Skip to main content
. 2017 Apr 4;14(4):e1002277. doi: 10.1371/journal.pmed.1002277

Table 4. Sensitivities and specificities corresponding to the suggested risk thresholds for the investigated risk prediction models and the National Lung Screening Trial criteria for 6-y lung cancer incidence in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial chest radiography arm.

Measure NLST criteria Bach model LLP model Simplified LLP model PLCOm2012 model Simplified PLCOm2012 model TSCE incidence model Knoke model TSCE CPS lung cancer death model TSCE NHS/HPFS lung cancer death model
Sensitivity 71.4% (68.0%–74.6%) 80.0% (76.9%–82.8%) 73.9% (70.6%–77.0%) 74.0% (70.7%–77.1%) 83.0% (80.2%–85.6%) 80.0% (76.9%–82.8%) 79.8% (76.8%–82.6%) 78.1% (75.0%–81.0%) 72.7% (69.3%–75.8%) 78.5% (75.4%–81.4%)
Specificity 62.2% (61.7%–62.7%) 62.4% (61.9%–62.8%) 62.2% (61.8%–62.7%) 62.6% (62.1%–63.1%) 62.5% (61.9%–62.8%) 62.4% (61.9%–62.8%) 62.3% (61.9%–62.8%) 62.2% (61.7%–62.7%) 62.2% (61.7%–62.7%) 62.3% (61.8%–62.8%)

Data given as percent (95% confidence interval).

CPS, Cancer Prevention Study; HPFS, Health Professionals Follow-up Study; LLP, Liverpool Lung Project; NHS, Nurses’ Health Study; NLST, National Lung Screening Trial; TSCE, Two-Stage Clonal Expansion.