Lessons Learned to Promote Lung Cancer Screening and Preempt Worsening Lung Cancer Disparities

Development of Lung Cancer Risk Prediction Machine Learning Models for Equitable Learning Health System: Retrospective Study

BACKGROUND

A significant proportion of young at-risk patients and nonsmokers are excluded by the current guidelines for lung cancer (LC) screening, resulting in low-screening adoption. The vision of the US National Academy of Medicine to transform health systems into learning health systems (LHS) holds promise for bringing necessary structural changes to health care, thereby addressing the exclusivity and adoption issues of LC screening.

OBJECTIVE

This study aims to realize the LHS vision by designing an equitable, machine learning (ML)-enabled LHS unit for LC screening. It focuses on developing an inclusive and practical LC risk prediction model, suitable for initializing the ML-enabled LHS (ML-LHS) unit. This model aims to empower primary physicians in a clinical research network, linking central hospitals and rural clinics, to routinely deliver risk-based screening for enhancing LC early detection in broader populations.

METHODS

We created a standardized data set of health factors from 1397 patients with LC and 1448 control patients, all aged 30 years and older, including both smokers and nonsmokers, from a hospital's electronic medical record system. Initially, a data-centric ML approach was used to create inclusive ML models for risk prediction from all available health factors. Subsequently, a quantitative distribution of LC health factors was used in feature engineering to refine the models into a more practical model with fewer variables.

RESULTS

The initial inclusive 250-variable XGBoost model for LC risk prediction achieved performance metrics of 0.86 recall, 0.90 precision, and 0.89 accuracy. Post feature refinement, a practical 29-variable XGBoost model was developed, displaying performance metrics of 0.80 recall, 0.82 precision, and 0.82 accuracy. This model met the criteria for initializing the ML-LHS unit for risk-based, inclusive LC screening within clinical research networks.

CONCLUSIONS

This study designed an innovative ML-LHS unit for a clinical research network, aiming to sustainably provide inclusive LC screening to all at-risk populations. It developed an inclusive and practical XGBoost model from hospital electronic medical record data, capable of initializing such an ML-LHS unit for community and rural clinics. The anticipated deployment of this ML-LHS unit is expected to significantly improve LC-screening rates and early detection among broader populations, including those typically overlooked by existing screening guidelines.

Digital Interventions to Support Lung Cancer Screening: A Systematic Review

INTRODUCTION

Lung cancer remains a leading cause of cancer-related deaths globally. Lung cancer screening (LCS) with low-dose computed tomography (LDCT) can reduce lung cancer mortality, but its adoption in the U.S. has been limited. Digital interventions have the potential to improve uptake of LCS. This systematic review aims to summarize the evidence for the effectiveness of digital interventions in promoting LCS.

METHODS

A systematic search of three electronic databases (PubMed, Embase, and Medline) was conducted to identify studies published between January 2014 and May 2023. Studies were reviewed and abstracted between February 2023 and July 2023. Outcomes related to knowledge, decision-making and screening were measured. Study quality was assessed using the Joanna Briggs Institute (JBI) critical appraisal tools.

RESULTS

Of 1,979 screened articles, 30 studies were included in this review. Digital interventions evaluated included decision aids (n=20), electronic health record (EHR)-based interventions (n=7), social media campaigns and mobile applications (n=3). Decision aids were the most commonly studied digital interventions, with most studies showing improved knowledge (13/13) and reduced decisional conflict (7/9) but most did not show a substantial change in screening use. Fewer studies tested clinician-facing or multi-level interventions.

DISCUSSION

Digital interventions, particularly decision aids, have shown promise in improving knowledge and the quality of decision-making around LCS. However, few interventions have been shown to substantially alter screening behavior and few clinician-facing or multi-level interventions have been rigorously tested. Further research is needed to develop effective tools for engaging patients in LCS, to compare the efficacy of different interventions, and evaluate implementation strategies in diverse healthcare settings.

Engaging veteran stakeholders to identify patient-centred research priorities for optimizing implementation of lung cancer screening

BACKGROUND

Patient engagement in research agenda setting is increasingly being seen as a strategy to improve the responsiveness of healthcare to patient priorities. Implementation of low-dose computed tomography (LDCT) screening for lung cancer is suboptimal, suggesting that research is needed.

OBJECTIVES

This study aimed to describe an approach by which a Veteran patient group worked with other stakeholders to develop a research agenda for LDCT screening and to describe the research questions that they prioritized.

METHODS

We worked with Veterans organizations to identify 12 Veterans or family members at risk for or having experience with lung cancer to form a Patient Advisory Council (PAC). The PAC met repeatedly from June 2018 to December 2020, both independently and jointly, with stakeholders representing clinicians, health administrators and researchers to identify relevant research topics. The PAC prioritized these topics and then identified questions within these areas where research was needed using an iterative process. Finally, they ranked the importance of obtaining answers to these questions.

RESULTS

PAC members valued the co-learning generated by interactions with stakeholders, but emphasized the importance of facilitation to avoid stakeholders dominating the discussion. The PAC prioritized three broad research areas-(1) the impact of insurance on uptake of LDCT; (2) how best to inform Veterans about LDCT; and (3) follow-up and impact of screening results. Using these areas as guides, PAC members identified 20 specific questions, ranking as most important (1) innovative outreach methods, (2) the impact of screening on psychological health, and (3) the impact of outsourcing scans from VA to non-VA providers on completion of recommended follow-up of screening results. The latter two were not identified as high priority by the stakeholder group.

CONCLUSIONS

We present an approach that facilitates co-learning between Veteran patients and providers, researchers and health system administrators to increase patient confidence in their ability to contribute important information to a research agenda. The research questions prioritized by the Veterans who participated in this project illustrate that for this new screening technology, patients are concerned about the practical details of implementation (e.g., follow-up) and the technology's impact on quality of life.

PATIENT OR PUBLIC CONTRIBUTION

Veterans and Veteran advocates contributed to our research team throughout the entire research process, including conceiving and co-authoring this manuscript.