Table 4.
Results from base-case and probabilistic cost-utility analysis for usual care, EULT, and robot-assisted training
| Unadjusted mean cost (98·3% CI) | Unadjusted mean QALYs (98·3% CI) | Adjusted*incremental QALY (98·3% CI) | Adjusted*incremental costs (98·3% CI) | Adjusted incremental cost-effectiveness ratio |
Probability that each therapy is cost effective at different willingness to pay thresholds† |
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| £0 | £10 000 | £20 000 | £30 000 | £50 000 | ||||||
| Robot-assisted training | £5387 (4777 to 5996) | 0·212 (0·195 to 0·229) | .. | .. | More expensive and less effective than EULT in both adjusted and unadjusted analyses | 0% | 0% | 0% | 0% | 0% |
| Enhanced upper limb therapy | £4451 (3548 to 5354) | 0·229 (0·213 to 0·244) | 0·010 (−0·005 to 0·025) | 741 (−461 to 1943) | £74 100 | 10% | 15% | 19% | 26% | 38% |
| Usual care | £3785 (2801 to 4770) | 0·212 (0·194 to 0·230) | .. | .. | .. | 90% | 85% | 81% | 74% | 62% |
Numbers of patients included in analyses were 178 in the usual care group, 259 in the EULT group, and 257 in the robot-assisted training group for the unadjusted cost calculation; 254 in the usual care group, 259 in the EULT group, and 257 in the robot-assisted training group for the unadjusted QALY calculation; and 171 in the usual care group, 254 in the EULT group, and 247 in the robot-assisted training group for the adjusted analyses. EULT=enhanced upper limb therapy. QALY=quality-adjusted life-year.
Adjusted analysis done using the seemingly unrelated regression (sureg) function on STATA, version 15; adjusted for centre, baseline ARAT score, time since stroke, baseline costs, and baseline utility score; performed for the comparison between usual care and EULT as the next best alternative.
The probabilistic sensitivity analysis includes all three therapies and was done for different threshold values of society's willingness to pay per QALY (£0, £10 000, £20 000, £30 000, and £50 000).