Washington, DC—Lung cancer false- positive rates varied 20-fold across the United States, raising questions about the potential impact on organized screening programs, according to data presented at the 2013 American Association for Cancer Research annual meeting.
A review of Medicare-related data showed that approximately 10% of patients who had surgery for diagnosed or suspected lung cancer turned out to have benign disease. However, the rate of false-positive results ranged from 1.3% to 25% across the 50 states. Moreover, 2% of patients found to have benign disease died during
The wide variation in false-positive rates emerged from a statistical patchwork that revealed no consistent pattern or factors associated with false results, reported Stephen A. Deppen, MS, a PhD candidate in the doctoral program in epidemiology at The Institute for Medicine and Public Health, Vanderbilt University, Nashville, TN.
“If the prevalence [of benign disease] differs by state or region, then a national lung cancer screening program may have varying results,” said Mr Deppen.
Interest in lung cancer screening has picked up in the aftermath of the National Lung Screening Trial, which showed that chest computed tomography (CT) scans afforded significant protection against lung cancer mortality compared with chest x-rays. However, 96% of positive CT scans proved to be false-positive. Moreover, 25% of patients who underwent surgery for suspected lung cancer ultimately were found to have benign disease.
Despite the high rate of false results, the mortality-lowering impact of chest CT has sparked interest in the development of a national lung cancer screening program. A better understanding of the origin and variation in rates of false-positive results could inform decision-making about planning and implementing a screening program.
In an effort to gain more insight into the accuracy of diagnostic imaging for lung cancer diagnosis, Mr Deppen and colleagues queried the Medicare Provider Analysis and Review Hospital National Limited Data Set for 2009, the most recent year with complete data.
The data showed that 25,362 patients underwent surgery for known or suspected lung cancer. Subsequently, 2312 (9.1%) patients were found to have benign disease: 8.5% among men and 9.8% among women. The records also showed that surgery was associated with an in-hospital mortality rate of 2.3%, including a rate of 2.1% among patients with benign disease.
Analysis of the results by geographic region showed that the overall false-positive rate provided an incomplete—and possibly misleading—picture. False-positive rates ranged from a low of 1.3% in Vermont to a high of 25% in Hawaii. Overall, the median rate was 8.8%, but rates varied significantly across the states (P <.01).
The Medicare data did not permit an evaluation of possible reasons for the wide geographic variation in false-positive rates. Mr Deppen speculated that one possible explanation involves differences in how quickly providers act on a positive image and send patients to surgery. Haste may increase the likelihood of false-positive rates. In addition, chronic fungal infections involving the lungs may lead to images that mistakenly suggest lung cancer.
Investigators have embarked on a multiyear data analysis to see whether the false-positive rate might have varied over time. In addition, Mr Deppen plans to examine information systems data from a geographic perspective to compare rates of false-positive lung cancer diagnoses with regional prevalences of fungal infections.