strengths and limitations of linked healthcare data for informing evidence-based decision-making.

Methods.

We will use de-identified data routinely recorded by TTad-services, secondary (hospital) care providers, and community pharmacies covering April-2021 to March-2024.

Routinely collected outcomes include patient-reported mental health (e.g., Patient Health Questionnaire-9 depression score), TTad-service metrics (e.g., recovery and reliable improvement), and service outcomes (e.g., referrals receiving 2+ treatment sessions).

We will use structural and analytical methods to reduce bias to produce causal effectiveness and cost-effectiveness evidence.

Structural methods include mimicking best RCT practice by pre-specifying our target trial design, using directed acyclic graphs to inform analytical confounder conditioning, and pre-specified analyses plans.

Analytical methods include g-methods such as inverse probability weights within marginal structural models, and quantitative bias analyses including E-values (evidence-for-causality values). Timelines for delivery. Six work-packages (WPs) over 36-months. WP1 (Project management), WP2 (PPI), and WP6 (Dissemination and capacity building) run throughout the 36-months.

WP3 (Secondary data use: ethics and information governance) and WP4 (Target Trial specification and analysis preparation) dominates the first 12-months to 18-months, respectively.

WP5 (Causal analyses of effectiveness and cost-effectiveness) includes data cleaning and structuring dominating Year-2, then analyses, assessing result robustness, and reporting and quality assurance processes dominating Year-3. Anticipated Impact and Dissemination. Initial impact will predominantly be for those with CMDs, TTad-services, broader NHS, and commissioners.

The findings will influence NICE clinical guidance, TTad-service delivery, methodological development and trust in RWE, leading to patient and service benefits.

Our plans and results will be shared widely including with patient groups, via a study website and newsletter, presented