Measuring fidelity in Cognitive Stimulation Therapy (CST) for dementia: individual and group evaluation

Cognitive Stimulation Therapy (CST) is an evidence-based and widely implemented cognitive therapy for dementia. The current study aimed to assess a novel measure of fidelity to CST through testing it in two contexts: virtual group CST (vCST) and individual CST (V-iCST).

A subsample of participants from prior feasibility randomised controlled trials of vCST and V-iCST were included in the current fidelity analysis (12 from vCST and 9 from V-iCST trials). Facilitators completed a self-report version of the measure after each CST session. Trained independent observers rated a purposive sample of video-recorded V-iCST and vCST sessions. Fidelity scores were compared across raters, and inter-rater agreement was assessed using weighted kappa and percentage agreement to evaluate the fidelity measure.

vCST and V-iCST interventions were delivered with moderate and high fidelity, respectively. The inter-rater agreement between facilitators and observers was moderate for V-iCST (weighted κ = 0.57, 65.5%) and vCST (weighted κ = 0.511, 60.7%).

The study found that the novel fidelity measure can be applied successfully to measure fidelity in a virtual CST setting. However, given the moderate inter-rater agreement, explicit conclusions regarding the reliability of self-report as a fidelity measure should be further investigated in larger trials. The findings highlight the importance of integrating a fidelity measurement to evaluate future implementations of CST. The self-report measure can be particularly useful in clinical practice, allowing for the monitoring of facilitation quality and as a training tool for clinicians delivering CST.

The study evaluated a novel fidelity measure for CST and supported its applicability to group and individual therapy settings. For clinical settings, our findings suggest the fidelity checklist and associated guidelines can be implemented as a training and self-evaluation tool for facilitators to assess service provision, and the self-reported version may be particularly advantageous as it reduces the need for observer assessment. In research settings, we have highlighted applicability of the measure to monitor the link between fidelity of CST delivery and treatment outcomes. We include the fidelity checklist and guidelines in the publication to share with researchers and clinicians.