ABSTRACT
Objective:To predict the financial impact of artificial intelligence (AI)-assisted community-based screening for primary angle-closure glaucoma (PACG) among the elderly on health expenses and medical insurance expenditures in remote areas of China.Methods:A total of 19 395 people aged more than 65 years old from Changjiang Li autonomous county, Hainan province were enrolled.A health economic model for glaucoma screening in Wenzhou, Zhejiang province was adopted.PACG was classified as suspected primary angle closure, primary angle closure, PACG, monocular blindness and binocular blindness according to its severity.The model was adjusted in combination with the epidemiology, metastasis probability and treatment compliance, screening and referral compliance, sensitivity and specificity of AI screening from glaucoma related literature and using data on local epidemiology and health costs of glaucoma from the Changjiang Hospital of Integrated Traditional Chinese and Western Medicine from 2016 to 2019.The health outcome, total health expenses and medical insurance expenditures of AI-assisted community-based screening and no screening for PACG in 15 years were compared and analyzed.Results:It is found that the AI-assisted community-based screening would prevent disease progression and reduce the number of PACG patients by 43% (134 cases) and blind patients by 50% (9 cases) in 15 years in comparison with no screening.Under the AI-assisted community-based screening system, the early diagnosis and intervention of positive PACG patients will cause the dramatic increase of both total health expenses and medical insurance expenditures compared with no screening in the short term.It was predicted that the increasing trend of medical insurance expenditure would turn to a downward trend and show cost saving advantage at the third year, whereas the declining trend of total health expense would not occur until 10 years later.Without screening, a large number of PACG patients would progress to advanced stages, so the medical costs would continue to rise and exceed the medical costs of the screening system at the 14th year.Conclusions:The AI-assisted community-based screening for PACG is cost-effective and can prevent disease progression, save health expenses and medical insurance expenditures in the long run in remote areas of China.