Table 3. Summary of results from scenario analyses.
| Scenario | Deterministic Results | Probabilistic Results | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Difference, PAZ vs. SUN | ICER (£) | NMB, by WTP for QALY (£) | Probability that Therapy is Dominant | Probability that Pazopanib is Cost-Effective, by WTP for QALY | ||||||||
| No. | Description | Costs (£) | QALYs | 20,000 | 30,000 | 50,000 | PAZ | SUN | £20,000 | £30,000 | £50,000 | |
| 1 | Base Case | −912 | 0.06 | Dominant | 2,101 | 2,696 | 3,886 | 51% | 1% | 96% | 95% | 90% |
| 2 | Time horizon set to 10 years. PFS and OS modelled using Kaplan–Meier to 5 years with Weibull extrapolation thereafter | −1,244 | 0.01 | Dominant | 1,366 | 1,428 | 1,550 | 28% | 4% | 76% | 67% | 59% |
| 3 | Time horizon set to 10 years. PFS and OS modelled using Weibull distribution for entire model time horizon | −1,164 | 0.07 | Dominant | 2,535 | 3,220 | 4,591 | 97% | 0% | 100% | 100% | 99% |
| 4 | Assuming PFS for sunitinib is equal to that for pazopanib | −747 | 0.06 | Dominant | 1,972 | 2,585 | 3,809 | 49% | 2% | 89% | 90% | 88% |
| 5 | Assuming OS for sunitinib is equal to that for pazopanib | −1,067 | 0.03 | Dominant | 1,624 | 1,903 | 2,460 | 46% | 2% | 87% | 84% | 78% |
| 6 | Assuming PFS and OS for sunitinib are equal to that for pazopanib | −902 | 0.03 | Dominant | 1,495 | 1,791 | 2,384 | 35% | 4% | 83% | 81% | 76% |
| 7 | Using IRC-assessed rather than investigator-assessed PFS | −875 | 0.06 | Dominant | 2,165 | 2,810 | 4,100 | 46% | 1% | 92% | 92% | 91% |
| 8 | Pazopanib RDIs set equal to those of sunitinib | −964 | 0.06 | Dominant | 2,154 | 2,748 | 3,938 | 54% | 1% | 95% | 95% | 91% |
| 9 | Sunitinib RDIs set equal to those of pazopanib | −976 | 0.06 | Dominant | 2,165 | 2,760 | 3,949 | 54% | 0% | 96% | 97% | 94% |
| 10 | Administration/ dispensing costs set to 50% of base-case value | −938 | 0.06 | Dominant | 2,128 | 2,722 | 3,912 | 55% | 1% | 94% | 95% | 91% |
| 11 | Administration/ dispensing costs set to 150% of base-case value | −886 | 0.06 | Dominant | 2,075 | 2,670 | 3,859 | 53% | 0% | 97% | 96% | 92% |
| 12 | Monthly costs associated with pazopanib and sunitinib set to 50% of base-case value | −580 | 0.06 | Dominant | 1,770 | 2,364 | 3,554 | 51% | 1% | 95% | 94% | 90% |
| 13 | Monthly costs associated with pazopanib and sunitinib set to 150% of base-case value | −1,244 | 0.06 | Dominant | 2,433 | 3,028 | 4,217 | 51% | 1% | 95% | 94% | 92% |
| 14 | Monthly costs associated with sunitinib set to those for pazopanib | -273 | 0.06 | Dominant | 1,463 | 2,057 | 3,247 | 54% | 1% | 96% | 95% | 91% |
| 15 | Monthly pre-progression routine cost set to 50% of base-case value | −900 | 0.06 | Dominant | 2,089 | 2,684 | 3,873 | 52% | 0% | 95% | 95% | 91% |
| 16 | Monthly pre-progression routine cost set to 150% of base-case value | −925 | 0.06 | Dominant | 2,114 | 2,709 | 3,898 | 52% | 0% | 96% | 95% | 91% |
| 17 | Monthly post-progression routine cost set to 50% of base-case value | −1,008 | 0.06 | Dominant | 2,197 | 2,792 | 3,981 | 53% | 1% | 96% | 95% | 91% |
| 18 | Monthly post-progression routine cost set to 150% of base-case value | −816 | 0.06 | Dominant | 2,006 | 2,601 | 3,790 | 52% | 1% | 95% | 95% | 91% |
| 19 | Decrement in utility for PFS vs. perfect health set to 50% of base-case value | −912 | 0.04 | Dominant | 1,621 | 1,975 | 2,683 | 51% | 1% | 95% | 94% | 91% |
| 20 | Decrement in utility for PFS vs. perfect health set to 150% of base-case value | −912 | 0.08 | Dominant | 2,582 | 3,417 | 5,088 | 42% | 1% | 94% | 90% | 83% |
| 21 | Decrement in utility for post- vs. pre-progression set to 50% of base-case value | −912 | 0.08 | Dominant | 2,561 | 3,386 | 5,035 | 61% | 0% | 97% | 97% | 96% |
| 22 | Decrement in utility for post- vs. pre-progression set to 150% of base-case value | −912 | 0.04 | Dominant | 1,642 | 2,006 | 2,736 | 54% | 1% | 96% | 95% | 91% |
| 23 | Assuming utility during PFS for sunitinib is equal to that for pazopanib | −912 | −0.01 | 171,379 | 806 | 752 | 646 | 47% | 1% | 96% | 96% | 91% |
| 24 | Assuming PFS and OS and utility during PFS for sunitinib are equal to that for pazopanib | −902 | 0.00 | Dominant | 902 | 902 | 902 | 38% | 2% | 89% | 86% | 78% |
| 25 | Time horizon set to 10 years. PFS and OS modelled using Weibull distribution for entire model time horizon, utility during PFS for sunitinib are equal to that for pazopanib | −1,164 | −0.01 | 170,156 | 1,027 | 959 | 822 | 20% | 6% | 69% | 64% | 58% |
| 26 | Time horizon set to 10 years. PFS and OS modelled using Weibull distribution for entire model time horizon, PFS and OS and utility during PFS for sunitinib are equal to that for pazopanib | −708 | −0.05 | 14,664 | −258 | −740 | −1,706 | 88% | 0% | 100% | 98% | 96% |
| 27 | Discount rate set to 0% | −1,020 | 0.06 | Dominant | 2,239 | 2,849 | 4,069 | 52% | 0% | 85% | 77% | 68% |
| 28 | Discount rate set to 6% | −841 | 0.06 | Dominant | 2,010 | 2,595 | 3,764 | 53% | 1% | 96% | 95% | 91% |
ICER, incremental cost-effectiveness ratio; IRC, independent review committee; NMB, net monetary benefit; OS, overall survival; PAZ, pazopanib; PFS, progression-free survival; QALY, quality-adjusted life-year; RDI, relative dose intensity; SUN, sunitinib; WTP, willingness to pay.