ORCID

Abstract

Abstract Background People in prison experience a range of physical and mental health problems. Evaluating the effectiveness and efficiency of prison-based interventions presents a number of methodological challenges. We present a case study of an economic evaluation of a prison-based intervention (“Engager”) to address common mental health problems. Methods Two hundred and eighty people were recruited from prisons in England and randomised to Engager plus usual care or usual care. Participants were followed up for 12 months following release from prison. The primary analysis is the cost per quality-adjusted life year (QALY) gained of Engager compared to usual care from a National Health Service (NHS) perspective with QALYs calculated using the CORE 6 Dimension. A cost-consequences analysis evaluated cross-sectoral costs and a range of outcomes. Results From an NHS perspective, Engager cost an additional £2737 per participant (95% of iterations between £1029 and £4718) with a mean QALY difference of − 0.014 (95% of iterations between − 0.045 and 0.017). For the cost-consequences, there was evidence of improved access to substance misuse services 12 months post-release (odds ratio 2.244, 95% confidence Interval 1.304–3.861). Conclusion Engager provides a rare example of a cost-utility analysis conducted in prisons and the community using patient-completed measures. Although the results from this trial show no evidence that Engager is cost-effective, the results of the cost-consequences analysis suggest that follow-up beyond 12 months post-release using routine data may provide additional insights into the effectiveness of the intervention and the importance of including a wide range of costs and outcomes in prison-based economic evaluations. Trial registration (ISRCTN11707331).

DOI

10.1007/s10198-021-01360-7

Publication Date

2021-08-05

Publication Title

European Journal of Health Economics

ISSN

1439-3972

Embargo Period

2021-09-21

Organisational Unit

Peninsula Medical School

Share

COinS