Overestimation of screening-related complications in the National Lung Screening Trial.

Authors

Kamel MK, Kariyawasam S, Stiles B

Affiliations

• Department of Surgery, Central Michigan University College of Medicine, Mt Pleasant, Mich.
• Department of Pulmonary and Critical Care Medicine, New York Presbyterian-Weill Cornell Medicine, New York, NY.
• Department of Cardiovascular and Thoracic Surgery, Albert Einstein School of Medicine, Montefiore Health System, New York, NY. Electronic address: brstiles@montefiore.org.

Abstract

BACKGROUND: Lung cancer screening-associated complications are often quoted as one of the major barriers for wider screening adoption. A detailed analysis of the National Lung Screening Trial dataset was performed to extrapolate the safety of lung cancer screening.

METHODS: Lung cancer screening-associated invasive procedures and their related complications were analyzed using the National Lung Screening Trial dataset. Factors associated with 90-day postprocedure complications per patient were analyzed with logistic-regression multivariable analysis.

RESULTS: Major complications rate in patients undergoing low-dose computed tomography screening who were diagnosed with lung cancer was 10.2% compared with only 0.04% for patients without lung cancer. Low-dose computed tomography screening, compared with chest radiography, led to major complications in an excess of only 3.5 per 10,000 patients without lung cancer. Among 25,633 patients without lung cancer who underwent low-dose computed tomography screening, 45 developed 90-day postprocedure complications (71 total complications). The most common were pneumothorax (n = 29; 41%), postprocedure hospitalization (n = 6; 8.5%), and infection/fever requiring antibiotics (n = 5; 7%). Cardiac/respiratory arrest occurred in less than 1 in 10,000 low-dose computed tomography-screened patients without lung cancer. On multivariable analysis, pulmonary comorbidity (confidence interval, 1.00-3.37) and procedure type (thoracoscopy [confidence interval, 2.04-10.64] or thoracotomy [confidence interval, 2.38-8.93]) were associated with postprocedure complications in patients without lung cancer. Randomization arm (low-dose computed tomography vs chest x-ray) was not a significant factor (confidence interval, 0.89-1.37).

CONCLUSIONS: It is more informative to report procedural complications in patients not found to have cancer as the true screening-associated risk. Only 4 in 10,000 of patients undergoing low-dose computed tomography screening but not found to have lung cancer will have major complications. Permanent or debilitating complications are exceedingly rare.