Analytical approach:
This economic evaluation was carried out alongside a single study with a two-year time horizon. The authors did not state explicitly which perspective was adopted in the study.
Effectiveness data:
Clinical inputs were derived from the Supported Work and Needs (SWAN) randomised clinical trial (RCT) that enrolled a final sample of 219 participants (109 in the intervention group and 110 in the treatment as usual group) from community mental health teams in two boroughs in south London. Researchers that assessed the clinical outcomes were blinded to allocation status. Follow-up was for two years. Thirty participants were lost to follow-up at two years (15 per group). Key endpoints were measures of psychosocial functioning such as Manchester Short Assessment (MANSA) version 2, brief modified versions of Lancashire Quality of Life Profile and the proportion of participants who found employment. Logistic and regression models were fitted to assess the potential impact of confounding factors.
Monetary benefit and utility valuations:
Not considered.
Measure of benefit:
The main endpoint used as the summary benefit measure was the proportion of participants in employment during the two-year follow-up period.
Cost data:
The economic analysis included costs of in-patient stay, visits to health care professionals and medications. Resource quantities were collected alongside the clinical trial. In particular, in-patient data were collected for the entire 24-month period. Other service data were collected for the three-month period prior to 12- and 24-month follow-up interviews. Costs of medications were derived from the British National Formulary. Other costs were based on official prices from the Personal Social Services Research Unit. Costs were in UK pounds sterling (£) and referred to 2006/2007 prices.
Analysis of uncertainty:
A cost-effectiveness acceptability curve was produced to show the probability that IPS was more cost-effective than treatment as usual for a range of values placed on a unit improvement in outcome (from £0 to £1,000). A sensitivity analysis considered the assumption that all patients who were not followed up had not worked.