Skip to main content
. 2016 Jun 1;7(2):95–104.

Table 2.

Multivariate Logistic Regression Models of Smoking Relapse Defined Prospectively in PLuSSa

Variables PR Non-PR OR 95% CI P Value
n 267 1142
Age (y, mean (SD)) 57.9 (6.2) 60.6 (6.9) 0.64b 0.50, 0.83 0.0005
Cigs/day 0.83 0.71, 0.98 0.026
    1-19 Cigs/day 79 (29.6) 291 (25.5)
    20-29 Cigs/day 120 (44.9) 458 (40.1)
    30-39 Cigs/day 46 (17.2) 223 (19.5)
    40 or more Cigs/day 22 (8.2) 170 (14.9)
Time since quit (y, median (Q1, Q3)) 1 (0, 4) 5 (2, 8)
    Spline 1 <0.0001
    Spline 2 0.039
    Spline 3 0.24
Number of symptomsc 1.21 1.01, 1.44 0.036
    None 76 (28.5) 408 (35.7)
    1 62 (23.2) 248 (21.7)
    2 or more 129 (48.3) 486 (42.6)
Baseline COPD (yes, %) 76 (28.5) 394 (34.5) 0.78 0.55, 1.10 0.16

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; OR, odds ratio; PLuSS, the Pittsburgh Lung Screening Study; PR, prospective relapse; SD, standard deviation.

a

Area under ROC curve=0.785.

b

Odds ratio for every 10 years.

c

Baseline symptoms, including phlegm, wheezing, shortness of breath, and ankle swelling.