Analytical approach:
This analysis was based on a single study with a one-year time horizon. The authors stated that the perspective of the service provider (including the health, social, voluntary, and criminal justice services) was adopted.
Effectiveness data:
The clinical data came from a prospective, parallel-group, randomised controlled trial (RCT) that enrolled 52 patients, with 28 (17 male) in the nidotherapy group and 24 (18 male) in the control group. Patients were assessed at baseline and then at six and 12 months by two researchers, who were blind to treatment allocation. The key endpoints were the length of hospital stay over one year and the change in clinical psychopathology, which was assessed using the Brief Psychiatric Rating Scale (BPRS).
Monetary benefit and utility valuations:
Not assessed.
Measure of benefit:
The change in the BPRS was the summary benefit measure.
Cost data:
The economic analysis included the costs of nidotherapy (time spent by the therapists plus overheads), hospital services, community health and social services, medications, and criminal justice services. The resource consumption was based on the Secure Facilities Service Use Schedule, which covers every possible service contact for this patient population (both in the community and in secure facilities). Resource use was assessed six months before randomisation and one year after randomisation. Costs were from the National Health Service (NHS) Reference Costs, the British Medical Association, Social Services Research Units, and the Legal Services Commission. They were in UK pounds sterling (£) and referred to the financial year 2004 to 2005. Tests were used to assess the statistical significance of cost differences, given the skewed distribution of the cost data.
Analysis of uncertainty:
Several one-way sensitivity analyses were carried out by varying the assumptions for costs or clinical data. A nonparametric bootstrapping approach was used to generate average estimates and confidence intervals for the expected costs and benefits. Cost-effectiveness acceptability curves were generated.