# Statistical and health economic analysis plan for a secure care hospital evaluation of manualised (interpersonal) art-psychotherapy: the SCHEMA randomized controlled trial

## TL;DR

This study plans to evaluate if art psychotherapy reduces aggression and distress in psychiatric patients with intellectual disabilities compared to usual care.

## Contribution

The paper presents a detailed statistical and health economic analysis plan for a randomized controlled trial of art psychotherapy in secure care settings.

## Key findings

- Aggressive behavior will be analyzed using a linear mixed-effect model adjusted for baseline and stratification factors.
- Health economic data will be used to estimate cost-effectiveness based on quality-adjusted life years over 38 weeks.
- Weekly MOAS scores will track changes in aggression from 19 to 38 weeks post-randomization.

## Abstract

The SCHEMA trial evaluates whether interpersonal art psychotherapy reduces the frequency/severity of aggressive incidents or patient distress associated with psychiatric symptoms, compared to usual care.

To describe the statistical and health economic analysis plan.

A multicentre, two-arm, parallel-group, single blind individually randomised controlled trial with 150 adults within NHS secure care who have borderline to mild/moderate intellectual disability. The primary outcome is the frequency/severity of aggressive behaviour, measured on the Modified Overt Aggression Scale (MOAS) 19 weeks post-randomisation, analysed using a linear mixed-effect model, adjusted for baseline MOAS and stratification by gender and psychosis diagnosis. Changes in aggressive behaviour will be evaluated using weekly MOAS scores between 19 and 38 weeks. Patient distress relating to psychiatric symptoms will be assessed using the Brief Symptom Inventory Positive Symptom Distress Index across baseline, 19, and 38 weeks. Health-related quality-of-life will be assessed using self- and proxy-reported EQ-5D three-level (EQ-5D-3L) and Recovering Quality of Life 10-item measures, the latter to estimate the ReQoL Utility Index, across baseline, 19, and 38 weeks. The self-reported EQ-5D-3L is collected using an adapted version for people with intellectual disabilities. Resource-use is collected based on secure care records, to estimate intervention and healthcare costs over 19 and 38 weeks. HRQoL and cost data will inform cost-effectiveness based on the incremental cost per quality-adjusted life year over 38 weeks.

This paper details the planned analyses and discusses recruitment challenges, sample size implications, and effect size assumptions. The plan was developed prior to database lock and unblinding to minimise analytical bias.

ISRCTN, ISRCTN57406593. Registered on 18/01/2023.

The online version contains supplementary material available at 10.1186/s13063-025-08934-3.