Highlights
- Integration of the MyLungHealth patient portal tool with clinician-facing decision support doubled the identification of patients newly eligible for lung cancer screening (4.7% vs 2.3%).
- The multi-center pragmatic trial involving over 31,000 participants demonstrated a statistically significant increase in low-dose CT (LDCT) ordering rates through personalized education.
- Direct-to-patient data collection effectively addressed critical electronic health record (EHR) gaps regarding smoking duration and cessation dates.
- While relative improvements were substantial (aOR 2.19 for eligibility), absolute increases remained modest, suggesting a need for more intensive implementation strategies.
Background
Lung cancer remains the leading cause of cancer-related mortality globally. Despite the proven efficacy of low-dose computed tomography (LDCT) in reducing mortality—as established by the National Lung Screening Trial (NLST) and the NELSON trial—screening uptake remains suboptimal. In the United States, screening rates for eligible individuals often hover below 15%, significantly trailing behind screening rates for breast or colorectal cancers.
Major barriers to implementation include the complexity of current screening criteria (e.g., USPSTF 2021 guidelines), incomplete smoking history documentation within EHRs, and the heavy cognitive load on primary care clinicians during brief office visits. Shared decision-making (SDM) is legally mandated for LCS reimbursement, yet clinicians often lack the time or tools to facilitate these discussions effectively. The MyLungHealth trial sought to bridge these gaps by shifting part of the data collection and educational burden directly to the patient via EHR-integrated portals.
Key Content
Study Architecture and Methodological Framework
The MyLungHealth trial was a pragmatic, unstratified, parallel-group randomized clinical trial conducted at two major academic health systems: University of Utah Health and NYU Langone Health. The study utilized a unique two-part design to address two distinct clinical hurdles:
- Study 1 (Eligibility Identification): Targeted 26,729 patients with “uncertain” eligibility due to missing or vague smoking history (e.g., 10–19 pack-years, unknown quit dates). The goal was to refine this data via patient input.
- Study 2 (Ordering Facilitation): Targeted 4,574 patients with documented eligibility (≥20 pack-years, current smoker or quit <15 years) to evaluate whether patient education would drive LDCT ordering.
Participants were randomized to either a control arm (clinician-facing Decision Precision+ tool only) or an intervention arm (Decision Precision+ plus the patient-facing MyLungHealth tool). The interventions were provided in both English and Spanish to ensure broader accessibility.
Bridging the EHR Data Gap: Study 1 Findings
Incomplete EHR documentation is a primary reason why eligible patients are overlooked. In Study 1, the MyLungHealth tool prompted patients to provide detailed smoking histories through their patient portals. This proactive approach resulted in a significant increase in the identification of screening-eligible patients. In the intervention arm, 4.7% (635/13,412) of participants were newly identified as eligible, compared to only 2.3% (308/13,317) in the control group. The adjusted odds ratio (aOR) of 2.19 (95% CI, 1.99-2.42; P < .001) underscores the power of patient-generated health data in correcting historical EHR inaccuracies.
Enhancing Clinical Action: Study 2 Findings
For patients already known to be eligible, the challenge is translating that knowledge into a clinical order. The MyLungHealth tool delivered personalized risk/benefit information and educational content to the patient prior to their visit. In Study 2, LDCT ordering rates were 20.5% in the intervention group versus 19.2% in the control group (aOR 1.16; 95% CI, 1.04-1.30; P = .008). While the difference was statistically significant, the narrow absolute margin suggests that patient education alone, even when paired with clinician reminders, may not be sufficient to overcome all barriers to ordering.
Translational and Technical Integration
Technically, the trial represents a milestone in “SMART on FHIR” (Substitutable Medical Applications, Reusable Technologies on Fast Healthcare Interoperability Resources) applications. By integrating the tool directly into the EHR workflow, the researchers minimized the friction typically associated with third-party software. The tool allowed for real-time updates to the patient’s record, ensuring that when the clinician opened the chart, the most accurate eligibility data—verified by the patient—was already present.
Expert Commentary
Addressing the ‘Modest’ Absolute Increase
The most critical takeaway from the MyLungHealth trial is the disparity between relative and absolute gains. A doubling of eligibility identification is a major success for health informatics; however, the fact that only ~20% of eligible patients had a scan ordered highlights the systemic inertia in preventive oncology. Experts suggest that several factors contribute to this: patient-level concerns regarding radiation or ‘overdiagnosis,’ clinician-level fatigue, and the logistical hurdles of scheduling specialized imaging.
The Role of the Patient Portal
A potential limitation of this study—and a point for future policy discussion—is the reliance on active patient portal accounts. As the trial required an active portal for the intervention, there is a risk of exacerbating the digital divide. Vulnerable populations with limited internet literacy or access may not benefit from these tools unless paired with human-mediated outreach, such as patient navigators or medical assistants.
Mechanistic Insights into Behavior Change
The success of Study 1 over Study 2 suggests that ‘data correction’ is a more straightforward problem to solve via informatics than ‘behavioral change.’ Correcting a quit date is a cognitive task for the patient; deciding to undergo a cancer screening involves emotional, financial, and physical considerations. Future tools might need to incorporate more robust shared decision-making features, perhaps utilizing AI-driven chatbots to address specific patient anxieties in real-time.
Conclusion
The MyLungHealth trial provides high-level evidence that EHR-integrated, patient-facing tools are effective at improving the accuracy of LCS eligibility screening and modestly increasing screening orders. By empowering patients to curate their own medical histories, health systems can overcome the chronic problem of ‘dirty data’ in smoking records. However, to reach the national targets for lung cancer screening, these digital tools must be viewed as one component of a larger, multi-modal strategy that includes clinician education, patient navigation, and potentially more aggressive automated ordering systems for high-risk individuals.
References
- Kukhareva PV, et al. Enhancement of Patient-Centered Lung Cancer Screening: The MyLungHealth Randomized Clinical Trial. JAMA Oncol. 2025;e255672. PMID: 41452617.
- Meza R, et al. Lung Cancer Predictions for the United States: 2021 USPSTF Recommendations. J Thorac Oncol. 2021;16(10):1612-1622. PMID: 34111584.
- The National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365(5):395-409. PMID: 21714641.
- de Koning HJ, et al. Reduced Lung-Cancer Mortality with CT Screening in a Randomized Trial. N Engl J Med. 2020;382(6):503-513. PMID: 31995683.