Presentation (CTI)

OBJECTIVES: Lung cancer (LC) is the most common cancer in Turkiye, accounting for 17.1% of new cancer diagnoses. It is also the leading cause of cancer-related deaths, with 38,505 deaths in 2022. Often asymptomatic in early stages, LC is typically detected as late-stage disease, limiting curative treatment options and resulting in low survival rates. Chest X-rays (CXR), the most widely used diagnostic modality, can play a critical role in incidental LC detection. qXR, an artificial intelligence (AI) software, can identify pulmonary and malignant nodules with high sensitivity, specificity, and accuracy. We developed a five-year budget impact model (BIM) to demonstrate the incremental costs and benefits of using AI-enabled CXR to identify incidental pulmonary nodules (IPN) within Turkiye's national healthcare system.
METHODS: A decision-tree model determined the number of patients identified incidentally per stage at diagnosis, with and without AI-enabled CXR. The model replicated the clinical pathway for lung cancer diagnosis, comparing AI-enabled CXR on routine imaging to current standard of care. An extensive range of costs was incorporated into the model, including set-up, diagnostics, healthcare resource use, treatment, and end-of-life costs.
RESULTS: In the first year of the model, the intervention resulted in additional expenditures due to an increased number of patients being diagnosed and undergoing treatment. After the initial investment, the cumulative budget impact turned negative as earlier-stage detection reduced treatment, healthcare resource use and end-of-life costs, yielding net cumulative savings nearing US$300 million by year five. Over the five-year time horizon, the model predicted over 6,000 premature deaths averted within the population.
CONCLUSIONS: The implementation of AI-enabled CXR in Turkiye is a cost-saving intervention over a five-year period. At population level, this intervention can help avert premature deaths due to lung cancer while saving healthcare costs.