Lung cancer remains the leading cause of cancer death in the US. Based on results from the National Lung Screening Trial (NLST), lung cancer screening (LCS) with low-dose computed tomography (LDCT) can reduce lung cancer mortality by 20% in a high-risk population.1 However, this benefit is predicated on high population-level adherence to annual LDCT examinations, which was more than 90% in clinical trials.1,2 Current research suggests lower rates of adherence in clinical practice,3 though studies have been limited by a focus on a single round of screening follow-up and challenges in measurement based on the differential follow-up recommended by the Lung CT Screening Reporting & Data System (Lung-RADS).4
Kim et al5 present results of longitudinal follow-up over 3 rounds of screening from the 5-site Population-Based Research to Optimize the Screening Process (PROSPR)–Lung Consortium. Their objectives were to evaluate adherence across 3 rounds of LCS (T0, T1, and T2) and to examine the association of adherence with lung cancer diagnosis and stage distribution. The cohort included individuals who underwent a baseline LDCT screening between 2015 and 2018 and had system follow-up data for at least 36 months. To account for differences in recommended timing of return to LDCT for patients with indeterminate findings on their baseline CT scan, Kim et al5 created a novel metric to define different adherence follow-up windows for patients with negative (Lung-RADS score of 1 or 2) and positive (Lung-RADS score of 3 or 4) screening results. The authors followed up lung cancer diagnoses in 3 screening periods: T0 (0-12 months), T1 (>12 to 24 months), and T2 (>24 to 36 months).
Of 10 170 participants who received baseline LDCT, 84.5% had negative screening results and 15.5% had positive screening results. Adherence at T1 and T2 was 61.2% and 50.5%, respectively, with those adherent at T1 being more likely to adhere at T2 (adjusted relative risk, 2.40; 95% CI, 2.06-2.79). Over 36 months, 2.7% of patients were diagnosed with lung cancer, with rates of 1.4%, 0.7%, and 0.8% during screening periods T0, T1, and T2, respectively. Diagnosis rates in the T1 and T2 periods (combined) were higher than those in the T0 period, aligning with NLST,1 despite lower adherence in this clinical cohort. Kim et al5 found that more lung cancers were diagnosed in individuals who were adherent to LDCT at T1 and T2. However, individuals diagnosed with lung cancer during the T2 period who were nonadherent to T2 LDCT were more likely to be diagnosed with advanced-stage disease. This difference in stage at diagnosis between adherent and nonadherent participants was first apparent in the T2 screening round (>24 to 36 months following the baseline CT) and would be expected to persist in further follow-up, as undetected incident cancers come to clinical attention at later stage.
Of the findings reported by Kim et al,5 one of the most consequential to LCS programs is that T1 adherence was associated with T2 adherence, suggesting that T1 adherence may predict adherence to ongoing follow-up. This is important as some individuals may be eligible for annual LCS for up to 30 years based on their age and smoking history and suggests that interventions need to focus on mitigating nonadherence to first follow-up after a baseline LDCT. Focusing on programmatic resources in the first follow-up period may provide a significant return on investment if individuals remain engaged over their screening lifetime. While not the focus of the study by Kim et al,5 individuals from rural and underserved settings, those of racial and ethnic minority populations, and those currently smoking are at higher risk of nonadherence to follow-up, and it is essential that these groups are included and emphasized in intervention development, adaptation, and evaluation.3,6 There are several ongoing trials to evaluate multilevel interventions to increase adherence to initial LDCT after baseline in diverse care settings, and we eagerly await these results.
Another key contribution to the LCS literature from the study by Kim et al5 is in their measurement of adherence. To examine follow-up adherence over multiple rounds of screening, the authors proposed a novel metric that accounts for variable follow-up for people with indeterminate findings at baseline LCS. At present, as the authors pointed out, there is substantial variability in how adherence is measured in both clinical research studies and by individual programs monitoring LCS quality. While standardization is helpful, we argue that some of this variability may be needed. In research, adherence may need to be defined differently based on specific study objectives. For example, a study reasonably looked at 1 cycle of adherence separately for individuals with negative and positive examination results to increase measurement precision and understand follow-up determinants, which may differ between routine screening and care for an indeterminate finding.7 To align with the objectives of this study, Kim et al5 should be commended for creating a single metric incorporating both follow-up for normal examination findings and return to usual screening care for patients with indeterminate findings after short-term follow-up. However, one weakness of this metric is that in measuring adherence to a subsequent cycle of screening (T2), the metric does not account for shifts in follow-up timing with an indeterminate T1 LCS result. With only 3 rounds of screening assessed by Kim et al,5 this likely did not result in significant misclassification but demonstrates the challenges of using a similar metric over more than 3 screening rounds, which would require increasingly complex branching logic.
For program monitoring, given current guidelines by which individuals may be eligible for ongoing LCS for 30 years, the study by Kim et al5 highlights that uniform metrics are needed, but the focus should likely be on longitudinal maintenance after occasional indeterminate findings rather than precise measures of adherence. It would be even better if these metrics could align with widely used performance measures, such as the Healthcare Effectiveness Data and Information Set (HEDIS), with standards currently under development for LCS. For other screening modalities, like mammography, HEDIS measures are cross-sectional and reflect a percentage of eligible patients who are up to date on screening. A parallel measure for LCS and a reasonable approach to summarize both uptake and maintenance would be the percentage of LCS-eligible individuals who have undergone at least 1 LDCT or diagnostic chest CT within the past 12 months. Programs can then examine data more granularly to determine whether deficiencies result from low uptake of a baseline examination, low initial follow-up (T1), or low ongoing maintenance screening (T2 and beyond).
The study by Kim et al5 is one of several from the PROSPR–Lung Consortium that have been essential in understanding early implementation of LCS in clinical settings, illuminating gaps that were not apparent in clinical trials. Their study demonstrates that adherence is an important barrier to be addressed to achieve the promised mortality benefits of LCS. Moreover, adherence rates in most clinical settings are likely even lower, as reflected in an estimated 22% annual adherence rate for programs reporting to the American College of Radiology registry.8 It will be essential to focus on high-need and underserved settings, including nonacademic, rural, and community practice settings, to further understand the multilevel barriers to LCS adherence and innovate solutions in partnership with communities and clinical programs.
Article Information
Published: March 18, 2025. doi:10.1001/jamanetworkopen.2025.0949
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2025 Duncan FC et al. JAMA Network Open.
Corresponding Author: Matthew Triplette, MD, MPH, Cancer Prevention Program, Public Health Sciences Division, Fred Hutchinson Cancer Center, 1100 Fairview Ave N, Seattle, WA 98109 (mtriplet@fredhutch.org).
Conflict of Interest Disclosures: Dr Triplette reported receiving grants from the National Cancer Institute, National Institute on Minority Health and Health Disparities, and Bristol Myers Squibb Foundation and consulting fees from the GO2 Foundation outside the submitted work. No other disclosures were reported.