# Logic model development through a feasibility RCT for a group-based weight management programme

## TL;DR

This paper evaluates a new group-based weight management program for severe obesity using a feasibility study to test its logic model and effectiveness.

## Contribution

The study introduces a realist evaluation approach to refine the logic model of a group-based psychological weight management intervention.

## Key findings

- The core therapeutic mechanisms of PROGROUP were largely valid, but delivery components and contextual assumptions required revision.
- The referral process could be simplified, and the intervention could be delivered by different professions.
- Realist evaluation of a feasibility study can strengthen the evidence base for psychological weight management interventions.

## Abstract

Clinical psychology interventions for reducing obesity have developed alongside pharmacological and surgical treatments, but usually as interventions for individual patients. Any healthcare intervention rests on a logic model: assumptions that through specific physical and social mechanisms, it will produce certain intended outcomes, provided that conducive background conditions (‘contexts’) exist. Using evidence from the feasibility trial preceding a full randomised controlled trial (RCT), this paper assesses the empirical validity of the initial logic model of a new group-based weight management intervention: PROGROUP, designed for patients with body mass index (BMI) ≥40 kg/m2 or ≥35 kg/m2 with comorbidities. We aimed to test whether:

PROGROUP’s programme components produce the intended outcomes at all, whatever their size and how.

The intervention can practicably be implemented (‘delivered’) as designed.

How the programme and delivery components affected each other.

Multimethod proof-of-concept study by means of realist evaluation of the initial PROGROUP logic model. We:

Elicited the logic model underlying the intervention design.

Compared these assumptions with data from a pre-RCT feasibility study in two English and one Welsh National Health Service (NHS) sites during 2021–2023.

Revised the logic model in light of the data, noting how much variation in delivery the programme components (therapeutic mechanisms) could tolerate.

Specialised ambulatory mental health services in the English NHS.

Adults with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidities).

Group-building techniques to enhance group members’ adoption of evidence-based methods of behaviour change affecting their dietary behaviour and physical activity.

Qualitative outcomes. What kinds of:

Mechanisms were established and triggered by the attempt to implement PROGROUP (secondary outcome, see objective 1 above).

Patient behaviours resulted and whether patients sustained them after the intervention ended (primary outcome; see objective 1 above); and what kinds of context affected that (objectives 2 and 3).

Quantitative measures not used.

The initial logic model assumed that the following sequence of mechanisms would produce weight loss: referral from GP to specialist weight management services; further referral to PROGROUP; preparatory individual consultation; facilitated group sessions produce a group identity; group identity reinforces weight management capability and motivation; further individual consultations adjust for individual circumstances; behaviour change outside the treatment setting, producing weight loss. Contexts necessary for these mechanisms to work included: sufficient catchment population; group size, continuity and membership retention; suitable location; facilitator training; and practical support outside the treatment setting.

The findings suggested revisions to the logic model, but more in the delivery components and contextual assumptions than the core therapeutic mechanisms. There was scope to simplify the referral mechanisms. Different professions could implement the model. A realist evaluation of a pre-RCT feasibility study can be used to make the intervention’s logic model more securely evidence-based, serving as a proof-of-concept test for the intervention. It indicated the conditions under which such group psychological interventions might be more widely used.

ISRCTN22088800.