We received several emails from our engaged readership about one of our front-page stories from the February issue. In brief, there were concerns raised about how CHEST Physician characterized the findings of the recent study by Huo et al in JAMA Internal Medicine. On my repeat review of our story and the Huo manuscript, as well as several conversations with content experts both within and outside of CHEST, I agree that we did mischaracterize the findings in our write-up. While the study was not necessarily poorly conducted, there were some methodological concerns that deserved more careful consideration before putting the findings into our publication. CHEST Physician Editorial Board member M. Patricia Rivera, MD, FCCP, and past CHEST President Gerard Silvestri, MD, MS, FCCP, have kindly put together a brief discussion of the potential problems with this paper; while we will further address this in our next issue to go to print (and will likely host further conversations about this manuscript down the road), I wanted to make this expert opinion available to the readership as soon as possible.
For those of you who took the time to write in, thanks so very much!
David A. Schulman, MD, FCCP
Editor in Chief, CHEST Physician
The cover story of the February 2019 edition of CHEST Physician titled “In real-world setting, LDCT screen is linked to high complication risk” erroneously interpreted a study by Huo and colleagues recently published in JAMA Internal Medicine. The cover story states that “the study included 174,702 individuals who underwent an invasive diagnostic procedure as a result of abnormal findings on lung cancer screening and 169,808 control subjects,” “the rates of complications associated with diagnostic procedures following LDCT for lung cancer screening were substantially higher than the rates reported in clinical trials of LDCT” and that “the findings emphasize the importance of discussing the risk of adverse events and cost as part of the shared decision-making process before LDCT screening.”
One wonders if the data reported by Huo and colleagues was skewed by the lens it was presented through or by the lens through which it was interpreted. Let us first elucidate that the study by Huo and colleagues titled “Complication Rates and Downstream Medical Costs Associated with Invasive Diagnostic Procedures for Lung Abnormalities in the Community Setting” was NOT a study of patients who underwent LDCT for lung cancer screening but rather a retrospective, database cohort study from 2008-2013 of patients within the age eligible for screening (age 55 to 77) WITHOUT lung cancer, who underwent similar invasive diagnostic procedures as those performed in the NLST in non–protocol-driven community practices.
Huo et al. hypothesized that the rates of complications after invasive diagnostic procedures observed among screen-eligible patients in the general population would be higher than those reported in the NLST and tested their hypothesis by estimating the complication rate of common invasive diagnostic procedures using data from a database of procedure codes. The database did not however, provide the clinical condition or indication for the procedures, define the number of procedures required to achieve a diagnosis, or define what was the most invasive procedure performed. The authors followed patients for 1 year after their procedure and reported any complication that occurred during that period as related to that procedure. This is not the standard in reporting complications from diagnostic bronchoscopic or radiologic procedures (usually occur within 24-48 hours, or maybe days) or thoracic surgery (30-90 days). As a significant number of the complications reported in the NLST were cardiac, it would be atypical to consider a cardiac complication occurring 1 year after an invasive diagnostic procedure as a complication related to the procedure.
Although the results of the study by Huo and colleagues may not be representative of complications from invasive diagnostic procedures in patients undergoing lung cancer screening, they do show that diagnostic procedures performed in the inpatient and outpatient setting for any pulmonary abnormalities (nodules, masses, adenopathy, infiltrates) are associated with a high risk of complications. In an era of advanced technologies and an increasing aging and chronic critically-ill population, clinicians need to carefully appraise the risks that may be incurred following a diagnostic procedure for a pulmonary lesion and equally, the benefit and diagnostic yield of the procedure. Multidisciplinary discussions, particularly in high-risk patients, can provide guidance to clinical decision-making regarding which procedure will be the least invasive, safest, and most likely to render a diagnosis for the individual patient. Furthermore, we need to take into account that complication rates following procedures are likely higher in centers with a low volume of diagnostic procedures or the inability to provide a less-invasive procedure that can still provide a diagnosis. While it is easy to be critical of large database analyses because of the inherent limitations associated with constructing cohorts that can provide meaningful data, we should not ignore the trends outlined in this article, particularly as the size of the cohort is substantial.
One cannot argue about the importance of discussing the risk of potential complications and cost as part of the shared decision-making process before LDCT screening, but the increased rate of complications reported by Huo et al. should not be interpreted as the complication rate from lung cancer screening in real-world setting, for this is inaccurate and has potential to create additional barriers in lung cancer screening, already beset by barriers on multiple levels. Moreover, we must emphasize that discussions of potential risks and cost from diagnostic pulmonary procedures should not be isolated to lung cancer screening.
M. Patricia Rivera, MD, FCCP
Professor of Medicine
Division of Pulmonary and Critical Care Medicine
Co-Director, Multidisciplinary Thoracic Oncology Program
Director, Multidisciplinary Lung Cancer Screening Program
Medical Director Bronchoscopy and PFT Laboratory
University of North Carolina at Chapel Hill
Chapel Hill, NC
Gerard A. Silvestri, MD, MS, FCCP
Hillenbrand Professor of Thoracic Oncology
Vice-Chair of Medicine for Faculty Development
Division of Pulmonary and Critical Care Medicine
Medical University of South Carolina