Objective: To perform a health economic evaluation of telemedicine diabetic retinopathy (DR) examination with a non-mydriatic fundus camera in China and to investigate the optimal examination interval. Methods: Based on 18 peer-reviewed articles related to epidemiology, clinical trial, and health economic evaluation of DR, surveys from 9 ophthalmologists in 3 tertiary hospitals in China, price lists for medical services in each province, and the negotiated price in 2021, a Markov model was conducted to evaluate the cost utility of telemedicine eye examination for diabetes mellitus patients aged 45 and older from the health system perspective. Separate analyses were performed for no examination and for examination intervals of every 1 to 5 years to predict the lifetime health gain, including cumulative days of blindness, cumulative life years, and quality-adjusted life years (QALYs), and costs for unilateral and bilateral direct medication with a 3.5% discount rate. Results: The cumulative days of blindness in the absence of a DR screening were 2 375.00 days, and ranged from 701.00 to 738.00 days for five different DR screening interval programs. The cumulative life years for no screening and five DR screening programs ranged from 27.120 34 to 28.005 00 years, with QALYs ranging from 9.502 96 to 9.875 02. The direct medication costs in the absence of a DR screening program were 72 785.00 yuan for both unilateral and bilateral scenarios. For the five DR screening intervals, the direct medication costs ranged from 52 065.00 to 52 408.00 yuan for unilateral and 79 100.00 to 79 603.00 yuan for bilateral. Comparing the incremental cost-effectiveness ratios between the DR screening intervals and no screening, the 1-to 5-year intervals were dominant in the unilateral scenario (between -56 368.54 and -55 523.75 yuan/QALY). In the bilateral scenario, the ratios ranged from 17 469.07 to 18 325.15 yuan/QALY. Using a willingness-to-pay threshold equal to the per capita GDP (80 976 yuan/QALY), the 1-year DR screening interval had an 85.9% probability of being cost-effective and a 55.2% probability of being dominant in the unilateral scenario. In the bilateral scenario, the 2-year interval held a 61.4% probability of being cost-effective. Conclusions: Analyses on the remote fundus consultation in diabetic patients and health economics based on the Markov model indicate that telemedicine DR examination through a non-mydriatic fundus camera can be effectively employed for diabetes mellitus patients in China. DR examination every two years is recommended for general diabetic patients, and DR examination every year may be chosen in developed areas.
目的： 探讨采用免散瞳眼底摄片法行糖尿病视网膜病变（DR）远程检查的卫生经济学价值和效果以及最具有成本效用的检查间隔时间方案。 方法： 基于18篇涉及DR流行病学、临床试验和卫生经济学的文献，全国3个三级甲等医院眼科9名医师的问卷调研资料和临床实际（全国各省医疗服务价格项目目录、2021年基本医疗保险谈判后的支付价格等），从中国医疗卫生服务体系角度构建Markov模型并设置参数，预测中国糖尿病患者从45岁开始无DR检查以及DR检查（均采用免散瞳眼底摄片法）间隔时间1、2、3、4、5年方案糖尿病患者的健康产出，包括累积失明天数、累积生命年、质量调整生命年（QALY，即健康效用值），以及相应的单眼、双眼直接医疗成本（采用3.5%贴现率），进行成本效用分析。 结果： 累积失明天数无DR检查方案为2 375.00 d，5种DR检查间隔时间方案为701.00~738.00 d。6种DR检查方案的累积生命年为27.120 34~28.005 00年，QALY为9.502 96~9.875 02。单眼直接医疗成本无DR检查方案为72 785.00元，5种DR检查间隔时间方案为52 065.00~52 408.00元。双眼直接医疗成本无DR检查方案为72 785.00元，5种DR检查间隔时间方案为79 100.00~79 603.00元。单眼治疗情况下，与无DR检查方案相比，DR检查间隔时间1、2、3、4、5年方案的增量成本效用比（ICER）分别为-55 523.75、-55 865.32、-56 034.66、-56 194.05、-56 368.54元/QALY，均具有绝对优势；双眼治疗情况下，与无DR检查方案相比，5种DR检查间隔时间方案的ICER为17 469.07~18 325.15元/QALY。以意愿支付阈值为1倍人均国内生产总值（80 976元/QALY），与无DR检查方案比较，单眼治疗情况下DR检查间隔时间1年方案的成本效用概率为85.9%，具有绝对优势的概率为55.2%；双眼治疗情况下，检查间隔时间2年方案具有成本效用的概率为61.4%。 结论： 以Markov模型对45岁以上糖尿病患者进行卫生经济学分析，提示免散瞳眼底摄片法具有较好的卫生经济学价值和效果，可作为我国DR远程检查方法；可采用DR检查间隔时间2年方案，即2年进行1次检查，有条件的地区可采用DR检查间隔时间1年方案，即每年进行1次检查。.