Abstract 289P
Background
Lung cancer can be treated curatively when diagnosed in the earlier stages. However, the symptoms are not commonly recognized by patients, resulting in late-stage diagnosis where survival rates are low. In low and middle income countries (LMICs) rates of lung cancer are increasing and improved methods of early detection are required to limit the burden of lung cancer. qXR® is an artificial intelligence (AI) software developed for the detection of abnormalities such as lung nodules on chest X-rays, currently the most commonly requested diagnostic test. The use of incidental pulmonary nodule detection could result in additional early-stage lung cancer diagnoses, requiring fewer resources compared to treatment in the late stages.
Methods
This study aimed to determine the impact of implementing AI at a population level for incidental pulmonary nodule detection on the budget of the healthcare system in Vietnam. A budget impact model with a time horizon of five years was developed, which compared the implementation of qXR to the current diagnostic process. A decision-tree design was used to determine the number of patients identified by qXR and their stage at diagnosis, while the national incidence and stage distribution are used for the current diagnostic process.
Results
The model predicts that the implementation of qXR results in 3,155 additional lung cancer diagnoses and 4,742 premature deaths prevented. While the additional diagnoses result in additional expenditures in the first year after implementation, the shift to early stages produce savings in subsequent years.
Conclusions
After five years the savings have counteracted the initial additional expenditures, making the implementation of qXR a costneutral intervention in Vietnam. In the future, this model will be used to assess the budget impact of qXR in other countries.
Legal entity responsible for the study
Qure.ai.
Funding
AstraZeneca Plc.
Disclosure
All authors have declared no conflicts of interest.