Autonomous artificial intelligence increases screening and follow-up for diabetic retinopathy in youth: the ACCESS randomized control trial

Photo credit to Andri Munazir

Autonomous artificial intelligence increases screening and follow-up for diabetic retinopathy in youth: the ACCESS randomized control trial

Last Updated: March 22, 2024By Tags: , , ,

Abstract

Diabetic retinopathy can be prevented with screening and early detection. We hypothesized that autonomous artificial intelligence (AI) diabetic eye exams at the point-of-care would increase diabetic eye exam completion rates in a racially and ethnically diverse youth population. AI for Children’s diabetic Eye Exams (NCT05131451) is a parallel randomized controlled trial that randomized youth (ages 8-21 years) with type 1 and type 2 diabetes to intervention (autonomous artificial intelligence diabetic eye exam at the point of care), or control (scripted eye care provider referral and education) in an academic pediatric diabetes center. The primary outcome was diabetic eye exam completion rate within 6 months. The secondary outcome was the proportion of participants who completed follow-through with an eye care provider if deemed appropriate. Diabetic eye exam completion rate was significantly higher (100%, 95%CI: 95.5%, 100%) in the intervention group (n = 81) than the control group (n = 83) (22%, 95%CI: 14.2%, 32.4%)(p < 0.001). In the intervention arm, 25/81 participants had an abnormal result, of whom 64% (16/25) completed follow-through with an eye care provider, compared to 22% in the control arm (p < 0.001). Autonomous AI increases diabetic eye exam completion rates in youth with diabetes.

Subject terms: Diabetes complications, Laboratory techniques and procedures, Population screening, Randomized controlled trials

Diabetic retinopathy is a complication of diabetes that can be prevented through screening, yet adherence is low. Here, the authors show that autonomous AI increases diabetic eye exam completion in a diverse cohort of youth with diabetes.

Introduction

Diabetic eye disease (DED) is a complication of diabetes that is the primary cause of blindness in working-age adults in the U.S.,. Early detection (‘screening’) and treatment can frequently prevent progression, but the majority of the 34 million people with diabetes in the US have a DED screening care gap, due to a lack of access, and education around the need for a diabetic eye exam,. This care gap is a major source of health disparity, with racial and ethnic minorities, and under-resourced communities, having worse outcomes, and a disproportionately higher prevalence of DED,.

While DED prevalence is lower in youth (defined as those aged <21 years) with diabetes, it affects approximately 4–9% of youth with type 1 diabetes (T1D) and 4-15% of youth with type 2 diabetes (T2D). The risk for DED increases with the duration of diabetes in T1D, and recent data from the TODAY2 follow-up study demonstrated a diabetic retinopathy prevalence rate of 49% at a mean diabetes duration of 12 years in youth onset T2D. The American Diabetes Association (ADA) and American Academy of Ophthalmology (AAO) recommend DED screening in T1D within 3–5 years of diagnosis and age greater than 11 years, and in T2D at the time of diagnosis in youth. However, only 35-72% of diabetic youth undergo recommended screening exams, with even higher care gap rates in minority and lower socioeconomic background youth,. Commonly reported barriers to screening include miscommunication regarding the need for a diabetic eye exam, time for an additional doctor’s visit, and transportation barriers,.

While the introduction of telemedicine over the last two decades has improved screening and facilitated early detection of diabetic eye disease, the development of diagnostic autonomous artificial intelligence (AI) systems for diagnosing DED has ushered in the next chapter of DED screening. Autonomous AI systems require a camera operator to obtain point-of-care fundus images, which are then interpreted by an AI algorithm to provide a diagnosis without human oversight. The regulatory approval of the first diagnostic autonomous AI system for diabetic eye exams was based on a pivotal trial against a prognostic standard, i.e., patient outcome, showing its safety, efficacy, and lack of racial and ethnic bias for diagnosing DED in adults with diabetes with 87% sensitivity and 91% specificity. In prior studies of youth with diabetes, we demonstrated that the diagnosability of autonomous AI in youth was 97.5%, with 85.7% sensitivity and 79.3% specificity in detecting DED, with no difference in diagnosability across demographic groups. We have also shown that this system can be implemented in a multidisciplinary diabetes clinic and has the potential to increase DED screening rates in underserved youth, while also being cost-savings to patients and caregivers,.

While diagnostic accuracy has been a focus of study of diagnostic AI systems,,, the effectiveness of autonomous AI to increase adherence and follow-up compared to traditional referral has not been evaluated in a rigorously designed randomized trial. We hypothesized that autonomous AI closes the diabetic eye exam care gap, and increases follow-up, compared to traditional eye care provider(ECP) referral in youth. To test this hypothesis, we designed ACCESS (AI for Childrens’ diabetiC Eye examS Study), a pre-registered, rigorously designed randomized control trial (RCT), to measure diabetic eye exam completion rates in a racially and ethnically diverse cohort of youth with T1D and T2D.

References

1. Centers for Disease Control and Prevention Diabetes and Vision Losshttps://www.cdc.gov/diabetes/managing/diabetes-vision-loss.html (December 19, 2022).
2. Zhang X, et al. Prevalence of diabetic retinopathy in the United States, 2005-2008. JAMA. 2010;304:649–656. doi: 10.1001/jama.2010.1111. [PMC free article] [PubMed] [CrossRef[]
3. Zhang X, et al. Vision health disparities in the United States by race/ethnicity, education, and economic status: findings from two nationally representative surveys. Am. J. Ophthalmol. 2012;154:S53–62.e51. doi: 10.1016/j.ajo.2011.08.045. [PMC free article] [PubMed] [CrossRef[]
4. Hale NL, Bennett KJ, Probst JC. Diabetes care and outcomes: disparities across rural America. J. Community Health. 2010;35:365–374. doi: 10.1007/s10900-010-9259-0. [PubMed] [CrossRef[]
5. Hill-Briggs F, et al. Social determinants of health and diabetes: a scientific review. Diabetes Care. 2020;44:258–279. doi: 10.2337/dci20-0053. [PMC free article] [PubMed] [CrossRef[]
6. Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial. Diabetes Control and Complications Trial Research Group. J. Pediatr125, 177–188 (1994). [PubMed]
7. Benoit SR, Swenor B, Geiss LS, Gregg EW, Saaddine JB. Eye care utilization among insured people with diabetes in the U.S., 2010-2014. Diabetes Care. 2019;42:427–433. doi: 10.2337/dc18-0828. [PubMed] [CrossRef[]
8. Shi Q, Zhao Y, Fonseca V, Krousel-Wood M, Shi L. Racial disparity of eye examinations among the U.S. working-age population with diabetes: 2002-2009. Diabetes Care. 2014;37:1321–1328. doi: 10.2337/dc13-1038. [PMC free article] [PubMed] [CrossRef[]
9. Patel, D. et al. Social determinants of health and impact on screening, prevalence, and management of diabetic retinopathy in adults: a narrative review. J. Clin. Med.11. 10.3390/jcm11237120 (2022). [PMC free article] [PubMed]
10. Wang SY, et al. Ophthalmic screening patterns among youths with diabetes enrolled in a large US managed care network. JAMA Ophthalmol. 2017;135:432–438. doi: 10.1001/jamaophthalmol.2017.0089. [PMC free article] [PubMed] [CrossRef[]
11. Dabelea D, et al. Association of type 1 diabetes vs type 2 diabetes diagnosed during childhood and adolescence with complications during teenage years and young adulthood. JAMA. 2017;317:825–835. doi: 10.1001/jama.2017.0686. [PMC free article] [PubMed] [CrossRef[]
12. Badaru A, et al. Correlates of treatment patterns among youth with type 2 diabetes. Diabetes Care. 2014;37:64–72. doi: 10.2337/dc13-1124. [PMC free article] [PubMed] [CrossRef[]
13. Jensen, ET. et al. Prevalence, Progression, and Modifiable Risk Factors for Diabetic Retinopathy in Youth and Young Adults With Youth-Onset Type 1 and Type 2 Diabetes: The SEARCH for Diabetes in Youth Study. Diabetes Care46, 1252–1260 (2023). [PMC free article] [PubMed]
14. American Diabetes Association. Children and adolescents: standards of medical care in Diabetes-2020. Diabetes Care43, S163–S182 (2020). [PubMed]
15. Bjornstad P, et al. Long-term complications in youth-onset type 2 diabetes. N. Engl J. Med. 2021;385:416–426. doi: 10.1056/NEJMoa2100165. [PMC free article] [PubMed] [CrossRef[]
16. American Diabetes Association. Children and adolescents: standards of medical care in diabetes-2021. Diabetes Care44, S180–s199 (2021). [PubMed]
17. Thomas CG, et al. Racial/ethnic disparities and barriers to diabetic retinopathy screening in youths. JAMA Ophthalmol. 2021;139:791–795. doi: 10.1001/jamaophthalmol.2021.1551. [PMC free article] [PubMed] [CrossRef[]
18. Bruggeman B, et al. Barriers to retinopathy screening in youth and young adults with type 1 diabetes. Pediatr. Diabetes. 2021;22:469–473. doi: 10.1111/pedi.13171. [PubMed] [CrossRef[]
19. Kirkizlar E, et al. Evaluation of telemedicine for screening of diabetic retinopathy in the Veterans Health Administration. Ophthalmology. 2013;120:2604–2610. doi: 10.1016/j.ophtha.2013.06.029. [PubMed] [CrossRef[]
20. Wilson C, Horton M, Cavallerano J, Aiello LM. Addition of primary care-based retinal imaging technology to an existing eye care professional referral program increased the rate of surveillance and treatment of diabetic retinopathy. Diabetes Care. 2005;28:318–322. doi: 10.2337/diacare.28.2.318. [PubMed] [CrossRef[]
21. Liu J, et al. Diabetic retinopathy screening with automated retinal image analysis in a primary care setting improves adherence to ophthalmic care. Ophthalmol. Retina. 2021;5:71–77. doi: 10.1016/j.oret.2020.06.016. [PMC free article] [PubMed] [CrossRef[]
22. Conlin PR, et al. Nonmydriatic teleretinal imaging improves adherence to annual eye examinations in patients with diabetes. J. Rehabil. Res. Dev. 2006;43:733–740. doi: 10.1682/JRRD.2005.07.0117. [PubMed] [CrossRef[]
23. Davis RM, et al. Telemedicine improves eye examination rates in individuals with diabetes: a model for eye-care delivery in underserved communities. Diabetes Care. 2003;26:2476. doi: 10.2337/diacare.26.8.2476. [PubMed] [CrossRef[]
24. Abramoff MD, Lavin PT, Birch M, Shah N, Folk JC. Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices. NPJ Digit. Med. 2018;1:39. doi: 10.1038/s41746-018-0040-6. [PMC free article] [PubMed] [CrossRef[]
25. Ipp E, et al. Pivotal evaluation of an artificial intelligence system for autonomous detection of referrable and vision-threatening diabetic retinopathy. JAMA Network Open. 2021;4:e2134254. doi: 10.1001/jamanetworkopen.2021.34254. [PMC free article] [PubMed] [CrossRef[]
26. Rajalakshmi R, Subashini R, Anjana RM, Mohan V. Automated diabetic retinopathy detection in smartphone-based fundus photography using artificial intelligence. Eye (Lond) 2018;32:1138–1144. doi: 10.1038/s41433-018-0064-9. [PMC free article] [PubMed] [CrossRef[]
27. Sosale B, Sosale AR, Murthy H, Sengupta S, Naveenam M. Medios- An offline, smartphone-based artificial intelligence algorithm for the diagnosis of diabetic retinopathy. Indian J. Ophthalmol. 2020;68:391–395. doi: 10.4103/ijo.IJO_1203_19. [PMC free article] [PubMed] [CrossRef[]
28. Takahashi H, Tampo H, Arai Y, Inoue Y, Kawashima H. Applying artificial intelligence to disease staging: Deep learning for improved staging of diabetic retinopathy. PLoS ONE. 2017;12:e0179790. doi: 10.1371/journal.pone.0179790. [PMC free article] [PubMed] [CrossRef[]
29. Abramoff MD, et al. Autonomous artificial intelligence increases real-world specialist clinic productivity in a cluster-randomized trial. NPJ Digit. Med. 2023;6:184. doi: 10.1038/s41746-023-00931-7. [PMC free article] [PubMed] [CrossRef[]
30. Wolf RM, et al. The SEE Study: safety, efficacy, and equity of implementing autonomous artificial intelligence for diagnosing diabetic retinopathy in youth. Diabetes Care. 2021;44:781–787. doi: 10.2337/dc20-1671. [PubMed] [CrossRef[]
31. Wolf RM, Channa R, Abramoff MD, Lehmann HP. Cost-effectiveness of autonomous point-of-care diabetic retinopathy screening for pediatric patients with diabetes. JAMA Ophthalmol. 2020;138:1063–1069. doi: 10.1001/jamaophthalmol.2020.3190. [PMC free article] [PubMed] [CrossRef[]
32. Verbraak FD, et al. Diagnostic accuracy of a device for the automated detection of diabetic retinopathy in a primary care setting. Diabetes Care. 2019;42:651–656. doi: 10.2337/dc18-0148. [PubMed] [CrossRef[]
33. Abràmoff MD, et al. Foundational considerations for artificial intelligence using ophthalmic images. Ophthalmology. 2022;129:e14–e32. doi: 10.1016/j.ophtha.2021.08.023. [PMC free article] [PubMed] [CrossRef[]
34. NCQA. HEDIS Measurement Year 2020 and Measurement Year 2021. Technical Specifications for Health Plans2L (NCQA, 2020).
35. Center for Medicare and Medicaid Services. Proposal to Establish Values for Remote Retinal Imaging (CPT code 92229). Pages 56ff (2021).
36. American Diabetes Association. Microvascular complications and foot care: standards of medical care in diabetes-2020. Diabetes Care43, S135–S151 (2020). [PubMed]
37. Mansberger SL, et al. Comparing the effectiveness of telemedicine and traditional surveillance in providing diabetic retinopathy screening examinations: a randomized controlled trial. Telemed. J. E. Health. 2013;19:942–948. doi: 10.1089/tmj.2012.0313. [PMC free article] [PubMed] [CrossRef[]
38. Lin DY, Blumenkranz MS, Brothers RJ, Grosvenor DM. The sensitivity and specificity of single-field nonmydriatic monochromatic digital fundus photography with remote image interpretation for diabetic retinopathy screening: a comparison with ophthalmoscopy and standardized mydriatic color photography. Am. J. Ophthalmol. 2002;134:204–213. doi: 10.1016/S0002-9394(02)01522-2. [PubMed] [CrossRef[]
39. Bhargava M, et al. Accuracy of diabetic retinopathy screening by trained non-physician graders using non-mydriatic fundus camera. Singapore Med. J. 2012;53:715–719. [PubMed[]
40. Zhou P, et al. Assessment of trained image grader performance in screening for retinopathy among youth with diabetes. J. Diabetes Sci. Technol. 2022;16:1580–1581. doi: 10.1177/19322968221120240. [PMC free article] [PubMed] [CrossRef[]
41. Lock, L. J., Channa, R., Brennan, M. B., Cao, Y. & Liu, Y. Effect of health system on the association of rurality and level of disadvantage with receipt of diabetic eye screening. BMJ Open Diabetes Res. Care10. 10.1136/bmjdrc-2022-003174 (2022). [PMC free article] [PubMed]
42. Braveman P. What are health disparities and health equity? We need to be clear. Public Health Reports (Washington, D.C.: 1974) 2014;129:5–8. doi: 10.1177/00333549141291S203. [PMC free article] [PubMed] [CrossRef[]
43. Abramoff, M., Tarver, E. M., Loyo-Barrios, N., Trujillo, S., Char, D. Considerations for addressing bias in artificial intelligence for health equity. NPJ Digit Med6, 170 (2023). [PMC free article] [PubMed]
44. Channa R, Wolf RM, Abràmoff MD, Lehmann HP. Effectiveness of artificial intelligence screening in preventing vision loss from diabetes: a policy model. NPJ Digit Med. 2023;6:53. doi: 10.1038/s41746-023-00785-z. [PMC free article] [PubMed] [CrossRef[]
45. Bresnick, G. et al. Adherence to ophthalmology referral, treatment and follow-up after diabetic retinopathy screening in the primary care setting. BMJ Open Diabetes Res. Care8. 10.1136/bmjdrc-2019-001154 (2020). [PMC free article] [PubMed]
46. Keenum Z, et al. Patients’ adherence to recommended follow-up eye care after diabetic retinopathy screening in a publicly funded county clinic and factors associated with follow-up eye care use. JAMA Ophthalmol. 2016;134:1221–1228. doi: 10.1001/jamaophthalmol.2016.3081. [PubMed] [CrossRef[]
47. Zangalli CS, et al. An education- and telephone-based intervention to improve follow-up to vision care in patients with diabetes: a prospective, single-blinded, randomized Trial. Am. J. Med. Qual. 2016;31:156–161. doi: 10.1177/1062860614552670. [PubMed] [CrossRef[]
48. Pedersen, E. R. et al. Redesigning clinical pathways for immediate diabetic retinopathy screening results. NEJM Catal. Innov. Care Delivery2. 10.1056/CAT.21.0096 (2021).
49. Mathenge, W. et al. Impact of artificial intelligence assessment of diabetic retinopathy on referral service uptake in a low-resource setting: The RAIDERS randomized trial. Ophthalmol. Sci. 10.1016/j.xops.2022.100168 (2022). [PMC free article] [PubMed]
50. Benjamin L, Shou KV. Predictors for non-diagnostic images in real world deployment of artificial intelligence assisted diabetic retinopathy screening. Investig. Ophthalmol. Vis. Sci. 2022;63:1157. []
51. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ (Clinical Research ed.) 2010;340:c332. doi: 10.1136/bmj.c332. [PMC free article] [PubMed] [CrossRef[]
52. Harris PA, et al. The REDCap consortium: building an international community of software platform partners. J. Biomed. Inform. 2019;95:103208. doi: 10.1016/j.jbi.2019.103208. [PMC free article] [PubMed] [CrossRef[]
53. Harris PA, et al. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J. Biomed. Inform. 2009;42:377–381. doi: 10.1016/j.jbi.2008.08.010. [PMC free article] [PubMed] [CrossRef[]

news via inbox

Subscribe here to get latest updates !