Health economic analysis of early screening for diabetic retinopathy in the elderly

doi:10.3877/cma.j.issn.2095-2007.2022.04.005

Abstract

Abstract:

Objective

To explore the cost-effectiveness of screening for diabetic retinopathy (DR) among people over 50 years old in China from social and health system perspectives.

Methods

CNKI, Wanfang Database, PubMed and Medline from the beginning of self-built database to December 2020 were searched, and parameters such as prevalence, transition rate, screening and treatment compliance, screening sensitivity and specificity, utility value and mortality rate of DR among people over 50 years old in China were collected. According to the international classification system, DR was be divided into four stages: non-sight-threatening DR, sight-threatening DR, diabetic macular edema and blindness caused by DR. The cost of screening program, referral examination and treatment was based on the Handan Eye Study and inpatients system of Beijing Tongren Hospital affiliated with Capital Medical University in 2020. TreeAgePro software was used to build a Markov model of DR screening, in which the research strategies were community screening, telescreening and artificial intelligence (AI)-based screening, the control strategy was no-screening, and the cohort population were people over 50 years old in rural and urban areas of China. The cost-effectiveness of different strategies was compared, and the cost-effectiveness of 1 to 5 years screening interval was compared. In 2020, China′s rural and urban per capita GDP was ￥49 000 and ￥84 000, respectively. Economic benefits were expressed in terms of incremental cost-effectiveness ratio (ICUR) and incremental cost-effectiveness ratio (ICER). The WHO defined interventions that cost between one to three times the per capita gross domestic product (GDP) was defined as cost-effective, and interventions that cost less than the per capita GDP as highly cost-effective, and interventions that cost more than three times the per capita GDP as not cost-effective. The sensitivity analysis was used to evaluate the robustness of the model.

Results

From the societal perspective, the ICUR of community, telemedicine and AI-based screening for people over 50 years old for rural elderly were ￥4217, ￥2772 and ￥3206, respectively. The ICER was ￥216 524, ￥152 726 and ￥164 745, respectively. The ICUR for urban elderly was ￥7133, ￥5061 and ￥5894, respectively. The ICER was ￥57 792, ￥37 632 and ￥55 839, respectively. Com-pared with elderly without screening , ICUR for rural and urban elderly and ICER for urban elderly were less than the per capita GDP. However, the ICER of 3 screening methods for over 50 years rural elderly was slightly higher than the cost-effectiveness threshold. From health system perspective, ICUR and ICER of 3 screening methods for rural and urban elderly were lower than the cost-effectiveness threshold. Community and AI-assisted screening were cost-effective for urban elderly at intervals of more than 2 years, while telescreening at intervals of more than 1 year was cost-effective; however, it was not cost-effective to screen rural residents frequently. The sensitivity analysis showed that when the cost-effectiveness threshold was 3 times per capita GDP, the cost-effectiveness probabilities of community, telemedicine and AI-assisted screening for rural population were 87.34%, 87.56% and 87.54%, and the probability of urban population 89.19%, 89.48% and 89.61%, respectively. When the cost-effectiveness threshold was 1 time per capita GDP, the cost-effectiveness probability of 3 screening methods for rural eldly was 86.03%, 86.84% and 86.55%, and the cost-effectiveness probability of urban elderly population 87.83%, 88.55% and 88.62%, respectively. It showed that the design and results of Markov model were robust and insensitive to the fluctuation of parameters.

Conclusions

DR screening for the elderly over 50 years old in China is cost-effective. Due to the different levels of economic development and disease burdens in rural and urban areas, appropriate DR screening models should be developed according to local conditions.

Key words: Diabetic retinopathy, Screening, Artificial intelligence, Telemedicine screening, Cost-effectiveness, Markov model

Ruyue Li, Minghua Li, Kaiwen Zhang, Yue Zhang, Dapeng Mou, Ningli Wang, Hanruo Liu. Health economic analysis of early screening for diabetic retinopathy in the elderly[J]. Chinese Journal of Ophthalmologic Medicine(Electronic Edition), 2022, 12(04): 216-221.

/ / Recommend

Add to citation manager EndNote|Ris|BibTeX

URL: https://zhykyxzz.cma-cmc.com.cn/EN/10.3877/cma.j.issn.2095-2007.2022.04.005

https://zhykyxzz.cma-cmc.com.cn/EN/Y2022/V12/I04/216

Figures/Tables 3

References 30

[1]	李筱荣，杨千惠. 美国眼科学会糖尿病视网膜病变临床指南解读[J]. 中华实验眼科杂志，2020，38(9)：795－798.
[2]	中华医学会糖尿病学分会. 中国2型糖尿病防治指南(2020年版)[J]. 国际内分泌代谢杂志，2021，41(5)：482－548.
[3]	Harding S, Greenwood R, Aldington S, et al. Grading and disease management in national screening for diabetic retinopathy in England and Wales[J]. Diabet Med, 2003, 20: 965－971.
[4]	UK National Screening Committee. Essential elements in devel－oping a diabetic retinopathy screening programme. National Screening Programme for Diabetic Retinopathy[M]. England: National Health Service, 2009.
[5]	Wu B, Li J, Wu HX. Strategies to screen for diabetic retinopathy in chinese patients with newly diagnosed type 2 diabetes: a cost－effectiveness analysis[J]. Medicine, 2015, 94: e1989.
[6]	Tang JJ, Liang YB, O′Neill C, et al. Cost－effectiveness and cost－utility of population－based glaucoma screening in China: a decision－analytic Markov model[J]. Lancet Glob Health, 2019, 7: e968－e978.
[7]	Liang YB, Friedman DS, Wong TY, et al. Rationale, design, methodology, and baseline data of a population－based study in rural China: the Handan Eye Study[J]. Ophthalmic Epidemiol, 2009, 16(2): 115－127.
[8]	Tah V, Orlans HO, Hyer J, et al. Anti－VEGF therapy and the retina: an update[J]. J Ophthalmol, 2015: 627674.
[9]	Wong TY, Sun J, Kawasaki R, et al. Guidelines on diabetic eye care: the international council of ophthalmology recommendations for screening, follow－up, referral, and treatment based on resource settings[J]. Ophthalmology, 2018, 125: 1608－1622.
[10]	Jin GM, Xiao W, Ding XH, et al. Prevalence of and risk factors for diabetic retinopathy in a rural Chinese population: the Yangxi eye study[J]. Invest Ophthalmol Vis Sci, 2018, 59(12): 5067－5073.
[11]	Xu TL, Wang BS, Liu H, et al. Prevalence and causes of vision loss in China from 1990 to 2019: findings from the Global Burden of Disease Study 2019[J] .Lancet Public Health, 2020, 5: e682－e691.
[12]	Lian JX, Gangwani RA, McGhee SM, et al. Systematic screening for diabetic retinopathy (DR) in Hong Kong: prevalence of DR and visual impairment among diabetic population[J]. Br J Ophthalmol, 2016, 100(2): 151－155.
[13]	Jin G, Ding X, Xiao W, et al. Prevalence of age－related macular degeneration in rural southern China: the Yangxi Eye Study[J]. Br J Ophthalmol, 2018, 102(5): 625－630.
[14]	Ye HH, Zhang Q, Liu XH, et al. Prevalence of age－related macular degeneration in an elderly urban Chinese population in China: the Jiangning Eye Study[J]. Invest Ophthalmol Vis Sci, 2014, 55(10): 6374－6380.
[15]	Li R, Yang Z, Zhang Y, et al. Cost－effectiveness and cost－utility of traditional and telemedicine combined population－based age－related macular degeneration and diabetic retinopathy screening in rural and urban China[J]. Lancet Reg Health West Pac, 2022, 23: e100435.
[16]	Rachapelle S, Legood R, Alavi Y, et al. The cost－utility of telemedicine to screen for diabetic retinopathy in India[J]. Ophthalmology, 2013, 120(3): 566－573.
[17]	Chan CK, Gangwani RA, McGhee SM, et al. Cost－effectiveness of screening for intermediate age－related macular degeneration during diabetic retinopathy screening[J]. Ophthalmology, 2015，122(11): 2278－2785.
[18]	Wang DD, Ding XH, He MG, et al. Use of eye care services among diabetic patients in urban and rural China[J]. Ophthalmology, 2010, 117(9): 1755－1762.
[19]	Ullah W, Pathan SK, Panchal A, et al. Cost－effectiveness and diagnostic accuracy of telemedicine in macular disease and diabetic retinopathy: A systematic review and meta－analysis[J]. Medicine (Baltimore), 2020, 99(25): e20306.
[20]	Ting DSW, Cheung CY, Lim G, et al. Development and validation of a deep learning system for diabetic retinopathy and related eye diseases using retinal images from multiethnic populations with diabetes[J]. JAMA, 2017, 318(22): 2211－2223.
[21]	Bawankar P, Shanbhag N, Smitha KS, et al. Sensitivity and specificity of automated analysis of single－field non－mydriatic fundus photographs by Bosch DR Algorithm－comparison with mydriatic fundus photography (ETDRS) for screening in undiagnosed diabetic retinopathy[J]. PLoS One, 2017, 12(12): e0189854.
[22]	Zhang W, Wei M. The evaluation of the mortality and life expectancy of Chinese population[J]. Popul J, 2016, 38: 18－28.
[23]	李茹月，刘含若. 卫生经济学评价常见致盲眼病筛查的进展[J/CD]. 中华眼科医学杂志(电子版)，2021，11(2)：94－98.
[24]	Kirkizlar E, Serban N, Sisson JA, et al. Evaluation of telemedicine for screening of diabetic retinopathy in the Veterans Health Administration[J]. Ophthalmology, 2013, 120: 2604－2610.
[25]	Tung TH, Shih HC, Chen SJ, et al. Economic evaluation of screening for diabetic retinopathy among Chinese type 2 diabetics: a community－based study in Kinmen, Taiwan[J]. Epidemiology, 2008, 18(5): 225－233.
[26]	彭金娟，邹海东，王伟伟，等. 上海市北新泾糖尿病视网膜病变社区远程筛查与传统筛查系统的成本比较[J]. 中华健康管理学杂志，2010，4(5)：297－298.
[27]	Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs[J]. JAMA, 2016, 316(22): 2402－2410.
[28]	Li ZX, Keel S, Liu C, et al. An automated grading system for detection of vision－threatening referable diabetic retinopathy on the basis of color fundus photographs[J]. Diabetes Care, 2018，41(12): 2509－2516.
[29]	Raumviboonsuk P, Krause J, Chotcomwongse P, et al. Deep learning versus human graders for classifying diabetic retinopathy severity in a nationwide screening program[J]. NPJ Digit Med, 2019, 2: e25.
[30]	王亚星，薛灿灿，李建军．眼科人工智能研究存在的主要问题及对策[J]. 眼科，2021，30(2)：81－84.

参数名称		农村		城市
参数名称		数值	来源文献	数值	来源文献
患病率(%)
	非STDR	0.56	[10]	2.92	[11]
	STDR	0.22	[10]	0.98	[11]
	DME	0.34	[10]	0.86	[11]
	DR致盲	0.05	[12]	0.05	[12]
依从性(%)
	社区筛查	90.7	[10，13]	82.5	[14]
	远程筛查	95.0	[15]	96.8	[10，14]
	AI辅助筛查	95.0	[15]	96.8	[10，14]
	全面医院检查	19.0	[6]	57.0	[6]
	治疗	50.0	[16]	60.0	[17]
	机会性筛查	18.9	[18]	45.1	[18]
社区筛查敏感性(%)
	非STDR	76	[5]	76	[5]
	STDR	95	[5]	95	[5]
	DME	82	[19]	82	[19]
远程筛查敏感性(%)
	非STDR	91	[20]	91	[20]
	STDR	91	[20]	91	[20]
	DME	59	[19]	59	[19]
AI辅助筛查敏感性(%)
	非STDR	91	[5，20，21，22]	91	[5，20，21，22]
	STDR	100	[5，20，21，22]	100	[5，20，21，22]
	DME	59	[5，20，21，22]	59	[5，20，21，22]
特异性(%)
	社区筛查	100	[5，20，21，22]	100	[5，20，21，22]
	远程筛查	97	[5，20，21，22]	97	[5，20，21，22]
	AI辅助筛查	91	[5，20，21，22]	91	[5，20，21，22]
效用值(%)
	正常	1.00	[6]	1.00	[6]
	非STDR	0.87	[5，6，16]	0.87	[5，6，16]
	STDR	0.70	[5，6，16]	0.70	[5，6，16]
	DME	0.83	[5，6，16]	0.83	[5，6，16]
	DR致盲	0.26	[5，6，16]	0.26	[5，6，16]
死亡率(%)
	50~54岁	0.364	[22]	0.364	[22]
	55~59岁	0.518	[22]	0.518	[22]
	60~64岁	0.854	[22]	0.854	[22]
	65~69岁	1.421	[22]	1.421	[22]
	70~74岁	3.149	[22]	3.149	[22]
	75~79岁	4.861	[22]	4.861	[22]
	80~84岁	8.932	[22]	8.932	[22]

参数名称		农村		城市
参数名称		数值	来源文献	数值	来源文献
患病率(%)
	非STDR	0.56	[10]	2.92	[11]
	STDR	0.22	[10]	0.98	[11]
	DME	0.34	[10]	0.86	[11]
	DR致盲	0.05	[12]	0.05	[12]
依从性(%)
	社区筛查	90.7	[10，13]	82.5	[14]
	远程筛查	95.0	[15]	96.8	[10，14]
	AI辅助筛查	95.0	[15]	96.8	[10，14]
	全面医院检查	19.0	[6]	57.0	[6]
	治疗	50.0	[16]	60.0	[17]
	机会性筛查	18.9	[18]	45.1	[18]
社区筛查敏感性(%)
	非STDR	76	[5]	76	[5]
	STDR	95	[5]	95	[5]
	DME	82	[19]	82	[19]
远程筛查敏感性(%)
	非STDR	91	[20]	91	[20]
	STDR	91	[20]	91	[20]
	DME	59	[19]	59	[19]
AI辅助筛查敏感性(%)
	非STDR	91	[5，20，21，22]	91	[5，20，21，22]
	STDR	100	[5，20，21，22]	100	[5，20，21，22]
	DME	59	[5，20，21，22]	59	[5，20，21，22]
特异性(%)
	社区筛查	100	[5，20，21，22]	100	[5，20，21，22]
	远程筛查	97	[5，20，21，22]	97	[5，20，21，22]
	AI辅助筛查	91	[5，20，21，22]	91	[5，20，21，22]
效用值(%)
	正常	1.00	[6]	1.00	[6]
	非STDR	0.87	[5，6，16]	0.87	[5，6，16]
	STDR	0.70	[5，6，16]	0.70	[5，6，16]
	DME	0.83	[5，6，16]	0.83	[5，6，16]
	DR致盲	0.26	[5，6，16]	0.26	[5，6，16]
死亡率(%)
	50~54岁	0.364	[22]	0.364	[22]
	55~59岁	0.518	[22]	0.518	[22]
	60~64岁	0.854	[22]	0.854	[22]
	65~69岁	1.421	[22]	1.421	[22]
	70~74岁	3.149	[22]	3.149	[22]
	75~79岁	4.861	[22]	4.861	[22]
	80~84岁	8.932	[22]	8.932	[22]

项目		农村			城市
项目		直接医疗	直接非医疗	间接医疗	直接医疗	直接非医疗	间接医疗
筛查成本
	社区筛查	30	8	98	30	8	168
	远程筛查	17	8	49	17	8	84
	AI辅助筛查	11	8	49	11	8	84
确诊成本		798	861	196	798	8	168
治疗成本
	STDR	14 014	3444	784	14 014	32	672
	DME	12 943	3444	784	12 943	32	672
	致盲	33 219	3983	25 145	33 219	3983	25 145
随诊成本
	STDR和DME	4975	861	196	4975	8	168
	致盲	0	3983	25 145	0	3983	25 145

项目		农村			城市
项目		直接医疗	直接非医疗	间接医疗	直接医疗	直接非医疗	间接医疗
筛查成本
	社区筛查	30	8	98	30	8	168
	远程筛查	17	8	49	17	8	84
	AI辅助筛查	11	8	49	11	8	84
确诊成本		798	861	196	798	8	168
治疗成本
	STDR	14 014	3444	784	14 014	32	672
	DME	12 943	3444	784	12 943	32	672
	致盲	33 219	3983	25 145	33 219	3983	25 145
随诊成本
	STDR和DME	4975	861	196	4975	8	168
	致盲	0	3983	25 145	0	3983	25 145

项目		农村成本效用			农村成本效果		城市成本效用			城市成本效果
项目		成本(元/人)	QALYs/人	ICURs(元/人)	致盲年/人	ICERs(元/人)	成本(元/人)	QALYs/人	ICURs(元/人)	致盲年/人	ICERs(元/人)
社会角度
	无筛查项目	5432	16.236 82		0.143 886		9072	16.064 29		0.197 916
	社区筛查	5558	16.267 12	4214^*	0.143 302	216 524	9240	16.088 68	7133^*	0.194 908	57 792^*
	远程筛查	5516	16.267 10	2772^*	0.143 339	152 726	9191	16.088 84	5061^*	0.194 616	37 632^*
	AI辅助筛查	5530	16.267 12	3206^*	0.143 296	164 745	9212	16.089 01	5894^*	0.194 313	55 839^*
卫生系统角度
	无筛查项目	1428	16.236 82		0.143 886		3906	16.064 29		0.197 916
	社区筛查	1470	16.267 12	1379^*	0.143 302	71 757^*	4011	16.088 68	4130^*	0.194 908	33 516^*
	远程筛查	1463	16.267 10	1120^*	0.143 339	61 768^*	4018	16.088 84	4487^*	0.194 616	33 376^*
	AI辅助筛查	1470	16.267 12	1204^*	0.143 296	61 873^*	4046	16.089 01	5404^*	0.194 313	37 079^*

Please choose a citation manager

Content to export