Screening Smarter in Lung Cancer: Letter to the Editor and Response

Putting forward the stories and writings that make up CHEST Advocates, I acknowledge that advocacy comes from many intentions, grounded in the complexity of humanity made up of morals, ethics, laws, and culture. Interpretation of actions, decisions, or, as is the case in science, data, may result in advocacy that differs from one individual to another for reasons grounded in those morals, ethics, laws, and culture.

In CHEST Advocates, we welcome such diverse opinions on how to best allocate our attention toward advocacy and attempt to bring in all sides in order to have a comprehensive understanding of the situation. As public health scientists, physicians, and citizens, meeting these diverse opinions and evolving interpretations is welcomed. In fact, it's desired.

In the most recent issue of CHEST Advocates, Otis Brawley, MD, former president of the American Cancer Society, discussed his views on lung cancer screening, with an emphasis on caution that is meant to protect the patient in their shared decision-making for the screening and its next steps. We have invited Otis Rickman, DO, FCCP, to provide another perspective around lung cancer screening, with the hopes that such caution does not result in inertia around lung cancer screening. Two great clinicians, advocating for patients.

Please enjoy these requested commentaries as we continue to advocate for patients. Only in such conversations and insights, where these dialogues are welcomed and reflected upon, can we aid in making medicine a public trust.

Panagis Galiatsatos, MD, MHS, FCCP
Editor in Chief, CHEST Advocates

To the Editor

Otis B. Rickman,
DO, FCCP, DAABIPCHEST Advocates recently released a piece called “Screening Smarter in Lung Cancer,” which raised several questions about the feasibility and efficacy of low-dose CT scan screening. It raises a legitimate concern: Rising CT scan utilization and the resulting radiation exposure warrant continued scrutiny. But the article’s central recommendation, that patients seek screening only at high-volume academic centers, deserves closer examination.

The mortality data is not in question. The National Lung Screening Trial (NLST) demonstrated a 20% relative reduction in lung cancer mortality with low-dose CT scan vs chest radiography.¹ NELSON, an independent randomized trial, corroborated this with a 24% reduction in lung cancer death among men at 10 years.² Notably, NLST’s benefit extended to all-cause mortality, a 6.7% relative reduction, which is the outcome least susceptible to misclassification of procedure-related deaths.¹ If screening-associated harm were meaningfully offsetting benefit, this end point would be the first to show it. It did not.

Where I would push back is on the article’s use of NLST’s 96.4% false-positive rate to support restricting screening access to select centers. That figure reflects the original 2011 trial protocol, before the Lung CT Screening Reporting and Data System (Lung-RADS) existed. When Lung-RADS was applied retrospectively to the same NLST cohort, the false-positive rate fell to 12.8% at baseline and 5.3% on subsequent screening rounds.³ Structured risk stratification, whether through Lung-RADS or nodule-specific calculators like the Brock model, has already done much of the work the article attributes to institutional prestige. In a contemporary real-world screening program applying these tools, only 15% of nodules taken to surgical resection proved benign.⁴

That benchmark is itself becoming outdated. A 2025 study from Cedars-Sinai directly compared traditional surgical wedge resection against a combined single-anesthesia pathway pairing robotic bronchoscopic biopsy, with resection when indicated, in patients with high pretest probability of malignancy. Wedge resection alone carried a 21.9% benign resection rate. The robotic bronchoscopy-first pathway reduced this to 7.6%, with no difference in time to treatment or procedural complications.⁵ This is the direction diagnostic technology in this field is moving, and it argues for expanding access to advanced navigation bronchoscopy rather than concentrating it, and the screening programs that feed it, within a small number of flagship institutions.

The article’s own cautionary example, a blind percutaneous needle biopsy causing pneumothorax and cardiac complications in a patient with limited cardiopulmonary reserve, illustrates the actual point of failure: the diagnostic pathway following a positive screen, not the screening program itself. Cone-bean CT scan-guided robotic bronchoscopy allows real-time confirmation of tool-in-lesion before tissue is obtained, addressing precisely this failure mode. A recent multicenter trial of this technology on nodules 20 mm or smaller reported a 99.4% tool-in-lesion rate and 89% diagnostic yield, with no pneumothorax events.⁶

I would also note that framing screening risk primarily as an access problem overlooks the more consequential access problem we actually have. National lung cancer screening uptake remains near 6% of eligible patients, with substantial regional variation.⁷ The clinical risk calculus this article describes, weighing screening’s benefits against its harms, is only relevant to the small fraction of eligible patients currently being screened at all. The larger opportunity for reducing lung cancer mortality lies in closing that gap, not in narrowing where screening and its downstream workup can safely occur.

I share the goal of this article: minimizing harm from an intervention we know saves lives. I would suggest that goal is better served by disseminating improved risk-stratification tools and modern bronchoscopic technology broadly, rather than by restricting access.

Otis B. Rickman, DO, FCCP, DAABIP
Professor of Clinical Education, University of Tennessee Health Science Center
Director, Interventional Pulmonology, Ascension Saint Thomas

Response

Otis Brawley, MDFirst, I want to thank Dr. Rickman for his gentlemanly letter with constructive criticisms. I am a medical oncologist trained as an epidemiologist specializing in cancer outcomes, especially cancer screening outcomes. As the former Chief Medical and Scientific Officer of the American Cancer Society, I am a coauthor of widely accepted cancer screening guidelines in the United States and Europe.

As I interpret lung cancer screening data, I am heavily influenced by the concept of efficacy in a controlled clinical trial, where patients or subjects tend to be healthier than the population at large, and those conducting the study are generally more expert and experienced in lung screening intervention. Effectiveness is the real-world data, and I am hard-pressed to name an instance where an effectiveness outcome has been better than an efficacy outcome.¹ 

The National Lung Screening Trial (NLST) and the Dutch Belgian Randomized Lung Cancer Screening Trial (NELSON) are two of the best-designed and best-conducted studies in the history of screening for any disease. These studies demonstrate both the benefits and the harms associated with lung screening when done in academic hospitals. The benefit is a 20% to 30% reduction in relative risk of death, which might increase with continued annual high-quality screening. The harms include false-positive results leading to additional imaging and potentially invasive procedures, adverse outcomes of those procedures, overdiagnosis, radiation exposure, incidental findings, and short-term increases in distress or anxiety associated with abnormal results. The NLST even demonstrated a small risk of death related to invasive diagnostics.  

The efficacy findings were judged by the US Preventive Services Task Force (USPSTF) as showing “moderate benefit and moderate harm.”² The findings led the USPSTF, the American Cancer Society (ACS), the American College of Chest Physicians (CHEST), and the American Thoracic Society (ATS) to recommend shared decision-making as part of the consenting process.^2-4 The USPSTF and the ACS also recommend that patients receive care at institutions with organized screening programs that provide high-quality services.^5,6 

Because of the documented harms associated with screening an asymptomatic population for lung cancer, the Medicare National Coverage Determination (NCD 210), first published in 2016 and modified in 2022, requires a counseling and shared decision-making session before the first screening of a CT scan for a Medicare recipient.⁷ Additionally, the US Centers for Medicare and Medicaid Services (CMS) mandates use of one or more approved decision aids.^8,9

Good effectiveness data was published by Rendle and colleagues in 2024.¹⁰ They assessed a registry of 9,266 patients undergoing baseline low-dose CT scans in five hospitals. Screening began after the original version of Lung CT Screening Reporting and Data System (Lung-RADS) was introduced for the assessment of nodules. In this study, the false-positive rate was 90.5%. I would argue this is not substantially different from the 96.4% false-positive rate in NLST. 

The Rendle study found that invasive diagnostic complications were appreciably more common in routine practice than in the NLST.¹⁰ In the real-world cohort, 30.6% of patients getting an invasive procedure had a complication; 20.6% had a major complication. In the NLST, the overall complication rate was 17.7%; 9.4% were major complications.¹⁰ 

Mine is not an argument that quality screening can only be done in academic centers. It is an argument for screening in organized screening, diagnostic, and treatment programs that keep records so that the program can be assessed. Some institutions keep records and have a fully staffed program for screening, diagnosis, and treatment. Some do not. Some do not conduct follow-up for outcomes; they literally offer the spiral CT scan and a list of places a person can go to get assessment of an abnormality. Some do not offer smoking cessation.⁶ 

Perhaps Dr. Rickman and I can agree that some community centers offer high-quality screening, diagnosis, and treatment.⁶ Also, some academic centers are less than optimal. I suspect few programs can meet the quality standards of NLST or NELSON, and it is unlikely that institutions in NLST or NELSON trials can currently achieve the same outcomes because they have lost the benefit of the healthy volunteer effect.

If I may, I’d like to pivot to focus on informed and shared decision-making in lung cancer screening. I worry that some of my colleagues may not fully appreciate the risk of harm, engage in shared decision-making consistently, or respect the subject/patient’s preferences and circumstances. I base this on comments like, “All eligible should get screened,” and “Telling the truth means people will not get screened,” which were posted to online forums such as LinkedIn.

The 2013 CHEST guideline and the 2015 ATS/CHEST policy statement express that shared decision-making is appropriate and is not designed to convince every eligible person to get screened.^4,5 The societies say that “full discussion of the harms and an individual’s right to decide against screening are integral to appropriate low-dose CT implementation and should not be viewed as an obstacle to increasing screening rates.” 

I am an advocate for patient autonomy and self-determination. I support shared decision-making in which the clinician and patient jointly consider the evidence and determine which option best fits the patient’s preferences and circumstances. I want those who qualify and want screening to have it in a high-quality screening program where we can minimize risks and optimize benefits. 

Telling people about the potential benefit then exaggerating that benefit or trivializing the risks does us no good. In shared decision-making, the physician doesn’t have to be neutral. The health care provider, who hopefully knows the data, can say, “Based on your risk and the evidence as I interpret it, I recommend screening,” while still acknowledging the potential for harm. 

If we as professionals do not appreciate these facts of screening, respect the ethical principle of patient autonomy, and clearly explain the potential benefits and potential harms, then we harm the physician-patient relationship, lose credibility, and justify and perpetuate medical distrust. 

Otis W. Brawley, MD, MACP, FRCP(L), FASCO, FAACR, FACE
Bloomberg Distinguish Professor of Oncology and Epidemiology
Johns Hopkins School of Medicine
Bloomberg School of Public Health
Sidney Kimmel Comprehensive Cancer Center