Are Primary Care Physicians Really Doing a Lousy Job of Screening for Lung Cancer? Hello, everyone. I'm Dr Kenny Lin, a family physician at Georgetown University Medical Center in Washington, DC, and I blog at Common Sense Family Doctor. Primary care physicians are doing a lousy job of referring patients for lung cancer screening. At least, that's the message from a study to be presented at the American Society of Clinical Oncology (ASCO) meeting, which suggested that less than 2% of eligible US smokers were screened for lung cancer with low-dose CT (LDCT) scans in 2016. The study's lead author contrasted this percentage with the 65% of women over age 40 who were screened for breast cancer with a mammogram in 2015, calling it a "disparity" and speculating that stigma around smoking may be an obstacle to patients getting screened. The president of ASCO called the 2% rate "very disappointing." This one-sided message that low-dose CT scans save lives and need to be done on more patients completely skips over the potential harms of lung cancer screening, which I described in a previous Medscape commentary: false-positive results, incidental findings, radiation exposure, and overdiagnosis. In approving coverage for LDCT in high-risk smokers in 2015, the Centers for Medicare & Medicaid Services required clinicians to have a shared decision-making discussion with patients using a decision support tool to allow them to carefully weigh the risks and benefits before deciding whether to undergo screening. Although the US Preventive Services Task Force (USPSTF) first recommended screening current and former heavy smokers aged 55-80 in 2013, the American Academy of Family Physicians has maintained that there is insufficient evidence to know whether the fewer lung cancer deaths and low complication rates observed in the National Lung Screening Trial (NLST) would be seen in community practice. In 2016, I discussed two community-based studies that found much higher rates of false positives and inappropriate screening in low-risk persons. And last year, the Veterans Health Administration (VA) reported the outcomes of a large demonstration study at eight of its hospitals: 60% of patients who underwent LDCT screening had nodules seen, and 40% had incidental findings—much higher rates than reported in the NLST. Although the vast majority of lung nodules detected turn out to be benign, a longitudinal study[4] found that more than half of persons with incidental pulmonary nodules experience at least mild emotional distress, and 25% continued to report distress after 2 years of surveillance. About 1 in 50 persons screened in the VA study underwent invasive diagnostic procedures for lung nodules, exposing them to serious physical harms. Another population-based study found that 7% of all persons who had a transthoracic needle lung biopsy suffered a pneumothorax requiring a chest tube, and 1% were complicated by hemorrhage. In this study, patients who had higher risks for complications were aged 60-69 years, smokers, and those with COPD—in other words, the population most likely to be eligible for lung cancer screening. Other researchers have pointed out that using a lung cancer risk prediction model threshold for offering lung cancer screening to patients, rather than the age and smoking pack-year history cutoffs used in the NLST and by the Task Force, could improve the benefit-to-harms ratio of LDCT. A recent analysis of nine lung cancer risk models suggested that despite substantial differences in the populations they targeted for screening at a 5-year risk threshold of 2%, the four best-performing models agreed on 73% of the persons chosen. The USPSTF plans to review these risk prediction models in order to update its 2013 recommendation in the next couple of years. The bottom line is that while it is important for family physicians to make eligible patients aware of the pros and cons of lung cancer screening, this is a complex, individualized decision in the context of a great deal of uncertainty. What proportion of patients choose to undergo LDCT should not be compared to rates of screening tests with much stronger evidence of net benefit. Source
