Sampling and correction scenario | Excluded | Sensitivity | Specificity |
Scenario 1 (all scans 60–65 min) | — | 96% (94%–97%) | 99% (98%–100%) |
Scenario 2 | — | 92% (90%–93%) | 99% (98%–100%) |
2 exclude out-of-range* | 20% | 94% (92%–95%) | 100% (99%–100%) |
2 corrected† | — | 95% (93%–96%) | 99% (99%–100%) |
Scenario 3 | — | 89% (86%–90%) | 96% (95%–97%) |
3 exclude out-of-range* | 44% | 93% (91%–95%) | 99% (98%–100%) |
3 corrected† | — | 93% (92%–95%) | 99% (98%–100%) |
Scenario 4 | — | 73% (70%–76%) | 91% (89%–92%) |
4 exclude out-of-range* | 64% | 82% (78%–86%) | 98% (95%–99%) |
4 corrected† | — | 87% (85%–89%) | 98% (97%–99%) |
↵* Out-of-range is defined as uptake time < 50 min or > 70 min, or difference between paired scans > 15 min.
↵† , where corr = corrected, mid = midpoint, and OBS = observed (11).
Sensitivity is probability of observing 30% or greater SUV decrease given true 40% decrease; specificity is probability of observing absolute change of <30% given no true change. n = 1,000 per scenario. Data in parentheses are 95% confidence intervals.