Table 4.

Multivariate Cox proportional hazards models for lung cancer risk associated with NCN characteristics for different time periods

Time period
0–23 mo24–59 mo60–84 mo
Spatial levelHR (95% CI)HR (95% CI)HR (95% CI)
Presence/absence of NCNsa
Person≥1 NCN (vs. no NCNs)9.8 (7.8–12.2)2.0 (1.6–2.5)1.6 (1.2–2.2)
Lung≥1 NCN (vs. no NCNs)17.1 (13.7–21.3)2.9 (2.3–3.6)2.3 (1.6–3.2)
NCN in other lung (vs. none)0.7 (0.6–0.9)0.9 (0.7–1.2)1.0 (0.8–1.2)
Lobe≥1 NCN (vs. no NCNs)40.4 (33–50)5.2 (4.1–6.6)3.0 (2.0–4.5)
NCN in another lobe (vs. none)0.6 (0.5–0.8)1.0 (0.8–1.3)1.3 (0.9–1.9)
LobeNCN characteristicsb
≥1 10+ mm NCN (vs. only 4–9 mm NCNs)12.8 (9.5–17.2)4.7 (2.9–7.5)N.S.
≥1 NCN with spiculated or poorly defined margins (vs. only NCNs with smooth margins)4.1 (3.0–5.5)2.3 (1.5–3.5)N.S.
≥1 Persistent NCN (vs. nonpersistent NCNs)N/Ac4.8 (2.8–8.3)N.S.
 All NCNs soft-tissue attenuationRefRefRef
≥1 NCN with ground-glass attenuation0.3 (0.2–0.4)N.S.3.1 (1.4–6.6)
≥1 NCN with mixed attenuationN.S.N.S.N.S.
  • aModel is for all lobes. Other factors in the model were age, sex, smoking status, pack-years, and COPD status.

  • bModel is for only those lobes with at least one NCN. Forward stepwise approach was used for variable selection for NCN characteristics; N.S. denotes that the factor was not selected for the final model in the stepwise process. The characteristic of multiple NCNs in a lobe was not selected for any time period. Other factors in the model were age, sex, smoking status, pack-years, and COPD status.

  • cNot applicable because lung cancer assessments begin immediately with no lag.