An EHR-based algorithm identified 21,446 patients eligible for lung cancer screening who had not yet initiated screening across Kaiser Permanente Washington clinics.
Implementing an EHR-based algorithm and decision support tools can effectively identify a large cohort of primary care patients eligible for lung cancer screening who have not yet been screened.
Abstract Rationale Despite US Preventive Services Task Force (USPSTF) lung cancer screening (LCS) recommendation for adults aged 50-80 years with 20 pack-year history and current use or quit 15 years ago, less than one-fifth of LCS-eligible adults have ever screened. LCS implementation in healthcare systems is complex, involving several key steps including risk assessment, initiation of screening, follow-up of findings, and lung cancer diagnosis. Kaiser Permanente Washington (KPWA) primary care providers (PCPs) currently do not have a defined LCS clinical workflow in the electronic health record (EHR) to identify those who are LCS eligible but have not yet screened. To address this, our team is piloting a quality improvement (QI) initiative to test improvements to the LCS pathway for PCPs. Methods The QI initiative includes developing an algorithm to identify LCS-eligible patients with no evidence of screening and multiple Plan, Do, Study, Act (PDSA) cycles where the first cycle includes two embedded EHR clinical decision support features for PCPs: a population health prompt available to all 25 KPWA-owned clinics, and a point-of-care advisory in two pilot clinics that includes a LCS navigator that steps a PCP through tobacco history, risk calculator, shared-decision making, diagnoses, and LDCT order. The first PDSA cycle launched October 22, 2025 and will run through mid-December. We describe results from the algorithm to identify eligible patients for the overall KPWA population, report rates of LCS prior to PDSA, and estimate the number of patients with visits in the two pilot clinics post-PDSA. Results As of August 13, 2025, across all clinics, 23,672 patients met age and tobacco-use criteria according to USPSTF guidelines and identified in the EHR, the majority of whom were current users of tobacco. Nearly 10% were already engaged in LCS and excluded. The final sample was 21,446 who might be LCS-eligible but had yet to initiate screening. Testing system-level quality improvements, we estimate 190 patients at clinic A and 45 patients at clinic B will have at least one in person ambulatory visit to prompt LCS discussion during PDSA cycle 1. Conclusions This QI initiative demonstrates the value of embedding research within healthcare systems to address complex implementation challenges like LCS. By integrating an evidence-based algorithm and EHR decision support tools into primary care, this project builds a foundation for scalable, data-driven improvement in screening uptake and care quality in PCP-driven LCS programs. This abstract is funded by: Eli Lilly
Justice et al. (Fri,) conducted a other in Lung cancer screening eligibility (n=21,446). EHR clinical decision support features was evaluated on Identification of LCS-eligible patients. An EHR-based algorithm identified 21,446 patients eligible for lung cancer screening who had not yet initiated screening across Kaiser Permanente Washington clinics.