Table 1.
Mean posterior parameter estimates with 95% credible intervals of the regression between brackets
| Scenario | α | β | α + β | Cost per QALY gained | Cost per life year gained |
|---|---|---|---|---|---|
| Base case scenario | −0.065 (−0.133/0.001) | −0.124 (−0.194/−0.052) | −0.190 (−0.309/−0.069) | 41,000 (25,900/110,400) | 30,000 (18,800/80,400) |
| Scenario 1 (prior α based on Gallet and Doucouliagos) | −0.096 (−0.150/−0.048) | −0.139 (−0.205/−0.072) | −0.236 (−0.331/−0.140) | 33,400 (24,200/55,200) | 24,300 (17,600/40,200) |
| Scenario 2 (prior α based on Claxton et al.) | −0.100 (−0.166/−0.034) | −0.141 (−0.213/−0.071) | −0.241 (−0.360/−0.123) | 32,700 (22,400/62,400) | 23,800 (16,300/45,400) |
| Scenario 3 (2001–2010 data only) | −0.088 (−0.181/0.005) | −0.117 (−0.202/−0.031) | −0.204 (−0.355/−0.054) | 38,400 (22,700/141,500) | 28,000 (16,500/103,000) |
| Scenario 4 (frequentist approach) | −0.066 (−0.134/0.002) | −0.125 (−0.198/−0.053) | −0.191 (−0.312/−0.070) | 40,800 (25,600/109,100) | 29,700 (18,600/79,500) |
| Scenario 5 (excluding lagged effect) | −0.013 (−0.070/0.047) | NA | NA | 690,000 (108.800/−154,500) | 502,500 (79,500/−112,600) |
Note. Costs per QALY and life year calculated by dividing mean costs by mean effects. 95% credible intervals calculated by setting the sum of α + β at 0.025 and 0.975 quantile. QALY, quality‐adjusted life year.