For the low-risk group, the 12-month EQ-5D score was 0.787 with the intervention and 0.773 with control; the total health care cost was £96.15 with the intervention, and £160.44 with control.
For the medium-risk group, the EQ-5D score was 0.687 with the intervention, and 0.635 with control; the total cost was £235.08 with the intervention, and £287.80 with control.
For the high-risk group, the EQ-5D score was 0.541 with the intervention, and 0.458 with control; the total cost was £383.88 with the intervention, and £357.47 with control.
After multiple regression to adjust for age, gender, duration of pain at baseline, and baseline scores on the RMDQ and EQ-5D, the cost per QALY of the intervention compared with control was £48,707 per QALY for the low-risk group. In this situation, where the intervention costs less, but is also less effective, the ratio refers to the cost savings per unit lost. So the intervention saved £48,707 per QALY lost, which was higher than the threshold acceptability level and, unlike the normal situation, a higher cost per unit lost is preferable.
The other two comparisons were more common, with both generating increased benefits at either an increased or a reduced cost. The cost per QALY gained for the high-risk group was £463; and for the medium-risk group the intervention dominated best practice, as it was more effective and less costly.
When the uncertainty was characterised within the model, the likelihood of the intervention being cost-effective at a willingness-to-pay threshold of £20,000 per QALY gained was 50% in the low-risk group, 99% in the medium-risk group, and 95% in the high-risk group.