Low-Dose CT Lung Screening More Sensitive Than Radiography
But positive predictive value lower; second study IDs cancer predictors among those with nodules on CT
WEDNESDAY, Sept. 4 (HealthDay News) -- Low-dose computed tomography (CT) lung screening is more sensitive than radiography, and predictors of cancer on low-dose CT have been identified, according to two studies published in the Sept. 5 issue of the New England Journal of Medicine.
Denise R. Aberle, M.D., from the David Geffen School of Medicine at the University of California in Los Angeles, and colleagues assessed results from the first two rounds of annual screenings (T1 and T2) with low-dose helical CT, compared with chest radiography. The researchers found that the sensitivity, specificity, positive predictive value, and negative predictive value were 94.4, 72.6, 2.4, and 99.9 percent, respectively, at round T1, with increased positive predictive value at T2 (5.2 percent). At round T1 in the radiography group, the corresponding values were 59.6, 94.1, 4.4, and 99.8 percent, with increased sensitivity and positive predictive value at T2.
Annette McWilliams, M.B., from the Vancouver General Hospital in Canada, and colleagues analyzed data from two cohorts of participants undergoing low-dose CT screening. The development data set included 1,871 participants with 7,008 nodules, while the validation data set included 1,090 persons with 5,021 nodules. The researchers found that the rate of cancer in the two data sets was 5.5 and 3.7 percent, respectively, among those with nodules. Older age, female sex, family history of lung cancer, emphysema, larger nodule size, location of the nodule, part-solid nodule type, lower nodule count, and spiculation were predictors of cancer. The final models showed excellent discrimination and calibration.
"Predictive tools based on patient and nodule characteristics can be used to accurately estimate the probability that lung nodules detected on baseline screening low-dose CT scans are malignant," McWilliams and colleagues write.