Green space and health benefits: a QALY and CEA of a mental health programme

Abstract

Green spaces are promoted as a means of improving the health-related quality of life (HRQoL) for people with health problems. This paper evaluates the Branching Out programme in Scotland, which provides patients with mental health problems, social, physical and recreational activities in an outdoor ‘community’ setting. Patients carry out group-based woodland activities over a 12-week period. Health agencies in the Strathclyde region of Scotland offer patients the opportunity to participate if they consider that the patient will benefit from Branching Out activities. The evaluation uses the SF-12 questionnaire to assess HRQoL of participants in the programme. From the SF-12 scores, quality adjusted life year (QALY) change is calculated. Cost per QALY is estimated in relation to economic costs of the program (staffing cost, facilities and other costs e.g. travel). Cost effectiveness analysis reveals Branching Out is comparable to other programmes oriented to social recovery. Its cost-effectiveness in terms of National Institute for Health & Care Excellence guidelines for medical treatments depends on the duration of the health improvement.

Keywords:

• QALYs
• mental health
• forests
• cost-effectiveness analysis

JEL classifications:

• I12
• I18
• H75

Acknowledgements

This study was funded by Forestry Commission Scotland (FCS). We wish to thank Hugh McNish, Jennifer Waterton and John Wildman for comments on the research. Views expressed in this paper are those of the authors and do not necessarily reflect those of FCS.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Extensive reviews of the medical evidence are given in Department of Health (2004b) and Pretty et al. (2005). English Nature (2003) has reviewed the positive impacts of nature on psychological well-being.

2. The analysis needs to assess the extent to which the green space HRQoL benefits are created rather than being diverted or autonomous.

3. Greater Glasgow and Clyde, Lothian, Forth Valley, Lanarkshire, Ayrshire and Arran, Borders and Tayside.

4. Traditionally, the economic value of outdoor recreation-type activities has been measured using a revealed preference (RP) e.g. a travel cost method (TCM); or through a stated preference (SP) technique. There are a number of issues in any application of a TCM: accurately estimating the cost incurred of travel to the site, and the value of travel time to reach the site. Stated preference (SP) has been used to elicit how individuals value health improvements, by directly asking them. SP techniques comprise contingent valuation (CV), and choice experiments (CEs). Both have been used to value health improvements (see Alberini 2012; Chanel and Luchini 2014) and outdoor recreation. Although CV and CEs could be used to value Branching Out, by asking participants their WTP this could be a challenging task where mental health patients are concerned. Moreover, asking participants would not reveal the benefit that society places on improving the health of mentally ill people.

5. Prior to 1 April 2013, NICE was called the National Institute for Health and Clinical Excellence. In Scotland, it is the Scottish Medicines Consortium which advises on medicines. Their advice on cost-effectiveness of treatments is analogous to NICE advice in England.

6. In Scotland, the Scottish Medicines Consortium advises on medicines. Their advice on cost-effectiveness is not categorical although the same range of £20,000– £30,000 per QALY is indicated as a guideline. http://www.scottishmedicines.org.uk/About_SMC/Policy_Statements/A_Guide_to_Quality_Adjusted_Life_Years

7. The Student's ‘t’-test is the usual parametric test employed which is based on the normal distribution. A paired sample ‘t’-test was used by Wilson (2009) to test the statistical significance of the pre- and post-HRQoL scores for 2007/2008 Branching Out participant sample, but no significant statistical difference was found in the health dimension scores. The Student's ‘t’-test may produce misleading results if the distribution is not normal; or if the data are not sampled independently from the two populations being compared, i.e. if the data are known to be dependently sampled. In such cases, the Student's ‘t- test may suggest there is a statistically significant difference between the two samples when in fact there is none.

8. Delivery costs for two of the 33 groups were not known at the time this report was written. Costs per DA were thus based on the groups for which there were complete data.

9. A ‘public good’ is a good that is non-rival and non-excludable in consumption. For example, recreation in forest X can be enjoyed by individual A without lessening the enjoyment of the same forest by individual B at a different time or in another area of the forest (i.e. it is non-rival: A's consumption of the good does not lessen the quantity of the good for individual B unless there is congestion), and because there is open public access, it is non-excludable and hence a price cannot be charged.

10. Curtis (2011). Unit costs of health and social care. Personal Social Services Research Unit, University of Kent. http://www.pssru.ac.uk/pdf/uc/uc2011/uc2011.pdf

11. Discounting costs was not relevant since they all took place within one year.

12. £429/0.0495. Using the pooled results which include those of 2007/2008 evaluation, the QALY change is 0.0195 and the cost per QALY is £21,840.

13. Based on the NICE maximum intervention cost of £30,000 per QALY.