Clinical Report: Can AI Clarify Lung Screening?

Overview

A study found that large language model (LLM)-generated summaries improved patient comprehension of lung cancer screening reports compared to standard reports. Participants reported higher clarity and satisfaction with the LLM-enhanced summaries, although anxiety levels remained unchanged.

Background

As low-dose computed tomography (LDCT) screening for lung cancer becomes more prevalent, the complexity of radiology reports can hinder patient understanding. Clear communication of screening results is essential for informed decision-making and patient engagement in their healthcare. This study explores the potential of AI-generated summaries to enhance comprehension and patient experience in lung cancer screening.

Data Highlights

The study utilized a randomized vignette survey involving US adults (sample size needed), comparing standard radiology reports with LLM-generated summaries. Key findings included:

Key Findings

• Participants receiving LLM-generated summaries reported higher self-reported comprehension.
• Perceived clarity and satisfaction were significantly greater in the summary group across various Lung-RADS categories.
• The improvement in comprehension persisted after adjusting for age, education level, and health literacy.
• Anxiety levels did not differ significantly between the two groups.
• The study's outcomes were based on self-reported measures rather than objective assessments, and the scenarios were hypothetical.

Clinical Implications

The findings suggest that integrating LLM-generated summaries into lung cancer screening reports may enhance patient understanding and satisfaction, though barriers to implementation in clinical practice should be considered.

Conclusion

LLM-generated plain-language summaries have the potential to improve patient comprehension of lung cancer screening results, warranting further evaluation in real-world clinical settings and mechanisms to ensure accuracy.

References

1. JAMA Network Open, 2023 -- Can AI Clarify Lung Screening?
2. conexiant — AI May Improve Lung Nodule Detection
3. the asco post — AI Integration in Global Programs of CT Screening for Lung Cancer and Other Tobacco-Related Illnesses
4. asco ai in oncology — LungIMPACT Explores AI Triage in Lung Cancer Detection
5. The ASCO Post — Machine Learning–Guided ‘Optical Biopsy’ Accurately Identifies Malignant Lung Nodules Intraoperatively
6. AI May Improve Lung Nodule Detection
7. AI Integration in Global Programs of CT Screening for Lung Cancer and Other Tobacco-Related Illnesses
8. LungIMPACT Explores AI Triage in Lung Cancer Detection
9. Final Recommendation Statement: Lung Cancer: Screening | United States Preventive Services Taskforce
10. Reduced Lung-Cancer Mortality with Low-Dose Computed Tomographic Screening - PMC