Background: Care farms (CFs), in which all or part of the farm is used for therapeutic purposes, show potential for improving well-being for disadvantaged groups. We assessed the feasibility of determining the cost-effectiveness of CFs in improving quality of life compared with comparator sites among probationers undertaking community orders (COs).
Objectives: (1) To conduct a systematic review of CF impacts and mechanisms in improving health and logic model development; (2) to inform future studies by estimating differences in quality of life and other outcomes, identifying factors driving CO allocation and ways to maximise recruitment and follow-up; and (3) to assess feasibility of cost-effectiveness analysis.
Review methods: A mixed-methods synthesis following Campbell Collaboration guidelines. Pilot study: three probation service regions in England, each with a CF and a comparator CO site. Participants were adult offenders (aged ≥ 18 years) serving COs. The primary outcome was quality of life [as measured via the Clinical Outcome in Routine Evaluation–Outcome Measure (CORE-OM)]. Other outcomes were health behaviours, mental well-being, connectedness to nature and reconvictions.
Data sources: In November 2014, we searched 22 health, education, environmental, criminal justice and social science electronic databases, databases of grey literature and care farming websites across Europe. There were no language restrictions. A full list of databases searched is given in Appendix 1; some examples include Web of Science, Cumulative Index to Nursing and Allied Health Literature (via EBSCOhost), The Campbell Library, Criminal Justice Abstracts (via EBSCOhost), MEDLINE (via Ovid) and Scopus (Elsevier B.V., Amsterdam, the Netherlands).
Results: Our systematic review identified 1659 articles: 14 qualitative, 12 quantitative and one mixed-methods study. Small sample sizes and poor design meant that all were rated as being at a high risk of bias. Components of CFs that potentially improve health are being in a group, the role of the farmer and meaningful work, and interaction with animals. There was a lack of quantitative evidence indicating that CFs improve quality of life and there was weak evidence of improved mental health, self-efficacy, self-esteem, affect and mood. In the pilot study we recruited 134 respondents, and only 21 declined; 37% were allocated to three CFs and the remainder to comparators. This was below our recruitment target of 300. Recruitment proved challenging as a result of the changes in probation (probation trusts were disbanded in 2014) and closure of one CF. We found significant differences between CFs and comparator users: those at CFs were more likely to be male, smokers, substance users, at higher risk of reoffending (a confounder) and have more missing CORE-OM questions. Despite these differences, the use of propensity analysis facilitated comparison. Participants consented to our team accessing, and we were able to link, probation and police reconviction data for 90% of respondents. We gained follow-up questionnaire data from 52% of respondents, including health and social care use cost data. We transformed CORE-OM into CORE-6D, allowing derivation of quality-adjusted life-years. As a pilot, our study was not powered to identify significant differences in outcomes. Qualitatively, we observed that within COs, CFs can be formally recognised as rehabilitative but in practice can be seen as punitive.
Limitations: Changes in probation presented many challenges that limited recruitment and collection of cost data.
Conclusions: Recruitment is likely to be feasible in a more stable probation environment. Retention among probationers is challenging but assessing reconvictions from existing data is feasible. We found worse health and risk of reoffending among offenders at CFs, reflecting the use of CFs by probation to manage challenging offenders.
Future work: A sufficiently powered natural experiment is feasible and of value. Using reconvictions (from police data) as a primary outcome is one solution to challenges with retention. Propensity analysis provides a viable method for comparison despite differences in participants at CFs and comparator sites. However, future work is dependent on stability and support for CFs within probation services.
Study registration: This study is registered as PROSPERO CRD42014013892 and SW2013–04 (the Campbell Collaboration).
Funding details: The National Institute for Health Research Public Health Research programme.
Copyright © Queen’s Printer and Controller of HMSO 2018. This work was produced by Elsey et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.