The Canadian researchers took a similar multiyear tack. They used their checklist to assess the potential hazard of lung nodules detected during the first CT scans, and then used two years of follow-up lung screenings to see whether their model worked.
A total of more than 12,000 lung cancer nodules observed on CTs of nearly 3,000 current and former smokers were examined.
The checklist runs counter to current standards, which mainly rely on the size of a nodule to determine whether more tests should be run.
But the team found that in one of five study participants, the largest nodule in their lungs did not prove malignant. Other dynamics like the shape of the mass, its location in the lung, and the individual person's risk factor must be accounted for to properly assess the hazard posed.
For example, nodules located in the upper lobes of the lungs increased the likelihood of cancer, the report says. However, more nodules found in a CT scan actually decreases cancer risk.
The new model is much better at predicting cancer than previous checklists, said Dr. Christine Berg, co-principal investigator of the National Lung Screening Trial for the U.S. National Cancer Institute.
"If you have a nodule you can predict it's positive almost 20 percent, which is a lot better than what we had," she said.
The model is even better at ruling out nonmalignant nodules, which could save patients from unnecessary repeat testing, lowering their radiation exposure and risk from procedures.
"If you say something is benign using the model, the chances of it being malignant are extremely small, about 99.6 percent," Berg said. "This is an excellent first cut of determining whether a nodule on your scan is malignant or not."
For more information on lung cancer screening, visit the U.S. National Cancer Institute.
Copyright © 2012 HealthDay. All rights reserved.