USING DELIBERATIVE METHODS TO GUIDE IMPLEMENTATION OF LUNG CANCER SCREENING

Purpose: Implementing lung cancer screening (LCS) is complex and involves value judgments about benefits and harms in screened populations. We tested a multi-level deliberative process to guide LCS implementation in an academic health system. 

Method: We convened three deliberative group sessions, a Network Panel (institutional subspecialty and primary care experts and leaders), a Practice Panel (primary care practitioners and quality improvement coaches) and a one-day Citizens’ Panel (screening guideline-eligible patients). Sessions included an informational component (summarizing evidence regarding LCS benefits and harms) and a deliberation component (soliciting perspectives about screening implementation). We analyzed session notes and transcripts for themes and points of consensus. The Citizens’ Panel evaluation also included pre-post assessment of LCS knowledge including mortality reduction, false positives, over-detection/over-treatment. 

Result: The Network Panel (n=10) reached consensus on key implementation principles, including having uniform institution-wide protocols for offering LCS, prioritizing population benefit, providing high-quality decision support and smoking cessation interventions, and standardizing lung nodule reporting and management. Practitioners (n=13) endorsed these principles and provided specific suggestions for protocols to identify screening-eligible patients and for a one-page decision aid. Both Network and Practice panelists were willing to limit active and systematic offering of LCS to higher-risk subgroups most likely to benefit. For the Citizens Panel (n=11: 5 women, 5 current/6 former smokers, ages 56-72 years, average 49 pack-years smoked), themes included surprise and concern regarding the small absolute benefit of LCS, the high false positive rate, the possibility of over-detection/over-treatment, and potential anxiety and costs associated with evaluating screen-detected nodules. Participants suggested having trained nurse-educators help patients understand the complex benefits and harms of LCS. Participants agreed with actively offering LCS to groups most likely to benefit; however, they were divided on whether the topic of LCS should be systematically raised with patients who were guideline-eligible but had low chances of benefiting (5 for, 6 against). Pre-post LCS knowledge scores increased by 2.7 (95%CI 1.7, 3.8) points on a 6-point scale. 

Conclusion: Using a multi-level deliberative process to bring informed perspectives to health-system LCS implementation is feasible. Deliberation yielded broad consensus on key implementation principles. Citizens (patients) found the population-level benefit to be smaller and the harms greater than they expected, and they were divided on offering LCS to guideline-eligible patients with low likelihood of benefitting.