|
Abstract:
|
We have applied a newly developed method on scheduling CT screening exam in lung cancer, based on an asymptomatic individual’s current age, screening history, risk tolerance and other parameters. The probability of incidence was a function of the next screening time, so that the future screening time can be found by limiting this risk to a small value p. That is, with probability (1-p), one will not be a clinical incident case before the next exam. We estimated the lead time distribution and probability of overdiagnosis if one would be diagnosed with cancer at the next screening, so that predictive information could be provided to individuals on how early the disease could be detected and the risk of overdiagnosis. We applied the methods to two cohorts (male and female heavy smokers) in the National Lung Screening Trial using low dose computerized tomography, comparing their future screening times, lead time and overdiagnosis by simulations. Under the same conditions, male heavy smokers need to schedule the next exam earlier than their female counterparts do; and older people should schedule it earlier than the younger ones. The mean lead time of female heavy smokers is longer than that of males; the risk of overdiagnosis is small for both cohorts, although it is a little bit higher in male than in female heavy smokers.
|
