Table 1.
Data
| Variable# | Researcher | Patient | Trial | |
|---|---|---|---|---|
|
Utilities$ |
|
|
|
|
| Success of experimental treatment within trial* |
95 (80–100) |
90 (50–100) |
|
|
| |
Assumed 100(1) in the equipoise model (inset) |
Assumed 100(1) in the equipoise model (inset) |
|
|
| |
(MC modeling: triangular distribution) |
(MC modeling: triangular distribution) |
|
|
| Failure of experimental treatment within trial* |
54 (10–100) |
16.3 (0–50) |
|
|
| |
Assumed 0 in the equipoise model (inset) |
Assumed 0 in the equipoise model (inset) |
|
|
| |
(MC modeling: triangular distribution) |
(MC modeling: triangular distribution) |
|
|
| Success of standard treatment within trial* |
70 (40–80) |
84 (50–1100) |
|
|
| |
Assumed 100 (1) in the equipoise model (inset) |
Assumed 100 (1) in the equipoise model (inset) |
|
|
| |
(MC modeling: triangular distribution) |
(MC modeling: triangular distribution) |
|
|
| Failure of Standard treatment within trial* |
44 (0–80) |
16.9 (0–50) |
|
|
| |
Assumed 0 in the equipoise model (inset) |
Assumed 0 in the equipoise model (inset) |
|
|
| |
(MC modeling: triangular distribution) |
(MC modeling: triangular distribution) |
|
|
|
Probabilities |
|
|
|
|
| Success of experimental treatment (e) |
|
|
0.41 (0.2-0.8)** |
|
| |
|
|
(0.5 for the equipoise model) |
|
| |
|
|
(MC modeling: binomial distribution; n = 450***) |
|
| Failure of experimental treatment |
|
|
1-e |
|
| Success of standard treatment (s) |
|
|
0.59 (0.2-0.8)** |
|
| |
|
|
(MC modeling: binomial distribution; n = 316***) |
|
| Failure of standard treatment |
|
|
1-s |
|
| Randomization (r) |
|
|
0.5 (0.2-0.8) |
|
| |
|
|
(0.5 for the equipoise model) |
|
| |
|
|
(MC modeling: triangular distribution) |
|
| Honoring trust (p) |
|
|
0.5 (0–1) |
|
|
Regret |
|
0.2 (0–1) |
|
|
| |
|
(MC modeling: triangular distribution) |
|
|
|
Guilt |
0.2 (0–1) |
|
|
|
| (MC modeling: triangular distribution) |
# - sensitivity analysis was performed for the values shown in Table but also for all extreme values (0–1)- the change in the assumptions did not affect the results of the analysis; $- based on the survey of 8 clinical investigators; * assumed to be the same within and outside of the trial; MC- Monte Carlo; **- actual values based on reference [41] is 0.41 (0.37-0.45) for experimental and 0.59 (0.55-0.62) for standard treatment, respectively ***-based on the reference [41].