Table 5.
Parameter varied from base-case analysisa | Mean decrease with user-testing (95% credible interval) | Incremental cost-saving | Net monetary benefitb | ||||
---|---|---|---|---|---|---|---|
Moderate-severe pADEs | All pADEs | QALY decrements | Mean (95% credible interval), £ | Probability user-testing cost-saving | Mean (95% credible interval), £ | Probability user-testing cost-effective | |
Base case (for comparison) | 157 (−13 to 363) | 411 (210–675) | 147.5 (−24.9 to 406.1) | 240,943 (43,527–491,576) | 0.99 | 3,190,064 (−346,709 to 8,480,665) | 0.96 |
Time horizon reduced to 1 year | 31 (−3 to 73) | 82 (42–136) | 31.2 (−5.2 to 87.1) | 46,648 (3,177–101,006) | 0.98 | 669,944 (−74,287 to 1,815,983) | 0.96 |
Time horizon increased to 10 years | 315 (−32 to 728) | 823 (420–1,356) | 265.8 (−46.5 to 742.3) | 449,553 (79,168–911,049) | 0.99 | 5,766,532 (−612,727 to 15,470,264) | 0.96 |
Number of doses of intravenous voriconazole administered using IMG per annum increased to 20,000 | 787 (−80 to 1,822) | 2,057 (1,039–3,402) | 730.5 (−133.4 to 2,041.3) | 1,231,112 (224,306–2,508,969) | 0.99 | 15,840,983 (−1,816,590 to 42,701,213) | 0.96 |
Undetected error frequency with current guidelines reduced to 32%, with unchanged relative reduction in error frequency after user-testing | 80 (−77 to 252) | 208 (36–421) | 75.4 (−64.7 to 269.0) | 127,725 (−57,148–335,690) | 0.92 | 1,636,052 (−1,253,565 to 5,611,961) | 0.88 |
Undetected error frequency with current guidelines reduced to 32% and relative effect of user-testing halved | 41 (−110 to 223) | 119 (−49–316) | 40.6 (−125.9 to 237.5) | 81,539 (−106,273–285,048) | 0.81 | 892,882 (−2,526,604 to 4,916,336) | 0.71 |
aSee Tables I to L of the Online Supplemental Material
bWillingness-to-pay threshold = £20,000 per QALY
pADE preventable adverse drug event, QALY quality-adjusted life year