Table 3.
Scenario analysis results
| Scenario | Utility generation mechanism (UGM) | True QALYs | Utility analysis method (UAM) | |||||
|---|---|---|---|---|---|---|---|---|
| % Mean error (ME) | % Mean absolute error (MAE) | |||||||
| Prog | TTD | Combo | Prog | TTD | Combo | |||
| Base case | Prog-derived | 3.4 | 0 | 0.2 | − 0.7 | 0.4 | 0.5 | 0.7 |
| TTD-derived | 3.3 | − 3.5 | − 0.4 | 5.8 | 3.5 | 0.6 | 5.9 | |
| Combo-derived | 3.1 | − 3.4 | − 0.2 | 5.5 | 3.4 | 0.6 | 5.5 | |
| Scenario 1 | Prog-derived | 3.4 | 0 | 0.2 | − 0.7 | 0.4 | 0.5 | 0.8 |
| N = 150 | TTD-derived | 3.3 | − 3.4 | − 0.4 | 6 | 3.5 | 0.7 | 6 |
| Combo-derived | 3.1 | − 3.4 | − 0.3 | 5.6 | 3.5 | 0.7 | 5.6 | |
| Scenario 2 | Prog-derived | 3.4 | 0 | 0.2 | − 0.7 | 0.2 | 0.3 | 0.7 |
| N = 300 | TTD-derived | 3.3 | − 3.4 | − 0.4 | 5.8 | 3.4 | 0.6 | 5.9 |
| Combo-derived | 3.1 | − 3.4 | − 0.3 | 5.4 | 3.5 | 0.5 | 5.5 | |
| Scenario 3 | Prog-derived | 4.5 | 0 | 0.3 | − 0.5 | 0.4 | 0.5 | 0.6 |
| Age = 55 years | TTD-derived | 4.5 | − 3.1 | − 0.4 | 4.3 | 3.1 | 0.7 | 4.3 |
| Combo-derived | 4.2 | − 3.1 | − 0.3 | 4 | 3.1 | 0.7 | 4 | |
| Scenario 4 | Prog-derived | 2.4 | 0 | 0 | − 1 | 0.4 | 0.4 | 1 |
| Age = 75 years | TTD-derived | 2.3 | − 4.6 | − 0.4 | 8.6 | 4.7 | 0.6 | 8.6 |
| Combo-derived | 2.2 | − 4.6 | − 0.3 | 8 | 4.7 | 0.6 | 8.1 | |
| Scenario 5 | Prog-derived | 3.4 | 0 | 0.4 | − 0.7 | 0.3 | 0.5 | 0.7 |
| Utility interval = 90 | TTD-derived | 3.4 | − 5.6 | − 0.4 | 5.3 | 5.6 | 1.1 | 5.5 |
| Combo-derived | 3.2 | − 5.6 | − 0.6 | 4.9 | 5.6 | 0.8 | 5.1 | |
| Scenario 6 | Prog-derived | 3.4 | 0 | 0 | − 0.7 | 0.5 | 0.5 | 0.8 |
| Utility interval = 180 | TTD-derived | 3.3 | − 1.2 | − 0.4 | 6.1 | 1.4 | 0.5 | 6.1 |
| Combo-derived | 3.1 | − 1.2 | − 0.1 | 5.7 | 1.4 | 0.6 | 5.7 | |
| Scenario 7 | Prog-derived | 3.4 | 0 | 0.7 | − 0.7 | 0.5 | 0.8 | 0.8 |
| Length = 18 months | TTD-derived | 3.3 | − 7.4 | − 0.4 | 5 | 7.4 | 1.8 | 5.4 |
| Combo-derived | 3.1 | − 7.7 | − 1 | 4.6 | 7.7 | 1.5 | 5 | |
| Scenario 8 | Prog-derived | 3.4 | 0 | 0.1 | − 0.7 | 0.4 | 0.4 | 0.8 |
| Length = 60 months | TTD-derived | 3.3 | − 2.5 | − 0.4 | 5.9 | 2.5 | 0.5 | 5.9 |
| Combo-derived | 3.2 | − 2.4 | − 0.2 | 5.5 | 2.5 | 0.6 | 5.5 | |
| Scenario 9 | Prog-derived | 3.3 | 0 | 0.1 | − 0.7 | 0.4 | 0.4 | 0.8 |
| Length = 60 months or progression | TTD-derived | 3.3 | − 2.5 | − 0.4 | 5.9 | 2.6 | 0.5 | 5.9 |
| Combo-derived | 3.1 | − 2.5 | − 0.2 | 5.5 | 2.6 | 0.6 | 5.6 | |
| Scenario 10 | Prog-derived | 3.4 | 0 | 0.1 | − 0.7 | 0.4 | 0.4 | 0.8 |
| Length = 60 months or progression + 30 days | TTD-derived | 3.3 | − 2.7 | − 0.4 | 5.9 | 2.7 | 0.5 | 5.9 |
| Combo-derived | 3.1 | − 2.6 | − 0.2 | 5.6 | 2.7 | 0.6 | 5.6 | |
| Scenario 11 | Prog-derived | 3.4 | 0 | 0.1 | − 0.7 | 0.4 | 0.5 | 0.8 |
| Missing data = 10% randomly MCAR | TTD-derived | 3.3 | − 1.8 | − 0.4 | 6 | 1.9 | 0.5 | 6 |
| Combo-derived | 3.1 | − 1.8 | − 0.1 | 5.6 | 1.9 | 0.6 | 5.6 | |
| Scenario 12 | Prog-derived | 3.4 | 0 | 0.2 | − 0.7 | 0.4 | 0.5 | 0.8 |
| Missing data = 10% of patients MCAR | TTD-derived | 3.3 | − 3.7 | − 0.4 | 5.8 | 3.7 | 0.7 | 5.8 |
| Combo-derived | 3.1 | − 3.6 | − 0.3 | 5.4 | 3.7 | 0.7 | 5.4 | |
| Scenario 13 | Prog-derived | 3.4 | 0.2 | 0.4 | − 0.5 | 0.5 | 0.5 | 0.7 |
| Missing data = proportional to utility (MNAR) | TTD-derived | 3.3 | − 2.7 | − 0.4 | 5.9 | 2.7 | 0.6 | 5.9 |
| Combo-derived | 3.2 | − 2.6 | − 0.1 | 5.7 | 2.7 | 0.6 | 5.7 | |
| Scenario 14 | Prog-derived | 3.4 | 0 | 0.2 | − 0.7 | 0.4 | 0.5 | 0.8 |
| Missingness increases closer to death | TTD-derived | 3.4 | − 3 | − 0.4 | 5.8 | 3 | 0.6 | 5.9 |
| Combo-derived | 3.2 | − 2.9 | − 0.2 | 5.4 | 2.9 | 0.6 | 5.5 | |
| Scenario 15 | Prog-derived | 1 | 0 | − 0.5 | − 2.5 | 0.4 | 0.6 | 2.5 |
| No long-term survivors | TTD-derived | 0.8 | − 4.5 | − 0.4 | 27.5 | 4.6 | 1.1 | 27.5 |
| Combo-derived | 0.7 | − 4.6 | − 0.5 | 25 | 4.8 | 1.3 | 25 | |
| Scenario 16 | Prog-derived | 2.2 | 0 | 0.1 | − 1.1 | 0.4 | 0.5 | 1.1 |
| Lower rate of long-term survivors | TTD-derived | 2.1 | − 6.5 | − 0.4 | 8.8 | 6.5 | 1 | 9.1 |
| Combo-derived | 2 | − 6.7 | − 0.7 | 8.1 | 6.7 | 1 | 8.5 | |
| Scenario 17 | Prog-derived | 4.3 | 0 | 0.2 | − 0.6 | 0.5 | 0.5 | 0.7 |
| Higher rate of long-term survivors | TTD-derived | 4.3 | − 1.4 | − 0.4 | 4.6 | 1.5 | 0.5 | 4.6 |
| Combo-derived | 4.1 | − 1.4 | − 0.1 | 4.3 | 1.5 | 0.5 | 4.3 | |
| Scenario 18 | Prog-derived | 3.4 | 0.1 | 0.2 | − 0.5 | 0.5 | 0.5 | 0.6 |
| 10% Pseudo-progression included | TTD-derived | 3.3 | − 10 | − 0.4 | 5.5 | 10 | 0.6 | 5.7 |
| Combo-derived | 3.2 | − 8.1 | − 0.2 | 5.4 | 8.1 | 0.6 | 5.5 | |
| Scenario 19 | Prog-derived | 3.3 | 0 | 0.2 | − 0.7 | 0.4 | 0.5 | 0.8 |
| Link between pre- and post-progression survival | TTD-derived | 3.2 | − 4.5 | − 0.4 | 6 | 4.5 | 0.7 | 6.1 |
| Combo-derived | 3 | − 4.4 | − 0.3 | 5.7 | 4.5 | 0.7 | 5.7 | |
| Scenario A | Prog-derived | 4 | 0.1 | 0.2 | − 0.4 | 0.2 | 0.3 | 0.4 |
| Ipilimumab melanoma | TTD-derived | 4 | − 14.2 | − 0.4 | 4.3 | 14.2 | 0.7 | 4.5 |
| Combo-derived | 3.7 | − 12 | − 0.4 | 4.3 | 12 | 0.6 | 4.4 | |
| Scenario B | Prog-derived | 3.7 | 0 | 0.4 | − 0.6 | 0.2 | 0.5 | 0.6 |
| Nivolumab RCC | TTD-derived | 3.7 | − 4.4 | − 0.4 | 5 | 4.4 | 0.9 | 5.1 |
| Combo-derived | 3.5 | − 4.5 | − 0.5 | 4.6 | 4.5 | 0.7 | 4.7 | |
| Scenario C | Prog-derived | 4.6 | 0 | 0.5 | − 0.5 | 0.2 | 0.5 | 0.5 |
| Pembrolizumab NSCLC | TTD-derived | 4.7 | − 3 | − 0.4 | 4.1 | 3 | 0.8 | 4.1 |
| Combo-derived | 4.4 | − 2.9 | − 0.3 | 3.8 | 2.9 | 0.7 | 3.8 | |
| Scenario D | Prog-derived | 3.5 | 0 | 0.5 | − 0.7 | 0.3 | 0.6 | 0.7 |
| Atezolizumab UCC | TTD-derived | 3.4 | − 5.3 | − 0.4 | 5.3 | 5.3 | 1.1 | 5.4 |
| Combo-derived | 3.2 | − 5.5 | − 0.5 | 4.9 | 5.5 | 0.9 | 5 | |