Delivery preferences for psychological intervention in cardiac rehabilitation: a pilot discrete choice experiment

BACKGROUND: Cardiac rehabilitation (CR) is a programme of care offered to people who recently experienced a cardiac event. There is a growing focus on home-based formats of CR and a lack of evidence on preferences for psychological care in CR. This pilot study aimed to investigate preferences for delivery attributes of a psychological therapy intervention in CR patients with symptoms of anxiety and/or depression. METHODS: A discrete choice experiment (DCE) was conducted and recruited participants from a feasibility trial. Participants were asked to choose between two hypothetical interventions, described using five attributes; intervention type (home or centre-based), information provided, therapy manual format, cost to the National Health Service (NHS) and waiting time. A separate opt-out was included. A conditional logit using maximum likelihood estimation was used to analyse preferences. The NHS cost was used to estimate willingness to pay for aspects of the intervention delivery. RESULTS: 35 responses were received (39% response rate). Results indicated that participants would prefer to receive any form of therapy compared with no therapy. Statistically significant results were limited, but included participants being keen to avoid not receiving information prior to therapy (β=−0.270; p=0.03) and preferring a lower cost to the NHS (β=−0.001; p=0.00). No significant results were identified for the type of psychological intervention, format of therapy/exercises and programme start time. Coefficients indicated preferences were stronger for home-based therapy compared with centre-based, but this was not significant. CONCLUSIONS: The pilot study demonstrates the feasibility of a DCE in this group, it identifies potential attributes and levels, and estimates the sample sizes needed for a full study. Preliminary evidence indicated that sampled participants tended to prefer home-based psychological therapy in CR and wanted to receive information before initiating therapy. Results are limited due to the pilot design and further research is needed.